In the field of ecology, there is a growing interest in machine (including deep) learning for processing and automatizing repetitive analyses on large amounts of images collected from camera traps, drones and smartphones, among others. These analyses include species or individual recognition and classification, counting or tracking individuals, detecting and classifying behavior. By saving countless times of manual work and tapping into massive amounts of data that keep accumulating with technological advances, machine learning is becoming an essential tool for ecologists. We refer to recent papers for more details on machine learning for ecology and evolution (Besson et al. 2022, Borowiec et al. 2022, Christin et al. 2019, Goodwin et al. 2022, Lamba et al. 2019, Nazir & Kaleem 2021, Perry et al. 2022, Picher & Hartig 2023, Tuia et al. 2022, Wäldchen & Mäder 2018).

In their paper, Nakagawa et al. (2023) conducted a systematic review of the literature on machine learning for wildlife imagery. Interestingly, the authors used a method unfamiliar to ecologists but well-established in medicine called rapid review, which has the advantage of being quickly completed compared to a fully comprehensive systematic review while being representative (Lagisz et al., 2022). Through a rigorous examination of more than 200 articles, the authors identified trends and gaps, and provided suggestions for future work. Listing all their findings would be counterproductive (you’d better read the paper), and I will focus on a few results that I have found striking, fully assuming a biased reading of the paper. First, Nakagawa et al. (2023) found that most articles used neural networks to analyze images, in general through collaboration with computer scientists. A challenge here is probably to think of teaching computer vision to the generations of ecologists to come (Cole et al. 2023). Second, the images were dominantly collected from camera traps, with an increase in the use of aerial images from drones/aircrafts that raise specific challenges. Third, the species concerned were mostly mammals and birds, suggesting that future applications should aim to mitigate this taxonomic bias, by including, e.g., invertebrate species. Fourth, most papers were written by authors affiliated with three countries (Australia, China, and the USA) while India and African countries provided lots of images, likely an example of scientific colonialism which should be tackled by e.g., capacity building and the involvement of local collaborators. Last, few studies shared their code and data, which obviously impedes reproducibility. Hopefully, with the journals’ policy of mandatory sharing of codes and data, this trend will be reversed. 

REFERENCES

Besson M, Alison J, Bjerge K, Gorochowski TE, Høye TT, Jucker T, Mann HMR, Clements CF (2022) Towards the fully automated monitoring of ecological communities. Ecology Letters, 25, 2753–2775. https://doi.org/10.1111/ele.14123

Borowiec ML, Dikow RB, Frandsen PB, McKeeken A, Valentini G, White AE (2022) Deep learning as a tool for ecology and evolution. Methods in Ecology and Evolution, 13, 1640–1660. https://doi.org/10.1111/2041-210X.13901

Christin S, Hervet É, Lecomte N (2019) Applications for deep learning in ecology. Methods in Ecology and Evolution, 10, 1632–1644. https://doi.org/10.1111/2041-210X.13256

Cole E, Stathatos S, Lütjens B, Sharma T, Kay J, Parham J, Kellenberger B, Beery S (2023) Teaching Computer Vision for Ecology. https://doi.org/10.48550/arXiv.2301.02211

Goodwin M, Halvorsen KT, Jiao L, Knausgård KM, Martin AH, Moyano M, Oomen RA, Rasmussen JH, Sørdalen TK, Thorbjørnsen SH (2022) Unlocking the potential of deep learning for marine ecology: overview, applications, and outlook†. ICES Journal of Marine Science, 79, 319–336. https://doi.org/10.1093/icesjms/fsab255

Lagisz M, Vasilakopoulou K, Bridge C, Santamouris M, Nakagawa S (2022) Rapid systematic reviews for synthesizing research on built environment. Environmental Development, 43, 100730. https://doi.org/10.1016/j.envdev.2022.100730

Lamba A, Cassey P, Segaran RR, Koh LP (2019) Deep learning for environmental conservation. Current Biology, 29, R977–R982. https://doi.org/10.1016/j.cub.2019.08.016

Nakagawa S, Lagisz M, Francis R, Tam J, Li X, Elphinstone A, Jordan N, O’Brien J, Pitcher B, Sluys MV, Sowmya A, Kingsford R (2023) Rapid literature mapping on the recent use of machine learning for wildlife imagery. EcoEvoRxiv, ver. 4 peer-reviewed and recommended by Peer Community in Ecology.  https://doi.org/10.32942/X2H59D

Nazir S, Kaleem M (2021) Advances in image acquisition and processing technologies transforming animal ecological studies. Ecological Informatics, 61, 101212. https://doi.org/10.1016/j.ecoinf.2021.101212

Perry GLW, Seidl R, Bellvé AM, Rammer W (2022) An Outlook for Deep Learning in Ecosystem Science. Ecosystems, 25, 1700–1718. https://doi.org/10.1007/s10021-022-00789-y

Pichler M, Hartig F Machine learning and deep learning—A review for ecologists. Methods in Ecology and Evolution, n/a. https://doi.org/10.1111/2041-210X.14061

Tuia D, Kellenberger B, Beery S, Costelloe BR, Zuffi S, Risse B, Mathis A, Mathis MW, van Langevelde F, Burghardt T, Kays R, Klinck H, Wikelski M, Couzin ID, van Horn G, Crofoot MC, Stewart CV, Berger-Wolf T (2022) Perspectives in machine learning for wildlife conservation. Nature Communications, 13, 792. https://doi.org/10.1038/s41467-022-27980-y

Wäldchen J, Mäder P (2018) Machine learning for image-based species identification. Methods in Ecology and Evolution, 9, 2216–2225. https://doi.org/10.1111/2041-210X.13075

The mean trophic level of catches in world fisheries has generally declined over the 20th century, a phenomenon called "fishing down the food web" (Pauly et al. 1998). Several mechanisms have been proposed to explain this decline including the collapse of, or decline in, higher trophic level stocks leading to the inclusion of lower trophic level stocks in the fishery. Fishing down the food web may lead to a reduction in the resilience, i.e., the capacity to rebound from change, of the fished community, which is concerning given the necessity of resilience in the face of climate change. 

The practice of adaptive harvesting, which involves fishing stocks based on their availability, can also result in a reduction in the average trophic level of a fishery (Branch et al. 2010). Adaptive harvesting, similar to adaptive foraging, can affect the resilience of fisheries. Generally, adaptive foraging acts as a stabilizing force in communities (Valdovinos et al. 2010), however it is not clear how including harvesters as the adaptive foragers will affect the resilience of the system.

Tromeur and Loeuille (2023) analyze the effects of adaptively harvesting a trophic community. Using a system of ordinary differential equations representing a predator-prey model where both species are harvested, the researchers mathematically analyze the impact of increasing fishing effort and adaptive harvesting on the mean trophic level and resilience of the fished community. This is achieved by computing the equilibrium densities and equilibrium allocation of harvest effort.  In addition, the researchers numerically evaluate adaptive harvesting in a tri-trophic system (predator, prey, and resource). The study focuses on the effect of adaptively distributing harvest across trophic levels on the mean trophic level of catches, the propensity for regime shifts to occur, the ability to return to equilibrium after a disturbance, and the speed of this return. 

The results indicate that adaptive harvesting leads to a decline in the mean trophic level of catches, resulting in “fishing down the food web”. Furthermore, the study shows that adaptive harvesting may harm the overall resilience of the system. Similar results were observed numerically in a tri-trophic community.

While adaptive foraging is generally a stabilizing force on communities, the researchers found that adaptive harvesting can destabilize the harvested community. One of the key differences between adaptive foraging models and the model presented here, is that the harvesters do not exhibit population dynamics. This lack of a numerical response by the harvesters to decreasing population sizes of their stocks leads to regime shifts. The realism of a fishery that does not respond numerically to declining stock is debatable, however it is very likely that there will a least be significant delays due to social and economic barriers to leaving the fishery, that will lead to similar results.

This study is not unique in demonstrating the ability of adaptive harvesting to result in “fishing down the food web”. As pointed out by the researchers, the same results have been shown with several different model formulations (e.g., age and size structured models). Similarly, this study is not unique to showing that increasing adaptation speeds decreases the resilience of non-linear predator-prey systems by inducing oscillatory behaviours. Much of this can be explained by the destabilising effect of increasing interaction strengths on food webs (McCann et al. 1998). 

By employing a straightforward model, the researchers were able to demonstrate that adaptive harvesting, a common strategy employed by fishermen, can result in a decline in the average trophic level of catches, regime shifts, and reduced resilience in the fished community. While previous studies have observed some of these effects, the fact that the current study was able to capture them all with a simple model is notable. This modeling approach can offer insight into the role of human behavior on the complex dynamics observed in fisheries worldwide.

References

Branch, T. A., R. Watson, E. A. Fulton, S. Jennings, C. R. McGilliard, G. T. Pablico, D. Ricard, et al. 2010. The trophic fingerprint of marine fisheries. Nature 468:431–435. https://doi.org/10.1038/nature09528

Tromeur, E., and N. Loeuille. 2023. Effects of adaptive harvesting on fishing down processes and resilience changes in predator-prey and tritrophic systems. bioRxiv 290460, ver 5 peer-reviewed and recommended by PCI Ecology. https://doi.org/10.1101/290460

McCann, K., A. Hastings, and G.R. Huxel. 1998. Weak trophic interactions and the balance of nature. Nature 395: 794-798. https://doi.org/10.1038/27427

Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres Jr. 1998. Fishing down marine food webs. Science 279:860–86. https://doi.org/10.1126/science.279.5352.860

Valdovinos, F.S., R. Ramos-Jiliberto, L. Garay-Naravez, P. Urbani, and J.A. Dunne. 2010. Consequences of adaptive behaviour for the structure and dynamics of food webs. Ecology Letters 13: 1546-1559. https://doi.org/10.1111/j.1461-0248.2010.01535.x

Reduce the impact the intensification of human activities has on the environmental is the challenge the humanity faces today, a major challenge that could be compared to climbing Everest without an oxygen supply. Indeed, over-population, pollution, burning fossil fuels, and deforestation are all evils which have had hugely detrimental effects on the environment such as climate change, soil erosion, poor air quality, and scarcity of drinking water to name but a few. In response to the ever-growing consumer demand, agriculture has intensified massively along with a drastic increase in the use of chemicals to ensure an adequate food supply while controlling crop pests. In this context, to address the disastrous effects of the intensive usage of pesticides on both human health and biodiversity, organic farming (OF) revealed as a miracle remedy with multiple benefits. Delattre et al. (2023) present a powerful modelling approach to decipher the crossed effects of the landscape structure and the OF expansion scenario on the pest abundance, both in organic and conventional (CF) crop fields. To this end, the authors ingeniously combined a grid-based landscape model with a spatially explicit predator-pest model. Based on an extensive in silico simulation process, they explore a diversity of landscape structures differing in their amount of semi-natural habitats (SHN) and in their fragmentation, to finally propose a ranking of various expansion scenarios according to the pest control methods in organic farming as well as to the pest and predators’ dissemination capacities. In total, 9 landscape structures (3 proportions of SHN x 3 fragmentation levels) were crossed with 3 expansion scenarios (RD = a random distribution of OF and CF in the grid; IP = isolated CF are converted; GP = CF within aggregates are converted), 4 pest management practices, 3 initial densities and 36 biological parameter combinations driving the predator’ and pest’s population dynamics. This exhaustive exploration of possible combinations of landscape and farming practices highlighted the main drivers of the various OF expansion scenarios, such as increased spillover of predators in isolated OF/CF fields, increased pest management efficiency in large patches of CF and the importance of the distance between OF and CF. In the end, this study brings to light the crucial role that landscape planning plays when OF practices have limited efficiency on pests. It also provides convincing arguments to the fact that converting to organic isolated CF as a priority seems to be the most promising scenario to limit pest densities in CF crops while improving predator to pest ratios (considered as a proxy of conservation biological control) in OF ones without increasing pest densities. Once further completed with model calibration validation based on observed life history traits data for both predators and pests, this work should be very helpful in sustaining policy makers to convince farmers of engaging in organic farming.

REFERENCES

Delattre T, Memah M-M, Franck P, Valsesia P, Lavigne C (2023) Best organic farming deployment scenarios for pest control: a modeling approach. bioRxiv, 2022.05.31.494006, ver. 2 peer-reviewed and recommended by Peer Community in Ecology. https://doi.org/10.1101/2022.05.31.494006

The temporal or spatial variability in species population sizes and interaction strength of animal and plant communities has a strong impact on aggregate community properties (for instance biomass), community composition, and species richness (Kokkoris et al. 2002). Early work on spatial and temporal variability strongly indicated that asynchronous population and environmental fluctuations tend to stabilise community structures and diversity (e.g. Holt 1984, Tilman and Pacala 1993, McCann et al. 1998, Amarasekare and Nisbet 2001). Similarly, trophic networks might be stabilised by spatial heterogeneity (Hastings 1977) and an asymmetry of energy flows along food chains (Rooney et al. 2006). The interplay between temporal, spatial, and trophic heterogeneity within the meta-community concept has got much less interest. In the recent preprint in PCI Ecology, Quévreux et al. (2023) report that Spatial heterogeneity of interaction strength has contrasting effects on synchrony and stability in trophic metacommunities. These authors rightly notice that the interplay between trophic and spatial heterogeneity might induce contrasting effects depending on the internal dynamics of the system. Their contribution builds on prior work (Quévreux et al. 2021a, b) on perturbed trophic cascades.

I found this paper particularly interesting because it is in the, now century-old, tradition to show that ecological things are not so easy. Since the 1930th, when Nicholson and Baily and others demonstrated that simple deterministic population models might generate stability and (pseudo-)chaos ecologists have realised that systems triggered by two or more independent processes might be intrinsically unpredictable and generate different outputs depending on the initial parameter settings. This resembles the three-body problem in physics. The present contribution of Quévreux et al. (2023) extends this knowledge to an example of a spatially explicit trophic model. Their main take-home message is that asymmetric energy flows in predator–prey relationships might have contrasting effects on the stability of metacommunities receiving localised perturbations. Stability is context dependent.

Of course, the work is merely a theoretical exercise using a simplistic trophic model. It demands verification with field data. Nevertheless, we might expect even stronger unpredictability in more realistic multitrophic situations. Therefore, it should be seen as a proof of concept. Remember that increasing trophic connectance tends to destabilise food webs (May 1972). In this respect, I found the final outlook to bioconservation ambitious but substantiated. Biodiversity management needs a holistic approach focusing on all aspects of ecological functioning. I would add the need to see stability and biodiversity within an evolutionary perspective.        

References

Amarasekare P, Nisbet RM (2001) Spatial Heterogeneity, Source‐Sink Dynamics, and the Local Coexistence of Competing Species. The American Naturalist, 158, 572–584. https://doi.org/10.1086/323586

Hastings A (1977) Spatial heterogeneity and the stability of predator-prey systems. Theoretical Population Biology, 12, 37–48. https://doi.org/10.1016/0040-5809(77)90034-X

Holt RD (1984) Spatial Heterogeneity, Indirect Interactions, and the Coexistence of Prey Species. The American Naturalist, 124, 377–406. https://doi.org/10.1086/284280

Kokkoris GD, Jansen VAA, Loreau M, Troumbis AY (2002) Variability in interaction strength and implications for biodiversity. Journal of Animal Ecology, 71, 362–371. https://doi.org/10.1046/j.1365-2656.2002.00604.x

May RM (1972) Will a Large Complex System be Stable? Nature, 238, 413–414. https://doi.org/10.1038/238413a0

McCann K, Hastings A, Huxel GR (1998) Weak trophic interactions and the balance of nature. Nature, 395, 794–798. https://doi.org/10.1038/27427

Quévreux P, Barbier M, Loreau M (2021) Synchrony and Perturbation Transmission in Trophic Metacommunities. The American Naturalist, 197, E188–E203. https://doi.org/10.1086/714131

Quévreux P, Pigeault R, Loreau M (2021) Predator avoidance and foraging for food shape synchrony and response to perturbations in trophic metacommunities. Journal of Theoretical Biology, 528, 110836. https://doi.org/10.1016/j.jtbi.2021.110836

Quévreux P, Haegeman B, Loreau M (2023) Spatial heterogeneity of interaction strength has contrasting effects on synchrony and stability in trophic metacommunities. hal-03829838, ver. 2 peer-reviewed and recommended by Peer Community in Ecology. https://hal.science/hal-03829838

Rooney N, McCann K, Gellner G, Moore JC (2006) Structural asymmetry and the stability of diverse food webs. Nature, 442, 265–269. https://doi.org/10.1038/nature04887

Tilman D, Pacala S (1993) The maintenance of species richness in plant communities. In: Ricklefs, R.E., Schluter, D. (eds) Species Diversity in Ecological Communities: Historical and Geographical Perspectives. University of Chicago Press, pp. 13–25.

​​In this article, the authors discuss the results of their study investigating the effects of heat stress and moisture stress on a terrestrial isopod Armadilldium vulgare, the common woodlouse [1]. Specifically, the authors have assessed how increased temperature or decreased moisture affects life history traits (such as growth, survival, and reproduction) as well as physiological traits (immune cell parameters and \( beta \)-galactosidase activity). This article quantitatively evaluates the effects of the two stressors on woodlouse. Terrestrial isopods like woodlouse are sensitive to thermal and moisture stress [2; 3] and are therefore good models to test hypotheses in global change biology and for monitoring ecosystem health.

​An important feature of this study is the combination of experimental, laboratory, and analytical techniques. Experiments were conducted under controlled conditions in the laboratory by modulating temperature and moisture, life history and physiological traits were measured/analyzed and then tested using models. Both stressors had negative impacts on survival and reproduction of woodlouse, and result in premature ageing. Although thermal stress did not affect survival, it slowed woodlouse growth. Moisture stress did not have a detectable effect on woodlouse growth but decreased survival and reproductive success. An important insight from this study is that effects of heat and moisture stressors on woodlouse are not necessarily linear, and experimental approaches can be used to better elucidate the mechanisms and understand how these organisms respond to environmental stress.

​This article is timely given the increasing attention on biological monitoring and ecosystem health.​

References:

[1] Depeux C, Branger A, Moulignier T, Moreau J, Lemaître J-F, Dechaume-Moncharmont F-X, Laverre T, Pauhlac H, Gaillard J-M, Beltran-Bech S (2022) Deleterious effects of thermal and water stresses on life history and physiology: a case study on woodlouse. bioRxiv, 2022.09.26.509512., ver. 3 peer-reviewd and recommended by PCI Ecology. https://doi.org/10.1101/2022.09.26.509512

[2] ​Warburg MR, Linsenmair KE, Bercovitz K (1984) The effect of climate on the distribution and abundance of isopods. In: Sutton SL, Holdich DM, editors. The Biology of Terrestrial Isopods. Oxford: Clarendon Press. pp. 339–367.​

[3] Hassall M, Helden A, Goldson A, Grant A (2005) Ecotypic differentiation and phenotypic plasticity in reproductive traits of Armadillidium vulgare (Isopoda: Oniscidea). Oecologia 143: 51–60.​ https://doi.org/10.1007/s00442-004-1772-3