In this post, Stephen Osakpolor talks about his review of process-based food web models for both aquatic and terrestrial ecosystem. These models are essential for understanding how stressors affect the interconnected environments.
Food webs, especially at the aquatic-terrestrial interface, are affected by multiple anthropogenic stressors. Anthropogenic stressor interactions can change over time and also spatial scale. However, most empirical studies focus on narrow temporal and spatial scales. Process-based food web models can help in predicting anthropogenic stressor interactions and food web responses over variable temporal and spatial scales. They can also contribute to the testing of hypotheses and mechanistic assumptions. Hence, they help to understand the functioning of meta-ecosystems, filling and identifying observation gaps, and developing management strategies.
The researchers reviewed process-based food web models applied to the aquatic-terrestrial interface (aquatic-terrestrial models) and theoretical meta-ecosystems (theoretical models).
In a nutshell, the study found that various process-based food web models so developed have led to a greater understanding of aquatic-terrestrial and meta-ecosystems in general. Specifically, the reviewed studies have provided predictions that can form the basis of experiments on the effects of environmental change on aquatic-terrestrial and meta-ecosystem food webs. However, relative to the theoretical models, the aquatic-terrestrial models provide a greater diversity of model state variables and complexity. Not surprisingly, owing to its generality, the theoretical models have been applied to study the effect of change in system components (e.g., change in cross-ecosystem flows, heterogeneity of ecosystems) not explicitly linked to an anthropogenic stressor on food webs in meta-ecosystems. In contrast, the aquatic-terrestrial models were also used to study anthropogenic stressors. Given below is an outline of the major research gaps that limit one’s understanding of food webs in an aquatic-terrestrial interface or meta-ecosystem context.
Flows and stressors from aquatic to terrestrial food webs
The reviewed aquatic-terrestrial models focused on how flows from terrestrial ecosystems affect aquatic food webs (Fig. 1), with none focusing on reciprocal flows. Similarly, only anthropogenic stressors propagating from terrestrial ecosystems to aquatic food webs were considered, for example, the effects of the reduction in fallen trees (caused by residential development and windstorm) on lake food webs. The focus on an aquatic ecosystem is not surprising as aquatic ecosystems generally receive higher cross-ecosystem flows caused by the difference in terrain profiles. Their concave terrain profile makes them spatial attractors of many flows from terrestrial ecosystems due to gravity, unlike the terrestrial ecosystems that exhibit convex terrain profiles. Consequently, the effects of terrestrial flows on aquatic ecosystems have been a major theme in ecosystem ecology for decades. However, empirical studies have shown that physically and biologically mediated flows from aquatic ecosystems overcome gravity to affect terrestrial food webs. Biologically mediated flows from aquatic to terrestrial ecosystems involve animal movements. Specifically, such flows include the emergence of aquatic insects and the distribution of fish-derived energy by foraging terrestrial predators in the terrestrial ecosystem. Physically mediated flows from the aquatic ecosystem involve the movements of nutrients, sediments and detritus through floods and sub-surface water flows into terrestrial ecosystems. Also, the relative importance of physically and biologically mediated pathways in the propagation of anthropogenic stressors from aquatic to terrestrial ecosystems remains unclear. For example, quantitative comparisons between biologically and physically mediated pathways of the pollutant transport from aquatic to terrestrial ecosystems and associated effects on food webs are lacking.
Therefore, process-based food web modeling studies (based on empirical evidence) focusing on flows and anthropogenic stressors propagating from aquatic to terrestrial ecosystems could help in understanding the relative effects of biologically and physically mediated pathways on terrestrial food webs through model simulations. They could also provide clarity on how aquatic-to-terrestrial anthropogenic stressors and effects on food webs change across various temporal and spatial scales. Moreover, such studies will enable freshwater scientists to convey the relevance of their work to terrestrial managers.
Anthropogenic stressors affect the time, space and quality characteristics of aquatic-terrestrial flows
The reviewed studies focused on how anthropogenic stressors affect the quantity of aquatic-terrestrial flows. However, anthropogenic stressors can also shift the timing, quality and spatial characteristics of such flows. This consequently affects the recipient’s food web. For example, rising temperature can lead to early emergence and faster development of merolimnic insects and conceivably lead to temporal mismatches with the dietary need of terrestrial consumers. Similarly, climate change can shift the distribution of species poleward (e.g., riparian vegetation), conceivably affecting recipient aquatic organisms (e.g., stream detritivores) through spatial mismatch. In terms of resource quality, climate warming of 2.5°◦C can reduce fatty acids (8.2% to 27.8%) in algae, which may indirectly affect terrestrial consumers via an altered nutritional quality of the aquatic subsidy in terms of merolimnic insects. Consequently, future process-based food web modeling studies should also focus on how anthropogenic stressors affect the above-mentioned characteristics of aquatic-terrestrial flows.
Publicly sharing code increases the reproducibility and transparency of scientific studies. Scientific journals are adopting guidelines that require scientists to publish their codes. This is crucial for models, because it avoids the inefficiency of “reinventing the wheel” and encourages their reusability. Recently, Culina et al. (2020) found that only 27% of eligible ecology publications shared their code, and they called for increased code availability. It matches approximately our case, as only 11 out of 27 models (approximately 41%) had their models/model codes publicly available. Hence, we second their call for an increase in code availability of process-based food web models for meta-ecosystems.
The study was authored by Stephen E. Osakpolor, Mira Kattwinkel, Jens Schirmel, Alexander Feckler, Alessandro Manfrin, and Ralf B. Schäfer here at the Institute for Environmental Sciences at the University of Landau.