16 September 2022 in Restoration Ecology
1 January 2022 in Journal of Environmental Management
Restoring stream ecosystem integrity by removing unused or derelict dams has become a priority for watershed conservation globally. However, efforts to restore connectivity are constrained by the availability of accurate dam inventories which often overlook smaller unmapped riverine dams. Here we develop and test a machine learning approach to identify unmapped dams using a combination of publicly available topographic and geospatial habitat data. Specifically, we trained a random forest classification algorithm to identify unmapped dams using digitally engineered predictor variables and known dam sites for validation. We applied our algorithm to two subbasins in the Hudson River watershed, USA, and quantified connectivity impacts, as well as evaluated a range of predictor sets to examine tradeoffs between classification accuracy and model parameterization effort. The random forest classifier achieved high accuracy in predicting dam sites (true positive rate = 89%, false positive rate = 1.2%) using a subset of variables related to stream slope and presence of upstream lentic habitats. Unmapped dams were prevalent throughout the two test watersheds. In fact, existing dam inventories underestimated the true number of dams by ∼80-94%. Accounting for previously unmapped dams resulted in a 62-90% decrease in dendritic connectivity indices for migratory fishes. Unmapped dams may be pervasive and can dramatically bias stream connectivity information. However, we find that machine learning approaches can provide an accurate and scalable means of identifying unmapped dams that can guide efforts to develop accurate dam inventories, thereby informing and empowering efforts to better manage them.
30 December 2024 in Proceedings of the National Academy of Sciences
CRClaudia Rudolf von Rohr
+6
20 August 2024 in Nature Climate Change
9 December 2025 in Nature Ecology & Evolution
Sodium (Na) is an essential nutrient for animals, but not for most plants. Consequently, herbivores may confront a mismatch between forage availability and metabolic requirement. Recent work suggests that larger-bodied mammals may be particularly susceptible to Na deficits, yet it is unknown whether Na availability constrains the density or distribution of large herbivores at broad scales. Here we show that plant-Na availability varies >1,000-fold across sub-Saharan Africa and helps explain continent-scale patterns of large-herbivore abundance. We combined field data with machine-learning approaches to generate high-resolution maps of plant Na, which revealed multi-scale gradients arising from sea-salt deposition, hydrology, soil chemistry and plant traits. Faecal Na concentration was positively correlated with modelled dietary Na, supporting the prediction that variation in plant Na is a major determinant of herbivore Na intake. Incorporating plant-Na availability improved model predictions of large-herbivore population density, especially for megaherbivore species, which are depressed in very-low-Na regions (<100 mg kg−1), consistent with Na limitation. Our study offers an explanation for the scarcity of megaherbivores in parts of Central and West Africa, which has major ecological ramifications given the strong influence of large herbivores on ecosystem functioning and the profound human-induced changes to Na availability in Africa and beyond.