Proposing efficient guidelines for biodiversity conservation often requires the use of forecasting tools. Species Distribution Models (SDM) are more and more used to predict how the distribution of a species will react to environmental change, including any large-scale management actions that could be implemented. Their use is also boosted by the increase of publicly available biodiversity databases[1]. The now famous aphorism by George Box "All models are wrong but some are useful"[2] very well summarizes that the outcome of a model must be adjusted to, and will depend on, the data that are used to parameterize it. The question of the reliability of using biodiversity databases to parameterize biodiversity models such as SDM –but the question would also apply to other kinds of biodiversity models, e.g. Population Viability Analysis models[3]– is key to determine the confidence that can be placed in model predictions. This point is often overlooked by some categories of biodiversity conservation stakeholders, in particular the fact that some data were collected using controlled protocols while others are opportunistic. 

In this study[4], the authors use a collection of databases covering a range of species as well as of geographic scales in France and using different data collection and validation approaches as a case study to evaluate the impact of data quality when performing Strategic Environmental Assessment (SEA). Among their conclusions, the fact that a large-scale database (what they call the “country” level) is necessary to reliably parameterize SDM. Besides this and other conclusions of their study, which are likely to be in part specific to their case study –unfortunately for its conservation, biodiversity is complex and varies a lot–, the merit of this work lies in the approach used to test the impact of data on model predictions.

References

1.  Feng, X. et al. A review of the heterogeneous landscape of biodiversity databases: Opportunities and challenges for a synthesized biodiversity knowledge base. Global Ecology and Biogeography 31, 1242–1260 (2022). https://doi.org/10.1111/geb.13497

2.  Box, G. E. P. Robustness in the Strategy of Scientific Model Building. in Robustness in Statistics (eds. Launer, R. L. & Wilkinson, G. N.) 201–236 (Academic Press, 1979). https://doi.org/10.1016/B978-0-12-438150-6.50018-2.

3.  Beissinger, S. R. & McCullough, D. R. Population Viability Analysis. (The University of Chicago Press, 2002).

4.  Ferraille, T., Kerbiriou, C., Bigard, C., Claireau, F. & Thompson, J. D. (2023) Integrating biodiversity assessments into local conservation planning: the importance of assessing suitable data sources. bioRxiv, ver. 3 peer-reviewed and recommended by Peer Community in Ecology.  https://doi.org/10.1101/2023.05.09.539999

Individuals can respond to environmental changes that they undergo directly (within-generation plasticity) but also through transgenerational plasticity, providing lasting effects that are transmitted to the next generations (Donelson et al. 2012; Munday et al. 2013; Kuijper & Hoyle 2015; Auge et al. 2017, Tariel et al. 2020). These parental effects can affect offspring via various mechanisms, notably via maternal transmission of hormones to the eggs or growing embryos (Mousseau & Fox 1998). While the effects of environmental quality may simply carry-over to the next generation (e.g., females in stressful environments give birth to offspring in poorer condition), parental effects may also be a mechanism that adjusts offspring phenotype in response to environmental variation and predictability, and thereby match offspring's phenotype to future environmental conditions (Gluckman et al. 2005; Marshall & Uller 2007; Dey et al. 2016; Yin et al. 2019), for example by preparing their offspring to an expected stressful environment.

When females experience stress during gestation or egg formation, elevations in glucocorticoids (GC) are expected to affect offspring phenotype in many ways, from the offspring's own GC levels, to their growth and survival (Sheriff et al. 2017). This is a well established idea, but how strong is the evidence for this? A meta-analysis on birds found no clear effect of corticosterone manipulation on offspring traits (38 studies on 9 bird species for corticosterone manipulation; Podmokła et al. 2018). Another meta-analysis including 14 vertebrate species found no clear effect of prenatal stress on offspring GC (Thayer et al. 2018). Finally, a meta-analysis on wild vertebrates (23 species) found no clear effect of GC-mediated maternal effects on offspring traits (MacLeod et al. 2021). As often when facing such inconclusive results, context dependence has been suggested as one potential reason for such inconsistencies, for exemple sex specific effects (Groothuis et al. 2019, 2020). However, sex specific measures on offspring are scarce (Podmokła et al. 2018). Moreover, the literature available is still limited to a few, mostly “model” species.

With their study, Amin et al. (2024) show the way to improve our understanding on GC transmission from mother to offspring and its effects in several aspects. First they used innovative non-invasive methods (which could broaden the range of species available to study) by quantifying cortisol metabolites from faecal samples collected from pregnant females, as proxy for maternal GC level, and relating it to GC levels from hairs of their neonate offspring. Second they used a free ranging large mammal (taxa from which literature is missing): the fallow deer (Dama dama). Third, they provide sex specific measures of GC levels. And finally but importantly, they are exemplary in their transparency regarding 1) the exploratory nature of their study, 2) their statistical thinking and procedure, and 3) the study limitations (e.g., low sample size and high within individual variation of measurements). I hope this study will motivate more research (on the fallow deer, and on other species) to broaden and strengthen our understanding of sex specific effects of maternal stress and CG levels on offspring phenotype and fitness.

References

Amin, B., Fishman, R., Quinn, M., Matas, D., Palme, R., Koren, L., & Ciuti, S. (2024). Sex differences in the relationship between maternal and foetal glucocorticoids in a free-ranging large mammal. bioRxiv, ver. 4 peer-reviewed and recommended by Peer Community in Ecology. https://doi.org/10.1101/2023.05.04.538920 

Auge, G.A., Leverett, L.D., Edwards, B.R. & Donohue, K. (2017). Adjusting phenotypes via within-and across-generational plasticity. New Phytologist, 216, 343–349. https://doi.org/10.1111/nph.14495

Dey, S., Proulx, S.R. & Teotonio, H. (2016). Adaptation to temporally fluctuating environments by the evolution of maternal effects. PLoS biology, 14, e1002388. https://doi.org/10.1371/journal.pbio.1002388

Donelson, J.M., Munday, P.L., McCormick, M.I. & Pitcher, C.R. (2012). Rapid transgenerational acclimation of a tropical reef fish to climate change. Nature Climate Change, 2, 30. https://doi.org/10.1038/nclimate1323

Gluckman, P.D., Hanson, M.A. & Spencer, H.G. (2005). Predictive adaptive responses and human evolution. Trends in ecology & evolution, 20, 527–533. https://doi.org/10.1016/j.tree.2005.08.001

Groothuis, Ton GG, Bin-Yan Hsu, Neeraj Kumar, and Barbara Tschirren. "Revisiting mechanisms and functions of prenatal hormone-mediated maternal effects using avian species as a model." Philosophical Transactions of the Royal Society B 374, no. 1770 (2019): 20180115. https://doi.org/10.1098/rstb.2018.0115

Groothuis, Ton GG, Neeraj Kumar, and Bin-Yan Hsu. "Explaining discrepancies in the study of maternal effects: the role of context and embryo." Current Opinion in Behavioral Sciences 36 (2020): 185-192. https://doi.org/10.1016/j.cobeha.2020.10.006 

Kuijper, B. & Hoyle, R.B. (2015). When to rely on maternal effects and when on phenotypic plasticity? Evolution, 69, 950–968. https://doi.org/10.1111/evo.12635   

MacLeod, Kirsty J., Geoffrey M. While, and Tobias Uller. "Viviparous mothers impose stronger glucocorticoid‐mediated maternal stress effects on their offspring than oviparous mothers." Ecology and Evolution 11, no. 23 (2021): 17238-17259.

Marshall, D.J. & Uller, T. (2007). When is a maternal effect adaptive? Oikos, 116, 1957–1963. https://doi.org/10.1111/j.2007.0030-1299.16203.x       

Mousseau, T.A. & Fox, C.W. (1998). Maternal effects as adaptations. Oxford University Press.

Munday, P.L., Warner, R.R., Monro, K., Pandolfi, J.M. & Marshall, D.J. (2013). Predicting evolutionary responses to climate change in the sea. Ecology Letters, 16, 1488–1500. https://doi.org/10.1111/ele.12185

Podmokła, Edyta, Szymon M. Drobniak, and Joanna Rutkowska. "Chicken or egg? Outcomes of experimental manipulations of maternally transmitted hormones depend on administration method–a meta‐analysis." Biological Reviews 93, no. 3 (2018): 1499-1517. https://doi.org/10.1111/brv.12406 

Sheriff, M. J., Bell, A., Boonstra, R., Dantzer, B., Lavergne, S. G., McGhee, K. E., MacLeod, K. J., Winandy, L., Zimmer, C., & Love, O. P. (2017). Integrating ecological and evolutionary context in the study of maternal stress. Integrative and Comparative Biology, 57(3), 437–449. https://doi.org/10.1093/icb/icx105

Tariel, Juliette, Sandrine Plénet, and Émilien Luquet. "Transgenerational plasticity in the context of predator-prey interactions." Frontiers in Ecology and Evolution 8 (2020): 548660. https://doi.org/10.3389/fevo.2020.548660 

Thayer, Zaneta M., Meredith A. Wilson, Andrew W. Kim, and Adrian V. Jaeggi. "Impact of prenatal stress on offspring glucocorticoid levels: A phylogenetic meta-analysis across 14 vertebrate species." Scientific Reports 8, no. 1 (2018): 4942. https://doi.org/10.1038/s41598-018-23169-w 

Yin, J., Zhou, M., Lin, Z., Li, Q.Q. & Zhang, Y.-Y. (2019). Transgenerational effects benefit offspring across diverse environments: a meta-analysis in plants and animals. Ecology letters, 22, 1976–1986. https://doi.org/10.1111/ele.13373

Worldwide, city expansion is happening at a fast rate and at the same time, urbanists are more and more required to make place for biodiversity. Choices have to be made regarding the area and spatial arrangement of suitable spaces for non-human living organisms, that will favor the long-term survival of their populations. To guide those choices, it is necessary to understand the mechanisms driving the effects of land management on biodiversity.

Research results on the effects of urbanization on genetic diversity have been very diverse, with studies showing higher genetic diversity in rural than in urban populations (e.g. Delaney et al. 2010), the contrary (e.g. Miles et al. 2018) or no difference (e.g. Schoville et al. 2013). The same is true for studies investigating genetic differentiation. The reasons for these differences probably lie in the relative intensities of gene flow and genetic drift in each case study, which are hard to disentangle and quantify in empirical datasets.

In their paper, Savary et al. (2024) used an elegant and powerful simulation approach to better understand the diversity of observed patterns and investigate the effects of dispersal limitation on genetic patterns (diversity and differentiation). Their simulations involved the landscapes of 325 real European cities, each under three different scenarios mimicking 3 virtual urban tolerant species with different abilities to move within cities while genetic drift intensity was held constant across scenarios. The cities were chosen so that the proportion of artificial areas was held constant (20%) but their location and shape varied. This design allowed the authors to investigate the effect of connectivity and spatial configuration of habitat on the genetic responses to spatial variations in dispersal in cities. 

The main results of this simulation study demonstrate that variations in dispersal spatial patterns, for a given level of genetic drift, trigger variations in genetic patterns. Genetic diversity was lower and genetic differentiation was larger when species had more difficulties to move through the more hostile components of the urban environment. The increase of the relative importance of drift over gene flow when dispersal was spatially more constrained was visible through the associated disappearance of the pattern of isolation by resistance. Forest patches (usually located at the periphery of the cities) usually exhibited larger genetic diversity and were less differentiated than urban green spaces. But interestingly, the presence of habitat patches at the interface between forest and urban green spaces lowered those differences through the promotion of gene flow. 

One other noticeable result, from a landscape genetic method point of view, is the fact that there might be a limit to the detection of barriers to genetic clusters through clustering analyses because of the increased relative effect of genetic drift. This result needs to be confirmed, though, as genetic structure has only been investigated with a recent approach based on spatial graphs. It would be interesting to also analyze those results with the usual Bayesian genetic clustering approaches. 

Overall, this study addresses an important scientific question about the mechanisms explaining the diversity of observed genetic patterns in cities. But it also provides timely cues for connectivity conservation and restoration applied to cities.  
 
References

Delaney, K. S., Riley, S. P., and Fisher, R. N. (2010). A rapid, strong, and convergent genetic response to urban habitat fragmentation in four divergent and widespread vertebrates. PLoS ONE, 5(9):e12767.
https://doi.org/10.1371/journal.pone.0012767
 
Miles, L. S., Dyer, R. J., and Verrelli, B. C. (2018). Urban hubs of connectivity: Contrasting patterns of gene flow within and among cities in the western black widow spider. Proceedings of the Royal Society B, 285(1884):20181224. https://doi.org/10.1098/rspb.2018.1224
 
Savary P., Tannier C., Foltête J.-C., Bourgeois M., Vuidel G., Khimoun A., Moal H., and Garnier S. (2024). How does dispersal shape the genetic patterns of animal populations in European cities? A simulation approach. EcoEvoRxiv, ver. 3 peer-reviewed and recommended by Peer Community in Ecology. https://doi.org/10.32942/X2JS41.
 
Schoville, S. D., Widmer, I., Deschamps-Cottin, M., and Manel, S. (2013). Morphological clines and weak drift along an urbanization gradient in the butterfly, Pieris rapae. PLoS ONE, 8(12):e83095.
https://doi.org/10.1371/journal.pone.0083095

Ancient human colonisation often had catastrophic consequences for native fauna. The North American Megafauna went extinct shortly after humans entered the scene and Madagascar suffered twice, before 1500 CE and around 1700 CE after the Malayan and European colonisation. Maoris colonised New Zealand by about 1300 and a century later the giant Moa birds (Dinornithiformes) sharply declined. But did they went extinct or are they an ancient example of Lazarus species, species thought to be extinct but still alive? Scattered anecdotes of late sightings of living Moas even up to the 20th century seem to suggest the latter. The quest for later survival has also a criminal aspect. Who did it, the Maoris or the white colonisers in the late 18th century?

The present work by Floe Foxon (2023) tries to settle this question. It uses a survival modelling approach and an assessment of the reliability of nearly 100 alleged sightings. The model favours the so-called overkill hypothesis, that Moas probably went extinct in the 15th century shortly after Maori colonisation. A small but still remarkable probability remained for survival up to 1770. Later sightings turned out to be highly unreliable.

The paper is important as it does not rely on subjective discussions of late sightings but on a probabilistic modelling approach with sensitivity testing prior applied to marsupials. As common in probabilistic approaches, the study does not finally settle the case. A probability of as much as 20% remained for late survival after 1450 CE. This is not improbable as New Zealand was sufficiently unexplored in those days to harbour a few refuges for late survivors. However, in this respect, it is a bit unfortunate that at the end of the discussion, the paper cites Heuvelmans, the founder of cryptozoology, and it mentions the ivory-billed woodpecker, which has recently been redetected. No Moa remains were found after 1450.

References

Foxon F (2023) The Moa the Merrier: Resolving When the Dinornithiformes Went Extinct. bioRxiv, 2023.08.07.552261, ver. 2 peer-reviewed and recommended by Peer Community in Ecology. https://doi.org/10.1101/2023.08.07.552261

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