Oeser, J. | bioRxiv Science

bioRxiv2024

The best of two worlds: using stacked generalization for integrating expert range maps in species distribution models

Species distribution models (SDMs) are powerful tools for assessing suitable habitat across large areas and at fine spatial resolution. Yet, the usefulness of SDMs for mapping species realized distributions is often limited, since data biases or missing information on dispersal barriers or biotic interactions hinder them from accurately delineating species range limits. One way to overcome this limitation is to integrate SDMs with expert range maps, which provide coarse-scale information on the extent of species ranges that is complementary to information offered by SDMs. Here, we propose a new approach for integrating expert range maps in SDMs based on an ensemble method called stacked generalization. Specifically, our approach relies on training a meta-learner regression model using predictions from one or more SDM algorithms alongside the distance of training points to expert-defined ranges as predictor variables. We demonstrate our approach with an occurrence dataset for 49 bat species covering four biodiversity hotspots in the Eastern Mediterranean, Western Asia, and Central Asia. Our approach offers a flexible method to integrate expert range maps with any combination of SDM modeling algorithms, thus facilitating the use of algorithm ensembles. In addition, it provides a novel, data-driven way to account for uncertainty in expert-defined ranges not requiring prior knowledge about their accuracy, which is often lacking. Our approach holds considerable promise for better understanding species distributions, and thus for biogeographical research and conservation planning. In addition, our work highlights the overlooked potential of stacked generalization as an ensemble method in species distribution modeling.

ecology↗

bioRxiv2022

Spatial variation in red deer density in relation to forest disturbance and ungulate management in a transboundary forest ecosystem

1. ContextForests in Europe are exposed to increasingly frequent and severe disturbances. The resulting changes in the structure and composition of forests can have profound consequences for the wildlife inhabiting them. Moreover, wildlife populations in Europe are often subjected to differential management regimes as they regularly extend across multiple national and administrative borders. The red deer (Cervus elaphus) population in the Bohemian Forest Ecosystem, straddling the Czech-German border, has experienced forest disturbances, primarily caused by windfalls and bark beetle (Ips typographus) outbreaks during the past decades. To adapt local management strategies to the changing environmental conditions and to coordinate them across the international border, reliable estimates of red deer density and abundance are highly sought-after by policymakers, wildlife managers, and stakeholders. 2. Approach and MethodsCovering a 1081-km2 study area, we conducted a transnational non-invasive DNA sampling study in 2018 that yielded 1578 genotyped DNA samples from 1120 individual red deer. Using spatial capture-recapture models, we estimated total and jurisdiction-specific abundance of red deer throughout the ecosystem and quantified the role of forest disturbance and differential management strategies in shaping spatial heterogeneity in red deer density. We hypothesized that (a) forest disturbances provide favourable habitat conditions (e.g., forage and cover), and (b) contrasting red deer management regimes in different jurisdictions create a differential risk landscape, ultimately shaping density distributions. 3. ResultsOverall, we estimated that 2851 red deer (95% Credible Intervals = 2609 - 3119) resided in the study area, with a relatively even overall sex ratio (1406 females, 1229 - 1612 and 1445 males, 1288 - 1626). The average red deer density was higher in Czechia (3.5 km-2, 1.2 - 12.3) compared to Germany (2 km-2, 0.2 - 11). The effect of forest disturbances on red deer density was context dependent. Forest disturbances had a positive effect on red deer density at higher elevations and a negative effect at lower elevations, which could be explained by partial migration and its drivers in this population. Density of red deer was generally higher in management units where hunting is prohibited. In addition, we found that sex ratios differed between administrative units and were more balanced in the non-intervention zones. 4. Synthesis and applicationsOur results show that the effect of forest disturbances on wild ungulates is modulated by additional factors, such as elevation and ungulate management practices. Overall density patterns and sex ratios suggested strong gradients in density between administrative units. With climate change increasing the severity and frequency of forest disturbances, population-level monitoring and management are becoming increasingly important especially for wide-ranging species as both wildlife and global change transcend administrative boundaries.

ecology↗