Mammals show a very low level of variation in vertebral count, both among and within species, in comparison to other vertebrates [1]. Jordan’s rule for fishes states that the vertebral number among species increases with latitude, due to ambient temperatures during development [2]. Temperature has also been shown to influence vertebral count within species in fish [3], amphibians [4], and birds [5]. However, in mammals the count appears to be constrained, on the one hand, by a possible relationship between the development of the skeleton and the proliferations of cell lines with associated costs (neural malformations, cancer etc., [6]), and on the other by the cervical origin of the diaphragm [7].
Knight et al. [8] investigate the effect of intrauterine temperature variation on skeletal morphology during development, and focus on a particular mammal, Dasypus novemcinctus, or nine-banded armadillo. Armadillos (Xenarthra) and are characterized by relatively low body temperatures and low basal rates of metabolism. Dasypus novemcinctus is the only xenarthran mammal to have naturally expanded its range into the middle latitudes of the U.S., and one of the few mammals that invaded North America from South America. It is one of few placentals that withstand considerable decrease of body temperature without torpor. It presents a resting body temperature that is low and variable for a placental mammal of its size [9] and is the only vertebrate that gives birth to monozygotic quadruplets. Among 42 monotreme, marsupial and placental genera, Dasypus novemcinctus shows the highest variation of thoracolumbar vertebral count [10].
The particularities of Dasypus novemcinctus regarding vertebral count variation and ability to withstand variable temperature qualify it as a target organism for study of the relationship between skeleton morphology and temperature in mammals.
Knight et al. [8] explored variability in vertebral count within Dasypus novemcinctus to understand whether temperature during development determines skeleton morphology. To this end they experimented with 22 armadillos (19 with data) and litters from 12 pregnant females, in two environments, for three years — an impressive effort and experimental setup. Moreover, they used a wide variety of advanced experimental and analytical techniques. For example, they implanted intra-abdominal, long-term temperature recorders, which recorded data every 6 to 120 minutes for up to several months. They analysed body temperature periodicity by approximation of the recordings with Fourier series, and they CT-scanned fetuses.
All 19 individuals (from which data could be gathered) exhibited substantial daily variation in body temperature. Several intriguing results emerged such as the counter-intuitive finding that the mammals’ body temperature fluctuates more indoors than outdoors. Furthermore, three females (out of 12) were found to have offspring with atypical skeletons, and two of these mothers presented an extremely low internal temperature early in pregnancy. Additionally, genetically identical quadruplets differed skeletally among themselves within two litters.
Results are not yet definitive about the relationship of temperature during development and vertebral count in Dasypus novemcinctus. However, Knight et al. [8] demonstrated that nine-banded armadillos survive with high daily internal temperature fluctuations and successfully bring to term offspring which vary in skeletal morphology among and within genetically identical litters despite major temperature extremes.

References

[1] Hautier L, Weisbecker V, Sánchez-Villagra MR, Goswami A, Asher RJ (2010) Skeletal development in sloths and the evolution of mammalian vertebral patterning. Proceedings of the National Academy of Sciences, 107, 18903–18908. doi: 10.1073/pnas.1010335107
[2] Jordan, D.S. (1892) Relations of temperature to vertebrae among fishes. Proceedings of the United States National Museum, 1891, 107-120. doi: 10.5479/si.00963801.14-845.107
[3] Tibblin P, Berggren H, Nordahl O, Larsson P, Forsman A (2016) Causes and consequences of intra-specific variation in vertebral number. Scientific Reports, 6, 1–12. doi: 10.1038/srep26372
[4] Peabody RB, Brodie ED (1975) Effect of temperature, salinity and photoperiod on the number of trunk vertebrae in Ambystoma maculatum. Copeia, 1975, 741–746. doi: 10.2307/1443326
[5] Lindsey CC, Moodie GEE (1967) The effect of incubation temperature on vertebral count in the chicken. Canadian Journal of Zoology, 45, 891–892. doi: 10.1139/z67-099
[6] Galis F, Dooren TJMV, Feuth JD, Metz JAJ, Witkam A, Ruinard S, Steigenga MJ, Wunaendts LCD (2006) Extreme selection in humans against homeotic transformations of cervical vertebrae. Evolution, 60, 2643–2654. doi: 10.1111/j.0014-3820.2006.tb01896.x
[7] Buchholtz EA, Stepien CC (2009) Anatomical transformation in mammals: developmental origin of aberrant cervical anatomy in tree sloths. Evolution and Development, 11, 69–79. doi: 10.1111/j.1525-142X.2008.00303.x
[8] Knight F, Connor C, Venkataramanan R, Asher RJ. (2020). Body temperatures, life history, and skeletal morphology in the nine-banded armadillo (Dasypus novemcinctus). PCI-Ecology. doi: 10.17863/CAM.50971
[9] McNab BK (1980) Energetics and the limits to a temperate distribution in armadillos. Journal of Mammalogy, 61, 606–627. doi: 10.2307/1380307
[10] Asher RJ, Lin KH, Kardjilov N, Hautier L (2011) Variability and constraint in the mammalian vertebral column. Journal of Evolutionary Biology, 24, 1080–1090. doi: 10.1111/j.1420-9101.2011.02240.x

Morand et al. (2024) designed convolutional neural networks to predict the occurrences of 38 marine animals worldwide. The environmental predictors were sea surface temperature, chlorophyll concentration, salinity and fifteen others. The time of some of the predictors was chosen to be as close as possible to the time of the observed occurrence.

This approach has previously only been applied to the analysis of the distribution of terrestrial plant species (Botella et al. 2018, Deneu et al. 2021), so the application here to very different marine ecosystems and organisms is a novelty worth highlighting and discussing.

A very interesting feature of PCI Ecology is that reviews are provided with the final manuscript and the present recommendation text.

In the case of the Morand et al. article, the reviewers provided very detailed and insightful comments that deserve to be published and read alongside the article.

The reviewers' comments question the ecological significance and implications of choosing fine temporal and spatial scales in CNN distribution modelling in order to obtain species distribution modelling (SDM).

The main question debated during the review process was whether the CNN modeling approach used here can be defined as a kind of niche modeling.

The fact is that most of the organisms studied here are mobile, and the authors have taken into account precise environmental information at dates close to those of species appearance (for example, "Temperature and chlorophyll values were also included 15 and 5 days before the occurrences"). In doing so, they took into account the fine spatial and temporal scales of species occurrences and environmental conditions, which can be influenced by both environmental preferences and the movement behaviors of individuals. The question then arises: does this approach really represent the ecological niches of the marine organisms selected? Given that most selected organisms may have specific seasonal movement dynamics, the CNN model also learns the individual movement behaviors of organisms over seasons and years. The ecological niche is a broader concept that takes into account all the environmental conditions that enable species to persist over the course of their lives and over generations. This differs from the case of sessile land plants, which must respond to the environmental context only at the points of appearance.

This is not a shortcoming of the methodology proposed here but rather an interesting conceptual issue to be considered and discussed. Modelling the occurrence of individuals at a given time and position can characterize not only the species' niche but also the dynamics of organisms' temporal movements. As a result, the model predicts the position of individuals at a given time, while the niche should also represent the role of environmental conditions faced by individuals at other times in their lives.
A relevant perspective would then be to analyze whether and how the neural network can help disentangle the ranges of environmental conditions defining the niche from those influencing the movement dynamics of individuals.

Another interesting point is that the CNN model is used here as a multi-species classifier, meaning that it provides the ranked probability that a given observation corresponds to one of the 38 species considered in the study, depending on the environmental conditions at the location and time of the observation. In other words, the model provides the relative chance of choosing each of the 38 species at a given time and place. Imagine that you are only studying two species that have exactly the same niche, a standard SDM approach should provide a high probability of occurrence close to 1 in localities where environmental conditions are very and equally suited to both species, while the CNN classifier would provide a value close to 0.5 for both species, meaning that we have an equal chance of choosing one or the other. Consequently, the fact that the probability given by the classifier is higher for a species at a given point than at another point does not (necessarily) mean that the first point presents better environmental conditions for that species but rather that we are more likely to choose it over one of the other species at this point than at another. In fact, the classification task also reflects whether the other 37 species are more or less likely to be found at each point. The classifier, therefore, does not provide the relative probability of occurrence of a species in space but rather a relative chance of finding it instead of one of the other 37 species at each point of space and time.

It is important that an ecologist designing a multi-species classifier for species distribution modelling is well aware of this point and does not interpret the variation of probabilities for a species in space as an indication of more or less suitable habitat for that specific species. On the other hand, predicting the relative probabilities of finding species to a given point at a given time gives an indication of the dynamics of their local co-occurrence. In this respect, the CNN approach is closer to a joint species distribution model (jSDM). As Ovaskainen et al. (2017) mention, "By simultaneously drawing on the information from multiple species, these (jSDM) models allow one to seek community-level patterns in how species respond to their environment". Let's return to the two species example we used above. The fact that the probabilities are 0.5 for both species actually suggests that both species can coexist at the same abundance at this location. In this respect, the CNN multi-species classifier offers promising prospects for the prediction of assemblages and habitats thanks to the relative importance of the most characteristic/dominant species from a species pool. The species pool comprises all classified species and must be sufficiently representative of the ecological diversity of species niches in the area.

Finally, CNN-based species distribution modelling is a powerful and promising tool for studying the distributions of multi-species assemblages as a function of local environmental features but also of the spatial heterogeneity of each feature around the observation point in space and time (Deneu et al. 2021). It allows acknowledging the complex effects of environmental predictors and the roles of their spatial and temporal heterogeneity through the convolution operations performed in the neural network. As more and more computationally intensive tools become available, and as more and more environmental data becomes available at finer and finer temporal and spatial scales, the CNN approach is likely to be increasingly used to study biodiversity patterns across spatial and temporal scales.

References

Botella, C., Joly, A., Bonnet, P., Monestiez, P., and Munoz, F. (2018). Species distribution modeling based on the automated identification of citizen observations. Applications in Plant Sciences, 6(2), e1029. https://doi.org/10.1002/aps3.1029

Deneu, B., Servajean, M., Bonnet, P., Botella, C., Munoz, F., and Joly, A. (2021). Convolutional neural networks improve species distribution modelling by capturing the spatial structure of the environment. PLoS Computational Biology, 17(4), e1008856. https://doi.org/10.1371/journal.pcbi.1008856

Morand, G., Joly, A., Rouyer, T., Lorieul, T., and Barde, J. (2024) Predicting species distributions in the open ocean with convolutional neural networks. bioRxiv, ver.3 peer-reviewed and recommended by PCI Ecology https://doi.org/10.1101/2023.08.11.551418

Ovaskainen, O., Tikhonov, G., Norberg, A., Guillaume Blanchet, F., Duan, L., Dunson, D., ... and Abrego, N. (2017). How to make more out of community data? A conceptual framework and its implementation as models and software. Ecology letters, 20(5), 561-576. https://doi.org/10.1111/ele.12757

The late eminent ecologist Gary Polis once stated that “most catalogued food-webs are oversimplified caricatures of actual communities” and are “grossly incomplete representations of communities in terms of both diversity and trophic connections.” Not content with that damning indictment, he went further by railing that “theorists are trying to explain phenomena that do not exist” [1]. The latter critique might have been push back for Robert May´s ground-breaking but ultimately flawed research on the relationship between food-web complexity and stability [2]. Polis was a brilliant ecologist, and his thinking was clearly influenced by his experiences researching desert food webs. Those food webs possess an uncommon combination of properties, such as frequent omnivory, cannibalism, and looping; high linkage density (L/S); and a nearly complete absence of apex consumers, since few species completely lack predators or parasites [3]. During my PhD studies, I was lucky enough to visit Joshua Tree National Park on the way to a conference in New England, and I could immediately see the problems posed by desert ecosystems. At the time, I was ruminating on the “harsh-benign” hypothesis [4], which predicts that the relative importance of abiotic and biotic forces should vary with changes in local environmental conditions (from harsh to benign). Specifically, in more “harsh” environments, abiotic factors should determine community composition whilst weakening the influence of biotic interactions. However, in the harsh desert environment I saw first-hand evidence that species interactions were not diminished; if anything, they were strengthened. Teddy-bear chollas possessed murderously sharp defenses to protect precious water, creosote bushes engaged in belowground “chemical warfare” (allelopathy) to deter potential competitors, and rampant cannibalism amongst scorpions drove temporal and spatial ontogenetic niche partitioning. Life in the desert was hard, but you couldn´t expect your competition to go easy on you.
If that experience colored my thinking about nature’s reaction to a capricious environment, then the seminal work by Robert Paine on the marine rocky shore helped further cement the importance of biotic interactions. The insights provided by Paine [5] brings us closer to the research reported in the preprint “The return of the trophic chain: fundamental vs realized interactions in a simple arthropod food web” [6], given that the authors in that study hold the environment constant and test the interactions between different permutations of a simple community. Paine [5] was able to elegantly demonstrate using the chief protagonist Pisaster ochraceus (a predatory echinoderm also known as the purple sea star) that a keystone consumer could exert strong top-down control that radically reshaped the interactions amongst other community members. What was special about this study was that the presence of Pisaster promoted species diversity by altering competition for space by sedentary species, providing a rare example of an ecological network experiment combining trophic and non-trophic interactions. Whilst there are increasing efforts to describe these interactions (e.g., competition and facilitation) in multiplex networks [7], the authors of “The return of the trophic chain: fundamental vs realized interactions in a simple arthropod food web” [6] have avoided strictly competitive interactions for the sake of simplicity. They do focus on two trophic forms of competition, namely intraguild predation and apparent competition. These two interaction motifs, along with prey switching are relevant to my own research on the influence of cross-ecosystem prey subsidies to receiving food webs [8]. In particular, the apparent competition motif may be particularly important in the context of emergent adult aquatic insects as prey subsidies to terrestrial consumers. This was demonstrated by Henschel et al. [9] where the abundances of emergent adult aquatic midges in riparian fields adjacent to a large river helped stimulate higher abundances of spiders and lower abundances of herbivorous leafhoppers, leading to a trophic cascade. The aquatic insects had a bottom-up effect on spiders and this subsidy facilitated a top-down effect that cascaded from spiders to leafhoppers to plants. The apparent competition motif becomes relevant because the aquatic midges exerted a negative indirect effect on leafhoppers mediated through their common arachnid predators.
In the preprint “The return of the trophic chain: fundamental vs realized interactions in a simple arthropod food web” [6], the authors have described different permutations of a simple mite community present in avocado orchards (Persea americana). This community comprises of two predators (Euseius stipulatus and Neoseiulus californicus), one herbivore as shared prey (Oligonychus perseae), and pollen of Carpobrotus edulis as alternative food resource, with the potential for the intraguild predation and apparent competition interaction motifs to be expressed. The authors determined that these motifs should be realized based off pairwise feeding trials. It is common for food-web researchers to depict potential food webs, which contain all species sampled and all potential trophic links based on laboratory feeding trials (as demonstrated here) or from observational data and literature reviews [10]. In reality, not all these potential feeding links are realized because species may partition space and time, thus driving alternative food-web architectures. In “The return of the trophic chain: fundamental vs realized interactions in a simple arthropod food web” [6], the authors are able to show that placing species in combinations that should yield more complex interaction motifs based off pairwise feeding trials fails to deliver – the predators revert to their preferred prey resulting in modular and simple trophic chains to be expressed. Whilst these realized interaction motifs may be stable, there might also be a tradeoff with function by yielding less top-down control than desirable when considering the potential for ecosystem services such as pest management. These are valuable insights, although it should be noted that here the fundamental niche is described in a strictly Eltonian sense as a trophic role [11]. Adding additional niche dimensions (sensu [12]), such as a thermal gradient could alter the observed interactions, although it might be possible to explain these contingencies through metabolic and optimal foraging theory combined with species traits. Nonetheless, the results of these experiments further demonstrate the need for ecologists to cross-validate theory with empirical approaches to develop more realistic and predictive food-web models, lest they invoke the wrath of Gary Polis´ ghost by “trying to explain phenomena that do not exist”.

References

[1] Polis, G. A. (1991). Complex trophic interactions in deserts: an empirical critique of food-web theory. The American Naturalist, 138(1), 123-155. doi: 10.1086/285208
[2] May, R. M. (1973). Stability and complexity in model ecosystems. Princeton University Press, Princeton, NJ, USA
[3] Dunne, J. A. (2006). The network structure of food webs. In Pascual, M., & Dunne, J. A. (eds) Ecological Networks: Linking Structure to Dynamics in Food Webs. Oxford University Press, New York, USA, 27-86
[4] Menge, B. A., & Sutherland, J. P. (1976). Species diversity gradients: synthesis of the roles of predation, competition, and temporal heterogeneity. The American Naturalist, 110(973), 351-369. doi: 10.1086/283073
[5] Paine, R. T. (1966). Food web complexity and species diversity. The American Naturalist, 100(910), 65-75. doi: 10.1086/282400
[6] Torres-Campos, I., Magalhães, S., Moya-Laraño, J., & Montserrat, M. (2018). The return of the trophic chain: fundamental vs realized interactions in a simple arthropod food web. bioRxiv, 324178, ver. 5 peer-reviewed and recommended by PCI Ecol. doi: 10.1101/324178
[7] Kéfi, S., Berlow, E. L., Wieters, E. A., Joppa, L. N., Wood, S. A., Brose, U., & Navarrete, S. A. (2015). Network structure beyond food webs: mapping non‐trophic and trophic interactions on Chilean rocky shores. Ecology, 96(1), 291-303. doi: 10.1890/13-1424.1
[8] Burdon, F. J., & Harding, J. S. (2008). The linkage between riparian predators and aquatic insects across a stream‐resource spectrum. Freshwater Biology, 53(2), 330-346. doi: 10.1111/j.1365-2427.2007.01897.x
[9] Henschel, J. R., Mahsberg, D., & Stumpf, H. (2001). Allochthonous aquatic insects increase predation and decrease herbivory in river shore food webs. Oikos, 93(3), 429-438. doi: 10.1034/j.1600-0706.2001.930308.x
[10] Brose, U., Pavao-Zuckerman, M., Eklöf, A., Bengtsson, J., Berg, M. P., Cousins, S. H., Mulder, C., Verhoef, H. A., & Wolters, V. (2005). Spatial aspects of food webs. In de Ruiter, P., Wolters, V., Moore, J. C., & Melville-Smith, K. (eds) Dynamic Food Webs. vol 3. Academic Press, Burlington, 463-469
[11] Elton, C. (1927). Animal Ecology. Sidgwick and Jackson, London, UK
[12] Hutchinson, G. E. (1957). Concluding Remarks. Cold Spring Harbor Symposia on Quantitative Biology, 22, 415-427. doi: 10.1101/sqb.1957.022.01.039

Different sources of uncertainty are known to affect our ability to predict ecological dynamics (Petchey et al. 2015). However, the consequences of uncertainty on prediction biases have been less investigated, especially when predictions are scaled up to higher levels of organisation as is commonly done in ecology for instance. The study of Orr et al. (2020) addresses this issue. It shows that, in complex systems, the uncertainty of unbiased predictions at a lower level of organisation (e.g. species level) leads to a bias towards underestimation of change at higher level of organisation (e.g. ecosystem level). This bias is strengthened by larger uncertainty and by higher dimensionality of the system.
This general result has large implications for many fields of science, from economics to energy supply or demography. In ecology, as discussed in this study, these results imply that the uncertainty of predictions of species’ change increases the probability of underestimation of changes of diversity and stability at community and ecosystem levels, especially when species richness is high. The uncertainty of predictions of species’ change also increases the probability of underestimation of change when multiple ecosystem functions are considered at once, or when the combined effect of multiple stressors is considered.
The consequences of species diversity on ecosystem functions and stability have received considerable attention during the last decades (e.g. Cardinale et al. 2012, Kéfi et al. 2019). However, since the bias towards underestimation of change increases with species diversity, future studies will need to investigate how the general statistical effect outlined by Orr et al. might affect our understanding of the well-known relationships between species diversity and ecosystem functioning and stability in response to perturbations.

References

Cardinale BJ, Duffy JE, Gonzalez A, Hooper DU, Perrings C, Venail P, Narwani A, Mace GM, Tilman D, Wardle DA, Kinzig AP, Daily GC, Loreau M, Grace JB, Larigauderie A, Srivastava DS, Naeem S (2012) Biodiversity loss and its impact on humanity. Nature, 486, 59–67. https://doi.org/10.1038/nature11148
Kéfi S, Domínguez‐García V, Donohue I, Fontaine C, Thébault E, Dakos V (2019) Advancing our understanding of ecological stability. Ecology Letters, 22, 1349–1356. https://doi.org/10.1111/ele.13340
Orr JA, Piggott JJ, Jackson A, Arnoldi J-F (2020) Why scaling up uncertain predictions to higher levels of organisation will underestimate change. bioRxiv, 2020.05.26.117200. https://doi.org/10.1101/2020.05.26.117200
Petchey OL, Pontarp M, Massie TM, Kéfi S, Ozgul A, Weilenmann M, Palamara GM, Altermatt F, Matthews B, Levine JM, Childs DZ, McGill BJ, Schaepman ME, Schmid B, Spaak P, Beckerman AP, Pennekamp F, Pearse IS (2015) The ecological forecast horizon, and examples of its uses and determinants. Ecology Letters, 18, 597–611. https://doi.org/10.1111/ele.12443

Estimating populations size of widespread, common species in a relatively large and heterogeneous country like France is difficult for several reasons, from having a sample covering well the diverse ecological gradients to accounting for detectability, the fact that absence of a species may represent a false negative, the species being present but not detected. Bird communities have been the focus of a very large number of studies, with some countries like the UK having long traditions of monitoring both common and rare species. Nabias et al. use a large, structured citizen science project to provide new estimates of common bird species, accounting for detectability and using different habitat and climate covariates to extrapolate abundance to non-sampled areas. About 2/3 of the species had estimates higher than what would have been expected using a previous attempt at estimating population size based in part on expert knowledge and projected using estimates of trends to the period covered by the citizen science sampling. Some species showed large differences between the two estimates, which could be in part explained by accounting for detectability.

This paper uses what is called model-based inference (as opposed to design-based inference, that uses the design to make inferences about the whole population; Buckland et al. 2000), both in terms of detectability and habitat suitability. The estimates obtained depend on how well the model components approximate the underlying processes, which in a complex dataset like this one is not easy to assess. But it clearly shows that detectability may have substantial implications for the population size estimates. This is of course not new but has rarely been done at this scale and using a large sample obtained on many species. Interesting further work could focus on testing the robustness of the model-based approach by for example sampling new plots and compare the expected values to the observed values. Such a sampling could be stratified to maximize the discrimination between expected low and high abundances, at least for species where the estimates might be considered as uncertain, or for which estimating population sizes is deemed important.

References

Buckland, S. T., Goudie, I. B. J., & Borchers, D. L. (2000). Wildlife Population Assessment: Past Developments and Future Directions. Biometrics, 56(1), 1-12. https://doi.org/10.1111/j.0006-341X.2000.00001.x

 Nabias, J., Barbaro, L., Fontaine, B., Dupuy, J., Couzi, L., et al. (2024) Reassessment of French breeding bird population sizes using citizen science and accounting for species detectability. HAL, ver. 2 peer-reviewed and recommended by Peer Community in Ecology. https://hal.science/hal-04478371