Seasonal dynamism or model realism: What drives better predictions of landscape connectivity?
Ecosphere · 2026-09 · 17 (9)
DOI: 10.1002/ecs2.70726 ↗Kurz gefasst
Fachartikel in der Zeitschrift Ecosphere, online erschienen am 15.09.2026. Verfasst von David D. Hofmann und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 99 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung.
Abstract (Originalsprache)
Connectivity modeling is central to conservation science, yet most studies assume static landscapes and fixed species‐habitat preferences, thus overlooking seasonal dynamics. Additionally, many rely on simplistic species‐habitat associations rather than realistic movement models that mechanistically capture animal behavior. These limitations risk masking key ecological processes, leading to poor agreement between predicted and observed movements. We propose and employ a conceptual framework illustrating that seasonality can be incorporated at three stages of the connectivity modeling pipeline: (1) when extracting spatial covariates for model fitting, (2) when specifying the selection model, and (3) when making predictions. This yields several model configurations varying in seasonal dynamism. Further, we compare the performance of alternative models that vary in their degree of biological realism, ranging from simplistic habitat‐based models to realistic formulations that consider both environmental and behavioral drivers of movement. Using GPS data from dispersing African wild dogs ( Lycaon pictus ) in the seasonally dynamic Okavango Delta, northern Botswana, we assess whether accounting for seasonality improves predictions of dispersal trajectories and connectivity. We fit integrated step‐selection functions and predict connectivity using individual‐based dispersal simulations while explicitly considering seasonal dynamism in both environmental covariates and the species' habitat and movement preferences. Furthermore, predictions from a simplistic habitat‐based model are contrasted with those from a more mechanistic and biologically realistic model enabled by individual‐based simulations. We evaluate predictive performance through cross‐validation across seasonal configurations and levels of model realism. We hypothesized that incorporating seasonality and using a realistic movement model would improve predictions. Contrary to expectations, seasonality modestly improved predictions when using a simplistic model, but had no added benefit when using the realistic model, which already performed well. Our findings offer a framework for integrating seasonality into connectivity studies and highlight the greater value of realistic movement models over simplistic ones. While seasonal dynamics are important, their benefits may be limited when biologically informed models are already in place. Connectivity studies should therefore prioritize mechanistic approaches to better capture species movement and landscape use.
Förderung
- Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
Zitieren
APA
Hofmann, D. D., Behr, D. M., McNutt, J. W., Ozgul, A., & Cozzi, G. (2026). Seasonal dynamism or model realism: What drives better predictions of landscape connectivity?. Ecosphere, 17(9). https://doi.org/10.1002/ecs2.70726
Vancouver
Hofmann DD, Behr DM, McNutt JW, Ozgul A, Cozzi G. Seasonal dynamism or model realism: What drives better predictions of landscape connectivity?. Ecosphere. 2026;17(9). doi:10.1002/ecs2.70726
BibTeX
@article{Hofmann2026,
author = {Hofmann, David D. and Behr, Dominik M. and McNutt, John W. and Ozgul, Arpat and Cozzi, Gabriele},
title = {Seasonal dynamism or model realism: What drives better predictions of landscape connectivity?},
journal = {Ecosphere},
year = {2026},
volume = {17},
number = {9},
doi = {10.1002/ecs2.70726}
}
EUVIMED