research paper on an animal

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Scientific research on animal biodiversity is systematically biased towards vertebrates and temperate regions

Roles Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliation Insect Ecology Group, University Museum of Zoology, Cambridge, Downing Street, Cambridge, United Kingdom

Roles Conceptualization, Writing – review & editing

Affiliation Biological Sciences, Institute for Life Sciences, University of Southampton, Southampton, United Kingdom

Roles Conceptualization, Funding acquisition, Methodology, Supervision, Writing – review & editing

• Mark A. Titley, 
• Jake L. Snaddon, 
• Edgar C. Turner

• Published: December 14, 2017
• https://doi.org/10.1371/journal.pone.0189577
• Reader Comments

Over the last 25 years, research on biodiversity has expanded dramatically, fuelled by increasing threats to the natural world. However, the number of published studies is heavily weighted towards certain taxa, perhaps influencing conservation awareness of and funding for less-popular groups. Few studies have systematically quantified these biases, although information on this topic is important for informing future research and conservation priorities. We investigated: i) which animal taxa are being studied; ii) if any taxonomic biases are the same in temperate and tropical regions; iii) whether the taxon studied is named in the title of papers on biodiversity, perhaps reflecting a perception of what biodiversity is; iv) the geographical distribution of biodiversity research, compared with the distribution of biodiversity and threatened species; and v) the geographical distribution of authors’ countries of origin. To do this, we used the search engine Web of Science to systematically sample a subset of the published literature with ‘biodiversity’ in the title. In total 526 research papers were screened—5% of all papers in Web of Science with biodiversity in the title. For each paper, details on taxonomic group, title phrasing, number of citations, study location, and author locations were recorded. Compared to the proportions of described species, we identified a considerable taxonomic weighting towards vertebrates and an under-representation of invertebrates (particularly arachnids and insects) in the published literature. This discrepancy is more pronounced in highly cited papers, and in tropical regions, with only 43% of biodiversity research in the tropics including invertebrates. Furthermore, while papers on vertebrate taxa typically did not specify the taxonomic group in the title, the converse was true for invertebrate papers. Biodiversity research is also biased geographically: studies are more frequently carried out in developed countries with larger economies, and for a given level of species or threatened species, tropical countries were understudied relative to temperate countries. Finally, biodiversity research is disproportionately authored by researchers from wealthier countries, with studies less likely to be carried out by scientists in lower-GDP nations. Our results highlight the need for a more systematic and directed evaluation of biodiversity studies, perhaps informing more targeted research towards those areas and taxa most depauperate in research. Only by doing so can we ensure that biodiversity research yields results that are relevant and applicable to all regions and that the information necessary for the conservation of threatened species is available to conservation practitioners.

Citation: Titley MA, Snaddon JL, Turner EC (2017) Scientific research on animal biodiversity is systematically biased towards vertebrates and temperate regions. PLoS ONE 12(12): e0189577. https://doi.org/10.1371/journal.pone.0189577

Editor: Bernd Schierwater, Tierarztliche Hochschule Hannover, GERMANY

Received: February 5, 2017; Accepted: November 29, 2017; Published: December 14, 2017

Copyright: © 2017 Titley et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The authors received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Since 1988, when the word was first used in a publication [ 1 ], the idea of ‘biodiversity’ has become integrated into both popular and scientific culture. The word produces more than 50 million hits on Google [ 2 ] and almost 90,000 in the scientific search engine and database Web of Science at the time of writing [ 3 ]. Moreover, systematic quantification of the number of papers studying biodiversity shows a marked increase over the last two decades ( Fig 1 ).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

A search for the word ‘biodiversity’ in Web of Science by year reveals the increase in biodiversity research over time (search date: 10 th February 2016).

https://doi.org/10.1371/journal.pone.0189577.g001

Biodiversity was formally defined at the 1992 United Nations Convention on Biological Diversity as ‘the variability among living organisms from all sources including, inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes of which they are part; this includes diversity within species, between species and of ecosystems’[ 4 ]. The most commonly used meaning is diversity at the species level, although despite being an intuitive concept, in practice definitions of what constitutes a species, and estimates of Earth’s species richness, remain uncertain and variable. Estimates for global species richness typically fall in the range of 3 million to 100 million species [ 5 ] although a working figure between 5 and 15 million is often suggested [ 6 ].

Contrary to this uncertainly, it is well established that diversity is not evenly distributed amongst taxa. Arthropods, and especially insects, account for most known eukaryote species: of the 1.2–2 million described species, approximately 925,000 are insects [ 7 , 8 ]. However, it has become clear that public perceptions of biodiversity do not reflect this invertebrate-dominated reality. In the UK, children asked to draw their ‘ideal rainforest’ over-represented mammals, reptiles and birds, and under-represented insects and annelids [ 9 ]. Such taxonomic chauvinism is by no means restricted to children, nor is it restricted to non-academics: 31% of papers published in 2001 in three prominent conservation journals focussed on birds and mammals [ 10 ]. Although this focus on larger species is understandable, owing to their greater apparency and potentially greater importance for ecosystem processes and vulnerability to environmental change [ 11 , 12 ], it does mean that invertebrate conservation issues and extinctions may go unreported or unacknowledged. This could hamper an overarching understanding of the state of the natural environment. For example, only 70 modern insect extinctions have been documented, despite thousands being estimated to have occurred [ 13 ].

Several previous studies have examined these taxonomic biases in journal articles. A survey of papers on vertebrates from nine high-impact journals reported a bias towards mammals and birds [ 14 ]. Furthermore, mammal and bird studies had more ‘narrowly framed’ introductions and mentioned the study organisms sooner than in studies on fishes, reptiles or amphibians. In a review of fifteen years of research from two leading conservation research journals ( Biological Conservation and Conservation Biology ), an over-representation of vertebrates and under-representation of invertebrates was revealed [ 15 ]. Within vertebrates, birds and mammals were over-represented, while other taxa were under-represented. A similar study analysed the research in three prominent conservation journals [ 10 ], finding once again a weighting in favour of vertebrates, as well as towards pristine landscapes and single species, rather than communities. Another study focussed on the research output of four ecological journals ( Journal of Animal Ecology , Journal of Applied Ecology , Oecologia , Ecology ) for the years 2006 and 2007 [ 16 ], and again highlighted the tendency to ignore invertebrates, in particular insects, in high-impact journals. Also reported was a preference in British Research Council NERC funding towards vertebrate ecologists (38%) compared with entomologists (13%).

Thus, the topic of taxonomic seems well studied, although these four papers all used a similar approach, focussing on the research output of a few selected journals. In the present article, we take a different, more wide-ranging approach, sampling across the published literature for papers whose title contains the word biodiversity. We therefore do not discriminate by journal (hence nor by impact factor), aiming to obtain a more holistic and longer-term view of taxonomic biases in global biodiversity research. In addition, we chose to investigate geographical biases, to assess whether biodiversity research is skewed towards certain regions and whether taxonomic biases are stronger in certain parts of the world.

Specifically, we first investigate whether reported taxonomic biases (towards vertebrates, and towards birds and mammals especially) pervade papers on biodiversity and whether this weighting has changed over time. Secondly, we investigate whether any bias differs between temperate and tropical regions. Thirdly, we investigate how the titles of papers on biodiversity are phrased. In particular, whether papers studying biodiversity differ in how likely they are to specify the study taxon in the title compared between papers on invertebrate and vertebrate biodiversity. This may reflect and promote a common (if subconscious) perception of which taxa represent biodiversity. Fourthly, we investigate the global distribution of biodiversity research, compared to the actual distribution of biodiversity, to assess how well research effort reflects biodiversity. We also compare it to the distribution of IUCN Red-Listed species and GDP, to assess how research effort reflects conservation priorities and wealth. Finally, we investigate the authors’ countries of origin relative to the study location, to assess whether there is a mismatch between the distribution of research on biodiversity and biodiversity researchers by country.

Materials & methods

Sample selection.

The scientific citation-indexing platform ‘Web of Science’ was used to sample research papers from the period 1995–2015, following a strict and repeatable search protocol. To be eligible for inclusion, papers’ title must have contained the word ‘biodiversity’, and also had to be a primary research article, in order to exclude review papers and other publication types such as books (which might have led to double-counting of studies). For each year, we then randomly selected 5% of all eligible articles using the random number generator www.random.org [ 17 ]. Five percent was an arbitrary figure that produced a sample size of 526 publications, which was quantifiable within the time frame of this project. This method may be cruder and return more irrelevant results than the careful examination of selected journals, but enabled us to easily generate a large sample size, and sample across a broad range of journals and disciplines over many years to obtain a comprehensive selection of biodiversity research. In this study we chose to focus on biases in animal biodiversity research, although we acknowledge that biases may also exist and be important across other taxonomic groups.

Data collection

For each of the 526 papers in our sample, we recorded the taxon/taxa studied; the climate zone (temperate or tropical) in which the study took place; whether or not the taxonomic group was specified in the title; the country in which the study took place; the country of origin of the paper’s authors; and the number of times that paper had been cited as recorded in Web of Science at the time of searching. Vertebrate studies were classified into one or more of five major vertebrate groups (Mammals, Birds, Reptiles, Amphibians and Fishes). Correspondingly, five major invertebrate groups were chosen because of their high species richness and because they are relatively well studied (Insects, Arachnids, Nematodes, Annelids, and Molluscs). Studies on invertebrates that could not be classified into these five groups were recorded as ‘Other invertebrates’. When recording the climate zone, we considered any studies taking place between the Tropics of Cancer and Capricorn (23.5°N and S respectively) as ‘tropical’. Since only six polar studies existed in the sample, there were not enough to include these as a separate climate zone. We therefore considered all studies taking place at latitudes higher than the tropics to be ‘temperate’. By this classification, studies in polar regions are also classified as temperate. For each author, their country of origin was recorded as the country of their affiliated institution. If a paper had multiple authors from different countries, multiple countries were recorded for the authors’ country of origin.

Data analysis

Statistical analyses were performed using R (version 3.0.2) [ 18 ]. To analyse the top 25% most-cited papers separately, the average number of citations per year was calculated (total citations to date divided by the time since publication). Chi-square tests were used to test for differences between temperate and tropical regions, and whether taxa were specified or not in the title. Wilcoxon rank-sum tests were used to test for differences between vertebrate and invertebrate residuals when comparing taxa for the proportion of studies versus proportion of described species as listed on the International Union for the Conservation of Nature (IUCN) database. Generalised linear models were used to test whether the number of biodiversity studies or authors in a country was related to Gross Domestic Product (GDP)–data from World Bank : World Development Indicators 2014 . Maps were created using QGIS (version 2.12.1) to visualise differences in research effort across countries worldwide. In particular, we mapped the number of biodiversity publications per 1000km 2 on vertebrates and invertebrates for each country, to visualise biases in research effort. We also mapped the number of authors relative to each country’s human population. By dividing the number of threatened species (data from IUCN [ 19 ]) by the number of biodiversity papers for each country, we also visualised countries that could be considered priorities for research (high numbers of threatened species relative to biodiversity research effort). Finally, analysis of covariance (ANCOVA) was used to test whether tropical and temperate regions differed in research effort for a given level of species or threatened species.

Taxonomic biases

Approximately half of the papers sampled studied vertebrates, and half studied invertebrates ( Fig 2 ). However, this is far from the true proportions of described species, where over 95% of species are invertebrates (see right-hand column of Fig 2 ). Furthermore, this focus on vertebrates has been roughly consistent over the last 20 years. Given their true species richness, vertebrates were significantly over-represented compared to invertebrates in the published literature (Wilcoxon rank-sum test, W = 24, N = 10, P<0.05) ( Fig 3 ). Invertebrate taxa were either slightly over-represented (annelids, molluscs, nematodes and ‘other invertebrates’) or under-represented (insects and arachnids). In addition, the taxonomic bias was greater in highly cited papers. Of the top 25% most cited papers in the sample, only 47% included invertebrates, compared with 57% of the entire sample.

The proportion of different taxonomic groups in the sample of papers with ‘biodiversity’ in the title is shown for 4 five-year periods since 1996. For comparison, the right-hand column illustrates the ‘true’ proportions of described species that each group makes up (data from IUCN [ 20 ]) Vertebrate and invertebrate taxa are separated by a grey line.

https://doi.org/10.1371/journal.pone.0189577.g002

The proportion of studies on each taxonomic group is plotted against the ‘actual’ proportion of described species [ 20 ] found in that taxon. Values were log transformed for clarity. The 1:1 line is shown (dotted); over-represented groups are found above the line while under-represented groups are below it. Vertebrate groups are shown in red and invertebrate groups are shown in blue.

https://doi.org/10.1371/journal.pone.0189577.g003

Comparing tropical and temperate regions

In terms of the proportion of studies, the bias towards vertebrates was greater in tropical regions than temperate regions (Chi-square test, X 2 = 30.65, N = 672, P<0.001) ( Fig 4 ). In tropical countries, 43% of studies included invertebrates, compared to 63% in temperate countries. General patterns of taxonomic over- or under-representation were similar in tropical and temperate regions, although arachnids were particularly under-represented in the tropics, and molluscs were under-represented in the tropics despite being over-represented in temperate studies.

The bias towards vertebrates is greater in tropical regions than temperate regions. The proportions of described species in different groups are shown in the right-hand column for comparison.

https://doi.org/10.1371/journal.pone.0189577.g004

Differences in title phrasing

The proportion of papers for which a taxonomic group was specified in the title differed between vertebrates and invertebrates (Chi-square test, X 2 = 103.45, N = 714, P<0.0001) ( Fig 5 ). Specifically, most papers that studied vertebrates did not specify the study taxon/taxa in the title, and instead referred to ‘biodiversity’ more generally. In contrast, the titles of studies on invertebrates usually specified which taxa were being studied. An exception to this pattern was studies on fishes, where the majority of studies specified the taxon in the title.

The majority of studies on vertebrates (with the exception of studies on fishes) do not mention the study taxon in the title. Conversely, for papers on invertebrates, the taxa being studied were specified more often than not.

https://doi.org/10.1371/journal.pone.0189577.g005

Geographic biases

Biodiversity research was more commonly carried out in developed countries with larger economies, for both vertebrate and invertebrate studies ( Fig 6 ). The United States of America had the highest number of studies of any country in the sample, but the density of biodiversity research appears to be generally highest in Western Europe. Most tropical areas had fewer studies and very little research was based in African countries. The number of biodiversity studies was positively related to countries’ nominal GDP (Poisson regression, z = 28.62, N = 232, P<0.0001) ( Fig 7 ).

The number of papers with ‘biodiversity’ in the title per 1000km 2 is shown, for a) papers that study vertebrates and b) papers that study invertebrates. Darker colours represent a higher density of studies.

https://doi.org/10.1371/journal.pone.0189577.g006

Nominal GDP in US$ is plotted against the number of biodiversity studies sampled from each country, revealing a positive relationship. The top ten countries for number of papers are labelled. Many countries with low GDP had no biodiversity papers identified from this sample.

https://doi.org/10.1371/journal.pone.0189577.g007

Certain counties had a higher number of threatened species relative to the biodiversity research effort (given by dividing the number of IUCN listed threatened species [ 19 ] by the number of research publications on biodiversity ( Fig 8 ). In particular, northern South America, Africa and SE Asia had a low relative number of publications. Note that large areas of Africa lacked any studies at all in our sample. We recorded a generally a positive relationship between the number of publications and the number of threatened and number of species recorded in the IUCN database [ 19 , 20 ] per country. However, for a given level of species or threatened species, tropical regions were understudied compared to temperate regions; interactions were significant between climate region and number of threatened species (F 3,227 = 36.06, p<0.0001) ( Fig 9A ) and between climate region and number of species (F 3 , 227 = 48.28, p<0.0001) ( Fig 9B ).

Dividing the number of animal species threatened with extinction [ 19 ] by the number of biodiversity studies reveals regions that are understudied given their number of threatened species. Countries in northern South America, Africa and SE Asia stand out as being relatively understudied; much of central Africa lacked studies altogether in this sample. Darker colours represent a higher number of listed threatened species per study.

https://doi.org/10.1371/journal.pone.0189577.g008

Scatterplots comparing the number of biodiversity papers against the number of threatened animal species (a) and species richness (b) listed in IUCN databases [ 19 , 20 ] per country. Temperate countries tend to have more biodiversity research than tropical countries for a given number of threatened species or a given species richness.

https://doi.org/10.1371/journal.pone.0189577.g009

As with the distribution of biodiversity research, the distribution of authors was heavily biased towards developed countries, particularly Western Europe ( Fig 10 ). Many countries in Africa, central Asia and South America lacked any authors on the papers in the sample; this is particularly true when looking at lead authors only ( Fig 10B ). The number of authors from a country was strongly related to wealth of that country as approximated by nominal GDP (Poisson regression, z = 69.91, N = 232, P < 0.0001). Furthermore, the GDP of authors’ countries of origin (median 2,066,902 million US$) was significantly higher than the GDP of study locations (median 1,453,770 million US$) (Wilcoxon rank-sum test, N = 513, W = 89086, P < 0.0001).

The number of authors (a) and lead authors (b) from each country relative to the country’s population. Many countries in Africa, central Asia and South America lacked authors on the papers in the sample.

https://doi.org/10.1371/journal.pone.0189577.g010

The taxonomic bias

These results clearly demonstrate more charismatic animal groups are over-represented in biodiversity research and have been since biodiversity first emerged as a research field. Mammals, which make up around 0.4% of known animal species [ 20 ], were studied in approximately 12% of papers with biodiversity in the title. The equivalent numbers for birds are 0.7% and 13%. In contrast, insects make up at least 70% of animal species [ 20 ] yet were studied in less than a quarter (23%) of papers. This result corroborates earlier findings [ 10 , 14 – 16 ], and extends the phenomenon to all biodiversity research rather than just that of selected journals. Due to the high proportion of species remaining to be described, particularly among the invertebrates, this figure is likely to be conservative. These results have implications for awareness of the natural world in the scientific community, particularly as this taxonomic bias was greater in the top quartile of most-cited papers, suggesting that the research with the highest impact and largest influence is even less representative of the real world in this regard.

The taxonomic bias was greater in tropical regions, where vertebrates were studied in more than half of papers, despite vertebrates comprising less than 5% of animal species. As tropical countries contain a higher total species number and are therefore likely to have a much higher proportion of undescribed species [ 5 ], particularly smaller taxa, this under-representation is likely to be even more marked in reality. Ensuring adequate research coverage across taxa in tropical regions has important conservation implications. Most species are found in the tropics [ 21 ] and tropical regions encompass many of the world’s conservation priority hotspots [ 22 ], but are currently experiencing habitat loss faster than any other region [ 23 ].

Not all invertebrate taxa were underrepresented however; in fact, four out of the six invertebrate groups were somewhat over-represented in scientific research. The overall lack of invertebrate studies is, more precisely, a dearth of global insect and arachnid research and tropical mollusc research. The fact that insects and arachnids were the least well represented groups in this study does not mean they are the least represented of all taxa, since there will be other poorly studied invertebrate groups included within the other invertebrates category, or within these groups at a finer taxonomic scale. However, since arachnids and insects are so speciose, the deficiency of research in these groups is perhaps most significant to understanding global biodiversity. Another key finding relating to taxonomic bias is that studies on vertebrates typically did not specify the taxon in the title, referring to ‘biodiversity’ more generally. This was not the case for invertebrate research, for which the study taxa were usually specified. This could reflect a general perception that vertebrates alone are sufficient to represent biodiversity.

This unequal coverage of research across taxa may have a complex combination of causes. Researchers themselves may find studying charismatic vertebrates more appealing. Alternatively, it could represent the increased challenges of working with more diverse taxa, particularly in terms species identification. This is despite studies showing that certain insect groups are informative indicators of biodiversity and cost effective taxa to sample [ 24 , 25 ]. General perceptions of biodiversity may also be influenced by journal editors publishing a disproportionate number of articles on vertebrates (consciously or subconsciously), because such articles may be more likely to gain traction within a scientific community that is already vertebrate-biased (especially if journals are under pressure to maintain a high impact factor driven by citations). Vertebrate-biased research may also appeal to the media who are catering for a vertebrate-preferring public audience [ 9 ]. The taxonomic bias could also be the product of funding bodies, which may preferentially award research grants for vertebrate studies if these are perceived to be more important, interesting or relevant to conservation and policy priorities. A few or all of these hypotheses may play a role in producing the biases reported in this study.

Taxonomic bias is not necessarily bad. A bias towards charismatic vertebrate taxa may be advantageous where such taxa have a disproportionately large role in ecosystem functioning (keystone species), in generating funds and support for conservation (flagship species), or when their protection also ensures the protection of much of their ecosystem (umbrella species) [ 26 , 27 ]. In addition, certain taxa may be used as surrogates for other harder-to-study groups [ 28 , 29 ], which may have a similar geographic distribution or show a similar response to disturbance. However, notwithstanding doubt over the prevalence of keystone species and the reliability of taxonomic surrogates [ 30 , 31 ], it is unlikely that the taxonomic bias we have observed has arisen as a result of deliberate decisions to select these taxa as indicators of other lesser-known animal groups.

In using the proportion of described species as a reference for many of our analyses, we implicitly make the assumption that all species are equal. However, clearly this is not the case in terms of ecosystem function or conservation priority. It would be interesting to investigate whether the proportion of research done on different taxonomic groups better reflects the distribution of ecological importance or conservation value among taxa (rather than the proportion of described species), but it remains a challenge to identify meaningful measures for these that are comparable across taxa and globally applicable [ 32 ].

The geographic bias

The distribution of biodiversity research and its authors’ countries of origin resemble the distribution of GDP, rather than that of actual biodiversity or numbers of threatened species. The distribution of research is skewed towards developed countries and particularly Western Europe. Furthermore, even when studies are carried out in lower GDP-countries, the authors tend to be based at institutions in wealthier nations. Tropical countries tend to have fewer biodiversity studies despite being where more biodiversity is found and where biodiversity is most threatened. Tropical regions were also where the taxonomic bias was greatest. Taken together, these findings have important implications for biodiversity conservation: the same areas that are most threatened and most diverse are the least studied [ 23 ] and where scientists research is most skewed towards less-speciose groups. Therefore, we are likely to continue to undervalue these under-studied groups, especially in parts of the world where they are most threatened, and perhaps allocate less funding to their protection. Moreover, given that conservation efforts will be more likely to succeed when we better understand the target organisms, there is a real possibility that we may be ill equipped to protect the majority of animal biodiversity. Research gaps may mean we are less likely to identify threatened invertebrates and notice their disappearance, and we may be less likely to identify underlying threats and their drivers. Furthermore, without a good understanding of invertebrate biodiversity loss, we may suffer a reduced ability to predict subsequent anthropogenic impacts on ecosystems worldwide. Given that funding and time are limited, biodiversity research should be focussed on certain taxa for scientifically justified reasons, rather than because of an underlying subjectivity in what we consider to be important. Crucially, conservationists need to be more aware of these unequal weightings to prevent biodiverse taxa being overlooked or understudied.

Redressing biases

Significant challenges remain in addressing the biases we found. One is to popularize these lesser-known taxa to allow recognition of their importance. This could be achieved through more targeted funding for these invertebrate groups (and under-represented countries). Another challenge is to ease the practical issues of identification and research on these taxa [ 33 ]. Opportunities may be found in novel techniques such as metagenomic sequencing [ 34 ], or the development of apps that aid easy identification worldwide [ 35 ]. The use of modern media may ease access to specimens digitally, and help to put researchers and taxonomic experts in touch. It will require a concerted effort to redress these research biases and to ensure the least studied taxa and countries do not remain so, thus ensuring that we maximise the contribution of biodiversity research to our understanding of nature, and minimise the further erosion of biodiversity in our increasingly imperiled world.

Supporting information

S1 dataset. data used in this study..

See ‘metadata’ sheet for more information.

https://doi.org/10.1371/journal.pone.0189577.s001

Acknowledgments

We would like to thank Xavier Bonnet and another anonymous reviewer for their very helpful comments on the manuscript.

• 1. Wilson EO. Biodiversity. National Academies Press; 1988.
• 2. Google [Internet]. 2016 [cited 16 Aug 2016]. Available: https://www.google.co.uk/#q=biodiversity
• 3. Web of Science [Internet]. 2016 [cited 6 Feb 2016]. Available: https://apps.webofknowledge.com
• 4. United Nations. Convention on Biological Diversity [Internet]. 1992. https://doi.org/10.1146/annurev.ento.48.091801.112645
• 5. Gaston KJ. Biodiversity. In: Sodhi N, Ehrlich PR, editors. Conservation Biology for All. Oxford University Press; 2010. https://doi.org/10.1093/acprof:oso/9780199554232.001.0001
• 6. May RM. The dimensions of life on Earth. In: Raven PH, Williams T, editors. Nature and Human Society: the Quest for a Sustainable World. Washington: National Academy; 2000. pp. 30–45.
• 7. Grimaldi D, Engel MS. Evolution of the Insects. Cambridge University Press; 2005.
• View Article
• Google Scholar
• PubMed/NCBI
• 17. RANDOM.ORG [Internet].
• 18. Core Team R. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2013. Available: http://www.r-project.org/
• 19. IUCN. Table 5: Threatened species in each country (totals by taxonomic group) [Internet]. 2016 [cited 14 Mar 2016]. Available: http://cmsdocs.s3.amazonaws.com/summarystats/2016-2_Summary_Stats_Page_Documents/2016_2_RL_Stats_Table_5_CORRECTED.pdf
• 20. IUCN. Table 1: Numbers of threatened species by major groups of organisms (1996–2014) [Internet]. 2014 [cited 14 Mar 2016]. Available: http://cmsdocs.s3.amazonaws.com/summarystats/2016-2_Summary_Stats_Page_Documents/2016_2_RL_Stats_Table_1.pdf

Advertisement

The impact of induced pluripotent stem cells in animal conservation

• Published: 16 January 2024
• Volume 48 , pages 649–663, ( 2024 )

Cite this article

• Yurou Wu 1 ,
• Chengwei Wang 1 ,
• Xinyun Fan 1 ,
• Yuxiao Ma 2 ,
• Zibo Liu 1 ,
• Chongyang Shen 3 &
• Chunjie Wu 4 , 5  

418 Accesses

1 Altmetric

Explore all metrics

It is widely acknowledged that we are currently facing a critical tipping point with regards to global extinction, with human activities driving us perilously close to the brink of a devastating sixth mass extinction. As a promising option for safeguarding endangered species, induced pluripotent stem cells (iPSCs) hold great potential to aid in the preservation of threatened animal populations. For endangered species, such as the northern white rhinoceros ( Ceratotherium simum cottoni ), supply of embryos is often limited. After the death of the last male in 2019, only two females remained in the world. IPSC technology offers novel approaches and techniques for obtaining pluripotent stem cells (PSCs) from rare and endangered animal species. Successful generation of iPSCs circumvents several bottlenecks that impede the development of PSCs, including the challenges associated with establishing embryonic stem cells, limited embryo sources and immune rejection following embryo transfer. To provide more opportunities and room for growth in our work on animal welfare, in this paper we will focus on the progress made with iPSC lines derived from endangered and extinct species, exploring their potential applications and limitations in animal welfare research.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Stem cells on regenerative and reproductive science in domestic animals

Naira Caroline Godoy Pieri, Aline Fernanda de Souza, … Fabiana Fernandes Bressan

Somatic Cells, Stem Cells, and Induced Pluripotent Stem Cells: How Do They Now Contribute to Conservation?

The use of induced pluripotent stem cells in domestic animals: a narrative review

Rachel A. Scarfone, Samantha M. Pena, … Thomas G. Koch

Data availability

Not applicable.

2018. DNA offers glimmer of hope for critically endangered rhino. Nature 558:164

Amano N, Wang YV, Boivin N, Roberts P (2021) “Emptying forests?” conservation implications of past human-primate interactions. Trends Ecol Evol 36:345–359

Article   PubMed   Google Scholar  

Anthony E, Lovell-Badge R, Morrison SJ (2021) New guidelines for stem cell and embryo research from the ISSCR. Cell Stem Cell 28:991–992

Article   CAS   PubMed   Google Scholar  

Bai C, Li X, Gao Y, Yuan Z, Hu P, Wang H, Liu C, Guan W, Ma Y (2016) Melatonin improves reprogramming efficiency and proliferation of bovine-induced pluripotent stem cells. J Pineal Res 61:154–167

Bao L, He L, Chen J, Wu Z, Liao J, Rao L, Ren J, Li H, Zhu H, Qian L, Gu Y, Dai H, Xu X, Zhou J, Wang W, Cui C, Xiao L (2011) Reprogramming of ovine adult fibroblasts to pluripotency via drug-inducible expression of defined factors. Cell Res 21:600–608

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ben-Nun IF, Montague SC, Houck ML, Tran HT, Garitaonandia I, Leonardo TR, Wang YC, Charter SJ, Laurent LC, Ryder OA, Loring JF (2011) Induced pluripotent stem cells from highly endangered species. Nat Methods 8:829–831

Ben-Nun IF, Montague SC, Houck ML, Ryder O, Loring JF (2015) Generation of induced pluripotent stem cells from mammalian endangered species. Methods Mol Biol 1330:101–109

Bernard L, Lindsay P (2009) Ethical issues in stem cell research. Endocr Rev 30:204–213

Article   Google Scholar  

Botigelli RC, Guiltinan C, Arcanjo RB, Denicol AC (2023) In vitro gametogenesis from embryonic stem cells in livestock species: recent advances, opportunities, and challenges to overcome. J Anim Sci 101:skad137

Article   PubMed   PubMed Central   Google Scholar  

Cai H, Xia X, Wang L, Liu Y, He Z, Guo Q, Xu C (2013) In vitro and in vivo differentiation of induced pluripotent stem cells into male germ cells. Biochem Biophys Res Commun 433:286–291

Calatayud NE, Jacobs LE, Williams CL, Steiner CC, Shier DM (2022) Recovering an endangered frog species through integrative reproductive technologies. Theriogenology 191:141–152

Camporesi S (2007) The context of embryonic development and its ethical relevance. Biotechnol J 2:1147–1153

Cao H, Yang P, Pu Y, Sun X, Yin H, Zhang Y, Zhang Y, Li Y, Liu Y, Fang F, Zhang Z, Tao Y, Zhang X (2012) Characterization of bovine induced pluripotent stem cells by lentiviral transduction of reprogramming factor fusion proteins. Int J Biol Sci 8:498–511

Ceballos G, Ehrlich PR, Raven PH (2020) Vertebrates on the brink as indicators of biological annihilation and the sixth mass extinction. Proc Natl Acad Sci U S A 117:13596–13602

Chanyandura A, Muposhi VK, Gandiwa E, Muboko N (2021) An analysis of threats, strategies, and opportunities for african rhinoceros conservation. Ecol Evol 11:5892–5910

Chen L, Tang L, Xiang H, Jin L, Li Q, Dong Y, Wang W, Zhang G (2014) Advances in genome editing technology and its promising application in evolutionary and ecological studies. Gigascience 3:24

Chen H, Zuo Q, Wang Y, Song J, Yang H, Zhang Y, Li B (2017) Inducing goat pluripotent stem cells with four transcription factor mRNAs that activate endogenous promoters. BMC Biotechnol 17:11

Comizzoli P (2015) Biotechnologies for wildlife fertility preservation. Anim Front 5:73–78

Comizzoli P, Holt WV (2019) Breakthroughs and new horizons in reproductive biology of rare and endangered animal species. Biol Reprod 101:514–525

Cowie RH, Bouchet P, Fontaine B (2022) The sixth mass extinction: fact, fiction or speculation? Biol Rev Camb Philos Soc 97:640–663

Cui YH, Chen W, Wu S, Wan CL, He Z (2023) Generation of male germ cells in vitro from the stem cells. Asian J Androl 25:13–20

De Los Angeles A, Ferrari F, Xi R, Fujiwara Y, Benvenisty N, Deng H, Hochedlinger K, Jaenisch R, Lee S, Leitch HG, Lensch MW, Lujan E, Pei D, Rossant J, Wernig M, Park PJ, Daley GQ (2015) Hallmarks of pluripotency. Nature 525:469–478

Déjosez M, Marin A, Hughes GM, Morales AE, Godoy-Parejo C, Gray JL, Qin Y, Singh AA, Xu H, Juste J, Ibáñez C, White KM, Rosales R, Francoeur NJ, Sebra RP, Alcock D, Volkert TL, Puechmaille SJ, Pastusiak A, Frost SDW, Hiller M, Young RA, Teeling EC, García-Sastre A, Zwaka TP (2023) Bat pluripotent stem cells reveal unusual entanglement between host and viruses. Cell 186:957-974.e928

Dejosez M, Zwaka TP (2012) Pluripotency and nuclear reprogramming. Annu Rev Biochem 81:737–765

Dutton LC, Dudhia J, Guest DJ, Connolly DJ (2019) Inducing pluripotency in the domestic cat (Felis catus). Stem Cells Dev 28:1299–1309

Endo Y, Kamei KI, Hasegawa K, Okita K, Ito H, Terada S, Inoue-Murayama M (2022) Generation and gene expression profiles of grevy’s zebra induced pluripotent stem cells. Stem Cells Dev 31:250–257

Evans MJ, Kaufman MH (1981) Establishment in culture of pluripotential cells from mouse embryos. Nature 292:154–156

Ezashi T, Yuan Y, Roberts RM (2016) Pluripotent stem cells from domesticated mammals. Annu Rev Anim Biosci 4:223–253

Folch J, Cocero MJ, Chesné P, Alabart JL, Domínguez V, Cognié Y, Roche A, Fernández-Arias A, Martí JI, Sánchez P, Echegoyen E, Beckers JF, Bonastre AS, Vignon X (2009) First birth of an animal from an extinct subspecies (Capra pyrenaica pyrenaica) by cloning. Theriogenology 71:1026–1034

Fredga K (1988) Aberrant chromosomal sex-determining mechanisms in mammals, with special reference to species with XY females. Philos Trans R Soc Lond B Biol Sci 322:83–95

Fu R, Yu D, Ren J, Li C, Wang J, Feng G, Wang X, Wan H, Li T, Wang L, Zhang Y, Hai T, Li W, Zhou Q (2020) Domesticated cynomolgus monkey embryonic stem cells allow the generation of neonatal interspecies chimeric pigs. Protein Cell 11:97–107

Fu B, Ma H, Liu D (2022) 2-cell-like cells: an avenue for improving SCNT efficiency. Biomolecules 12:1611

Fuet A, Montillet G, Jean C, Aubel P, Kress C, Rival-Gervier S, Pain B (2018) NANOG is required for the long-term establishment of avian somatic reprogrammed cells. Stem Cell Reports 11:1272–1286

Fujie Y, Fusaki N, Katayama T, Hamasaki M, Soejima Y, Soga M, Ban H, Hasegawa M, Yamashita S, Kimura S, Suzuki S, Matsuzawa T, Akari H, Era T (2014) New type of Sendai virus vector provides transgene-free iPS cells derived from chimpanzee blood. PLoS ONE 9:e113052

Gafni O, Weinberger L, Mansour AA, Manor YS, Chomsky E, Ben-Yosef D, Kalma Y, Viukov S, Maza I, Zviran A, Rais Y, Shipony Z, Mukamel Z, Krupalnik V, Zerbib M, Geula S, Caspi I, Schneir D, Shwartz T, Gilad S, Amann-Zalcenstein D, Benjamin S, Amit I, Tanay A, Massarwa R, Novershtern N, Hanna JH (2013) Derivation of novel human ground state naive pluripotent stem cells. Nature 504:282–286

Gómez MC, Jenkins JA, Giraldo A, Harris RF, King A, Dresser BL, Pope CE (2003) Nuclear transfer of synchronized african wild cat somatic cells into enucleated domestic cat oocytes. Biol Reprod 69:1032–1041

Gorczyca G, Wartalski K, Wiater J, Samiec M, Tabarowski Z, Duda M (2021) Anabolic steroids-driven regulation of porcine ovarian putative stem cells favors the onset of their neoplastic transformation. Int J Mol Sci 22:11800

Gross M (2018) Last call to save the rhinos. Curr Biol 28:R1–R3

Guan J, Wang G, Wang J, Zhang Z, Fu Y, Cheng L, Meng G, Lyu Y, Zhu J, Li Y, Wang Y, Liuyang S, Liu B, Yang Z, He H, Zhong X, Chen Q, Zhang X, Sun S, Lai W, Shi Y, Liu L, Wang L, Li C, Lu S, Deng H (2022) Chemical reprogramming of human somatic cells to pluripotent stem cells. Nature 605:325–331

Gubbay J, Collignon J, Koopman P, Capel B, Economou A, Münsterberg A, Vivian N, Goodfellow P, Lovell-Badge R (1990) A gene mapping to the sex-determining region of the mouse Y chromosome is a member of a novel family of embryonically expressed genes. Nature 346:245–250

Hayashi K, Ohta H, Kurimoto K, Aramaki S, Saitou M (2011) Reconstitution of the mouse germ cell specification pathway in culture by pluripotent stem cells. Cell 146:519–532

Hayashi M, Zywitza V, Naitou Y, Hamazaki N, Goeritz F, Hermes R, Holtze S, Lazzari G, Galli C, Stejskal J, Diecke S, Hildebrandt TB, Hayashi K (2022) Robust induction of primordial germ cells of white rhinoceros on the brink of extinction. Sci Adv 8:eabp9683

Hermes R, Hildebrandt TB, Göritz F, Fasel NJ, Holtze S (2019) First cryopreservation of phyllostomid bat sperm. Theriogenology 131:28–31

Herrick JR (2019) Assisted reproductive technologies for endangered species conservation: developing sophisticated protocols with limited access to animals with unique reproductive mechanisms. Biol Reprod 100:1158–1170

Hikabe O, Hamazaki N, Nagamatsu G, Obata Y, Hirao Y, Hamada N, Shimamoto S, Imamura T, Nakashima K, Saitou M, Hayashi K (2016) Reconstitution in vitro of the entire cycle of the mouse female germ line. Nature 539:299–303

Hildebrandt TB, Hermes R, Walzer C, Sós E, Molnar V, Mezösi L, Schnorrenberg A, Silinski S, Streich J, Schwarzenberger F, Göritz F (2007) Artificial insemination in the anoestrous and the postpartum white rhinoceros using GnRH analogue to induce ovulation. Theriogenology 67:1473–1484

Hildebrandt TB, Hermes R, Colleoni S, Diecke S, Holtze S, Renfree MB, Stejskal J, Hayashi K, Drukker M, Loi P, Göritz F, Lazzari G, Galli C (2018) Embryos and embryonic stem cells from the white rhinoceros. Nat Commun 9:2589

Hildebrandt TB, Hermes R, Goeritz F, Appeltant R, Colleoni S, de Mori B, Diecke S, Drukker M, Galli C, Hayashi K, Lazzari G, Loi P, Payne J, Renfree M, Seet S, Stejskal J, Swegen A, Williams SA, Zainuddin ZZ, Holtze S (2021) The ART of bringing extinction to a freeze - history and future of species conservation, exemplified by rhinos. Theriogenology 169:76–88

Hildebrandt TB, Holtze S, Colleoni S, Hermes R, Stejskal J, Lekolool I, Ndeereh D, Omondi P, Kariuki L, Mijele D, Mutisya S, Ngulu S, Diecke S, Hayashi K, Lazzari G, de Mori B, Biasetti P, Quaggio A, Galli C, Goeritz F (2023) In vitro fertilization program in white rhinoceros. Reproduction 166:383–399

Honda A (2018) Applying iPSCs for preserving endangered species and elucidating the evolution of mammalian sex determination. BioEssays 40:e1700152

Honda A, Choijookhuu N, Izu H, Kawano Y, Inokuchi M, Honsho K, Lee AR, Nabekura H, Ohta H, Tsukiyama T, Ohinata Y, Kuroiwa A, Hishikawa Y, Saitou M, Jogahara T, Koshimoto C (2017) Flexible adaptation of male germ cells from female iPSCs of endangered Tokudaia osimensis. Sci Adv 3:e1602179

Hori T, Hashizaki F, Narushima E, Komiya T, Orima H, Tsutsui T (2006) A trial of intrauterine insemination using a fiberscope in the giant panda (Ailuropoda melanoleuca). J Vet Med Sci 68:987–990

Hou Z, An L, Han J, Yuan Y, Chen D, Tian J (2018) Revolutionize livestock breeding in the future: an animal embryo-stem cell breeding system in a dish. J Anim Sci Biotechnol 9:90

Hu Y, Yang Y, Tan P, Zhang Y, Han M, Yu J, Zhang X, Jia Z, Wang D, Yao K, Pang H, Hu Z, Li Y, Ma T, Liu K, Ding S (2023) Induction of mouse totipotent stem cells by a defined chemical cocktail. Nature 617:792–797

Huang Y, Li D, Zhou Y, Zhou Q, Li R, Wang C, Huang Z, Hull V, Zhang H (2012) Factors affecting the outcome of artificial insemination using cryopreserved spermatozoa in the giant panda (Ailuropoda melanoleuca). Zoo Biol 31:561–573

Ishikura Y, Ohta H, Sato T, Murase Y, Yabuta Y, Kojima Y, Yamashiro C, Nakamura T, Yamamoto T, Ogawa T, Saitou M (2021) In vitro reconstitution of the whole male germ-cell development from mouse pluripotent stem cells. Cell Stem Cell 28:2167-2179.e2169

Jebb D, Huang Z, Pippel M, Hughes GM, Lavrichenko K, Devanna P, Winkler S, Jermiin LS, Skirmuntt EC, Katzourakis A, Burkitt-Gray L, Ray DA, Sullivan KAM, Roscito JG, Kirilenko BM, Dávalos LM, Corthals AP, Power ML, Jones G, Ransome RD, Dechmann DKN, Locatelli AG, Puechmaille SJ, Fedrigo O, Jarvis ED, Hiller M, Vernes SC, Myers EW, Teeling EC (2020) Six reference-quality genomes reveal evolution of bat adaptations. Nature 583:578–584

Jeon Y, Nam YH, Cheong SA, Kwak SS, Lee E, Hyun SH (2016) Absence of nucleolus formation in raccoon dog-porcine interspecies somatic cell nuclear transfer embryos results in embryonic developmental failure. J Reprod Dev 62:345–350

Jewgenow K, Zahmel J (2020) Preservation of female genetic resources in feline species. Theriogenology 156:124–129

Katayama M, Hirayama T, Tani T, Nishimori K, Onuma M, Fukuda T (2018) Chick derived induced pluripotent stem cells by the poly-cistronic transposon with enhanced transcriptional activity. J Cell Physiol 233:990–1004

Katayama M, Fukuda T, Kaneko T, Nakagawa Y, Tajima A, Naito M, Ohmaki H, Endo D, Asano M, Nagamine T, Nakaya Y, Saito K, Watanabe Y, Tani T, Inoue-Murayama M, Nakajima N, Onuma M (2022) Induced pluripotent stem cells of endangered avian species. Commun Biol 5:1049

King NM, Perrin J (2014) Ethical issues in stem cell research and therapy. Stem Cell Res Ther 5:85

Klitzman R (2010) The use of eggs and embryos in stem cell research. Semin Reprod Med 28:336–344

Kogut I, McCarthy SM, Pavlova M, Astling DP, Chen X, Jakimenko A, Jones KL, Getahun A, Cambier JC, Pasmooij AMG, Jonkman MF, Roop DR, Bilousova G (2018) High-efficiency RNA-based reprogramming of human primary fibroblasts. Nat Commun 9:745

Komori M, Kikuchi O, Sakuma T, Funaki J, Kitada M, Kamataki T (1992) Molecular cloning of monkey liver cytochrome P-450 cDNAs: similarity of the primary sequences to human cytochromes P-450. Biochim Biophys Acta 1171:141–146

Korody ML, Ford SM, Nguyen TD, Pivaroff CG, Valiente-Alandi I, Peterson SE, Ryder OA, Loring JF (2021) Rewinding extinction in the northern white rhinoceros: genetically diverse induced pluripotent stem cell Bank for Genetic Rescue. Stem Cells Dev 30:177–189

Kumar D, Anand T, Vijayalakshmy K, Sharma P, Rajendran R, Selokar NL, Yadav PS, Kumar D (2019) Transposon mediated reprogramming of buffalo fetal fibroblasts to induced pluripotent stem cells in feeder free culture conditions. Res Vet Sci 123:252–260

Kuroiwa A, Handa S, Nishiyama C, Chiba E, Yamada F, Abe S, Matsuda Y (2011) Additional copies of CBX2 in the genomes of males of mammals lacking SRY, the Amami spiny rat (Tokudaia osimensis) and the tokunoshima spiny rat (tokudaia tokunoshimensis). Chromosome Res 19:635–644

Lanza RP, Cibelli JB, Diaz F, Moraes CT, Farin PW, Farin CE, Hammer CJ, West MD, Damiani P (2000) Cloning of an endangered species (Bos gaurus) using interspecies nuclear transfer. Cloning 2:79–90

Lee BR, Yang H, Byun SJ, Park TS (2023) Research note: development of a chicken experimental model platform for induced pluripotent stem cells by using CRISPR/Cas9-mediated NANOG knock-in reporter DF1 cells. Poult Sci 102:102425

Li P, Hu H, Yang S, Tian R, Zhang Z, Zhang W, Ma M, Zhu Y, Guo X, Huang Y, He Z, Li Z (2013) Differentiation of induced pluripotent stem cells into male germ cells in vitro through embryoid body formation and retinoic acid or testosterone induction. Biomed Res Int 2013:608728

PubMed   Google Scholar  

Li X, Zhang P, Jiang S, Ding B, Zuo X, Li Y, Cao Z, Zhang Y (2018) Aging adult porcine fibroblasts can support nuclear transfer and transcription factor-mediated reprogramming. Anim Sci J 89:289–297

Liu G, David BT, Trawczynski M, Fessler RG (2020) Advances in pluripotent stem cells: history, mechanisms, technologies, and applications. Stem Cell Rev Rep 16:3–32

Liu M, Zhao L, Wang Z, Su H, Wang T, Yang G, Chen L, Wu B, Zhao G, Guo J, Yang Z, Zhang J, Hao C, Ma T, Song Y, Bao S, Zuo Y, Li X, Cao G (2021) Generation of sheep induced pluripotent stem cells with defined DOX-inducible transcription factors via piggyBac transposition. Front Cell Dev Biol 9:785055

Liu F, Wang J, Yue Y, Li C, Zhang X, Xiang J, Wang H, Li X (2023) Derivation of arbas cashmere goat induced pluripotent stem cells in LCDM with trophectoderm lineage differentiation and interspecies chimeric abilities. Int J Mol Sci 24:14728

Lovell-Badge R, Anthony E, Barker RA, Bubela T, Brivanlou AH, Carpenter M, Charo RA, Clark A, Clayton E, Cong Y, Daley GQ, Fu J, Fujita M, Greenfield A, Goldman SA, Hill L, Hyun I, Isasi R, Kahn J, Kato K, Kim JS, Kimmelman J, Knoblich JA, Mathews D, Montserrat N, Mosher J, Munsie M, Nakauchi H, Naldini L, Naughton G, Niakan K, Ogbogu U, Pedersen R, Rivron N, Rooke H, Rossant J, Round J, Saitou M, Sipp D, Steffann J, Sugarman J, Surani A, Takahashi J, Tang F, Turner L, Zettler PJ, Zhai X (2021) ISSCR guidelines for stem cell research and clinical translation: the 2021 update. Stem Cell Reports 16:1398–1408

Lu Y, West FD, Jordan BJ, Jordan ET, West RC, Yu P, He Y, Barrios MA, Zhu Z, Petitte JN, Beckstead RB, Stice SL (2014) Induced pluripotency in chicken embryonic fibroblast results in a germ cell fate. Stem Cells Dev 23:1755–1764

Marchetto MCN, Narvaiza I, Denli AM, Benner C, Lazzarini TA, Nathanson JL, Paquola ACM, Desai KN, Herai RH, Weitzman MD, Yeo GW, Muotri AR, Gage FH (2013) Differential L1 regulation in pluripotent stem cells of humans and apes. Nature 503:525–529

Martello G, Smith A (2014) The nature of embryonic stem cells. Annu Rev Cell Dev Biol 30:647–675

Merling RK, Sweeney CL, Choi U, De Ravin SS, Myers TG, Otaizo-Carrasquero F, Pan J, Linton G, Chen L, Koontz S, Theobald NL, Malech HL (2013) Transgene-free iPSCs generated from small volume peripheral blood nonmobilized CD34+ cells. Blood 121:e98-107

Mochiduki Y, Okita K (2012) Methods for iPS cell generation for basic research and clinical applications. Biotechnol J 7:789–797

Moradi S, Mahdizadeh H, Šarić T, Kim J, Harati J, Shahsavarani H, Greber B, Moore JBT (2019) Research and therapy with induced pluripotent stem cells (iPSCs): social, legal, and ethical considerations. Stem Cell Res Ther 10:341

Müller K, Eder S, Jakop U, Schiller J, Müller P, Bashawat M (2020) Assisted reproduction for felid species conservation-sperm competences at risk. Reprod Domest Anim 55(Suppl 2):55–60

Muñoz E, Castro M, Aguila L, Contreras MJ, Fuentes F, Arias ME, Felmer R (2023) Standardization of a sex-sorting protocol for stallion spermatozoa by means of absolute RT-qPCR. Int J Mol Sci 24:11947

Nagaoka SI, Nakaki F, Miyauchi H, Nosaka Y, Ohta H, Yabuta Y, Kurimoto K, Hayashi K, Nakamura T, Yamamoto T, Saitou M (2020) ZGLP1 is a determinant for the oogenic fate in mice. Science 367:eaaw4115

Nagy K, Sung HK, Zhang P, Laflamme S, Vincent P, Agha-Mohammadi S, Woltjen K, Monetti C, Michael IP, Smith LC, Nagy A (2011) Induced pluripotent stem cell lines derived from equine fibroblasts. Stem Cell Rev Rep 7:693–702

Naniwa Y, Sakamoto Y, Toda S, Uchiyama K (2019) Bovine sperm sex-selection technology in Japan. Reprod Med Biol 18:17–26

Oikawa M, Kobayashi H, Sanbo M, Mizuno N, Iwatsuki K, Takashima T, Yamauchi K, Yoshida F, Yamamoto T, Shinohara T, Nakauchi H, Kurimoto K, Hirabayashi M, Kobayashi T (2022) Functional primordial germ cell-like cells from pluripotent stem cells in rats. Science 376:176–179

Okita K, Matsumura Y, Sato Y, Okada A, Morizane A, Okamoto S, Hong H, Nakagawa M, Tanabe K, Tezuka K, Shibata T, Kunisada T, Takahashi M, Takahashi J, Saji H, Yamanaka S (2011) A more efficient method to generate integration-free human iPS cells. Nat Methods 8:409–412

Olivera R, Moro LN, Jordan R, Luzzani C, Miriuka S, Radrizzani M, Donadeu FX, Vichera G (2016) In vitro and in vivo development of horse cloned embryos generated with iPSCs, mesenchymal stromal cells and fetal or adult fibroblasts as nuclear donors. PLoS ONE 11:e0164049

Olivera R, Moro LN, Jordan R, Pallarols N, Guglielminetti A, Luzzani C, Miriuka SG, Vichera G (2018) Bone marrow mesenchymal stem cells as nuclear donors improve viability and health of cloned horses. Stem Cells Cloning 11:13–22

CAS   PubMed   PubMed Central   Google Scholar  

Onozato D, Yamashita M, Fukuyama R, Akagawa T, Kida Y, Koeda A, Hashita T, Iwao T, Matsunaga T (2018) Efficient generation of cynomolgus monkey induced pluripotent stem cell-derived intestinal organoids with pharmacokinetic functions. Stem Cells Dev 27:1033–1045

Otake T, Kuroiwa A (2016) Molecular mechanism of male differentiation is conserved in the SRY-absent mammal. Tokudaia Osimensis Sci Rep 6:32874

Paul R (2018) World’s last male northern white rhino dies in kenyan reserve. Int Environ Rep: Reference File 41:433

Google Scholar  

Pennington PM, Marshall KL, Capiro JM, Howard L, Durrant BS (2020) Pregnancies following long luteal phases in southern white rhinoceros (Ceratotherium simum simum). Zoo Biol 39:141–144

Pessôa LVF, Bressan FF, Freude KK (2019a) Induced pluripotent stem cells throughout the animal kingdom: availability and applications. World J Stem Cells 11:491–505

Pessôa LVF, Pires PRL, Del Collado M, Pieri NCG, Recchia K, Souza AF, Perecin F, da Silveira JC, de Andrade AFC, Ambrosio CE, Bressan FF, Meirelles FV (2019b) Generation and miRNA characterization of equine induced pluripotent stem cells derived from fetal and adult multipotent tissues. Stem Cells Int 2019:1393791

Pieri NCG, de Souza AF, Botigelli RC, Pessôa LVF, Recchia K, Machado LS, Glória MH, de Castro RVG, Leal DF, Fantinato Neto P, Martins S, Dos Santos Martins D, Bressan FF, de Andrade AFC (2022) Porcine primordial germ cell-like cells generated from induced pluripotent stem cells under different culture conditions. Stem Cell Rev Rep 18:1639–1656

Plotnick RE, Smith FA, Lyons SK (2016) The fossil record of the sixth extinction. Ecol Lett 19:546–553

Post Y, Puschhof J, Beumer J, Kerkkamp HM, Clevers H (2020) Snake venom gland organoids. Cell 180:233-247.e221

Prescott SL, Srinivasan R, Marchetto MC, Grishina I, Narvaiza I, Selleri L, Gage FH, Swigut T, Wysocka J (2015) Enhancer divergence and cis-regulatory evolution in the human and chimp neural crest. Cell 163:68–83

Rahmani F, Movahedin M, Mazaheri Z, Soleimani M (2019) Transplantation of mouse iPSCs into testis of azoospermic mouse model: in vivo and in vitro study. Artif Cells Nanomed Biotechnol 47:1585–1594

Ramaswamy K, Yik WY, Wang XM, Oliphant EN, Lu W, Shibata D, Ryder OA, Hacia JG (2015) Derivation of induced pluripotent stem cells from orangutan skin fibroblasts. BMC Res Notes 8:577

Rawat N, Singh MK, Sharma T, Vats P, Nagoorvali D, Palta P, Chauhan MS, Manik RS (2021) Media switching at different time periods affects the reprogramming efficiency of buffalo fetal fibroblasts. Anim Biotechnol 32:155–168

Ren J, Pak Y, He L, Qian L, Gu Y, Li H, Rao L, Liao J, Cui C, Xu X, Zhou J, Ri H, Xiao L (2011) Generation of hircine-induced pluripotent stem cells by somatic cell reprogramming. Cell Res 21:849–853

Rossant J, Fu J (2023) Why researchers should use human embryo models with caution. Nature 622:454–456

Ruan W, Han J, Li P, Cao S, An Y, Lim B, Li N (2011) A novel strategy to derive iPS cells from porcine fibroblasts. Sci China Life Sci 54:553–559

Sakai Y, Nakamura T, Okamoto I, Gyobu-Motani S, Ohta H, Yabuta Y, Tsukiyama T, Iwatani C, Tsuchiya H, Ema M, Morizane A, Takahashi J, Yamamoto T, Saitou M (2020) Induction of the germ cell fate from pluripotent stem cells in cynomolgus monkeys†. Biol Reprod 102:620–638

Samiec M, Skrzyszowska M (2005) Molecular conditions of the cell nucleus remodelling/reprogramming process and nuclear transferred embryo development in the intraooplasmic karyoplast injection technique: a review. Czeh J Anim Sci 50:185–195

Samiec M, Skrzyszowska M (2012) Roscovitine is a novel agent that can be used for the activation of porcine oocytes reconstructed with adult cutaneous or fetal fibroblast cell nuclei. Theriogenology 78:1855–1867

Samiec M, Romanek J, Lipiński D, Opiela J (2019) Expression of pluripotency-related genes is highly dependent on trichostatin A-assisted epigenomic modulation of porcine mesenchymal stem cells analysed for apoptosis and subsequently used for generating cloned embryos. Anim Sci J 90:1127–1141

Saragusty J, Diecke S, Drukker M, Durrant B, Friedrich Ben-Nun I, Galli C, Göritz F, Hayashi K, Hermes R, Holtze S, Johnson S, Lazzari G, Loi P, Loring JF, Okita K, Renfree MB, Seet S, Voracek T, Stejskal J, Ryder OA, Hildebrandt TB (2016) Rewinding the process of mammalian extinction. Zoo Biol 35:280–292

Sartori C, DiDomenico AI, Thomson AJ, Milne E, Lillico SG, Burdon TG, Whitelaw CB (2012) Ovine-induced pluripotent stem cells can contribute to chimeric lambs. Cell Reprogram 14:8–19

Sasaki K, Yokobayashi S, Nakamura T, Okamoto I, Yabuta Y, Kurimoto K, Ohta H, Moritoki Y, Iwatani C, Tsuchiya H, Nakamura S, Sekiguchi K, Sakuma T, Yamamoto T, Mori T, Woltjen K, Nakagawa M, Yamamoto T, Takahashi K, Yamanaka S, Saitou M (2015) Robust in vitro induction of human germ cell fate from pluripotent stem cells. Cell Stem Cell 17:178–194

Schweinfurth MK, Call J (2019) Revisiting the possibility of reciprocal help in non-human primates. Neurosci Biobehav Rev 104:73–86

Segers S (2023) The IVG “relatedness paradox”: researchers should mind speculation. Trends Biotechnol 41:1220–1222

Shimada T, Mimura M, Inoue K, Nakamura S, Oda H, Ohmori S, Yamazaki H (1997) Cytochrome P450-dependent drug oxidation activities in liver microsomes of various animal species including rats, Guinea pigs, dogs, monkeys, and humans. Arch Toxicol 71:401–408

Sinclair AH, Berta P, Palmer MS, Hawkins JR, Griffiths BL, Smith MJ, Foster JW, Frischauf AM, Lovell-Badge R, Goodfellow PN (1990) A gene from the human sex-determining region encodes a protein with homology to a conserved DNA-binding motif. Nature 346:240–244

Siqueira da Fonseca SA, Abdelmassih S, de Mello CintraLavagnolli T, Serafim RC, Clemente Santos EJ, Mota Mendes C, de Souza Pereira V, Ambrosio CE, Miglino MA, Visintin JA, Abdelmassih R, Kerkis A, Kerkis I (2009) Human immature dental pulp stem cells’ contribution to developing mouse embryos: production of human/mouse preterm chimaeras. Cell Prolif 42:132–140

Song H, Li H, Huang M, Xu D, Gu C, Wang Z, Dong F, Wang F (2013) Induced pluripotent stem cells from goat fibroblasts. Mol Reprod Dev 80:1009–1017

Stanton MM, Tzatzalos E, Donne M, Kolundzic N, Helgason I, Ilic D (2019) Prospects for the use of induced pluripotent stem cells in animal conservation and environmental protection. Stem Cells Transl Med 8:7–13

Steinle H, Weber M, Behring A, Mau-Holzmann U, von Ohle C, Popov AF, Schlensak C, Wendel HP, Avci-Adali M (2019) Reprogramming of urine-derived renal epithelial cells into iPSCs using srRNA and consecutive differentiation into beating cardiomyocytes. Mol Ther Nucleic Acids 17:907–921

Stoops MA, Campbell MK, DeChant CJ, Hauser J, Kottwitz J, Pairan RD, Shaffstall W, Volle K, Roth TL (2016) Enhancing captive indian rhinoceros genetics via artificial insemination of cryopreserved sperm. Anim Reprod Sci 172:60–75

Sukparangsi W, Thongphakdee A, Karoon S, Suban Na Ayuthaya N, Hengkhunthod I, Prakongkaew R, Bootsri R, Sikaeo W (2022) Establishment of fishing cat cell biobanking for sustainable conservation. Front Vet Sci 9:989670

Sumer H, Liu J, Malaver-Ortega LF, Lim ML, Khodadadi K, Verma PJ (2011) NANOG is a key factor for induction of pluripotency in bovine adult fibroblasts. J Anim Sci 89:2708–2716

Summers PM, Shephard AM, Hodges JK, Kydd J, Boyle MS, Allen WR (1987) Successful transfer of the embryos of przewalski’s horses (Equus przewalskii) and Grant’s zebra (E. burchelli) to domestic mares (E. caballus). J Reprod Fertil 80:13–20

Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K, Yamanaka S (2007) Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131:861–872

Tan T, Wu J, Si C, Dai S, Zhang Y, Sun N, Zhang E, Shao H, Si W, Yang P, Wang H, Chen Z, Zhu R, Kang Y, Hernandez-Benitez R, Martinez Martinez L, Nuñez Delicado E, Berggren WT, Schwarz M, Ai Z, Li T, Rodriguez Esteban C, Ji W, Niu Y, Izpisua Belmonte JC (2021) Chimeric contribution of human extended pluripotent stem cells to monkey embryos ex vivo. Cell 184:2020-2032.e2014

Taura D, Noguchi M, Sone M, Hosoda K, Mori E, Okada Y, Takahashi K, Homma K, Oyamada N, Inuzuka M, Sonoyama T, Ebihara K, Tamura N, Itoh H, Suemori H, Nakatsuji N, Okano H, Yamanaka S, Nakao K (2009) Adipogenic differentiation of human induced pluripotent stem cells: comparison with that of human embryonic stem cells. FEBS Lett 583:1029–1033

Teeling EC, Vernes SC, Dávalos LM, Ray DA, Gilbert MTP, Myers E (2018) Bat biology, genomes, and the Bat1K project: to generate chromosome-level genomes for all living bat species. Annu Rev Anim Biosci 6:23–46

Turner L (2021) ISSCR’s guidelines for stem cell research and clinical translation: supporting development of safe and efficacious stem cell-based interventions. Stem Cell Reports 16:1394–1397

Umegaki-Arao N, Pasmooij AM, Itoh M, Cerise JE, Guo Z, Levy B, Gostyński A, Rothman LR, Jonkman MF, Christiano AM (2014) Induced pluripotent stem cells from human revertant keratinocytes for the treatment of epidermolysis bullosa. Sci Transl Med 6:264ra164

Uno Y, Hosaka S, Matsuno K, Nakamura C, Kito G, Kamataki T, Nagata R (2007) Characterization of cynomolgus monkey cytochrome P450 (CYP) cDNAs: is CYP2C76 the only monkey-specific CYP gene responsible for species differences in drug metabolism? Arch Biochem Biophys 466:98–105

Verma R, Holland MK, Temple-Smith P, Verma PJ (2012) Inducing pluripotency in somatic cells from the snow leopard (Panthera uncia), an endangered felid. Theriogenology 77:220-228.e2282

Verma R, Liu J, Holland MK, Temple-Smith P, Williamson M, Verma PJ (2013) Nanog is an essential factor for induction of pluripotency in somatic cells from endangered felids. Biores Open Access 2:72–76

Wang J, Liu X, Yang J, Guo H, Li J, Huo L, Zhao H, Wang X, Yan X, Li B, Sun Y (2021) Effects of small-molecule compounds on fibroblast properties in golden snub-nosed monkey (rhinopithecus roxellana). J Med Primatol 50:323–331

Weeratunga P, Harman RM, Van de Walle GR (2023) Induced pluripotent stem cells from domesticated ruminants and their potential for enhancing livestock production. Front Vet Sci 10:1129287

Wen B, Wang G, Li E, Kolesnichenko OA, Tu Z, Divanovic S, Kalin TV, Kalinichenko VV (2022) In vivo generation of bone marrow from embryonic stem cells in interspecies chimeras. Elife 11:e74018

Wernig M, Meissner A, Foreman R, Brambrink T, Ku M, Hochedlinger K, Bernstein BE, Jaenisch R (2007) In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state. Nature 448:318–324

Wu Z, Chen J, Ren J, Bao L, Liao J, Cui C, Rao L, Li H, Gu Y, Dai H, Zhu H, Teng X, Cheng L, Xiao L (2009) Generation of pig induced pluripotent stem cells with a drug-inducible system. J Mol Cell Biol 1:46–54

Wu Y, Zhang Y, Mishra A, Tardif SD, Hornsby PJ (2010) Generation of induced pluripotent stem cells from newborn marmoset skin fibroblasts. Stem Cell Res 4:180–188

Wu J, Okamura D, Li M, Suzuki K, Luo C, Ma L, He Y, Li Z, Benner C, Tamura I, Krause MN, Nery JR, Du T, Zhang Z, Hishida T, Takahashi Y, Aizawa E, Kim NY, Lajara J, Guillen P, Campistol JM, Esteban CR, Ross PJ, Saghatelian A, Ren B, Ecker JR, Izpisua Belmonte JC (2015) An alternative pluripotent state confers interspecies chimaeric competency. Nature 521:316–321

Wüstner LS, Klingenstein M, Frey KG, Nikbin MR, Milazzo A, Kleger A, Liebau S, Klingenstein S (2022) Generating iPSCs with a high-efficient, non-invasive method-an improved way to cultivate keratinocytes from plucked hair for reprogramming. Cells 11:1955

Xu J, Yu L, Guo J, Xiang J, Zheng Z, Gao D, Shi B, Hao H, Jiao D, Zhong L, Wang Y, Wu J, Wei H, Han J (2019) Generation of pig induced pluripotent stem cells using an extended pluripotent stem cell culture system. Stem Cell Res Ther 10:193

Yamashiro C, Sasaki K, Yabuta Y, Kojima Y, Nakamura T, Okamoto I, Yokobayashi S, Murase Y, Ishikura Y, Shirane K, Sasaki H, Yamamoto T, Saitou M (2018) Generation of human oogonia from induced pluripotent stem cells in vitro. Science 362:356–360

Yang S, Yuan Q, Niu M, Hou J, Zhu Z, Sun M, Li Z, He Z (2017) BMP4 promotes mouse iPS cell differentiation to male germ cells via Smad1/5, Gata 4, Id1 and Id2. Reproduction 153:211–220

Ye L, Muench MO, Fusaki N, Beyer AI, Wang J, Qi Z, Yu J, Kan YW (2013) Blood cell-derived induced pluripotent stem cells free of reprogramming factors generated by Sendai viral vectors. Stem Cells Transl Med 2:558–566

Yoshino T, Suzuki T, Nagamatsu G, Yabukami H, Ikegaya M, Kishima M, Kita H, Imamura T, Nakashima K, Nishinakamura R, Tachibana M, Inoue M, Shima Y, Morohashi KI, Hayashi K (2021) Generation of ovarian follicles from mouse pluripotent stem cells. Science 373:eabe0237

Yu P, Lu Y, Jordan BJ, Liu Y, Yang JY, Hutcheson JM, Ethridge CL, Mumaw JL, Kinder HA, Beckstead RB, Stice SL, West FD (2014) Nonviral minicircle generation of induced pluripotent stem cells compatible with production of chimeric chickens. Cell Reprogram 16:366–378

Zhang J, Zhi M, Gao D, Zhu Q, Gao J, Zhu G, Cao S, Han J (2022) Research progress and application prospects of stable porcine pluripotent stem cells†. Biol Reprod 107:226–236

Zhao L, Long C, Zhao G, Su J, Ren J, Sun W, Wang Z, Zhang J, Liu M, Hao C, Li H, Cao G, Bao S, Zuo Y, Li X (2022) Reprogramming barriers in bovine cells nuclear transfer revealed by single-cell RNA-seq analysis. J Cell Mol Med 26:4792–4804

Zhou T, Benda C, Dunzinger S, Huang Y, Ho JC, Yang J, Wang Y, Zhang Y, Zhuang Q, Li Y, Bao X, Tse HF, Grillari J, Grillari-Voglauer R, Pei D, Esteban MA (2012) Generation of human induced pluripotent stem cells from urine samples. Nat Protoc 7:2080–2089

Zhou Q, Wang M, Yuan Y, Wang X, Fu R, Wan H, Xie M, Liu M, Guo X, Zheng Y, Feng G, Shi Q, Zhao XY, Sha J, Zhou Q (2016) Complete meiosis from embryonic stem cell-derived germ cells in vitro. Cell Stem Cell 18:330–340

Zhu Y, Hu HL, Li P, Yang S, Zhang W, Ding H, Tian RH, Ning Y, Zhang LL, Guo XZ, Shi ZP, Li Z, He Z (2012) Generation of male germ cells from induced pluripotent stem cells (iPS cells): an in vitro and in vivo study. Asian J Androl 14:574–579

Zvick J, Tarnowska-Sengül M, Ghosh A, Bundschuh N, Gjonlleshaj P, Hinte LC, Trautmann CL, Noé F, Qabrati X, Domenig SA, Kim I, Hennek T, von Meyenn F, Bar-Nur O (2022) Exclusive generation of rat spermatozoa in sterile mice utilizing blastocyst complementation with pluripotent stem cells. Stem Cell Reports 17:1942–1958

Zywitza V, Frahm S, Krüger N, Weise A, Göritz F, Hermes R, Holtze S, Colleoni S, Galli C, Drukker M, Hildebrandt TB, Diecke S (2022) Induced pluripotent stem cells and cerebral organoids from the critically endangered Sumatran rhinoceros. iScience 25:105414

Zywitza V, Rusha E, Shaposhnikov D, Ruiz-Orera J, Telugu N, Rishko V, Hayashi M, Michel G, Wittler L, Stejskal J, Holtze S, Göritz F, Hermes R, Wang J, Izsvák Z, Colleoni S, Lazzari G, Galli C, Hildebrandt TB, Hayashi K, Diecke S, Drukker M (2022b) Naïve-like pluripotency to pave the way for saving the northern white rhinoceros from extinction. Sci Rep 12:3100

Download references

Acknowledgements

Not applicable

This work was supported by the National Natural Science Fund of China (No. 82174226), the Tianfu laboratory transfer payment project (No. 2021ZYCD012).

Author information

Authors and affiliations.

School of Pharmacy/School of Modem Chinese Medicine Industry, Chengdu University of Traditional Chinese Medicine, Chengdu, 611137, People’s Republic of China

Yurou Wu, Chengwei Wang, Xinyun Fan, Zibo Liu & Xun Ye

Department of Biology, New York University, New York, NY, USA

School of Basic Medical Sciences, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, People’s Republic of China

Chongyang Shen

Innovative Institute of Chinese Medicine and Pharmacy/Academy for Interdiscipline, Chengdu Univesity of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, People’s Republic of China

Sichuan Engineering Research Center for Endangered Medicinal Animals, Chengdu, China

You can also search for this author in PubMed   Google Scholar

Contributions

CJW : Conceptualization, Methodology, Supervision. CYS : Conceptualization, Methodology, Supervision. YRW : Writing–original draft, preparation, visualization; CWW : writing—review & editing. XYF: visualization. YXM : writing and harmonizing the manuscript. ZBL : writing—review & editing. XY : writing—review & editing. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Chongyang Shen or Chunjie Wu .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

The authors declare that they have no affiliations with or involvement in any organization or entity with any financial interest in the subject matter or.

materials discussed in this manuscript.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Wu, Y., Wang, C., Fan, X. et al. The impact of induced pluripotent stem cells in animal conservation. Vet Res Commun 48 , 649–663 (2024). https://doi.org/10.1007/s11259-024-10294-3

Download citation

Received : 02 November 2023

Accepted : 04 January 2024

Published : 16 January 2024

Issue Date : April 2024

DOI : https://doi.org/10.1007/s11259-024-10294-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Induced pluripotent stem cells
• Endangered and extinct species
• Animal conservation
• Find a journal
• Publish with us
• Track your research

• Open access
• Published: 06 March 2021

Research perspectives on animal health in the era of artificial intelligence

• Pauline Ezanno   ORCID: orcid.org/0000-0002-0034-8950 1 ,
• Sébastien Picault 1 ,
• Gaël Beaunée 1 ,
• Xavier Bailly 2 ,
• Facundo Muñoz 3 ,
• Raphaël Duboz 3 , 4 ,
• Hervé Monod 5 &
• Jean-François Guégan 3 , 6 , 7  

Veterinary Research volume  52 , Article number:  40 ( 2021 ) Cite this article

14k Accesses

30 Citations

10 Altmetric

Metrics details

Leveraging artificial intelligence (AI) approaches in animal health (AH) makes it possible to address highly complex issues such as those encountered in quantitative and predictive epidemiology, animal/human precision-based medicine, or to study host × pathogen interactions. AI may contribute (i) to diagnosis and disease case detection, (ii) to more reliable predictions and reduced errors, (iii) to representing more realistically complex biological systems and rendering computing codes more readable to non-computer scientists, (iv) to speeding-up decisions and improving accuracy in risk analyses, and (v) to better targeted interventions and anticipated negative effects. In turn, challenges in AH may stimulate AI research due to specificity of AH systems, data, constraints, and analytical objectives. Based on a literature review of scientific papers at the interface between AI and AH covering the period 2009–2019, and interviews with French researchers positioned at this interface, the present study explains the main AH areas where various AI approaches are currently mobilised, how it may contribute to renew AH research issues and remove methodological or conceptual barriers. After presenting the possible obstacles and levers, we propose several recommendations to better grasp the challenge represented by the AH/AI interface. With the development of several recent concepts promoting a global and multisectoral perspective in the field of health, AI should contribute to defract the different disciplines in AH towards more transversal and integrative research.

1 Introduction

Artificial intelligence (AI) encompasses a large range of theories and technologies used to solve problems of high logical or algorithmic complexity. It crosses many disciplines, including mechanistic modelling, software engineering, data science, and statistics (Figure  1 ). Introduced in the 1950s, many AI methods have been developed or extended recently with the improvement of computer performance. Recent developments have been fuelled by the interfaces created between AI and other disciplines, such as bio-medicine, as well as massive data from different fields, particularly those associated with healthcare [ 1 , 2 ].

Interactions between animal health (AH), artificial intelligence (AI), and closely related research domains. This illustration is pinpointing only the links between AH (in blue), AI and its main subfields (in red), and other related fields of research (in black). It can be naturally complexified through the interactions between AH and other research topics (e.g., human medicine) or between core disciplines (e.g., statistics and physics).

AI addresses three challenges that also make sense in animal health (AH): (1) understanding a situation and its dynamics, e.g., epidemic spread; (2) the perception of the environment, which corresponds in AH to the detection of patterns (e.g., repeated sequence of observations), forms (e.g., of a protein) and signals (e.g., increased mortality compared to a baseline) at different scales; (3) computer-based decision making, or, more realistically, human decision support (e.g., expert systems, diagnostic support, resource allocation).

To answer these challenges, a wide range of concepts and methods are developed in AI. This includes machine learning (ML), a widely known AI method nowadays, which has been developing since the 1980s [ 3 ]. Since the 2000s, deep learning is developing with the rise of big data and the continuous increasing of computing capacities, enabling the exploration of massive amount of information that cannot be processed by conventional statistical methods. In addition, this also includes methods and algorithms for solving complex problems, automating tasks or reasoning, integrating information from heterogeneous sources, or decision support (Figure  1 ). These methods are now uprising in the human health sector, but are still rarely used to study animal health issues that they would help to revisit.

Part of the scientific challenges faced in AH can be approached from a new perspective by using some of these AI methods to analyse the ever-increasing data collected on animals, pathogens, and their environment. AH research benefits from advances in machine and deep learning methods, e.g., in predictive epidemiology, individual-based precision medicine, and to study host–pathogen interactions [ 2 , 4 ]. These methods contribute to disease diagnosis and individual case detection, to more reliable predictions and reduced errors, to speed-up decisions and improved accuracy in risk analysis, and to better targeting interventions in AH [ 5 ]. AH research also benefits from scientific advances in other domains of AI. Knowledge representation and modelling of reasoning [ 6 ] allow more realistic representations of complex socio-biological systems such as those encountered in AH. Examples include processes related to decision-making and uncertainty management [ 7 , 8 ], as well as of patient life courses like in human epidemiology [ 9 ]. This contributes to making them more readable by noncomputer experts. In addition, advances in problem solving under constrained resource allocation [ 10 ], in autonomous agents [ 11 ], multi-agent systems [ 12 ], and multi-level systems [ 13 ], as well as on automatic computer code generation [ 14 ] can be mobilised to enhance efficient and reliable epidemiological models. Interestingly, this may aid to anticipate the effect of control and management decisions at different spatial and temporal scales (animal, herd, country…).

Conducting research at the AH/AI interface also leads to identify new challenges for AI, on themes common with human health but considering different contexts and perspectives [ 15 ]. First, taking into account the particular agro- and socio-economic conditions of production systems is crucial when dealing with AH. Animal production systems depend on human activities and decisions. They can be a source of income (e.g., livestock) or labour forces and source of food in family farming. Citizens have also high expectations in terms of ethics and animal welfare [ 16 ]. Conventional measures to control animal diseases may no longer be acceptable by society (e.g., mass culling during outbreaks [ 17 ], antimicrobial usage, [ 18 ]). Alternatives must be identified and assessed, and AI can contribute. For example, individual-based veterinarian medicine is emerging, mobilising both AI methods and new AH data streams, these data differing from data in human health [ 19 ]. The integration of data from deep sequencing in AH, including emerging technologies for studying the metabolome and epigenome, is also a challenge [ 20 , 21 ]. Second, interactions between animal species, in particular between domestic animals and wildlife, lead to specific infectious disease risks (e.g., multi-host pathogens such as for African swine fever, pathogens crossing the species barrier facilitated by frequent contacts and promiscuity). The intensity of such interactions could increase due to separate or synergistic actions of environmental (e.g., landscape homogenisation, land use change for agriculture development, climate change), demographic (e.g., increasing global demand for animal production) and societal (e.g., outdoor livestock management) pressures. In addition, working on multi-species disease networks provides crucial information on the underlying molecular mechanisms favouring interspecific transmission [ 22 ]. Third, animal populations are governed by recurrent decision-making that also impacts health management (e.g., trade, control measures). Economic criteria as consequences on livestock farmers’ incomes are therefore essential indicators for evaluating AH control strategies, which can sometimes be misunderstood or may be at odds with societal expectations. These specificities make the AH/AI interface a theme of interest to stimulate new methodological work and to solve some of old and current locks faced by AH research today. With the development of new concepts in health such as One Health, Ecohealth and Planetary Health, promoting multidisciplinarity, stakeholders’ participation, data sharing, and tackling the complexity of health issues (e.g., multi-host pathogen transmission, short and long-term climatic impacts on disease patterns [ 23 ]), AI could participate in this new development by making it possible to technically solve some of the complex problems posed.

Mobilising the literature published at the AH/AI interface between 2009 and 2019 (Additional file 1 A), focusing our literature search on mainly livestock and wildlife, as well as interviews conducted with French researchers positioned at this interface (Additional file 1 B), we identified the main research areas in AH in which AI is currently involved country-wide. We explored how AI methods contribute to revisiting AH questions and may help remove methodological or conceptual barriers within the field. We also analysed how AH questions interrogate and stimulate new AI technical or scientific developments. In this paper, we first discuss issues related to data collection, organisation and access (Section  1 ), then we discuss how AI methods contribute to revisiting our understanding of animal epidemiological systems (Section  2 ), to improving case detection and diagnosis at different scales (Section  3 ), and to anticipating pathogen spread and control in a wide range of scenarios in order to improve AH management, facilitate decision-making and stimulate innovation (Section  4 ). Finally, we present the possible obstacles and levers to the development of AI in modern AH (Section  5 ), before making recommendations to best address the new challenges represented by this AH/IA interface (Section  6 ).

2 Collect, organise and make accessible quality data

A central point for research in AH remains the quality and availability of data, at the different organization levels of living systems and therefore at different spatial and temporal scales [ 24 ]. Data of interest are diverse. They can be obtained thanks to molecular analysis (e.g., genomic, metagenomics, or metabolic data), from observational data on individuals (e.g., body temperature, behaviour, milk production and composition, weight, feed intake), or from the production system (e.g., herd structure, breeding, management of sanitary issues). They can also be obtained at larger scales, beyond herds or local groups of animals (e.g., epidemiological data, demographic events, commercial movements, meteorological data, land-use occupation).

Even though the acquisition of these massive and heterogeneous data remain challenging (e.g., metabolome data), a large and diverse amount is already collected: (i) through mandatory reporting in accordance with regulations (e.g., commercial movements of cattle, epidemio-surveillance platform), (ii) by automatic devices (e.g., sensors, video surveillance systems), and (iii) on an ad hoc basis as part of research programs. This leads to a very wide diversity of data properties, and therefore of their management, access and possible uses. These data can be specifically obtained for certain animals or herds (e.g., during cohort monitoring programs) or by private companies (e.g., pig trade movements such as in France, milk collection). This can limit accessibility to academics and public research. Globalisation and large-scale animal trade may generate the need to use data obtained at worldwide scale in AH, especially for quantitative epidemiology (e.g., transcontinental spread of pathogens, animal genetics and breed management) leading to standardisation issues [ 25 ].

Consideration should be given to future systems for observing, collecting and managing these data [ 26 ], and to practices aimed at better collaboration between stakeholders. While data management has always been an important element in applied research to facilitate their use and valorisation, it is now a strategic issue both in theoretical and more applied research, coupled with a technical and algorithmic challenge [ 27 , 28 , 29 ]. Indeed, producing algorithms to manage massive data flows and stocks, by optimising calculations, is a challenge, particularly in real time. It seems also necessary to make heterogeneous data sources interoperable, requiring dedicated methodological developments [ 25 ]. In addition, much of the data is private, with ownership often heterogeneous (e.g., multiple owners, non-centralised data, closed data) and sometimes unclear (e.g., lack of knowledge of the real owner of the data between, for example, farmers, the data collector or the farmers’ union). All this tends to considerably complicate access to the data, raises questions about intellectual property, and raises questions in relation to regulations with regards to data protection, e.g., the adaptation of regulation to AH while respecting the confidentiality of the personal data mobilised. What is the relevant business model for data collection or access to existing databases? What about the openness of AH data (e.g., duality between the notion of public good and the private nature of certain data) to make it possible to experiment in real situations and compare the performance of AI algorithms? Answering these questions would facilitate the collection and sharing of ad hoc data. AI, particularly when combining a participatory framework with expert systems and multi-agent systems, helps to build realistic representations of complex socio-biological systems. Thus, it proves to be an effective tool to promote the collaboration of different stakeholders in collective and optimised decision-making, and to assess of the impact of changes in uses and practices [ 30 ].

Encouraging experimentation of AI technologies at a territorial scale becomes crucial to favour their development, validate their performance, and measure their predictive quality. In AH, simplified access to data-generating facilities would allow innovative solutions to be tested on a larger scale and would accelerate their development and evaluation. Substantial expertise exists (e.g., epidemiological data platform, large cohorts, experimental farms) that could be put to good use. In addition, AI could help to revisit sampling methods for field data collection in AH and epidemiological surveillance, by better and more dynamically targeting the data to be collected while avoiding redundant collinear, non-necessary data.

3 Contribution of AI to better understand animal epidemiological systems

Recent technological advances involving AI approaches have made it possible to obtain vast quantities of measurements and observations, as well as to store and share these data more efficiently. This has resulted in an increasing requirement for appropriate data analytical methods. AI methods emerged as the response of the computer-science community to these requirements, leveraging the exponential improvements in computational power. In parallel, statistical methods have greatly evolved in the last few decades as well, e.g., with regards to dimensionality-reduction in the spaces of variables and parameters, variable selection, and model comparison and combination. The rise in computational power has unleashed the development of Bayesian inference through simulation or approximate methods [ 31 ]. Bayesian methods have, in turn, facilitated the integration of data from diverse sources, the incorporation of prior knowledge and allowed for inference on more complex and realistic models while changing the paradigm of statistical inference [ 32 , 33 , 34 ].

3.1 Better understanding the evolution of AH and socio-ecological systems in a One Health context

Learning methods can be used to do phylogenetic reconstructions, contributing in particular to new evolutionary scenarios of pathogens and their transmission pathways. For example, phylogenetic models offer an interesting perspective for identifying environmental bacterial strains with high infectious potentiality [ 35 ], or for predicting the existence of putative host reservoirs or vectors [ 36 ]. The analysis of pathogen sharing among hosts has been used to classify the potential reservoirs of zoonotic diseases using machine learning [ 37 ]. The analysis of pathogen genomes can also be used to identify genotypes of animal pathogens that are more likely to infect humans [ 38 ].

Using phenomenological niche models that rely on data distribution more than on hypotheses about ecological processes at play, disease occurrence data or retrospective serological data coupled with environmental variables can be related to the risk of being exposed to a pathogen. Thus, they can help monitor potential spillovers and emerging risks and anticipate the epidemic pathogen spread [ 39 ]. For instance, Artificial Neural Networks (ANN) have identified the level of genetic introgression between wild and domesticated animal populations in a spatialized context [ 40 ], which may help to understand gene diffusion in host × pathogen systems involving multiple host species, and characterise specimen pools at higher risk to act as pathogen spreaders or sinks. Other AI approaches such as multi-agent models, a more mechanistic approach, have been used in an explicit spatial context for vector-borne pathogen transmissions, and proved to be sufficiently versatile to be adapted to several other particular contexts [ 12 ].

It should be noted here that several studies reveal the relatively ancient nature of AI research in AH. Such AI methods have often made it possible to identify signals (e.g., genetic introgression) or even particular patterns or properties (e.g., importance of density-dependence in the vector-borne transmission) that are less visible or hardly detectable by more conventional statistical treatments.

All these approaches contribute to better understand pathogen transmission in complex system networks as generally observed for emerging infections in tropical, developing regions of the world. On this matter, an improved knowledge is key for protecting humans against these new threats, and AI/AH interfaces development and training in cooperation with the poorest countries would facilitate synergistic effects and actions to predict and tackle new disease threats.

3.2 Reliability, reproducibility and flexibility of mechanistic models in AH

Better understanding and predicting pathogen spread often requires an explicit and integrative representation of the mechanisms involved in the dynamics of AH systems, irrespective of the scale (within-host: [ 41 ]; along a primary production chain: [ 42 ]; in a territory: [ 43 , 44 ]; over a continent: [ 45 ]).

Mathematical (equations) or computer-based (simulations) models can be used. Such mechanistic models (i.e., which represent the mechanisms involved in the infection dynamics), when sufficiently modular to represent contrasted situations, make it possible to anticipate the effects of conventional but also innovative control measures (e.g., new candidate molecules, sensors, genomic selection; [ 46 ]).

However, to assess realistic control measures, mechanistic epidemiological models require the integration of observational data and knowledge from biology, epidemiology, evolution, ecology, agronomy, sociology or economics. Their development can rapidly face challenges of reliability, transparency, reproducibility, and usage flexibility. Moreover, these models are often developed de novo, making little use of previous models from other systems. Finally, these models, even based on realistic biological hypotheses, may be considered negatively as black boxes by end users (health managers), because the underlying assumptions often became hidden in the code or equations.

The integration of multiple modelling perspectives (e.g., disciplines, points of view, spatio-temporal scales) is an important question in the modelling-simulation field. Epidemiological modelling could benefit from existing tools and methods developed in this field [ 47 , 48 , 49 ]. Although essential, good programming practices alone [ 50 ] cannot meet these challenges [ 51 ]. Scientific libraries and platforms accelerate the implementation of the complex models often needed in AH. For example, the R library SimInf [ 52 ] helps integrate observational data into mechanistic models. The BROADWICK framework [ 53 ] provides reusable software components for several scales and modelling paradigms, but still requires modellers to write large amounts of computer code.

New methods at the crossroads between software engineering and AI can enhance transparency and reproducibility in mechanistic modelling, fostering communication between software scientists, modellers and AH researchers throughout the modelling process (e.g., assumption formulation, assessment, and revision). Knowledge representation methods from symbolic AI, formalised using advanced software engineering methods such as domain-specific languages (DSL, e.g., in KENDRICK for differential equation models: [ 54 ]), makes model components accessible in a readable structured text file instead of computer code. Hence, scientists from various disciplines and field managers can be more involved in the model design and evaluation. Scenario exploration and model revision also no longer require rewriting the model code.

Other AI methods can improve model flexibility and modularity. Autonomous software agents enable to represent various levels of abstraction and organisation [ 55 ], helping modellers go more easily back and forth within small and larger scales, and ensure that all relevant mechanisms are adequately formalised at proper scales (i.e., scale-dependency of determinants and drivers in hierarchical living systems). Combining knowledge representation (through a DSL) and such a multi-level agent-based simulation architecture (e.g., in EMULSION, Figure  2 , [ 56 ]) enables to encompass several types of models (e.g., compartmental, individual-based) and scales (e.g., individual, population, territory), and it tackles simultaneously the recurring needs for transparency, reliability and flexibility in modelling contagious diseases. This approach should also facilitate in the future the production of support decision tools for veterinary and public health managers and stakeholders.

AI at the service of mechanistic epidemiological modelling (adapted from [ 51 ] ) . A . Modellers develop each epidemiological model de novo, producing specific codes not easily readable by scientists from other disciplines or by model end-users. B . Using AI approaches to combine a domain-specific language and an agent-based software architecture enhances reproducibility, transparency, and flexibility of epidemiological models. A simulation engine reads text files describing the system to automatically produce the model code. Complementary add-ons can be added if required. Models are easier to transfer to animal health managers as decision support tools.

3.3 Extracting knowledge from massive data in basic AH biology

Supervised, unsupervised and semi-supervised learning methods facilitate basic research development in biology and biomedicine, for example by using morphological analyses to study cell mobility [ 57 ]. The use of classification approaches and smart filters allows nowadays to sort massive molecular data (e.g., data from high throughput sequencing and metagenomics). Metabolic, physiological and immunological signalling pathways are explored, and metabolites are identified and quantified in complex biological mixtures, which was before a major challenge [ 58 ]. In addition, diagnostic time may be reduced by developing image analysis processing (e.g., accelerated detection of clinical patterns; [ 59 , 60 ]), often necessary to study host–pathogen interactions in animal pathology. For example, the use of optimisation methods has improved the understanding of the fragmentation of prion assemblages, contributing to a significant reduction in the time required to diagnose neurodegenerative animal diseases, thus paving the way for identifying potential therapeutic targets [ 61 ]. In livestock breeding, there is a methodological transition underway from traditional prediction strategies to more advanced machine learning approaches including artificial neural networks, deep learning and Bayesian networks which are being used to improve the reliability of genetic predictions and further the understanding of phenotypes biology. [ 62 ].

In human health, new disciplines have emerged in the second half of the 20 th century at the interface between AI and flagship disciplines, such as cell biology and immunology. Interface disciplines have developed, e.g., computational biology and immunology, which today must spread to AH. Current human immunology is based on the description of fine molecular and cellular mechanisms (e.g., the number of known interleukins has increased considerably compared to the 1970s). The desire to understand the processes underlying immune responses has led to a revolution by inviting this discipline to focus on complex systems biology and AI-based approaches [ 63 ]. However, the imbalance between the numbers of immunologists and immunology modellers is hampering the fantastic growth of this new discipline.

As an additional level of complexity, the hierarchical nature of biological systems makes that at the individual level, animals including humans must be considered as holobionts made of myriads of hosted microbial forms that form discrete ecological units (i.e., infracommunities). The potential of AI to grasp such diversity and complexity (e.g., tissue-specific microbiotes) and to scaling-up to higher levels of organization (e.g., component and compound communities of microbes, including pathogens, circulating in herd and in a given region) is certainly tremendous and should be studied with the same vigour as recent development in computational biology and immunology [ 40 ].

4 Revisiting AH case detection methods at different scales

Managing livestock health issues requires effective case detection methods, at the individual or even infra-individual (organ) scale, at the group/herd scale, or at larger scales (e.g., territories, countries). Machine learning methods allow detecting patterns and signals in massive data, e.g., in spatial data or time-series of health syndromes and disease cases, contributing to the development of smart agriculture and telemedicine (Figure  3 ). Alerts can be produced, and contribute to management advice in numerical agriculture [ 64 ] and veterinary practices [ 65 ]. AI may contribute to an earlier detection of infected cases and the rationalisation of treatments (including antimicrobials) in farm animals, by analysing data collected from connected sensors [ 66 ], by targeting individuals or groups of animals [ 59 ], or even by using mechanistic models to predict the occurrence of case detections and their treatment [ 67 ]. Also, machine learning methods enable to discriminate pathogen strains and thus to better understand their respective transmission pathways if different [ 68 ]. Finally, therapeutic strategies can be reasoned through multi-criteria optimisation, by identifying whom to treat in a herd, when, according to what protocol and for how long, in order to maximise the probability of cure while minimising both the risk of drug resistance and the volume or number of doses that are necessary (i.e., individual-based and precision medicine).

Extracting information from massive data to monitor animal health and better rationalise treatments.

Nevertheless, alert quality depends on the quality and representativeness of the datasets used by the learning algorithms. Numerous biases (e.g., hardware, software, human) can affect prediction accuracy. Moreover, alerts produced after training necessarily reflect the specificities of the system from which the data originates (e.g., area, period, rearing practices). Thus, result transposition to other epidemiological systems or to the same system subjected to environmental or regulatory changes remains risky. Furthermore, while machine learning methods (e.g., classification, image analysis, pattern recognition, data mining) provide solutions for a wide range of biomedical and bio-health research questions, it is crucial to demonstrate the performance of these methods by measuring their predictive quality and comparing them to alternative statistical methods whenever possible [ 69 ].

At the population level, case detection is based on direct (detection of syndromes) or indirect surveillance, mobilising syndrome proxies. Hence, the emergence of some animal diseases can be detected by syndromic surveillance, by detecting abnormal or rare signals in routine data (e.g., mortality, reproduction, abortion, behaviour, milk production, increased drug use; [ 70 ]). Also, serological data can be used retrospectively to identify individual characteristics related to a risk of being exposed to a pathogen, and thus orientate management efforts (e.g., in wildlife; [ 71 ]). Statistics and AI are largely complementary to address such issues. Both mobilise the wide range of available data, which are highly heterogeneous, massive and mostly sparse, to detect signals that are often weak or scarce [ 28 , 72 , 73 ]. Such signals can be proxy records (e.g., emergence of infectious diseases following environmental disturbances), health symptoms and syndromes, or even metabolic pathways in cascades which can be precursors of chronic or degenerative diseases. AI also includes methods to mobilise information available on the web. For example, semi-automatic data mining methods enabled to identify emerging signals for international surveillance of epizooties [ 74 ] or to analyse veterinary documents such as necropsy reports [ 75 , 76 ]. Methods from the field of natural language processing can compensate the scarcity of data by extracting syntactic and semantic information from textual records, triggering alerts on new emerging threats that could have been missed otherwise.

On a large to very large scale (i.e., territory, country, continent, global), data analysis of commercial animal movements between farms makes it possible to predict the associated epidemic risk [ 77 , 78 ]. These movements are difficult to predict, particularly since animal trade is based on many factors associated with human activities and decisions. Methods for recognising spatio-temporal patterns and methodological developments for the analysis of oriented and weighted dynamic relational graphs are required in this field because very few of the existing methods allow large-scale systems to be studied, whereas datasets are often very large (e.g., several tens or even hundreds of thousands of interacting operations).

On this topic, the specific frontier between learning methods of AI and statistics is relatively blurred, lying most on the relative prominence of the computational performance of algorithms versus mathematics, probability and rigorous statistical inference. While machine learning methods are more empirical, focused on improving their predictive performance, statistics is more concerned with the quantification and modelling of uncertainties and errors [ 79 , 80 ]. In the last decade, both communities have started to communicate and to mix together. Methods have cross-fertilised, giving birth to statistical models using synthetic variables generated by AI methods, or AI algorithms optimising statistical measures of likelihood or quality. New research areas, such as Probabilistic Machine Learning, have emerged at the interface between the two domains [ 1 , 80 , 81 ]. Meanwhile, machine learning and statistics have kept their specific interests and complementarity; machine learning methods are especially well-suited to processing non-standard data types (e.g., images, sounds), while statistics can draw inference and model processes for which only few data are available, or where the quantities of interest are extreme events.

5 Targeted interventions, model of human decisions, and support of AH decisions

5.1 choosing among alternatives.

A challenge for animal health managers is to identify the most relevant combinations of control measures according to local (e.g., farm characteristics, production objectives) and territorial (e.g., available resources, farm location, management priorities) specificities. They have to anticipate the effects of health, environmental and regulatory changes, and deliver quality health advice. The question also arises of how to promote innovation in AH, such as to anticipate the required characteristics of candidate molecules in vaccine strategies or drug delivery [ 82 , 83 ], or to assess the competitive advantage of new strategies (e.g., genomic selection of resistant animals, new vaccines) over more conventional ones. Private (e.g., farmers, farmers’ advisors) and collective managers (e.g., farmer groups, public authorities) need support decision tools to better target public incentives, identify investments to be favoured by farmers [ 46 ] and target the measures as effectively as possible: who to target (which farms, which animals)?; with which appropriate measure(s)?; when and for how long? These questions become essential to reasoning about input usage (e.g., antimicrobials, pesticides, biocides) within the framework of the agro-ecological transition.

The use of mechanistic modelling is a solution to assess, compare and prioritise ex ante a wide range of options (Figure  4 ; [ 84 ]). However, most of the available models do not explicitly integrate human decision-making, while control decisions are often made by farmers (e.g., unregulated diseases), with sometimes large-scale health and decision-making consequences (e.g., pathogen spread, dissemination of information and rumours, area of influence). Recent work aims to integrate humans and their decisions by mobilising optimal control and adaptive strategies from AI [ 7 , 85 ] or health economics methods [ 86 , 87 ]. A challenge is to propose clear and context-adapted control policies [ 88 ]. Such research is just starting in AH [ 46 ] and must be extended as part of the development of agro-ecology, facing current societal demand for product quality and respect for ecosystems and their biodiversity on one side, animal well-being and ethics on the other side, and more generally international health security.

Identifying relevant strategies to control bovine paratuberculosis at a regional scale (adapted from [ 76 ] ) . Classically, identifying relevant strategies means defining them a priori and comparing them, e.g., by modelling. Only a small number of alternatives can be considered. If all alternatives are considered as in the figure, it results in a multitude of scenarios whose analysis becomes challenging. Here, each point corresponds to the epidemiological situation after 9 years of pathogen spread over a network of 12 500 dairy cattle herds for a given strategy (asterisk: no control). Initially, 10% of the animals are infected on average in 30% of the herds. The blue dots correspond to the most favourable strategies. Mobilizing AI approaches in such a framework, especially optimization under constraints, would facilitate the identification of relevant strategies by exploring the space of possibilities in a more targeted manner.

5.2 Accounting for expectations and fears of animal health managers

Animal health managers should have access to model predictions in a time frame compatible with management needs, which is problematic in the face of unpredictable emerging events (e.g., new epidemiological systems, transmission pathways, trade patterns, control measures). Developing a library of models included in a common framework would strengthen the responsiveness of modellers in animal epidemiology. Relevant models would be developed more quickly and would gain accuracy from real-time modelling as epidemics progress [ 89 , 90 ]. However, if this makes move more quickly from concepts (knowledge and assumptions) to simulations and support decision tools, a gain in performance is still required to perform analyses at a very large scale. The automatic generation of high-performance computer code could be a relevant solution, which however remains a crucial methodological lock to be addressed in AI. In addition, it is often required to perform a very large number of calculations or to analyse very large datasets, which call for a rational use of computing resources. Software transferred to health managers sometimes require the use of private cloud resources (i.e., it does not run on simple individual computers), highlighting the trade-offs between simulation cost, service continuity (e.g., failure management) and time required to obtain simulation results [ 91 ]. These questions are currently related to computer science research, and collaborations are desirable between these researchers and those from AH.

Managers also wish to rely on accurate predictions from realistic representations of the biological systems. Before being used, model behaviour should be analysed, which raises the questions of exploring the space of uncertainties and data, and of optimization under constraints. This often requires intensive simulations, which would benefit from optimization algorithms to explore more efficiently the space of possibilities. In turn, this would allow, for example, the automatic identification of how to achieve a targeted objective (e.g., reducing the prevalence of a disease below an acceptable threshold) while being constrained in resource allocation. While this issue finds solutions in modern statistics for relatively simple systems, it represents a science front for complex systems (e.g., large scale, multi-host/multi-pathogen systems) that are becoming the norm. In addition, optimization goals specific to AH may generate ad hoc methodological needs [ 92 ]. The needs in abstraction and analysis capacity are massive and could benefit from complementarities between AI (e.g., reasoned exploration, intelligent use of computer resources, optimized calculations) and statistics to extract as much information as possible from the data: (1) explore, analyse, predict; (2) infer processes and emergent properties. Methodological developments are still required and would benefit many health issues, particularly in relation to the currently evolving concepts of reservoir-host, edge-host and species barrier [ 93 ]. Furthermore, methodological developments and dissemination of existing methods should be reinforced.

Finally, three barriers have been identified to the development of support decision tools for health managers, related to the societal issue of the acceptability of AI sensu lato, as a major factor of progress. First, ethical issues, which are obvious when it comes to human health, are just as important to consider in AH. Which AI-based tools do we want for modern animal husbandries and trades, and for which objectives? Are these tools not likely to lead to discrimination against farms according to their health status, even when this status cannot be managed by the farmer alone? Second, in AH too, there is a fear that AI-tools may replace human expertise. However, automating does not mean replacing human, his expertise and decision [ 94 ], but rather supporting his capacities for abstraction and analysis, accelerating the global process, making predictions more reliable, guiding complementary research. Nevertheless, a significant development of computer resources and equipment is not without impacting the environment in terms of carbon footprint (e.g., energy-intensive servers, recycling of sensors), which must also be accounted for. Third, the very high complexity of analysing results and acculturating end-users with knowledge issued from academic research, particularly AI, is an obstacle to the appropriation of AI-tools by their users. This may lead to the preference for simpler and more easily accessible methods. However, the latter may not always be the most relevant or reliable. Citizen science projects, also known as community participation in human epidemiology, enable AH to co-design and co-construct the AI-tools of tomorrow with their end-users [ 95 ], to better meet their expectations and needs, and to increase their confidence in the predictions of sometimes obscure research models, especially when they are hard to read (e.g., lines of code). Similarly, these AI-tools could be developed together with public decision-makers, livestock farmers, agro-food industries and sectoral trade unions. Co-construction gives time to explain the science behind the tools and makes it more transparent and useful. This citizen participation, which is nowadays supported in many countries, guarantees decisions more in line with citizens’ expectations and corresponds to a general trend towards structured decision-making. AI must contribute to this democratisation of aid in public decision-making in AH.

6 Barriers to the development of research at the AI/AH interface

Research conducted at the interface between AI and AH requires strong interactions between biological disciplines (e.g., infectiology, immunology, clinical sciences, genetics, ecology, evolution, epidemiology, animal and veterinary sciences) and more theoretical disciplines (e.g., modelling, statistics, computer science), sometimes together with sociology and economics. Conducting research at this interface requires strengthening the few teams already positioned in Western Europe, but also bringing together teams working around the concepts of One Health, Ecohealth and Planetary Health to benefit from recent achievements in infectious disease ecology and modelling, plant health and environmental health [ 96 ]. This work must be based on a wide range of methodological skills (e.g., learning methods, data mining, information systems, knowledge representation, multi-agent systems, problem solving, metamodeling, optimisation, simulation architecture, model reduction, decision models). The need for research, training and support are crucial issues at national, European and international levels. Also, a facilitated and trusted connection is required between academics, technical institutes, and private partners, who are often the holders or collectors of data of interest to solve AH research questions through AI approaches. The construction of better inter-sectoral communication and coordination must be done at supra-institutional level, as this theme seems hyper-competitive and as some current divisions still go against information and data sharing.

An acculturation of researchers to AI, its methods and potential developments, but also its limitations, must be proposed to meet the challenges of 21 st century agriculture. Indeed, there are obstacles to conducting research on this scientific front. Establishing the new collaborations required between teams conducting methodological work and teams in the fields of application remains difficult given the low number of academic staff on these issues, their very high current mobilisation and their low availability to collaborate on new subjects, as well as the difficulty of understanding and mastering these methods. There is a need for watching and training on AI methods available or under development, new softwares/packages, and their applicability. To develop key collaborations and establish a strategic positioning, an interconnection can also be made via transversal teams which appears as a preferential path. Solutions must also be found to encourage method percolation in the community and the development of scientific and engineering skills.

Finally, AI methods, such as classification, machine learning, data mining, and the innovations in AH to which these methods can lead are rarely discussed in veterinary high school education, whereas these students represent the future professionals of AH [ 96 ]. Similarly, there is a quasi-absence of sentinel networks of veterinarians, even if it is developing, although AH questions can arise on a large and collective scale. The scientific community would also benefit from further increasing its skills and experience in the valuation, transfer and protection of intellectual property on these AI methods and associated outcomes.

7 Levers to create a fruitful AI/AH interface

7.1 data sharing and protection.

No innovation at the interface between AI and AH is possible without strong support for the organisation of data storage, management, analysis, calculation, and restitution. The major risk is that demands for AI developments inflate without being supported by available human resources. In addition, an expertise in law, jurisdiction and ethics is required with regard to the acquisition, holding, use and protection of data in AH. This question must be considered at least at the inter-institutional/national level, and could benefit from a similar thinking already engaged in human health. The issue is to be able to support any change with regard to data traceability to their ownership, whether being from public or private domains.

New data are rich and must be valued as much as possible, not by each owner separately, but through data sharing and the mobilisation of multi-disciplinary skills to analyse such heterogeneous and complex data. Hence, data interoperability skills are required and must be developed. Models for making federated data sustainable over decades are required [ 97 ]. In addition, further encouraging the publication of data papers as valuable research products can help to develop the necessary culture of sharing, documentation and metadata.

Finally, to be able to launch ambitious experiments with AI methods on real data, it is necessary to (1) remove unauthorised access to data by negotiating with owners at large scale; (2) analyse and understand the related effect on methodological developments; and (3) if necessary, extend such initiatives to other areas, at national scale, or even across European countries.

7.2 Attract the necessary skills

An undeniable barrier to conduct such research comes from human resources, in particular the current insufficient capacity of supervision by permanent scientists. Collaborations are a solution to attract new skills. However, initiating collaborations at the AH/AI interface becomes very complicated because the qualified teams are already overwhelmed. Skill development at this interface must be supported, the cross-fertilisation of disciplines being essential. A watch on methods must also be carried out, accompanied by explanations for application fields, to train researchers and engineers. Financial incentives for scientist internships in specialised laboratories would increase skill capitalisation in advanced methods, while facilitating future national or international collaborations. In a context of limited resources as observed in many countries nowadays (e.g., new opened positions in national institutions) and limited experts pool (e.g., skills), facilitating post-doctoral fellows and continuing education of researchers becomes crucial. Finally, to consolidate the pool of future researchers in AH, promoting basic AI education in initial training of AH researchers, engineers and veterinarians is paramount.

More specifically concerning current research in immunology, cell biology and infectiology, the contribution of AI has been more widely considered in human health, which could feed a similar reflection in AH as locks and advances are not very specific. Before embarking on the fronts of science (e.g., emerging epigenomics and metabolomics in AH), a few persons from these biological disciplines should acculturate into AI, or even acquire autonomy in the use of methods [ 98 ], which internationally tends to be the trend [ 63 ]. This can be done through the sharing of experiences and basic training on existing methods, their advantages and limitations compared to other methods coming from statistics and mathematical modelling.

7.3 Encourage the development of AH/AI projects

Projects at the AH/AI interface, like any interdisciplinary project, must mobilise teams from both groups of disciplines and allow everyone to progress in their own discipline. However, identifying the issues shared between the most relevant disciplines requires a good acculturation of the disciplines between them, as well as an otherness aimed at better understanding each other [ 95 ], which is not yet the case at the AH/AI interface.

In terms of funding, European project calls offer interesting opportunities, but a significant imbalance persists between the ability to generate data and analyse complex issues, and the availability of human resources and skills to address such issues through AI methods or other modern methods in statistics, mathematics and computer science. The major international foundations (e.g., Bill and Melinda Gates) can also be mobilised on emerging infectious diseases at the animal/human interface (e.g., characterisation of weak signals, phenologies, emergence precursors), with a more significant methodological value. However, risk-taking is rarely allowed by funding agencies, although it is crucial to initiate interdisciplinary work. Dedicated incentive funding would support projects in their initial phase and make larger projects emerge after consolidation of the necessary disciplinary interactions.

Finally, these projects are generally based on the use of significant computing resources. Thus, research institutes and private partners should contribute in a financial or material way to the shared development of digital infrastructures, data centres, supercomputing centres on a national scale, as well as support recognised open-source software platforms on which a large part of the research conducted is based (e.g., Python, R ).

7.4 Promoting innovation and public–private partnership

Encouraging public–private partnership would promote a leverage effect on public funding and would make it possible to place AI research and development on a long-term basis in AH. Mapping the highly changing landscape of companies in the AH/AI sector, whether international structures or start-ups, would provide a better understanding of the possible interactions. Similarly, mapping academic deliverables produced at this interface would increase their visibility and highlight their potential for valorisation or transfer. Finally, considering the production of documented algorithms as scientific deliverables, along with publications, would help support this more operational research. More broadly, it would be advisable to initiate a communication and education/acculturation policy around AI and its development in AH (e.g., links with the society, farmers, agricultural unions, public services).

8 Conclusion

The use of AI methods (e.g., machine learning, expert systems, analytical technologies) converges today with the collecting of massive and complex data, and allows these fields to develop rapidly. However, it is essential not to perceive massive data and AI as the same trend, because the accumulation of data does not always lead to an improvement in knowledge. Nevertheless, the more data are numerous and representative of working concepts and hypotheses, the more important results can be obtained from AI applications. The underlying ethical, deontological and legal aspects of data ownership, storage, management, sharing and interoperability also require that a reflection be undertaken nationally and internationally in AH to better manage these data of multi-sectoral origin and their various uses. Moreover, while the effort to acquire such data is impressive, the development of AI skills within the AH community remains limited in relation to the needs. Opportunities for collaborations with AI teams are limited because these teams are already in high demand. To ensure that AH researchers are well aware of the opportunities offered by AI, but also of the limits and constraints of AI approaches, a training effort must be provided and generalized. Finally, the current boom in AI now makes it possible to integrate the knowledge and points of view of the many players in the field of animal health and welfare further upstream. However, this requires that AI and its actors accept to deal with the specificity and complexity of AH, which is not a simple library of knowledge that can be digitised to search for sequences or informative signals.

Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL (2018) Artificial intelligence in radiology. Nat Rev Cancer 18:500–510

Article   CAS   Google Scholar  

Karczewski KJ, Snyder MP (2018) Integrative omics for health and disease. Nat Rev Genet 19:299–310

Murphy KP (2012) Machine learning: a probabilistic perspective. In: Adaptive computation and machine learning series. MIT Press, USA

Zhang W, Chien J, Yong J, Kuang R (2017) Network-based machine learning and graph theory algorithms for precision oncology. NPJ Precis Oncol 1:25

Article   Google Scholar  

Saria S, Butte A, Sheikh A (2018) Better medicine through machine learning: what’s real, and what’s artificial? PLoS Med 15:e1002721. https://doi.org/10.1371/journal.pmed.1002721

Article   PubMed   PubMed Central   Google Scholar  

Bedi G, Carrillo F, Cecchi GA, Fernández Slezak D, Sigman M, Mota NB, Ribeiro S, Javitt DC, Copelli M, Corcoran CM (2015) Automated analysis of free speech predicts psychosis onset in high-risk youths. Schizophrenia 1:15030

Maclachlan MJ, Springborn MR, Fackler PL (2017) Learning about a moving target in resource management: optimal Bayesian disease control. Am J Agri Econ 99:140–162. https://doi.org/10.1093/ajae/aaw033

Lynn LA (2019) Artificial intelligence systems for complex decision-making in acute care medicine: a review. Patient Saf Surg 13:6. https://doi.org/10.1186/s13037-019-0188-2

Pinaire J, Azé J, Bringay S, Landais P (2017) Patient healthcare trajectory. An essential monitoring tool: a systematic review. Health Inf Sci Syst 5:1

Vrakas D, Vlahavas IPL (2008) Artificial intelligence for advanced problem solving techniques. Information Science Reference, Hershey, PA, pp. 369

Shakshuki E, Reid M (2015) Multi-agent system applications in healthcare: current technology and future roadmap. Proc Comput Sci 52:252–261. https://doi.org/10.1016/j.procs.2015.05.071

Roche B, Guégan JF, Bousquet F (2008) Multi-agent systems in epidemiology: a first step for computational biology in the study of vector-borne disease transmission. BMC Bioinform 9:435. https://doi.org/10.1186/1471-2105-9-435

Picault S, Huang Y-L, Sicard V, Ezanno P (2017) Enhancing Sustainability of Complex Epidemiological Models through a Generic Multilevel Agent-based Approach. In: Proceedings of the 26 th International Joint Conference on Artificial Intelligence (IJCAI). pp. 374–380, AAAI. https://doi.org/10.24963/ijcai.2017/53p

Russell S, Norvig P (2010) Artificial intelligence a modern approach. 3 rd edn. Upper Saddle River, New Jersey, pp. 1132

Google Scholar  

Ducrot C, Bed’Hom B, Béringue V, Coulon JB, Fourichon C, Guérin JL, Krebs S, Rainard P, Schwartz-Cornil I, Torny D, Vayssier-Taussat M, Zientara S, Zundel E, Pineau T (2011) Issues and special features of animal health research. Vet Res 42:96

Clark B, Stewart GB, Panzone LA, Kyriazakis I, Frewer LJ (2016) A systematic review of public attitudes, perceptions and behaviours towards production diseases associated with farm animal welfare. J Agric Environ Ethics 29:455–478. https://doi.org/10.1007/s10806-016-9615-x

Miguel E, Grosbois V, Caron A, Pople D, Roche B, Donnelly C (2020) A systemic approach to assess the potential and risks of wildlife culling for infectious disease control. Commun Biol 3:353. https://doi.org/10.1038/s42003-020-1032-z

Hur B, Hardefeldt LY, Verspoor K, Baldwin T, Gilkerson JR (2019) Using natural language processing and VetCompass to understand antimicrobial usage patterns in Australia. Aust Vet J 97:298–300. https://doi.org/10.1111/avj.12836

Article   CAS   PubMed   Google Scholar  

Behmann J, Hendriksen K, Mueller U, Buescher W, Pluemer L (2016) Support vector machine and duration-aware conditional random field for identification of spatio-temporal activity patterns by combined indoor positioning and heart rate sensors. Geoinformatica 20:693–714. https://doi.org/10.1007/s10707-016-0260-3

Suravajhala P, Kogelman LJA, Kadarmideen HN (2016) Multi-omic data integration and analysis using systems genomics approaches: methods and applications in animal production, health and welfare. Genet Sel Evol 48:38. https://doi.org/10.1186/s12711-016-0217-x

Article   CAS   PubMed   PubMed Central   Google Scholar  

Goldansaz SA, Guo AC, Sajed T, Steele MA, Plastow GS, Wishart DS (2017) Livestock metabolomics and the livestock metabolome: a systematic review. PLoS One 12:e0177675. https://doi.org/10.1371/journal.pone.0177675

Anvar SY, Tucker A, Vinciotti V, Venema A, van Ommen GJ, van der Maarel SM, Raz V, ’t Hoen PA (2011) Interspecies translation of disease networks increases robustness and predictive accuracy. PLoS Comput Biol 7:e1002258. https://doi.org/10.1371/annotation/fc0b4192-6427-4fb3-b347-c66651adf855

Morand S, Guégan J-F, Laurans Y (2020) From One Health to Ecohealth, mapping the incomplete integration of human, animal and environmental health. Iddri, Issue Brief No. 04/20

Ezenwa VO, Prieur-Richard A-H, Roche B, Bailly X, Becquart P, Garcia-Peña GE, Hosseini PR, Keesing F, Rizzoli A, Suzán GA, Vignuzzi M, Vittecoq M, Mills JN, Guégan J-F (2015) Interdisciplinarity and infectious diseases: an Ebola case study. PLoS Pathog 11:e1004992. https://doi.org/10.1371/journal.ppat.1004992

Van Boeckel TP, Takahashi S, Liao Q, Xing W, Lai S, Hsiao V, Liu F, Zheng Y, Chang Z, Yuan C, Metcalf CJE, Yu H, Grenfell BT (2016) Hand, foot, and mouth disease in China: critical community size and spatial vaccination strategies. Sci Rep 6:25248. https://doi.org/10.1038/srep25248

Holmstrom LK, Beckham TR (2017) Technologies for capturing and analysing animal health data in near real time. Rev Sci Tech 36:525–538

Neethirajan S (2017) Recent advances in wearable sensors for animal health management. Sens Biosensing Res 12:15–29

Perez AM, Zeng D, Tseng CJ, Chen H, Whedbee Z, Paton D, Thurmond MC (2009) A web-based system for near real-time surveillance and space-time cluster analysis of foot-and-mouth disease and other animal diseases. Prev Vet Med 91:39–45. https://doi.org/10.1016/j.prevetmed.2009.05.006

Article   PubMed   Google Scholar  

Wilkinson MD, Dumontier M, Aalbersberg IJ, Appleton G, Axton M, Baak A, Blomberg N, Boiten J-W, da Silva Santos LB, Bourne PE, Bouwman J, Brookes AJ, Clark T, Crosas M, Dillo I, Dumon O, Edmunds S, Evelo CT, Finkers R, Gonzalez-Beltran A, Gray AJG, Groth P, Goble C, Grethe JS, Heringa J, ’t Hoen PAC, Hooft R, Kuhn T, Kok R, Kok J, et al (2016) The FAIR guiding principles for scientific data management and stewardship. Sci Data 3:160018. https://doi.org/10.1038/sdata.2016.18

Binot A, Duboz R, Promburom P, Phimpraphai W, Cappelle J, Lajaunie C, Goutard FL, Pinyopummintr T, Figuié M, Roger FL (2015) A framework to promote collective action within the One Health community of practice: using participatory modelling to enable interdisciplinary, cross-sectoral and multi-level integration. One Health 1:44–48. https://doi.org/10.1016/j.onehlt.2015.09.001

Robert CP (2014) Bayesian computational tools. Annu Rev Stat Appl 1:153–177. https://doi.org/10.1146/annurev-statistics-022513-115543

Dunson DB (2001) Commentary: practical advantages of Bayesian analysis of epidemiologic data. Am J Epidemiol 153:1222–1226. https://doi.org/10.1093/aje/153.12.1222

Uusitalo L (2007) Advantages and challenges of Bayesian networks in environmental modelling. Ecol Model 203:312–318. https://doi.org/10.1016/j.ecolmodel.2006.11.033

Fokoué E (2019) On the ubiquity of the Bayesian paradigm in statistical machine learning and data science. Math Appl 8:189–209. https://doi.org/10.13164/ma.2019.12

Bailly X (2017) Hidden Markov phylogenetic models offer an interesting perspective to identify “high risk lineages” of environmental pathogens. Infect Genet Evol 55:45–47. https://doi.org/10.1016/j.meegid.2017.08.007

Babayan SA, Orton RJ, Streicker DG (2018) Predicting reservoir hosts and arthropod vectors from evolutionary signatures in RNA virus genomes. Science 362:577–580. https://doi.org/10.1126/science.aap9072

Wardeh M, Sharkey KJ, Baylis M (2020) Integration of shared-pathogen networks and machine learning reveals the key aspects of zoonoses and predicts mammalian reservoirs. Proc Biol Sci 287:20192882. https://doi.org/10.1098/rspb.2019.2882

Li J, Zhang S, Li B, Hu Y, Kang X-P, Wu X-Y, Huang M-T, Li Y-C, Zhao Z-P, Qin C-F, Jiang T (2020) Machine learning methods for predicting human-adaptive influenza A viruses based on viral nucleotide compositions. Mol Biol Evol 37:1224–1236. https://doi.org/10.1093/molbev/msz276

Peters DPC, McVey DS, Elias EH, Pelzel-McCluskey AM, Derner JD, Burruss ND, Schrader TS, Yao J, Pauszek SJ, Lombard J, Rodriguez LL (2020) Big data-model integration and AI for vector-borne disease prediction. Ecosphere 11:e03157. https://doi.org/10.1002/ecs2.3157

Lek S, Guégan J-F (2000) Artificial neuronal networks. In: Application to ecology and evolution. Springer, Berlin. https://doi.org/10.1016/j.it.2016.11.006

Go N, Touzeau S, Islam Z, Belloc C, Doeschl-Wilson A (2019) How to prevent viremia rebound? Evidence from a PRRSv data-supported model of immune response. BMC Syst Biol 13:15

Ferrer Savall J, Bidot C, Leblanc-Maridor M, Belloc C, Touzeau S (2016) Modelling Salmonella transmission among pigs from farm to slaughterhouse: interplay between management variability and epidemiological uncertainty. Intern J Food Microbiol 229:33–43. https://doi.org/10.1016/j.ijfoodmicro.2016.03.020

Widgren S, Engblom S, Bauer P, Frössling J, Emanuelson U, Lindberg A (2016) Data-driven network modelling of disease transmission using complete population movement data: spread of VTEC O157 in Swedish cattle. Vet Res 47:81

Qi L, Beaunée G, Arnoux S, Dutta BL, Joly A, Vergu E, Ezanno P (2019) Neighbourhood contacts and trade movements drive the regional spread of bovine viral diarrhoea virus (BVDV). Vet Res 50:30. https://doi.org/10.1186/s13567-019-0647-x

Buhnerkempe MG, Tildesley MJ, Lindström T, Grear DA, Portacci K, Miller RS, Lombard JE, Werkman M, Keeling MJ, Wennergren U, Webb CT (2014) The impact of movements and animal density on continental scale cattle disease outbreaks in the United States. PLoS One 9:e91724. https://doi.org/10.1371/journal.pone.0091724

Ezanno P, Andraud M, Beaunée G, Hoch T, Krebs S, Rault A, Touzeau S, Vergu E, Widgren S (2020) How mechanistic modelling supports decision 1 making for the control of enzootic infectious diseases. Epidemics 32:100398

Garira W (2018) A primer on multiscale modelling of infectious disease systems. Infect Dis Model 3:176–191. https://doi.org/10.1016/j.idm.2018.09.005

Traoré M, Zacharewicz G, Duboz R, Zeigler B (2018) Modeling and simulation framework for value-based healthcare systems. Simulation 95:481–497. https://doi.org/10.1177/0037549718776765

Childs LM, El Moustaid F, Gajewski Z, Kadelka S, Nikin-Beers R, Smith JW Jr, Walker M, Johnson LR (2019) Multi-scale models and data for infectious diseases: a systematic review. PeerJ Preprints 7:e27485v1. https://doi.org/10.7287/peerj.preprints.27485v1

Sandve GK, Nekrutenko A, Taylor J, Hovig E (2013) Ten simple rules for reproducible computational research. PLoS Comput Biol 9:e1003285. https://doi.org/10.1371/journal.pcbi.1003285

Leek JT, Peng RD (2015) Opinion: reproducible research can still be wrong: adopting a prevention approach. Proc Natl Acad Sci USA 112:1645–1646. https://doi.org/10.1073/pnas.1421412111

Widgren S, Bauer P, Eriksson R, Engblom S (2016) SimInf: an R package for data-driven stochastic disease spread simulations. ArXiv160501421 Q-Bio Stat. http://arxiv.org/abs/1605.01421

O’Hare A, Lycett SJ, Doherty TM, Salvador LC, Kao RR (2016) Broadwick: a framework for computational epidemiology. BMC Bioinform 17:65. https://doi.org/10.1186/s12859-016-0903-2

Bui TMA, Stinckwich S, Ziane M, Roche B, Ho TV (2015) KENDRICK: a domain specific language and platform for mathematical epidemiological modelling. In: proc. IEEE RIVF International Conference on Computing and Communication Technologies, Research, Innovation, and Vision for the Future. pp. 132–7. https://doi.org/10.1109/RIVF.2015.7049888

Mathieu P, Morvan G, Picault S (2018) Multi-level agent-based simulations: four design patterns. Simul Model Pract Theory 83:51–64. https://doi.org/10.1016/j.simpat.2017.12.015

Picault S, Huang Y-L, Sicard V, Arnoux S, Beaunée G, Ezanno P (2019) EMULSION: transparent and flexible multiscale stochastic models in human, animal and plant epidemiology. PLoS Comput Biol 15:e1007342. https://doi.org/10.1371/journal.pcbi.1007342

Sebag AS, Plancade S, Raulet-Tomkiewicz C, Barouki R, Vert J-P, Walter T (2015) Inferring an ontology of single cell motions from high-throughput microscopy data. In: Proc. IEEE International Symposium on Biomedical Imaging, Apr. 2015, New-York, USA, pp. 160–163. https://doi.org/10.1109/ISBI.2015.7163840

Tardivel P, Canlet C, Lefort G, Tremblay-Franco M, Debrauwer L, Concordet D, Servien R (2017) ASICS: an automatic method for identification and quantification of metabolites in complex 1D 1H NMR spectra. Metabolomics 13:109

Dórea FC, Muckle CA, Kelton D, McClure JT, McEwen BJ, McNab WB, Sanchez J, Revie CW (2013) Exploratory analysis of methods for automated classification of laboratory test orders into syndromic groups in veterinary medicine. PLoS One 8:e57334. https://doi.org/10.1371/journal.pone.0057334

Gandia P, Jaudet C, Chatelut E, Concordet D (2017) Population pharmacokinetics of tracers: a new tool for medical imaging? Clin Pharmacokinet 56:101–106

Chyba M, Coron J-M, Mileyko Y, Rezaei H (2016) Optimization of prion assemblies fragmentation. In: Proc. IEEE Conference on Decision and Control (CDC), Las Vegas, USA, 6

Nayeri S, Sargolzaei M, Tulpan D (2019) A review of traditional and machine learning methods applied to animal breeding. Anim Health Res Rev 20:31–46. https://doi.org/10.1017/S1466252319000148

Bassaganya-Riera J, Hontecillas R (2016) Introduction to computational immunology. In: Bassaganya-Riera J (ed) Computational immunology: models and tools. pp. 1–8

Liakos KG, Busato P, Moshou D, Pearson S, Bochtis D (2018) Machine learning in agriculture: a review. Sensors 18:2674. https://doi.org/10.3390/s18082674

Jones-Diette JS, Dean RS, Cobb M, Brennan ML (2019) Validation of text-mining and content analysis techniques using data collected from veterinary practice management software systems in the UK. Prev Vet Med 167:61–67. https://doi.org/10.1016/j.prevetmed.2019.02.015

Morota G, Ventura RV, Silva FF, Koyama M, Fernando SC (2018) Big data analytics and precision animal agriculture symposium: machine learning and data mining advance predictive big data analysis in precision animal agriculture. J Anim Sci 96:1540–1550. https://doi.org/10.1093/jas/sky014

Picault S, Ezanno P, Assié S (2019) Combining early hyperthermia detection with metaphylaxis for reducing antibiotics usage in newly received beef bulls at fattening operations: a simulation-based approach. In: Society of veterinary epidemiology and preventive medicine (SVEPM), pp. 13. Utrecht, The Netherland, 27-30/3/2019

Esener N, Green MJ, Emes RD, Jowett B, Davies PL, Bradley AJ, Dottorini T (2018) Discrimination of contagious and environmental strains of Streptococcus uberis in dairy herds by means of mass spectrometry and machine-learning. Sci Rep 8:17517. https://doi.org/10.1038/s41598-018-35867-6

Hepworth PJ, Nefedov AV, Muchnik IB, Morgan KL (2012) Broiler chickens can benefit from machine learning: support vector machine analysis of observational epidemiological data. J R Soc Interface 9:1934–1942. https://doi.org/10.1098/rsif.2011.0852

Marceau A, Madouasse A, Lehébel A, van Schaik G, Veldhuis A, Van der Stede Y, Fourichon C (2014) Can routinely recorded reproductive events be used as indicators of disease emergence in dairy cattle? An evaluation of 5 indicators during the emergence of bluetongue virus in France in 2007 and 2008. J Dairy Sci 97:6135–6150. https://doi.org/10.3168/jds.2013-7346

Fountain-Jones NM, Machado G, Carver S, Packer C, Recamonde-Mendoza M, Craft ME (2019) How to make more from exposure data? An integrated machine learning pipeline to predict pathogen exposure. J Anim Ecol 88:1447–1461. https://doi.org/10.1111/1365-2656.13076

Charras-Garrido M, Azizi L, Forbes F, Doyle S, Peyrard N, Abrial D (2013) On the difficulty to delimit disease risk hot spots. Int J Appl Earth Obs 22:99–105. https://doi.org/10.1016/j.jag.2012.04.005

Forbes F, Charras-Garrido M, Azizi L, Doyle S, Abrial D (2013) Spatial risk mapping for rare disease with hidden Markov fields and variational EM. Annals Appl Stat 7:1192–1216

Arsevska E, Valentin S, Rabatel J, de Goër de Hervé J, Falala S, Lancelot R, Roche M (2018) Web monitoring of emerging animal infectious diseases integrated in the French Animal Health Epidemic Intelligence System. PLoS One 13:0199960. https://doi.org/10.1371/journal.pone.0199960

Küker S, Faverjon C, Furrer L, Berezowski J, Posthaus H, Rinaldi F, Vial F (2018) The value of necropsy reports for animal health surveillance. BMC Vet Res 14:191. https://doi.org/10.1186/s12917-018-1505-1

Bollig N, Clarke L, Elsmo E, Craven M (2020) Machine learning for syndromic surveillance using veterinary necropsy reports. PLoS One 15:e0228105. https://doi.org/10.1371/journal.pone.0228105

Hoscheit P, Geeraert S, Beaunée G, Monod H, Gilligan CAG, Filipe J, Vergu E, Moslonka-Lefebvre M (2016) Dynamical network models for cattle trade: towards economy-based epidemic risk assessment. J Complex Netw 5:604–624. https://doi.org/10.1093/comnet/cnw026

Moslonka-Lefebvre M, Gilligan CA, Monod H, Belloc C, Ezanno P, Filipe JAN, Vergu E (2016) Market analyses of livestock trade networks to inform the prevention of joint economic and epidemiological risks. J R Soc Interface 13:20151099. https://doi.org/10.1098/rsif.2015.1099

Efron B (2020) Prediction, estimation, and attribution. J Am Stat Ass 115:636–655. https://doi.org/10.1080/01621459.2020.1762613

Ghahramani Z (2012) Probabilistic modelling, machine learning, and the information revolution. MIT Computer Science and Artificial Intelligence Lab, http://mlg.eng.cam.ac.uk/zoubin/talks/mit12csail.pdf , Accessed 17 Oct 2019

Hastie T, Tibshirani R, Friedman JH (2009) The elements of statistical learning: data mining, inference, and prediction. 2 nd edn. Springer Series in Statistics. Springer

Goodswen SJ, Kennedy PJ, Ellis JT (2017) On the application of reverse vaccinology to parasitic diseases: a perspective on feature selection and ranking of vaccine candidates. Int J Parasitol 47:779–790. https://doi.org/10.1016/j.ijpara.2017.08.004

Schneider G (2019) Mind and machine in drug design. Nat Mach Intell 1:128–130. https://doi.org/10.1038/s42256-019-0030-7

Beaunée G, Vergu E, Joly A, Ezanno P (2017) Controlling bovine paratuberculosis at a regional scale: towards a decision modeling tool. J Theor Biol 435:157–183. https://doi.org/10.1016/j.jtbi.2017.09.012

Viet A-F, Krebs S, Rat-Aspert O, Jeanpierre L, Belloc C, Ezanno P (2018) A modelling framework based on MDP to coordinate farmers’ disease control decisions at a regional scale. PLoS One 13:e0197612. https://doi.org/10.1371/journal.pone.0197612

Wang T, Hennessy DA (2015) Strategic interactions among private and public efforts when preventing and stamping out a highly infectious animal disease. Am J Agri Econ 97:435–451. https://doi.org/10.1093/ajae/aau119

Tago D, Hammitt JK, Thomas A, Raboisson D (2016) The impact of farmers’ strategic behavior on the spread of animal infectious diseases. PLoS One 11:e0157450. https://doi.org/10.1371/journal.pone.0157450

Probert WJM, Lakkur S, Fonnesbeck CJ, Shea K, Runge MC, Tildesley MJ, Ferrari MJ (2019) Context matters: using reinforcement learning to develop human-readable, state-dependent outbreak response policies. Phil Trans R Soc B 374:20180277. https://doi.org/10.1098/rstb.2018.0277

Liang R, Lu Y, Qu X, Su Q, Li C, Xia S, Liu Y, Zhang Q, Cao X, Chen Q, Niu B (2020) Prediction for global African swine fever outbreaks based on a combination of random forest algorithms and meteorological data. Transbound Emerg Dis 67:935–946. https://doi.org/10.1111/tbed.13424

Salje H, Tran Kiem C, Lefrancq N, Courtejoie N, Bosetti P, Paireau J, Andronico A, Hozé N, Richet J, Dubost C-L, Le Strat Y, Lessler J, Levy Bruhl D, Fontanet A, Opatowski L, Boelle P-Y, Cauchemez S (2020) Estimating the burden of SARS-CoV-2 in France. Science 369:208–211

Parlavantzas N, Pham LM, Morin C, Arnoux S, Beaunée G, Qi L, Gontier P, Ezanno P (2019) A service-based framework for building and executing epidemic simulation applications in the cloud. Concurr Comp Pract Exper 32:e5554. https://doi.org/10.1002/cpe.5554

Shah N, Malensek M, Shah H, Pallickara S, Pallickara SL (2019) Scalable network analytics for characterization of outbreak influence in voluminous epidemiology datasets. Concurr Comp Pract Exper 31:e4998. https://doi.org/10.1002/cpe.4998

Han BA, Majumdar S, Calmon FP, Glicksberg BS, Horesh R, Kumar A, Perer A, von Marschall EB, Wei D, Mojsilović A, Varshney KR (2019) Confronting data sparsity to identify potential sources of Zika virus spillover infection among primates. Epidemics 27:59–65. https://doi.org/10.1016/j.epidem.2019.01.005

Reddy S, Fox J, Purohit MP (2019) Artificial intelligence-enabled healthcare delivery. J R Soc Med 112:22–28

Duboz R, Echaubard P, Promburom P, Kilvington M, Ross H, Allen W, Ward J, Deffuant G, de Garine-Wichatitsky M, Binot A (2018) Systems thinking in practice: participatory modelling as a foundation for integrated approaches to health. Front Vet Sci 5:303. https://doi.org/10.3389/fvets.2018.00303

Van der Waal K, Morrison RB, Neuhauser C, Vilalta C, Perez AM (2017) Translating big data into smart data for veterinary epidemiology. Front Vet Sci 4:110. https://doi.org/10.3389/fvets.2017.00110

Reichman OJ, Jones MB, Schildhauer MP (2011) Challenges and opportunities of open data in ecology. Science 331:703–705. https://doi.org/10.1126/science.1197962

Schultze JL (2015) Teaching ‘big data’ analysis to young immunologists. Nat Immunol 16:902–905

Download references

Acknowledgements

This work has benefited from interactions with many French researchers (Additional file 1 B) interested in the AI/AH interface, for which we thank them here. We also thank Stéphane Abrioux, Didier Concordet and Human Rezaie for participating to the discussions.

PE is supported by the French Research Agency (project CADENCE: ANR-16-CE32-0007). XB is involved in the project “MOnitoring Outbreak events for Disease surveillance in a data science context” supported by the EU Framework Programme for Research and Innovation H2020 (H2020-SC1-BHC-2018–2019, Grant 874850). JFG is supported by both an “Investissement d’Avenir” managed by the French Research Agency (LABEX CEBA: ANR-10-LABX-25-01) and a US NSF-NIH Ecology of infectious diseases award (NSF#1911457), and is also supported by IRD, INRAE, and Université of Montpellier. The funding bodies had no role in the study design, data analysis and interpretation, and manuscript writing.

Author information

Authors and affiliations.

INRAE, Oniris, BIOEPAR, Nantes, France

Pauline Ezanno, Sébastien Picault & Gaël Beaunée

INRAE, EpiA, Theix, France

Xavier Bailly

ASTRE, Univ Montpellier, CIRAD, INRAE, Montpellier, France

Facundo Muñoz, Raphaël Duboz & Jean-François Guégan

Sorbonne Université, IRD, UMMISCO, Bondy, France

Raphaël Duboz

Université Paris-Saclay, INRAE, Jouy-en-Josas, MaIAGE, France

Hervé Monod

MIVEGEC, IRD, CNRS, Univ Montpellier, Montpellier, France

Jean-François Guégan

Comité National Français Sur Les Changements Globaux, Paris, France

You can also search for this author in PubMed   Google Scholar

Contributions

PE carried out the literature review and analysed the interviews. PE and JFG conducted the interviews, drafted and wrote the manuscript. SP, GB, FM, RD, HM provided complementary views and references in their respective disciplines. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Pauline Ezanno .

Ethics declarations

Competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1..

Systematic literature review, interviews, previous publication in French.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Ezanno, P., Picault, S., Beaunée, G. et al. Research perspectives on animal health in the era of artificial intelligence. Vet Res 52 , 40 (2021). https://doi.org/10.1186/s13567-021-00902-4

Download citation

Received : 16 June 2020

Accepted : 20 January 2021

Published : 06 March 2021

DOI : https://doi.org/10.1186/s13567-021-00902-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Animal disease
• Artificial intelligence
• Decision support tool

Veterinary Research

ISSN: 1297-9716

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals

Animal behaviour articles from across Nature Portfolio

Animal behaviour is the scientific study of the behaviour of animals. The discipline covers study under experimental conditions, behaviourism, or natural conditions, ethology.

Birds convey complex signals in simple songs

The quality of a bird’s song during courtship can influence whether a male is selected as a mate. An innovative approach using machine learning offers a way to analyse the characteristics of birdsong.

• Kate T. Snyder
• Nicole Creanza

Whales make waves in the quest to discover why menopause evolved

Why do several species of whale experience menopause, and why does the phenomenon occur at all? Analysing whale data might help to answer these questions and shed light on why menopause evolved in humans.

• Rebecca Sear

Latest Research and Reviews

Carcass and meat characteristics of nellore young bulls fed diet using cottonseed cake as a replacer of the forage fiber source.

• Angelo Herbet Moreira Arcanjo
• Luís Carlos Vinhas Ítavo
• Rafael Goes Cardoso Paro

Nutritional ecology of a prototypical generalist predator, the red fox ( Vulpes vulpes )

• A. Balestrieri
• S. Gigliotti

Dynamics of evolutionary succession and coordination between opposite adaptations in cuckoo hosts under antagonistic coevolution

This study revealed that cuckoo hosts coordinated two evolutionarily successive but opposing behaviours by reshaping the instinctive egg retrieval reaction to different flexible reactions during their coevolution with parasitic cuckoos.

• Canchao Yang

Uniquely human intelligence arose from expanded information capacity

Theories of how human cognition differs from that of non-human animals often posit domain-specific advantages. In this Perspective, Cantlon and Piantadosi posit that differences in domain-general information capacity underlie uniquely human capacities.

• Jessica F. Cantlon
• Steven T. Piantadosi

A GPS assisted translocation experiment to study the homing behavior of red deer

• Václav Silovský
• Lukas Landler
• Miloš Ježek

Efficacy of an inflatable deterrent for reducing New World vulture human-wildlife conflict

• Bryan M. Kluever
• Betsy A. Evans
• Eric A. Tillman

News and Comment

Pregnancy’s effect on ‘biological’ age, polite birds, and the carbon cost of home-grown veg

We round up some recent stories from the Nature Briefing.

• Benjamin Thompson
• Flora Graham

Meet the real-life versions of Dune ’s epic sandworms

A Dune -loving worm palaeontologist makes the case that worms have been just as important on Earth as they are in the blockbuster film.

• Julian Nowogrodzki

Bumblebees show behaviour previously thought to be unique to humans

The bees learn how to complete a two-step puzzle too complex to solve independently, by watching other bees.

Bees and chimpanzees learn from others what they cannot learn alone

It has been argued that human culture rests on a unique ability to learn from others more than we could possibly learn alone in a lifetime. Two studies show that we share this ability with bumblebees and chimpanzees.

• Alex Thornton

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Ethical care for research animals

WHY ANIMAL RESEARCH?

The use of animals in some forms of biomedical research remains essential to the discovery of the causes, diagnoses, and treatment of disease and suffering in humans and in animals., stanford shares the public's concern for laboratory research animals..

Many people have questions about animal testing ethics and the animal testing debate. We take our responsibility for the ethical treatment of animals in medical research very seriously. At Stanford, we emphasize that the humane care of laboratory animals is essential, both ethically and scientifically.  Poor animal care is not good science. If animals are not well-treated, the science and knowledge they produce is not trustworthy and cannot be replicated, an important hallmark of the scientific method .

There are several reasons why the use of animals is critical for biomedical research: 

••  Animals are biologically very similar to humans. In fact, mice share more than 98% DNA with us!

••  Animals are susceptible to many of the same health problems as humans – cancer, diabetes, heart disease, etc.

••  With a shorter life cycle than humans, animal models can be studied throughout their whole life span and across several generations, a critical element in understanding how a disease processes and how it interacts with a whole, living biological system.

The ethics of animal experimentation

Nothing so far has been discovered that can be a substitute for the complex functions of a living, breathing, whole-organ system with pulmonary and circulatory structures like those in humans. Until such a discovery, animals must continue to play a critical role in helping researchers test potential new drugs and medical treatments for effectiveness and safety, and in identifying any undesired or dangerous side effects, such as infertility, birth defects, liver damage, toxicity, or cancer-causing potential.

U.S. federal laws require that non-human animal research occur to show the safety and efficacy of new treatments before any human research will be allowed to be conducted.  Not only do we humans benefit from this research and testing, but hundreds of drugs and treatments developed for human use are now routinely used in veterinary clinics as well, helping animals live longer, healthier lives.

It is important to stress that 95% of all animals necessary for biomedical research in the United States are rodents – rats and mice especially bred for laboratory use – and that animals are only one part of the larger process of biomedical research.

Our researchers are strong supporters of animal welfare and view their work with animals in biomedical research as a privilege.

Stanford researchers are obligated to ensure the well-being of all animals in their care..

Stanford researchers are obligated to ensure the well-being of animals in their care, in strict adherence to the highest standards, and in accordance with federal and state laws, regulatory guidelines, and humane principles. They are also obligated to continuously update their animal-care practices based on the newest information and findings in the fields of laboratory animal care and husbandry.  

Researchers requesting use of animal models at Stanford must have their research proposals reviewed by a federally mandated committee that includes two independent community members.  It is only with this committee’s approval that research can begin. We at Stanford are dedicated to refining, reducing, and replacing animals in research whenever possible, and to using alternative methods (cell and tissue cultures, computer simulations, etc.) instead of or before animal studies are ever conducted.

Organizations and Resources

There are many outreach and advocacy organizations in the field of biomedical research.

• Learn more about outreach and advocacy organizations

Stanford Discoveries

What are the benefits of using animals in research? Stanford researchers have made many important human and animal life-saving discoveries through their work. 

• Learn more about research discoveries at Stanford

Animal Law Research

Primary sources: cases, statutes, regulations and treaties, secondary sources: books, articles, news, current awareness, research and advocacy, getting help, credit and cc license.

Animal Law is concerned with the rights and welfare of nonhuman animals, as well as the requirements, responsibilities and liabilities associated with keeping or interacting with them.  Under this umbrella are wild animals as well as animals used for food and research, in entertainment, and as companions, pets or service animals.  This guide contains some research recommendations, highlighting key primary sources, secondary sources and current awareness sources. 

Know that you may not find "animal law" as a discrete topic area in research databases.  Instead, you might look to elements of property law, contract law, tort law, criminal law, environmental law, and agriculture and food law.

"farm animals"  by  lboren2687

Federal legislation

These are among the most researched and cited of animal laws at the federal level:

• Animal Welfare Act (USDA)
• Humane Methods of Slaughter Act (USDA)
• Horse Protection Act (USDA)
• Twenty-Eight Hour Law (USDA)

Congressional Research Service (CRS) and U.S. Government Accountability Office (GAO) reports provide additional context on the federal legislation.

• CRS Reports relating to Animal Agriculture Congressional Research Service reports organized by the National Agricultural Law Center
• GAO Reports on the Humane Methods of Slaughter Act GAO 10-203: Actions are Needed to Strengthen Enforcement
• GAO Report on the Animal Welfare Act GAO 10-945: Oversight of Dealers of Random Source Dogs and Cats Would Benefit from Additional Management Information and Analysis (2010)

State legislation

• Massachusetts Law About Animals A compilation of MA laws, regulations, cases and web sources on animal law from the Massachusetts Trial Court Law Libraries.
• NCSL Environmental and Natural Resources State Bill Tracking Database National Conference of State Legislatures tracks environment and natural resource bills introduced in the 50 states, territories and Washington, DC. Search here for wildlife bills, including invasive wildlife species and pollinators.
• National AgLaw Center - State Animal Cruelty Statutes A compilation from the National Agricultural Law Center of the animal cruelty statutes across the 50 states.

Applicable U.S. Government Agencies

• USDA, Animal and Plant Health Inspection Service
• FSIS (Department of Agriculture, Food Safety and Inspection Service) Part of the USDA.
• Fish and Wildlife Service (FWS)
• US Dept of Health and Human Services: National Institutes of Health, Office of Laboratory Animal Welfare

Some Relevant International Agreements

• Convention on the Conservation of Migratory Species of Wild Animals
• Convention on International Trade in Endangered Species of Wild Fauna and Flora
• Agreement on the Conservation of Small Cetaceans of the Baltic and North Seas (ASCOBANS)

Using Secondary Sources

Secondary sources are a great place to begin if you're new to animal law research, or to consult later in your research for legal interpretation and analysis. To learn more about different types of secondary sources and how best to use them, visit the following guide:

• Secondary Sources: ALRs, Encyclopedias, Law Reviews, Restatements, & Treatises by Catherine Biondo Last Updated Sep 12, 2023 3609 views this year

Selected Treatises and Other Texts

Tips on Finding Materials on Animal Law in Hollis

Try the following Library of Congress subject searches in the HOLLIS online catalog  to find additional materials. You can also substitute another country's name or region of the world (such as "Latin America")  where "United States" appears.

Animal welfare --  Law   and   legislation  --  United   States  -- Legal research. ; Animal rights --  United   States  -- Legal research. ; Animal industry --  Law   and   legislation  --  United   States  -- Legal research. ; Animal experimentation --  Law   and   legislation  --  United   States  -- Legal research. ; Laboratory  animals  --  Law   and   legislation  --  United   States  -- Legal research. ; Working  animals  --  Law   and   legislation  --  United   States  -- Legal research. ; Domestic  animals  --  Law   and   legislation  --  United   States  -- Legal research. ; Animals  in the performing arts --  Law   and   legislation  --  United   States  -- Legal research.

Legal blogs (or "blawgs") are a good way to tap into current conversation.  Here are links to two blog listings:

• Justia Blawg Search - Animal and Dog Law Blawgs
• ABA Journal Animal Law Blog Index

Research and Advocacy

• Harvard Law School - Animal Law & Policy Program Started in 2014, the Brooks McCormick Jr. Animal Law & Policy Program at HLS is "Committed to analyzing and improving the treatment of animals through the legal system"
• Animal Law Resource Center A site for current information on animal law and advocacy maintained by the National Anti-Vivisection Society, with assistance from Chicago-area law students.

Contact Us!

  Ask Us!  Submit a question or search our knowledge base.

Chat with us!  Chat   with a librarian (HLS only)

Email: [email protected]

 Contact Historical & Special Collections at [email protected]

  Meet with Us   Schedule an online consult with a Librarian

Hours  Library Hours

Classes  View  Training Calendar  or  Request an Insta-Class

 Text  Ask a Librarian, 617-702-2728

 Call  Reference & Research Services, 617-495-4516

Thank you to Stephen Wiles and Terri Saint-Amour for their work on the initial version of this guide.

This guide is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License .

You may reproduce any part of it for noncommercial purposes as long as credit is included and it is shared in the same manner. 

• Last Updated: Sep 12, 2023 10:46 AM
• URL: https://guides.library.harvard.edu/law/animallawresearch

Harvard University Digital Accessibility Policy

How and why animals can live alongside humans

New study suggests animals can live alongside humans--if they are risk-analysis experts.

New research suggests animals can thrive in human-dominated environments by being expert judges of risk. Alexis Breen from the Max Planck Institute for Evolutionary Anthropology in Leipzig, and Dominik Deffner from the Max Planck Institute for Human Development in Berlin, examined the behaviour of great-tailed grackles, a bird species successfully invading much of urban North America, showing the sex that leads this charge -- the dispersing males -- shy away from risk, which is a characteristic the researchers show is well-suited to chaotic environments like cities. These findings provide unique insight into how and why animals and humans can coexist.

"For animals, living alongside humans is 'risky business'. But some species, like grackles, are clearly coping better in human-dominated environments, even seeking them out. We wanted to find out the secret to grackles' urban-invasion success story," said Breen.

The research is based on new analyses of grackles' feeding behaviour. Deffner explained: "Cities are chaotic; they may have more cafes where animals like grackles can nab food, but they're also filled with unpredictable people and their pets. To help manage this uncertainty, we thought grackles might use a specific strategy when trying to find food."

Across three different populations, the researchers first examined how quickly grackles learned food was hidden in one particular place over another. Next, when the location of the food was swapped, the researchers examined how quickly the grackles relearned where to find it.

Better safe than sorry when leading an urban invasion

"Our key behavioural finding is that -- across all three populations -- male grackles were faster than female grackles at relearning the location of an out-of-sight treat. This robust result means male grackles are more efficient foragers in uncertain environments," said Breen.

Under uncertainty, how do male grackles 'outlearn' female grackles? "Unlike females, males exhibit pronounced risk-sensitive learning. That is, males pay close attention to whether they recently found food, and, if so, they pretty much stick to feeding from that location, instead of gambling on exploring another location," explained Deffner. The researchers said they were able to infer this strategy from grackles' feeding behaviour via cognitive modelling.

"This sex differences in grackles' learning makes biological sense," said Breen, adding: "In this species, males are the ones that disperse and move into new territories; in other words, they lead their species' urban invasion. So, as urban-invasion leaders, male grackles should proceed with caution -- new neighbourhoods will pose new challenges." The authors said they thought later-arriving females could overcome these same challenges by learning from the already established, and therefore presumably 'knowledgeable', males.

Risk-sensitive learners are winners in unpredictable environments

On their computers, the researchers also artificially simulated evolution, to examine the kinds of learning strategies that emerge victorious from unpredictable environments like urban environments. Deffner explained: "In this urban-like environment, pretend animals need to learn to find food. The learning strategy they use to find food determines how much they get to eat. And how much they get to eat determines whether they can have babies who also learn roughly the same way. Over many generations, then, the animals with the best learning strategy will come to dominate the urban-like environment. Importantly, these 'winners' will give us an idea of how animals in general can thrive in the Anthropocene."

Which learning strategy do unpredictable urban-like environments prefer? "Strikingly, in times of uncertainty, we found risk-sensitive learners were more likely to dominate over learners with other strategies. This result implies risk-sensitive learners like male grackles are better adapted to cope in chaotic settings, human-induced or otherwise," said Breen.

Breen concluded: "Our study offers compelling evidence for how and why at least grackles are thriving in unpredictable urban environments. We link sex differences in foraging grackles' learning strategies to sex differences in who leads their species' invasion, and we further link the learning strategy used by these urban-invasion leaders to likely being a generally good one for any animal navigating a life shared with humans."

To help facilitate future similar study on human-animal coexistence, the researchers created an online repository where scientists and the public alike can freely access their custom-built modelling tools. In reference to the repository, Deffner said: "We hope this open-science resource proves useful to others."

• Behavioral Science
• Extreme Survival
• Wild Animals
• Rainforests
• Environmental Awareness
• Sustainability
• Exotic Species
• Urban planning
• Columbian White-tailed Deer
• Great Ape language
• American Mink

Story Source:

Materials provided by Max Planck Institute for Evolutionary Anthropology . Note: Content may be edited for style and length.

Journal Reference :

• Alexis J Breen, Dominik Deffner. Risk-sensitive learning is a winning strategy for leading an urban invasion . eLife , 2024; 12 DOI: 10.7554/eLife.89315

Cite This Page :

Explore More

• Drug Development Made Easier
• RNA That Doesn't Age
• 'Rainbow' Detected On an Exoplanet
• Spears and Throwing Sticks 300,000 Years Old
• High Carbon Impact of Tourism at Yellowstone
• Extreme Starburst Galaxy
• Asthma: Disease May Be Stoppable
• Stellar Collisions and Zombie-Like Survivors
• Tiny Robot Swarms Inspired by Herd Mentality
• How the Brain Regulates Emotions

Trending Topics

Strange & offbeat.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Animals (Basel)

Animal Experiments in Biomedical Research: A Historical Perspective

Simple summary.

This article reviews the use of non-human animals in biomedical research from a historical viewpoint, providing an insight into the most relevant social and moral issues on this topic across time, as well as to how the current paradigm for ethically and publically acceptable use of animals in biomedicine has been achieved.

The use of non-human animals in biomedical research has given important contributions to the medical progress achieved in our day, but it has also been a cause of heated public, scientific and philosophical discussion for hundreds of years. This review, with a mainly European outlook, addresses the history of animal use in biomedical research, some of its main protagonists and antagonists, and its effect on society from Antiquity to the present day, while providing a historical context with which to understand how we have arrived at the current paradigm regarding the ethical treatment of animals in research.

1. Introduction

Animal experimentation has played a central role in biomedical research throughout history. For centuries, however, it has also been an issue of heated public and philosophical discussion. While there are numerous historical overviews of animal research in certain fields or time periods, and some on its ethical controversy, there is presently no comprehensive review article on animal research, the social controversy surrounding it, and the emergence of different moral perspectives on animals within a historical context. This perspective of animal use in the life sciences and its moral and social implications from a historical viewpoint is important to gauge the key issues at stake and to evaluate present principles and practices in animal research.

This review aims to provide a starting point for students and scholars—either in the life sciences or the humanities—with an interest in animal research, animal ethics, and the history of science and medicine. The reader interested in a more in-depth analysis on some of the topics reviewed is referred to the reference list for suggestions of further reading.

2. From Antiquity to the Renaissance

Humans have been using other vertebrate animal species (referred to henceforth as animals) as models of their anatomy and physiology since the dawn of medicine. Because of the taboos regarding the dissection of humans, physicians in ancient Greece dissected animals for anatomical studies [ 1 ]. Prominent physicians from this period who performed “vivisections” ( stricto sensu the exploratory surgery of live animals, and historically used lato sensu as a depreciative way of referring to animal experiments) include Alcmaeon of Croton (6th–5th century BCE) [ 2 , 3 ], Aristotle, Diocles, Praxagoras (4th century BCE), Erasistratus, and Herophilus (4th–3rd century BCE) [ 1 , 3 , 4 ]. The latter two were Hellenic Alexandrians who disregarded the established taboos and went on to perform dissection and vivisection on convicted criminals, benefiting from the favorable intellectual and scientific environment in Alexandria at the time [ 1 ]. All of these authors had a great influence on Galen of Pergamon (2nd–3rd century CE), the prolific Roman physician of Greek ethnicity who developed, to an unprecedented level, the techniques for dissection and vivisection of animals [ 3 , 5 ] and on which he based his many treatises of medicine. These remained canonical, authoritative, and undisputed until the Renaissance [ 1 , 6 ].

For most ancient Greeks, using live animals in experiments did not raise any relevant moral questions. The supposed likeliness of humans to their anthropomorphic deities granted them a higher ranking in the scala naturae (“the chain of being”), a strict hierarchy where all living and non-living natural things—from minerals to the gods—were ranked according to their proximity to the divine. This view of humans as superior would later influence and underline the Judeo-Christian perspective of human dominion over all nature, as represented by texts by Augustine of Hippo (IV century) and Thomas Aquinas (XIII Century), the most influential Christian theologians of the Middle Ages. For Augustine, animals were part of a natural world created to serve humans (as much as the “earth, water and sky”) and humankind did not have any obligations to them. For Thomas Aquinas, the mistreatment of another person’s animal would be sinful, not for the sake of the animal in itself, but because it is someone else’s property. Cruelty to animals was nevertheless condemned by Aquinas, as it could lead humans to develop feelings and actions of cruelty towards other humans. Also, for this theologian, one could love irrational creatures for the sake of charity, the love of God and the benefit of fellow humans (for selected texts, see reference [ 7 ]).

The belief amongst ancient Greek physicians that nature could be understood by means of exploration and experiment—and the medical knowledge thus obtained to be of clinical relevance in practice—would be replaced by other schools of medical thought. Most notably, the Empiric school (3rd century BCE–4th century) would reject the study of anatomy and physiology by dissection of cadavers or by vivisection, not only on the grounds of cruelty and the established taboos, but also for its uselessness. Empiricists believed pain and death would distort the normal appearance of internal organs and criticized the speculative nature of the conclusions drawn from experiments. Indeed, and despite taking an experimental approach to understand the human body and illness, the interpretations of physiological processes made by ancient Greeks who performed vivisections were often inaccurate. The theoretical frameworks by which physicians interpreted their experiments more often than not led them to misguided conclusions. Observations would be understood in light of such paradigms as the Hippocratic theory of the four humors or the Pythagorean theory of the four elements, along with others of natural or supernatural basis, and to which they added their own theoretical conceptions and observational errors [ 1 , 4 , 6 , 8 , 9 ]. The study of human or animal anatomy and physiology was hence deemed irrelevant for clinical practice. Beginning with the decline of the Roman Empire and continuing throughout the Middle Ages, physiological experiments—along with scientific activity in general—would fall almost entirely into disuse and medical knowledge would become dogmatic. In an increasingly Christianized Europe, there was little motivation to pursue scientific advancement of medical knowledge, as people became more concerned with eternal life than with worldly life, and returned to Pre-Hippocratic beliefs in supernatural causes for disease and in the healing power of faith and superstition. Therefore, and despite medieval physicians’ reverence for Galen and his predecessors, the experimental approach used by these classical authors had been sentenced to oblivion [ 3 , 8 , 9 , 10 , 11 ].

The use of animal experiments to satisfy scientific enquiry would only re-emerge in the Renaissance. Flemish anatomist Vesalius (1514–1564), through the course of his work as a physician and surgeon, realized that many anatomical structures thought to exist in humans—on account of them being present in other animals—were in fact absent [ 6 ]. This led him to break the established civil and religious rules and dissect illegally obtained human cadavers, and publish very accurate descriptions of the human anatomy, which challenged the authority of the classical authors. As Herophilus did centuries before (but not carried on by his successors) [ 1 ] Vesalius would also examine the similarities and differences between the internal structure of humans and other animals, thus setting the foundations of modern comparative anatomy.

Alongside the progress in anatomical knowledge made possible by experimenters defying the Catholic Church’s opposition to the dissection of human bodies, the Renaissance period also witnessed the resurgence of vivisection as a heuristic method for the understanding of animal physiology. Vesalius would again recognize the value of physiological experiments on animals as both a learning and teaching resource—he would vivisect animals for medical students as the finishing touch at the end of his courses—a view shared by his contemporary, and presumable student and rival, Realdo Colombo (1516–1559) [ 3 ]. Later, Francis Bacon (1561–1626), considered by many the founder of modern scientific methodology, would also approve of the scientific relevance of vivisection, stating that “the inhumanity of anatomia vivorum was by Celsus justly reproved; yet in regard of the great use of this observation, the inquiry needed (…) might have been well diverted upon the dissection of beasts alive, which notwithstanding the dissimilitude of their parts may sufficiently satisfy this inquiry” [ 12 ].

3. Seventeenth Century and the Dawn of the Enlightenment

Physiological experiments on animals carried on throughout the seventeenth century, in the period favorable to scientific progress now known as the Age of Enlightenment. René Descartes’s (1596–1650) description of animals as “machine-like” [ 13 ] was heavily criticized by many of his contemporaries, but nevertheless provided scientists a way to justify what would now be considered extremely gruesome experiments [ 3 , 14 , 15 , 16 ] in a time when anesthesia, for humans and animals alike, was not available. It has been argued, however, that Descartes’s views on animals were misinterpreted [ 17 , 18 ]—misconstructions that may not always have been free from malice, either by his contemporaries [ 19 ] or present-day critics [ 20 ]—as he did not explicitly state that animals were incapable of feeling pain and indeed recognized them to be able to do so insofar as it depends on a bodily organ, and even admitted animals to be capable of such sentiments as fear, anger, hope or joy [ 13 ]. Nonetheless, regardless of it being misinterpreted or not (for a discussion see [ 21 ]), Cartesian machinism would be recurrently evoked in defense of vivisection in the 17th and 18th centuries [ 14 , 15 , 16 ]. Malebranche, following his interpretation of Descartes, would explicitly justify vivisection on the grounds of it only being “apparently harmful” to animals [ 3 , 15 , 22 ]. Also, as someone deeply interested in physiology and medicine [ 23 , 24 ], and a “man of his time,” Descartes performed vivisections himself [ 15 , 16 , 21 ], an activity for which his—perhaps more apologetic than wholehearted—view of animals as soulless, senseless automata “absolved man from the suspicion of crime” [ 25 ].

As for other contemporary philosophers, Baruch Spinoza (1632–1677) did not deny animals’ ability to feel, but considered we should nevertheless “use them as we please, treating them in a way which best suits us; for their nature is not like ours” [ 26 ], whereas John Locke (1632–1704) fully recognized that animals could feel and stated that children should be brought up to abhor the killing or torturing of any living thing in order to prevent them from later becoming capable of cruel actions to fellow humans [ 27 ]. Immanuel Kant (1724–1804) would reject Cartesian mechanistic views, thus acknowledging sentience to other animals. However, Kant would not extend his concept of human intrinsic and inalienable dignity to other species. In his Of Duties to Animals and Spirits , and mirroring Thomas Aquinas’s views on the subject, he observed that “all animals exist only as means, and not for their own sakes, in that they have no self-consciousness, whereas man is the end (…) it follows that we have no immediate duties to animals; our duties towards them are indirect duties to humanity” [ 28 ]. Kant believed his anthropocentric philosophy provided the moral tradition and contemporary thought of his society; it was a philosophical underpinning, rather than an abstraction distant from the thoughts and feelings of the ordinary man [ 29 ]. Indeed, his argument that cruelty against animals would lead to cruelty to humans was—as it continues to be—popular amongst the public and scholars (e.g., [ 30 ]). In Duties to Animals, Kant would refer to William Hogarth’s (1697–1764) popular series “The Four Stages of Cruelty” ( Figure 1 ), a set of four engravings that depicted how cruel actions against animals could lead to moral degradation and crime. Regarding animal use in research, Kant would state that “Vivisectionists, who use living animals for their experiments, certainly act cruelly, although their aim is praiseworthy, and they can justify their cruelty, since animals must be regarded as man’s instruments; but any such cruelty for sport cannot be justified” [ 28 ]. While he believed actions that offended human intrinsic dignity were unacceptable—no matter how laudable their ultimate purpose should be—when it came to animals it would not be the actions themselves, but rather their justification that defined the acceptability of those actions. While the Enlightenment marked the beginning of the departure from Christian theocentrism, in the new anthropocentric view, animals continued to have no moral standing on their own. In perspective, it should be noted this was a time in which the slave market thrived and women were seen as inferior. However, the recognition of animals’ sentience in the new philosophical thought would later be instrumental for new ethical perspectives to arise on the moral status of animals.

“First Stage of Cruelty” by William Hogarth (1750), the first plate from “The Four Stages of Cruelty” series, which describes the escalating violent behavior that follows childhood cruelty to animals to an adulthood of criminal life. In this scene, two boys plunge an arrow into the rectum of a dog, while another boy, most likely the pet’s owner, pleads with them to stop. Meanwhile, some boys are burning the eye of a bird, while others tie bones to a dog’s tail. Also, some boys play “cock-throwing” (a popular sport in eighteenth-century England, consisting of throwing stones or bottles at a cockerel tied to a stake) while others hang fighting cats, and others even throw animals from windows. Source: © Victoria and Albert Museum, London.

Amidst the list of notable Western seventeenth-century physiologists using animals, the most noteworthy was undoubtedly William Harvey (1578–1657), physician to kings James I and Charles I, and one of the founders of modern science. In 1628, his groundbreaking Exercitatio Anatomica de Motu Cordis et Sanguinis in Animalibus (“An Anatomical Exercise on the Motion of the Heart and Blood in Living Beings”) was published, in which he provided the most accurate description of blood circulation and heart function of his time [ 31 , 32 , 33 , 34 ]. Using the results of meticulously planned experiments on live animals, as well as their interpretation through mathematics and physics, in this treatise, Harvey disproved many of Galen’s fifteen-hundred-year-old ideas [ 35 , 36 ]. In the tradition of his own academic lineage (he studied in Parma with the renowned anatomist Fabricius, a pupil of Colombo), Harvey was also a prolific and skilled comparative anatomist, whose studies on the anatomy of animals included species of several taxa, including mammals, fish, amphibians, reptiles and even insects [ 37 ].

Harvey’s De Motu Cordis was highly criticized, since his experimental observations did not fit the prevalent theories of Western natural philosophy of his time (for an insight on the social, scientific and academic context surrounding Harvey see [ 33 , 37 , 38 ]), still heavily grounded on Galenic principles. Harvey’s findings would challenge firmly established beliefs, such as blood being continuously produced in the liver and transported through the veins to be consumed by other organs, while arteries were thought to be filled with air; the heart was believed to have a heating—rather than pumping—function, and blood was thought to flow between the ventricles across a permeable septum; the vascular system as a whole was thought to be open; the arterial and venous bloods were believed not to mix; and the mere concept of blood circulation was virtually unknown (however, his teacher Fabricius might already have envisaged the concept of blood circulation [ 34 ]. Also, blood circulation was already known in Chinese medicine sixteen centuries before Harvey [ 39 ]). From an epistemological point of view, such opposition also reflected a dispute between the empiricist and the rationalist approach to the understanding of nature, for Harvey professed “to learn and teach anatomy not from books but from dissection, not from the tenets of Philosophers but from the fabric of Nature” (from De Motus Cordi , cited in [ 32 ]). Not surprisingly, Descartes—although a researcher himself—disagreed emphatically with most of Harvey’s findings, since he believed that theories forged through philosophical reflection on metaphysics were superior to those resulting from experimental observation, thus considering experiments or interpretations that did not confirm his own natural philosophy as flawed [ 40 , 41 ]. He nevertheless praised Harvey’s discovery of circulation and the method of experiment and observation that had led to it, a support that would actually help to turn the tide amidst scholars in favor of Harvey’s observation-over-doctrine ideas and methodological approach on experimental physiology, thereby setting the ground for further developments in physiological knowledge [ 38 ].

Further advancements in physiology would be prompted by questions left unsolved by Harvey, many of them addressed by an ensemble of his colleagues and followers at Oxford who applied Harvey’s principle that life should be interpreted in light of new findings in physics in their physiological experiments on animals [ 42 , 43 , 44 , 45 ]. The Oxford Group included polymaths like Robert Hooke (1635–1703), John Locke (1632–1704), John Mayou (1640–1679), Richard Lower (1631–1691), Thomas Willis (1621–1675), Robert Boyle (1627–1691) and Christopher Wren (1632–1723), amidst several others. Most physiologists did not expect direct therapeutic applications to result from their experiments [ 45 ]. There were, however, a few exceptions, such as Lower’s attempts at intra and inter-species blood transfusions having in mind their medical application, or Johann Wepfer’s (1620–1695) use of animals as a proxy to humans to infer the toxicity of several substances [ 3 ], a practice that is still carried out to this day. Seventeenth-century physiology would mark the dawn of modern scientific inquiry in the life sciences. Animal experiments were now proving to be more informative and relevant for obtaining scientifically sound knowledge on basic biological processes than ever before. These advancements would eventually diminish the importance of Galenic dogmatic medicine—although some of its principles would still endure for many years—and ultimately pave the way for today’s evidence-based medicine.

The seventeenth century would also witness the advent of skepticism towards experiments on animals on scientific grounds. Physicians like Jean Riolan, Jr. (1580–1657) and Edmund O’Meara (1614–1681) began to question the validity of physiological experiments carried out on animals in such an extremely altered state as one endured under vivisection, although their hidden agenda was to restore the credibility of Galenic medicine [ 3 , 46 , 47 ]. This dispute between critics and advocates of the informative value of animal models of human physiology still echoes today, e.g., [ 48 ].

The moral acceptability of inducing suffering in animals on the physiologist’s workbench would also become an issue raised in opposition of vivisection before the end of the seventeenth century [ 3 ]. However, the acceptance of the animal-machine paradigm by many physiologists reassured them that their scientific undertakings were not cruel. Furthermore, even the many who acknowledged that animals suffered a great deal with experiments, nevertheless defended themselves against the accusation of cruelty by alleging that the suffering inflicted was not unjustified, but rather for the sake of humankind, in the same line of reasoning by which today animal research is still justified. Nevertheless, these scientists were often overwhelmed by the extreme ill treatment they forced themselves to carry out on fully conscious animals [ 3 , 45 , 49 ]. One such investigator was Robert Boyle, whose infamous experiments on live animals on an air pump (conceived by him and developed by Robert Hooke) consisted in registering how animals responded to increasingly rarefied air. While only two animal experiments in Boyle’s “pneumatic chamber” are described in his New Experiments Physico-Mechanical Touching the Spring of the Air and its Effects (1660)—he would nevertheless go on to publish further animal studies on physiology [ 50 , 51 ]. Public demonstrations of this experiment would become very popular in the eighteenth century, although it bore more of an entertaining, rather than educational, nature ( Figure 2 ).

“An Experiment on a Bird in an air pump”, by Joseph Wright of Derby (detail) (1768). In this brilliant artwork, the artist captures the multiple reactions elicited by the use of live animals as experimental subjects in eighteenth-century Britain, for which we can find a parallel in present day’s diverse attitudes on this topic, including shock, sadness, appreciation, curiosity and indifference. Currently in The National Gallery , London. Source: Wikimedia Commons .

4. Eighteenth Century and the Rise of Moral Consideration for Animals

Amongst the many remarkable physiologists of the eighteenth century, polymaths Stephen Hales (1677–1761) and Albrecht von Haller (1708–1777) stood out. Hales was responsible for the first measurement of pressure in the blood vessels, and for other important insights into cardiovascular and respiratory physiology [ 52 , 53 , 54 ]. He also gave landmark contributions to public health and other medical breakthroughs, including the invention of forceps. Von Haller was arguably the most prolific physiologist of his time, better known for his groundbreaking work on inflammation, neurophysiology, heart function, and hemodynamics [ 55 , 56 , 57 , 58 , 59 , 60 ]. Both researchers were disgusted by the gruesomeness of their own experiments and were concerned about their moral justification, but nevertheless carried on, certain of the need for the use of live animals for the comprehension of many basic physiological processes, which were yet far from being understood [ 3 , 49 , 61 , 62 ]. Other relevant landmarks of eighteenth-century biomedical science based on animal studies included the foundation of experimental pharmacology [ 63 ], electrophysiology [ 64 , 65 ], and modern embryology [ 66 ]. Despite these advancements in biological knowledge, the clinical relevance of animal studies continued to be challenged [ 3 , 61 , 62 ] and, indeed, direct benefits to human health from animal experiments would remain elusive throughout the eighteenth century [ 45 , 55 ] and well into the following century.

Opposition to vivisection had raised its tone since the beginning of the eighteenth century, prompted by the popularization of public displays of experiments on live animals—in particular the notorious demonstrations of Boyle’s notorious air pump experiments [ 3 , 61 , 62 ], which were seen as purposeless, and thus inherently cruel—but became more prominent in the second half of the century, particularly in northern Europe [ 3 , 61 , 62 , 67 ]. Anthropocentric views on human duties to animals began to become increasingly challenged by philosophers, from Voltaire’s (1694–1778) criticism of Cartesian machinism and the gruesomeness of animal experiments [ 68 ] to Jean-Jacques Rousseau’s (1712–1778), Jeremy Bentham’s (1748–1832) and Arthur Schopenhauer’s (1788–1860) criticism of those who viewed animals as mere “means to an end.” By referring to sentience rather than intelligence to grant animals inherent worth, these philosophers proposed a shift from an anthropocentric justification for our duties of kindness to animals, to human obligations towards other animals for the sake of the animals themselves [ 69 , 70 , 71 ]. Rousseau proposed that despite animals being unable to understand the concept of natural law or rights, they should nonetheless, as a “consequence of the sensibility with which they are endowed (…) partake of natural right.” While Bentham found the concept of natural right “nonsense” [ 72 ], he sanctioned the idea of granting animals moral standing for the sake of their sentience. As he would famously state: “The question is not, Can they reason? Nor, Can they talk? But, Can they suffer?” [ 71 ]. From his utilitarian philosophy standpoint ( i.e. , that a moral action is that which results in the highest overall wellbeing for all stakeholders), he deemed animal research acceptable, provided the experiment had “a determinate object, beneficial to mankind, accompanied with a fair prospect of the accomplishment of it,” thus admitting that humans had precedence over other animals, limited by the due consideration for their suffering [ 73 ]. Bentham’s utilitarianism continues to exert a great deal of influence in today’s debate on animal use in the life sciences.

Among philosophers and physiologists alike, the issue of discussion was now not if animals could feel or not and to what extent, but rather whether vivisection was justifiable based on the benefit for human beings derived from it. Thus, even when researchers had strong misgivings about the inflicted suffering of animals, benefit to humans remained a valid justification for them to pursue their scientific goals through vivisection [ 61 ]. While knowledge of bodily functions and pathology was still incipient at that time, eighteen-century physiologists differed from their seventeenth-century predecessors, as they believed that medical improvements could one day be achieved through advancing knowledge by the means of animal experimentation [ 62 ]. The same rationale—that human interests took precedence over animal suffering—would also be used by nineteenth-century physicians as an ethical justification for the use of animals.

5. The Nineteenth-Century Medical Revolution and the Upsurge of the Antivivisection Societies

By the beginning of the nineteenth century, medicine was undergoing a major revolution. The organization of medical practice was changing, with the construction of hospitals, the university training of medical doctors, and the invention of new instruments and methods for the medical profession [ 74 ]. There was also a growing acknowledgement by the medical community that most medical practice, up to that period, was based on unproven traditions and beliefs and that most therapies were not only ineffective but often worsened the patient’s condition. As a result, medical practice increasingly began to focus more on understanding pathology and disease progression, pursuing more accurate diagnosis and prognosis, and thus providing reliable and useful information to patients and families, as they realized this was often the best they could do at the time. This paradigm shift would help give more credit and recognition to medical doctors and scientists, who, at that time, were often viewed with disdain and suspicion by the general public. This gain in medical knowledge would, however, sometimes be at the expense of unapproved trials, invasive procedures, and no respect for what we would now call patients’ rights [ 75 , 76 , 77 ].

Another kind of medical revolution was taking place in the laboratories, one that would ultimately provide the consistent basic science on which twentieth-century modern medicine would set its foundations. This scientific revolution began with a political one. The French Revolution of the late eighteenth century would later, in the first-half of nineteenth century, set the grounds for the establishment of the Académie Royale de Médecine , a thriving academic environment where science—and physiology, chemistry, and pharmacy, in particular—would finally be incorporated into medicine. The acknowledgment of the great knowledge gap in physiology and pathology, and the openness to positivist views on scientific knowledge, led to the definitive abandonment of the quasi-esoteric and, up to that time, dominant vitalistic theories in physiology, which stated that a vital principle, the “soul”, was the main source of living functions in organisms, rather than biochemical reactions. This led to a generalization of the understanding of all bodily processes as an expression of physical and chemical factors, and to a greater relevance given to animal experiments for answering scientific questions ( Figure 3 ). At the Académie , animal experiments were being increasingly prompted by existing clinical problems, and carried out with the ultimate goal of developing new therapeutic approaches to tackle these issues. Importantly, the integration of veterinarians in the Académie was deemed valuable for their insight on such experiments [ 57 , 78 , 79 ]. Amidst many other prominent scientists, two physician–physiologists stood out for their contributions to experimental physiology, François Magendie (1783–1855) and, most notably, Magendie’s disciple, Claude Bernard (1813–1878) [ 67 , 80 , 81 , 82 , 83 , 84 ]. Bernard’s experimental epistemology, unlike his tutor’s more exploratory approach, advocated that only properly controlled and rigorously conducted animal experiments could provide reliable information on physiology and pathology of medical relevance, setting the landmark of experimental medicine [ 85 , 86 , 87 , 88 ]. Conciliating Descartes’s rationalism with Harvey’s empiricism, Bernard acknowledged the importance of ideas and theories for the formulation of hypotheses, safeguarding, however, that these were only useful if testable and only credible if substantiated through experimentation [ 80 , 89 ]. He seemed to have been aware of how important and groundbreaking his approach to medical knowledge would become, when in his opening remarks to medical students in his very first lecture he quoted himself from his seminal “Introduction to the Study of Experimental Medicine,” stating: The scientific medicine that I’m responsible to teach does not yet exist. We can only prepare the materials for future generations by founding and developing the experimental physiology which will form the basis of experimental medicine” [ 89 ].

“A physiological demonstration with vivisection of a dog,” by Émile-Édouard Mouchy. This 1832 oil painting—the only secular painting known of the artist—illustrates how French scholars valued physiological experimentation in service of scientific progress [ 90 ]. Notice how the struggling of the animal does not seem to affect the physiologist or his observers. Currently part of the Wellcome Gallery collection, London. Source: Wellcome Library .

From the 1830s and throughout the second half of the century, the concept of scientific medicine would also flourish amidst a distinct group of German/Prussian physiologists. Following the rationale that biology could be understood through the means of chemistry and physics, and through their pivotal animal experiments and the use of microscopy, these scientists vastly contributed to the development of anatomy, histology, pathology, embryology, neurophysiology, physiology and physics. The setting for this scientific and epistemological progress was the Anatomisches Museum in Berlin, where anatomist, zoologist, and physiologist Johannes Müller (1801–1858) offered workspace and supervision to brilliant students whose independent research he wished to encourage. Although lacking the money, space, and instruments available in the great German laboratories founded after 1850, the museum provided these young scientists—notably Theodor Schwann (1810–1882), Robert Remak (1815–1865), and Friedrich Henle (1809–1885) in the 1830s, and Carl Ludwig (1816–1895), Emil du Bois-Reymond (1818–1896), Ernst Brücke (1819–1892), Hermann von Helmholtz (1821–1894), and Rudolph Virchow (1821–1902) in the 1840s—a singular intellectual atmosphere for research. Henle and Virchow would become leaders of the 1840s’ medical revolution in Germany, promoting the reform of medicine by providing it with a scientific basis, while Brücke, Helmholtz, and Bois-Raymond’s focused on the development of physiology as an autonomous science [ 83 , 91 , 92 , 93 , 94 , 95 ]. Their contributions to medical knowledge through the nineteenth century, along with Magendie’s and Bernard’s pivotal works, would deeply influence their counterparts across the Western world in the latter decades of the nineteenth century. Thousands of students flocked to attend medical schools in Germanic universities (and French institutes, although to a lesser extent), many of them from across the Atlantic [ 85 , 88 , 91 , 96 , 97 ]. This, in turn, would lead to an unprecedented rise in animal research-based advancement in biological and medical knowledge in the late nineteenth century—with important consequences for public health and quality of life—as further discussed later in this text.

While the second half of the nineteenth century marked the beginning of scientifically meaningful and medically relevant animal research, this period also saw opposition to vivisection becoming a more widespread idea in Europe, especially in Britain. Although animal experiments were not yet regulated in the first half of the century, the development of British physiology research in the Victorian Era was losing pace to Germany and France, where unprecedented progress in medical knowledge was taking place. The openly antivivisectionist positions of influential jurists, politicians, literary figures, clergymen, distinguished members of the medical community, and even Queen Victoria, contributed to an unfriendly environment for animal-based medical research [ 90 , 91 ]. There was, however, also a matter of divergence of opinion between British anatomists and French physiologists on which was the best approach for obtaining medical knowledge. Taking advantage of the rising antivivisection trend, British anatomists explored the (undoubted) gruesomeness of Magendie’s experiments, along with some nationalistic partisanship and xenophobic feelings against France, in their defense of anatomical observation as the primary method for advancing physiology, to the detriment of experiment through vivisection. However, they seldom disclosed their own positive (or at least ambivalent) views on animal experiments as a means to corroborate findings achieved through anatomical exploration [ 90 , 98 , 99 ]. Magendie would become the arch-villain of the antivivisection movement. Despite the broad recognition of his contributions to science by most peers, he was also amongst the most infamous of his time for the disdain he held for his experimental subjects. This contestation was louder outside of France, where many of his fellow scientists, even those who approved of animal experimentation, described him as an exceptionally cruel person who submitted animals to needless torture [ 85 , 90 , 100 ]. His public presentations became the most notorious, particularly one he performed in England when he dissected a dog’s facial nerves while the animal was nailed down by each paw, and was left overnight for further dissection the following day [ 82 ]. A description of Magendie’s classes to medical students by an American physician added further to the widespread disgust directed towards his work:

This surgeon’s spring course of experimental physiology commenced in the beginning of April. I seldom fail of “assisting” at his murders. At his first lecture, a basketful of live rabbits, 8 glass receivers full of frogs, two pigeons, an owl, several tortoises and a pup were the victims ready to lay down their lives for the good of science! His discourse was to explain the function of the fifth pair of nerves. The facility was very striking with which the professor could cut the nerve at its origin, by introducing a sharp instrument through the cranium, immediately behind and below the eye. M. Magendie drew the attention of the class to several rabbits in which the fifth pair of nerves had been divided several days before. They were all blind of one eye, a deposition of lymph having taken place in the comes, from inflammation of the eye always following the operation alluded to, although the eye is by this section deprived of all its sensibility. Monsieur M. has not only lost all feeling for the victims he tortures, but he really likes his business. When the animal squeaks a little, the operator grins; when loud screams are uttered, he sometimes laughs outright. The professor has a most mild, gentle and amiable expression of countenance, and is in the habit of smoothing, fondling and patting his victim whilst occupied with preliminary remarks, and the rabbit either looks him in the face or ‘licks the hand just raised to shed his blood. During another lecture, in demonstrating the functions of the motive and sensitive fibers of the spinal nerves, he laid bare the spinal cord in a young pup, and cut one bundle after another of nerves. (…) Living dissection is as effectual a mode of teaching as it is revolting, and in many cases the experiments are unnecessarily cruel and too frequently reiterated; but so long as the thing is going on, I shall not fail to profit by it, although I never wish to see such experiments repeated. cit in Olmsted, 1944 [ 101 ]

All of Magendie’s experiments were carried out without anesthesia or analgesia (and animals would be left in agony for hours, or for students’ “hands-on” anatomical studies. While, in fairness, it should be recognized that anesthetics had not yet been discovered when Magendie performed the bulk of his work, even after this technique had become available, he and nearly all of his students continued to forgo anesthesia in their experiments [ 102 ]. Moreover, animal studies on the effects of anesthetics themselves (Bernard was responsible for significant contributions to the understanding of the physiology of anesthesia: for an overview, see references [ 103 , 104 ]) were performed, as well as anatomical studies that could well have been conducted with cadavers, with no need for animals to be exposed to such prolonged suffering. Magendie was so ill famed in Britain that his experiments were referenced in the House of Commons by Richard Martin (1754–1834) when he presented a bill for the abolition of bear-baiting and that would become the “Cruel Treatment of Cattle Act” of 1822, one of the first animal protection laws. He would be again evoked in the report favoring the regulation of animal experiments that led to the “Cruelty to Animals Act” of 1876, the first piece of legislation ever to regulate animal experiments. By that time, Magendie had been dead for over twenty years [ 82 , 90 , 100 ].

After Magendie’s death, the focus of antivivisectionists’ attention moved to Bernard’s works, which included cutting open conscious animals under the paralyzing effects of curare , or slowly “cooking” animals in ovens for his studies on thermoregulation [ 105 ]. Bernard’s line of work would eventually have a heavy personal cost. Tired of her husband’s atrocious experiments, his wife would divorce him—taking with her his two daughters, who grew up to hate him—and, joining the antivivisectionists’ ranks, set up rescue shelters for dogs. Even Bernard’s cause of death is attributed to years of work in a humid, cramped, and poorly ventilated laboratory. He would, however, die a national hero, being given the first state funeral ever to be granted to a scientist in France. In his later years, he would collect the highest academic and political honors, including a seat in the French senate [ 88 , 102 , 106 , 107 ].

Despite their utter disregard for animal suffering, Magendie and Bernard did not see themselves as the immoral senseless villains portrayed by their detractors, but rather as humanists. Indeed, their view that animals did not deserve the same moral consideration as humans made them condemn experiments in humans without previous work on animals, the general principle on which the use of animal models in biomedical science is still grounded. In a time when proper dosage, administration, and monitoring of anesthesia were still largely unknown, often leading to serious side effects and accidental deaths, Magendie would state, on the use of anesthetics in humans without previous and thorough tests on animals: “That is what I do not find moral, since we do not have the right to experiment on our fellows” [ 5 , 108 , 109 ]. The amorality of human experiments prior to animal testing in animals was also an ethical argument raised in favor of vivisection by Bernard [ 89 ], who wrote:

No hesitation is possible, the science of life can be established only by experiment, and we can save living beings from death only by sacrificing others. Experiments must be made either on man or on animals. Now I think physicians already make too many dangerous experiments on man, before carefully studying them on animals. I do not admit that it is moral to try more or less dangerous or active remedies on patients, without first experimenting with them on dogs; for I shall prove, further on, that results obtained on animals may all be conclusive for man when we know how to experiment properly. If it is immoral, then, to make an experiment on man when it is dangerous to him, even though the result may be useful to others, it is essentially moral to do experiments on an animal, even though painful and dangerous to him, if they may be useful to man.

British physiologists often refrained from experimenting on mammals, mostly on account of the public’s opposition to the gruesomeness of continental physiologists’ experiments. However, with the publication of Bernard’s book (1868) and John Burdon-Sanderson’s Handbook for the Physiological Laboratory (1873), the scientific relevance of animal experiments became increasingly acknowledged, providing a utilitarian justification for vivisection, despite the harm endured by animals, eventually resulting in the rise in animal studies in medical schools in Britain in the 1870s [ 5 , 88 , 99 ]. Furthermore, by this time, anesthetics were already available and used by British physiologists, leading RSPCA secretary John Colam to state that “laboratory practices in England were very different indeed from [those] of foreign physiologists.” While the usefulness of anesthetics to chemically restrain animals was certainly advantageous for researchers, pain relief was most likely the major reason behind their ready adoption by many physiologists in Britain, as the paralyzing properties of curare were already known and used for this purpose. In fact, even before the solidifying of the antivivisectionist struggle, British physiologists had set themselves guidelines for responsible research [ 110 , 111 ]. Nevertheless, many researchers still found the analgesic and anesthetic effect of these volatile agents to be a source of undesired variability, thus avoiding their use altogether [ 99 , 105 ].

The upsurge of animal research in Britain was accompanied by an intensification of the antivivisectionist struggle. In 1875, the first animal protection society with the specific aim of abolishing animal experiments was founded: the Victoria Street Society for the Protection of Animals Liable to Vivisection (later known as the National Anti-Vivisection Society), led by Irish feminist, suffragist, and animal advocate Frances Power Cobbe (1822–1904). Vivisection became a matter of public debate, only matched in Great Britain that century by the controversy around the 1859 publication of Charles Darwin’s (1809–1882) On the Origin of Species , in which he presented a strong scientific rationale for the acknowledgement of our close kinship with the rest of the animal world, giving both physiologists and antivivisectionists a strong argument for their cause, depending on the perspective.

As the original argument of antivivisectionists that animal research was inacceptable because it did not provide useful medical knowledge began to lose strength (however, it remained a recurrent accusation against animal research, see, for instance, [ 112 ]), the discussion shifted towards preventing unnecessary harm, rather than questioning the scientific value of animal experiments [ 99 ]. On the other hand, the use of anesthetics now allowed British scientists to argue that most physiological experiments involved little, if any, pain [ 105 , 110 , 113 ]. While this made some antivivisectionists ponder about their own standing on the use of animals in research—namely those who opposed vivisection on the grounds that the intense and prolonged suffering endured by animals on the physiologist table was intolerable—many others felt that the most relevant value at stake was the preservation of each animal life in itself, questioning if human benefit was sufficient reason for sacrificing animals [ 99 , 110 ]. Moreover, the claim that animals were rendered senseless to pain gave carte blanche to many physiologists to use as many animals as they pleased for research, teaching, and demonstrations, despite anesthesia often being improperly administered, thus failing to prevent suffering for more than the brief initial moments. A famous quote by George Hoggan (a former vivisectionist who was appalled to witness Bernard’s experiments and who would later co-found the Victoria Street Society ) illustrates the relevance of the new ethical issues that emerged: “I am inclined to look upon anaesthetics as the greatest curse to vivisectible animals” [ 5 , 99 ].

In the last decades of the nineteenth century, all of today’s most relevant arguments on the debate surrounding the use of animals for scientific purposes were already in place, as well as most of the rhetoric and means of action in defense of each position. These views included outright abolitionism and, on the opposite pole, scientists demanding to be allowed to work without restrictions; non-scientists accusing researchers to be self-biased and unable to think ethically about their work and, on the other side of the barricade, researchers disdaining the authority of non-scientists to criticize their work; the benefit for humankind argument vs . the questioning of the scientific and medical value of animal research on scientific grounds; public demand for stronger regulation vs . researchers’ appeals for more autonomy, freedom, and public trust; advocates of the justifiability of only applied research (but not basic research) vs. apologists of the value of all scientific knowledge, see [ 105 , 112 , 113 ].

Just like today, there were also those who valued both animal protection and scientific progress and, recognizing that each side had both relevant and fallacious arguments, found themselves in the middle-ground, where they sought ways for compromise and progress. Amongst these, the most notable was Charles Darwin, known for his affection to animals and abhorrence for any kind of cruelty, but also for his commitment to scientific reasoning and progress [ 111 , 114 , 115 ]. Additionally, Joseph Lister (1827–1912), one of the most influential physicians of his time, would decline a request by Queen Victoria in 1875 for him to speak out against vivisection. Lister was one of the few British surgeons that carried out vivisection, albeit only occasionally, and was acquainted with some of the most eminent continental physiologists. In his response letter to the Queen, he pointed out the importance of animal experiments for the advancement of medical knowledge, stressed that anesthetics should be used at all times, and also denounced the ill treatment of animals in sports, cruel training methods, and artificial fattening of animals for human consumption as being more cruel than their use in research [ 116 ].

With the controversy assuming growing complexity and relevance, two opposing bills were presented to the British parliament in 1875: the “Henniker bill,” named after its sponsor Lord Henniker and promoted by Frances Cobbe, and the “Playfair bill,” named after scientist and Member of Parliament, Lyon Playfair, and promoted by Charles Darwin himself, along with fellow scientists and friends. Despite coming from opposite ends , both bills proposed reasonable regulation of animal experiments, rather than demanding severe restriction or granting scientists unlimited rights to use animals. Somewhat surprisingly, the Playfair bill drafted by researchers was, in some aspects, more restrictive than Henniker’s by proposing, for instance, that animal experiments should only be performed for the advancement of physiology and not for teaching purposes. The crucial difference was that the Henniker bill called for all researchers and all kinds of experiments to be properly licensed and supervised, as it is today in Great Britain, while the Playfair bill proposed that the law should only be applied to painful experiments. In the absence of parliamentary consensus for either one or the other bill, a Royal Commission—properly balanced to include members of the RSPCA and a few eminent scientists, including T.H. Huxley—was appointed that same year to address this issue, which would result in the 1876 amendment of the 1835 Cruelty to Animals Act in order to regulate the use of animals for scientific purposes, being the first case of this kind of legislation in the world [ 99 , 111 , 117 , 118 ]. This bill would endure for 110 years, until the enactment of the 1986 Animals (Scientific Procedures) Act, and remain the only known legislation to regulate animal experiments for nearly 50 years, despite some attempts to pass similar bills in other Western countries, where antivivisectionism was growing, particularly in Germany, Switzerland, Sweden, and North America [ 14 , 119 ].

The recrudescence and spread of antivivisection feelings in the late nineteenth century was coincidental with the long-awaited beginning of direct clinical and public health benefit from animal research. Before the end of the century, the germ theory of infectious diseases— i.e. , that pathogenic microbes were the causative agent of such diseases, rather than internal causes, “miasmas” in the air or water, or even more esoteric explanations—would gain broad recognition by the medical community, mostly on account of the work of Louis Pasteur (1822–1895) and his German counterpart, Robert Koch (1843–1910), which was largely based on animal experimentation. This knowledge would have an immediate, profound, and enduring effect on public health, surgery and medicine. Although it had been earlier proposed by Ignaz P. Semmelweis (1818–1865) that puerperal fever was caused by infections resulting from poor hygiene of physicians [ 120 ], only after Joseph Lister’s paper On the Antiseptic Principle of the Practice of Surgery (1867)—prompted by Pasteur’s findings—was the importance of hand-washing and instrument sterilization before surgical procedures and child delivery finally acknowledged, leading to a drastic drop in deaths from puerperal fever and post-surgical sepsis. Until then, previous efforts to make hand-washing a standard procedure had been ridiculed by the medical class.

Pasteur, a professor of chemistry with a doctoral thesis on crystallography, would turn his attention to biology in 1848 [ 121 ]. He began by unraveling the biological nature of fermentation (the inhibiting effect of oxygen on fermentation is still called the “Pasteur effect”), moving on to devise solutions of great economic value by tackling wine and beer spoiling, as well as silkworm disease, all of which he properly identified as being caused by microbes. Together with Claude Bernard, a close friend, he would later develop the process of pasteurization to destroy microorganisms in food. Pasteur began hypothesizing that microbes could also be the causative agents of many diseases affecting humans and other animals. Together with his disciples, most notably Emile Roux (1853–1933), he would go on to identify Staphylococcus , Streptococcus , the “septic vibrio” (now Clostridium septicum ), the causative agents of anthrax ( Bacillus anthracis ) and chicken cholera ( Pasteurella multocida ), being the first to develop vaccines for these zoonotic diseases, as well as for Swine Erypselas, thus setting the foundations of modern immunology [ 122 ]. However, it would be Pasteur’s successful use of a therapeutic vaccine against rabies in humans that would grant him international celebrity status [ 107 , 122 , 123 , 124 ].

Pasteur’s work required the experimental infection of numerous animals, as well as inflicting surgical wounds to test antiseptic techniques and disinfectant products, which made him a prime target of antivivisectionists. Either by genuine conviction or pragmatic convenience, amongst the ranks of Pasteur’s critics for his use of animals, one could easily find opponents of vaccination and the germ theory. Pasteur would frequently receive hate letters and threats, mostly for his infection studies on dogs, although he also used chickens, rabbits, rodents, pigs, cows, sheep, and non-human primates ( Figure 4 ). Pasteur was, however, more sensitive to animal suffering than most of his French counterparts. Not only was he uneasy with the experiments conducted—although sure of their necessity—he would also always insist animals be anesthetized whenever possible to prevent unnecessary suffering. He would even use what we now call “humane endpoints” (for a definition, see [ 125 ]): in a detailed description of his method for the prophylactic treatment of rabies (from 1884), the protocol for infecting rabbits with the rabies virus (for ulterior extraction of the spinal cord to produce a vaccine), he stated that: “The rabbit should begin to show symptoms on the sixth or seventh day, and die on the ninth or tenth. Usually the rabbit is not allowed to die, but is chloroformed on the last day in order to avoid terminal infections and unnecessary suffering” [ 126 ]. Furthermore, he would become directly responsible for saving countless animals from the burden of disease and subsequent culling [ 5 , 107 , 113 , 127 , 128 ].

This full-page illustration of Pasteur in his animal facility was published in Harper’s Weekly in the United States, on 21 June 1884. At this time, there was moderate curiosity on Pasteur’s work in the US, which would intensify after his first successful human trials of a therapeutic vaccine for rabies in 1885. In the article, the reader is reassured that the use of dogs is both humane and justified in the interest of mankind. The use of other species, however, is barely mentioned [ 5 ]. Source: Images from the History of Medicine , U.S. National Library of Science.

Robert Koch, a practicing rural physician, would follow the tradition of the great German/Prussian physiologists of his time (and indeed was a student to many of them), providing invaluable contributions to medical knowledge through animal research, mainly in the field of bacteriology and pathology. His famous “Koch postulates” would play an important role in microbiology Along with his associates, Koch developed from scratch methods that are still used today, such as microphotography of organisms, solid medium culture, and staining or microbe quantification. They would go on to identify the causative agents of tuberculosis ( Micobacterium tuberculosis , also known as the “Koch bacillus”), cholera ( Vibrio cholera , albeit 30 years after Filippo Pacini, 1812–1883 [ 129 ]), and anthrax. The overlapping interest of Pasteur and Koch on anthrax would trigger a bitter rivalry between the two, fuelled by their different approaches to microbiology, as well as chauvinistic Germany–France rivalry [ 130 , 131 ]. Koch’s own school of microbiology housed many of the leading late-nineteenth, early-twentieth-century medical researchers. This included Emil von Behring (1854–1917) and Paul Ehrlich (1854–1915), both responsible for the first antitoxin for treatment of diphtheria—developed from horse serum—for which von Behring received the Nobel Prize in 1901. Von Behring would also develop an antitoxin for immunization against tetanus, along with Shibasaburo Kitasato (1853–1931), who had also studied under Koch. In 1908, Ehrlich would also be awarded the Nobel Prize for contributions to immunology, and would yet again be nominated for his contributions to chemotherapy and the development of Salvasaran (an effective treatment against syphilis), in particular [ 132 , 133 , 134 , 135 ]. The development and production of vaccines and antitoxins led to a dramatic increase in the number of animals used in research. The number of animals used by physiologists in the nineteenth century would be negligible in comparison with the several hundred used by Pasteur to develop, test, and produce vaccines, the thousands of mice used by Paul Ehrlich for the production of Salvasaran—his syphilis drug—and the millions of primates that would be used to produce Polio vaccines in the 1950s [ 5 ].

6. The Twentieth-Century Triumph of Science-Based Medicine

By the end of the nineteenth and beginning of the twentieth century, the pharmacopeia had effective, scientifically tested drugs, a landmark that allowed for an increasing number of people to understand the importance and validity of scientifically sound medical knowledge and, with it, the relevance of animal-based research (see [ 113 , 136 , 137 , 138 ]). One could still find as far as the end of the nineteenth century, however, physicians who disregarded the ideals of scientific medicine and vigorously stood by their traditional epistemological view of medicine and clinical practice, which they saw as more of a form of art than as a science. Many such physicians also opposed experiments on live animals and were members of antivivisection societies [ 77 , 139 , 140 , 141 ]. Nonetheless, the medical profession, medicine itself, and human health had now been irreversibly changed by science, and would continue to be pushed forward throughout the twentieth century to now.

The twentieth century would witness astonishing advances in medical knowledge and the treatment of disease. The discovery of vitamins, hormones, antibiotics, safe blood transfusion, new and safer vaccines, insulin, hemodialysis, chemo and radiotherapy for cancer, the eradication of small pox (and the near eradication of poliomyelitis), advanced means of diagnostic and new surgical techniques are but a very few examples of twentieth century’s medical achievements that have not only saved millions of lives—human and non-human—but also allowed countless humans and animals to live a “life worth living,” by the relief of disease-induced suffering. The advances of biomedical research to human health since the dawn of the past century are countless, with animal research playing a role in a number of important discoveries (for an overview, see [ 142 ]). Of the 103 Nobel Prizes in physiology or medicine given since 1901, on 83 occasions work conducted on vertebrate species (other than human) was awarded, while in another four instances, research relied heavily on results obtained from animal experiments in vertebrates conducted by other groups [ 143 ]. Another indirect measure of the impact that biomedical progress had on the twentieth century was the increase in life expectancy, which in some developed countries doubled between 1900 and 2000, and is still on the rise today [ 144 , 145 , 146 ].

By the 1910–1920s, antivivisection groups were fighting an increasingly difficult war for the public’s support. The argument that no medical progress could be obtained through animal research was becoming increasingly difficult to uphold and, as researchers pledged to avoid animal suffering whenever possible, criticism of animal experiments on the grounds of cruelty toned down. However, not all scientists had sufficient, if any, consideration for animal suffering, and research would continue to be unregulated in most countries. Nevertheless, the exaggerated claims, radical abolitionist views, and scientific denialism by more inflexible antivivisectionists would make them lose support from the general public and more moderate animal protection groups, leading to a decline—albeit not an end—to the antivivisection movement, until its resurgence in the 1970s. Confronted with a general lack of support, moreover in a period that would witness two great world wars and a serious economic recession—which would push the interests of animals further to the background—the line of action of antivivisectionists through most of the twentieth century focused on banning the use of dogs and other companion animals [ 5 , 147 , 148 , 149 , 150 ].

The toning down of the opposition to animal use in the life sciences had also something to do with the emergence of rodent species as a recurrent animal model in research. Unlike dogs or horses, rodents like mice and rats were seen as despicable creatures by most of the public, and therefore less worthy of moral consideration, which in turn deemed their use in research more acceptable [ 147 ]. While this came as an advantage to researchers, it is hard to say, however, if the actual weight of the public’s misgivings about the use of domestic animals was a relevant contributing factor for the ready adoption of rodent models, especially when considering their other numerous advantages as experimental animals when compared to other species. Firstly, they are small, easy to handle, and relatively cheap to house. Secondly, they are highly resistant to successive inbreeding and have a short lifespan and rapid reproduction rate [ 151 , 152 ].

Domesticated rats ( Rattus norvegicus ) were the first rodent species to be used for scientific purposes. Their use in physiological research dates to as early as 1828, but only in the first decades of the twentieth century did they become a preferred tool in research, after the development in 1909 of the first standard rat strain, the Wistar Rat , from which half of all rats used in laboratories today are estimated to have descended (for a historical perspective, see [ 153 , 154 ]) ( Figure 5 ). The mouse ( Mus musculus ) had also been used in the nineteenth century, famously by Gregor Mendel in his 1850s studies on heredity of coat color, until the local bishop censored mouse rearing as inappropriate for a priest, which made him turn to using peas instead [ 155 ]. The mouse would be again picked up in the beginning of the nineteenth century by Lucien Cuénot (1866–1951) to demonstrate that mammals also possessed “genes” (a vague concept at the time) that followed the laws of Mendelian inheritance, and would since then become a privileged model in the study of genetics, a field that would grow exponentially after the discovery of the DNA structure in 1953 by James Watson (born 1928) and Francis Crick (1916–2004). In 1980 John Gordon and Franck Ruddle developed the first transgenic mouse [ 156 ], and in 1988, the first gene knockout model was produced, which granted Mario R. Capecchi (born 1937), Martin J. Evans (born 1941), and Oliver Smithies (born 1925) the 2007 Nobel Prize. In 2002, the mouse became the second mammal, after humans, to have its whole genome sequenced. These, along with other technologies, have opened unlimited possibilities for the understanding of gene function and their influence in several genetic and non-genetic diseases, and have made the mouse the most commonly used animal model in our day (for a historical overview of the use of the mouse model in research, see [ 157 , 158 ]), with prospects being that it will continue to have a central role in biomedicine in the foreseeable future.

Two outbred laboratory rats, of the Lister Hooded (Long–Evans) strain. Rodents are the most commonly used laboratory animals, making up nearly 80% of the total of animals used in the European Union, followed by cold-blooded animals (fish, amphibian and reptiles, making up a total of 9.6%) and birds (6.3%) [ 159 ] Photo: Francis Brosseron, reproduced with permission.

7. Animal Liberation and the Pathway for a More Humane Science

Opposition to animal experiments resurged in the second half of the twentieth century, in particular after the 1975 publication of Animal Liberation by Australian philosopher Peter Singer (born 1946) [ 160 ]. Singer offered a strong philosophical grounding for the animal rights movement, by arguing that the use of animals in research—as well as for food, clothing or any other purpose—is mostly based on the principle of speciesism (coined by Richard Ryder in 1970 [ 161 ]), under which animals are attributed a lower moral value on the sole basis of belonging to a different species [ 162 ], which he considers to be no less justifiable than racism or sexism. His argument, however, does not stem from the premise that animals have intrinsic rights. As a preference utilitarian—and unlike hedonistic utilitarians like Bentham and Mill who argued we should act in order to maximize net happiness—Singer proposed that our actions should aim to do what on balance “furthers the interests of those affected” [ 163 ]. Holding that the interest of all sentient beings to both avoid pain and have positive experiences deserves equal consideration, he thus argues that it is difficult to justify animal research, since it generally does not hold to Bentham’s “ Each to count for one and none for more than one ” postulate. Furthermore, it is usually unfeasible to prospectively quantify how many may benefit directly from a given animal experiment. According to Singer, by using the principle of equal consideration of interests, one should give priority to relieving the greater suffering. Singer does not propose we should assume different species suffer similarly under the same conditions but, on the contrary, that care should be taken when comparing the interests of different species as, for instance, a human cancer patient, for his higher cognitive skills, can suffer a great deal more than a mouse with the same disease [ 164 ]. For this reason, he does not consider animal research to always be morally wrong in principle, and even admits that in certain occasions it may be justifiable, albeit these situations are, in his view, exceptional [ 165 ].

The animal rights movement would, however, receive from American philosopher Tom Regan (born 1938) a more uncompromising view of our duties to animals than Singer’s utilitarianism, one that would question the use of animals in research—or in any other way—altogether, regardless of the purpose of research. In Regan’s book The Case for Animal Rights (1983), he proposed we should extend the Kantian concept of intrinsic value to all sentient beings. This perspective inherently affords vertebrates rights, despite their incapacity to understand or demand such rights, as it is also the case—Regan argues—of small infants and the severely mentally handicapped. Hence, the respect for the life and wellbeing of sentient animals should be taken as absolute moral values, which can only be violated in very specific and extreme cases—such as self-defense. Regan’s moral philosophy hence only allows for an abolitionist view on animal research—since no “ends” can justify the “evil means” of sacrificing an animal in the face of the inviolable dignity of sentient beings [ 166 ]—and has become the main theoretical reference for the animal rights movement.

From the impact of Singer’s and Regan’s works in society and the academic world, “animal ethics” would emerge as a whole new field of philosophical and bioethical studies, and, with it, new and diverse ethical views on animals—including on animal research—and of our duties towards them. However, despite the diversity of philosophical views on the use of animals, the public debate on animal research would become polarized between animal rights activists and animal research advocates. While the first would uphold an uncompromising abolitionist stand, one could also find on the opposite side several persons who did not regard animal research as a moral issue at all [ 167 ]. Furthermore, and despite the debate in the philosophical ground remained civilized—even between diametrically opposed perspectives, see, for example, [ 168 ]—in the “real world” the antagonism began to build up. In the 1970s, animal rights extremist groups began resorting to terrorist actions, thus becoming a serious problem for researchers and authorities in several Western countries still today. These actions more often consist of trespassing, raiding animal facilities and laboratories, damage to property, harassment and death threats to researchers, their families and neighbors. It has also sometimes escalated into kidnapping, car and mail bombings, arson of homes and research facilities, mailing of AIDS-contaminated razorblades, and violence against scientists and their family members [ 169 , 170 ]. These actions, which have been classified as unjustifiable and damaging to the animal rights cause by Tom Regan himself [ 171 ], made researchers close themselves within their community and avoid speaking publicly about their work [ 172 , 173 , 174 ], which in turn left pro-research advocacy to emotion-appealing campaigns, of the likes of the Foundation for Biomedical Research’s “Will I be alright, Doctor?” film [ 175 ], or the advertisement depicted in Figure 6 .

A large advertisement published in the 13 May 1991 edition of The Hour (p. 9), and part of a campaign in defense of animal research, sponsored by the United States Surgical Corporation. While the value of Pasteur’s work is undeniable, there is, however, no scientific grounding for the claim that only by experimenting on dogs would a vaccine for rabies have been developed, or that other animal models or even non-animal methods could not have been used to achieve this in over a century. These dramatic and biased portraits of animal research are now more uncommon, as an increasing number of scientists acknowledge the need to be more candid and open to objective discussion over the possibilities and limitations of animal research, and of the scientific process altogether.

In spite of the emergence of the animal rights movement, animal research for biomedical purposes was—as it continues to be—seen as morally acceptable by the majority of the public [ 176 , 177 ]. It became, however, increasingly evident that animal suffering was morally and socially relevant, and that an ethical balance between the benefits brought about by biomedical progress and the due consideration to animal wellbeing should be sought. However, whilst antivivisection movements would only re-emerge in the late 1970s [ 5 , 178 ], the need for a more humane science had already been acknowledged and addressed within the scientific community as early as the 1950s.

Following the first edition of the Universities Federation for Animal Welfare’s Handbook on the Care and Management of Laboratory Animals (1954), the organization’s founder Charles Hume commissioned that same year a general study on humane techniques in animal experimentation to zoologist and classicist (and overall polymath) William Russell (1925–2006) and microbiologist Rex Burch (1926–1996), under a project chaired by immunologist Peter Medawar (1915–1987), Nobel Prize laureate in 1960 [ 179 , 180 , 181 ]. From this work, Russell and Burch would develop the tenet of the “Three Rs”— Replacement, Reduction, Refinement —principles that would be extensively developed in their seminal book, The Principles of Humane Experimental Technique [ 182 ]. In this book, the authors argued “humane science” to be “best science,” going so far as to state that “If we are to use a criterion for choosing experiments to perform, the criterion of humanity is the best we could possibly invent.” Replacement was defined as “any scientific method employing non-sentient material [to] replace methods which use conscious living vertebrates”; Reduction as the lowering of “the number of animals used to obtain information of a given amount and precision”; and Refinement as the set of measures undertaken to “decrease in the incidence or severity of […] procedures applied to those animals which have to be used,” later including also the full optimization of the wellbeing of laboratory animals, also seen as a basic requirement for the quality of science [ 179 ]. They also challenged the commonly held belief that vertebrate animals—and mammals in particular—are always the most suitable models in biomedical research, a reasoning they called the high-fidelity fallacy . Despite receiving a warm welcome, Russell and Burch’s work would remain largely ignored well into the 1970s. In 1978, physiologist David Henry Smyth (1908–1979) would again bring the Three Rs to the light of day and encompass them under the concept of alternatives [ 67 , 183 ], which he defined as “all procedures which can completely replace the need for animal experiments, reduce the numbers of animals required, or diminish the amount of pain or distress suffered by animals in meeting the essential needs of man and other animals” [ 184 ]. More than a restatement of the Three Rs, this definition had the added value of placing onto researchers the burden of providing convincing evidence for the necessity of using animals [ 183 ], a particularly important statement from the then-president of the UK’s Research Defence Society.

The Three Rs approach would provide an ethically and scientifically sound framework on which a reformist approach to the use of animals in biomedicine could be grounded. It would also set the stage for a more moderate advocacy of animal rights to appear: while remaining incompatible with an abolitionist animal rights perspective, this paradigm grants animals something like a right to protection from suffering, or at least certain suffering beyond a defined threshold [ 185 ], preserving the central idea that there are absolute and non-negotiable limits to what can be done to animals. This welfarist perspective stems from a utilitarian view that animals can be used as means to an end as long as their interests—as far as they can be ascertained—are taken into account, but also accepting that the lives and wellbeing of human beings must be granted greater consideration than animals’. Utilitarian philosopher Raymond G. Frey (1925–2012) offered a philosophical view compatible with the current paradigm, by acknowledging that what we do to animals matters morally, since animals’ sentience and ability to control their lives grants them moral standing and a rightful place in the “moral community.” However, when weighing the interests of humans against animals’ interests (or between animals, or humans), he held that the main question should not lie on one who has moral standing or not, or to which degree, but rather on whose life may be more valuable. In Frey’s view, the value of life “is a function of its quality, its quality of its richness, and its richness of its capacities and scope for enrichment.” Hence, as a result of their higher cognitive capabilities, human lives are typically richer than animal lives, being therefore generally more valuable [ 186 ].

A “welfarist–reformist” approach has been accepted as a compromise by some prominent animal rights advocates who, while maintaining the long-term goal of a full end to all animal experiments, believe that it is by successive short-term improvements of the status quo that their goal can be achieved; see [ 178 , 187 , 188 ]. This position—also endorsed by influential animal advocacy groups like the Humane Society of the United States, or the UK’s FRAME—has, however, been highly criticized by less compromising animal rights advocates, like Regan and Gary Francione (born 1954), who believe reformist attitudes validate and perpetuate the exploitation of animals [ 171 , 189 , 190 ].

The 1980s and the 1990s would witness considerable progress in the development and acknowledgment of the Three Rs, to the satisfaction of William Russell and Rex Burch, who lived to see the “rediscovery” of their principles and the emergence of a whole new field of research inspired by their groundbreaking work [ 179 , 191 ]. As Peter Medawar had predicted in the 1960s, the number of animals used in research would peak in the 1970s and start to decline thereafter, although the number of biomedical papers has since then more than doubled [ 181 , 192 , 193 , 194 , 195 , 196 ]. This data is, however, limited to the Western world, as statistics on animal use in emerging countries such as India and China are unavailable [ 197 ], and there is no way to assess if (and, if so, to what extent) the decline in numbers of animals used in Western countries may be attributed to the outsourcing of animal experiments to these emerging countries. In recent years, the rise in the use of genetically modified animals has led to the stabilization of what would otherwise be a continuously downward trend [ 198 , 199 ] ( Figure 7 ).

This schematic illustration (adapted with permission from an original by Professor Bert van Zupthen) attempts to describe trends in the use of animals for scientific purposes in the Western world across time. It depicts the emergence of the first vivisection studies by classical Greek physicians, the absence of animal-based research—along with most medical and scientific research—across the Middle Ages, its resurgence in the Renaissance onwards, and the rapid increase in animal studies following the rise of science-based physiology and medicine in the nineteenth century. The curves represented are nevertheless conjectural, as there are no reliable statistics on animal use for most of the period covered. Even nowadays it is hard to estimate trends in animal research, as data from several developed countries is insufficient (for instance, in the United States, rodents, fish and birds are not accounted for in the statistics). The available data, however, suggest that the number of animals used in research and testing in the Western world peaked in the 1970s, and decreased until the late 1990s, or early 2000s, to about half the number of 30 years earlier, and stabilizing in recent years. While many, if not most, researchers do not foresee an end to animal experiments in biomedicine, the European Commission has nevertheless set full replacement of animal experiments as an ultimate goal [ 204 ], and the Humane Society of the United States has the optimistic goal of full replacement by the year 2050 [ 192 ].

In 1999, the Declaration of Bologna, signed in the 3rd World Congress on Alternatives and Animal Use in the Life Sciences, would reaffirm that “ humane science is a prerequisite for good science, and is best achieved in relation to laboratory animal procedures by the vigorous promotion and application of the Three Rs ” [ 200 ]. The Three Rs would also become the overarching principle of several legislative documents regulating animal use in science since the 1980s (including the latest European legislation [ 201 ]). Most recently, biomedical researchers in both industry and academia have also acknowledged the central importance of the Three Rs and the need for more transparency regarding animal use in biomedical research through the Basel Declaration [ 202 , 203 ]. More important, there are currently thousands of scientists devoted to the progress of animal welfare and development of alternatives to animal use in the life sciences.

8. Conclusion

The historical controversy surrounding animal research is far from being settled. While the key arguments in this debate have not differed significantly since the rise of antivivisectionism in nineteenth-century England—and even before—we have since then moved a long way forward in regards to the protection of animals used in research and transparency regarding such use. While animal experiments have played a vital role in scientific and biomedical progress and are likely to continue to do so in the foreseeable future, it is nonetheless important to keep focusing on the continuous improvement of the wellbeing of laboratory animals, as well as further development of replacement alternatives for animal experiments.

Acknowledgments

The author thanks Francis Brosseron ( Lycée Français du Porto ) for the photograph in Figure 5 , Bert van Zutphen (Emeritus Professor, Utrecht University) for the original picture that has been adapted for Figure 7 , and I. Anna S. Olsson, Manuel Sant’Ana (IBMC, University of Porto) and four anonymous referees for their valuable comments on this manuscript.

Conflict of Interest

The author declares no conflict of interest.

References and Notes

New Research

Microplastics Are Contaminating Ancient Archaeological Sites

New research suggests plastic particles may pose a threat to the preservation of historic remains

Aaron Boorstein

Staff Contributor

Today, microplastics are found almost everywhere: oceans , food , the atmosphere and even human lungs , blood and placenta s. But while they’re thought of as a modern problem, plastic particles are now appearing where one might least expect: ancient archaeological sites.

Researchers found microplastics in soil deposits 7.35 meters (24.11 feet) below the ground, according to a study published this month in the journal Science of the Total Environment . The soil samples date to the first or early second century C.E. and were sourced from two archaeological sites in York, England. Some were excavated in the late 1980s, while others were contemporary samples.

The scientists then used an imaging technique called μFTIR , which can detect microplastics’ quantities, size and composition. Across all samples, they found 66 particles consisting of 16 polymer types.

“This feels like an important moment, confirming what we should have expected: that what were previously thought to be pristine archaeological deposits, ripe for investigation, are in fact contaminated with plastics,” says John Schofield , an archaeologist at the University of York, in a statement .

Microplastics are fragments of plastic that are smaller than five millimeters long, the diameter of a standard pencil eraser . They come from a variety of sources, including laundry, landfills, beauty products and sewage sludge.

“In the last not even 100 years—mostly since the 1950s—we as humans have produced eight billion tons of plastic, and the estimate is only about 10 percent of that has been recycled,” Leigh Shemitz, president of the climate education group SoundWaters, told Yale Sustainability in 2020.

Microplastics have been found in soil samples before. In fact, almost one-third of all plastic waste ends up in soil or freshwater, according to the United Nations Convention to Combat Desertification .

But the new study provides “the first evidence of [microplastic] contamination in archaeological sediment (or soil) samples,” write the researchers. These findings could change how archaeologists protect historic sites.

“While preserving archaeological remains in situ has been the favored approach in recent years, the new findings could trigger a change in approach, as microplastic contamination could compromise the remains’ scientific value,” writes CNN ’s Jack Guy.

In situ , Latin for “in the place,” is the term used to describe archaeological objects that have not been moved from their original locations. Leaving remains in situ helps prevent site and artifact damage, preserves contextual setting and allows future researchers to gather information.

“The presence of microplastics can and will change the chemistry of the soil, potentially introducing elements which will cause the organic remains to decay,” says David Jennings , chief executive of York Archaeology, in the statement. “If that is the case, preserving archaeology in situ may no longer be appropriate.”

Now, the researchers will shift their attention toward better understanding the implications of their findings. They know microplastics could threaten the integrity of archaeological samples, but what exactly does that harm look like?

“To what extent this contamination compromises the evidential value of these deposits and their national importance is what we'll try to find out next,” says Schofield.

Get the latest stories in your inbox every weekday.

Aaron Boorstein | READ MORE

Aaron Boorstein is an intern with  Smithsonian magazine.