Ecology Letters最新文献

Plants can ‘cry for help’ in response to herbivory as well as anticipate herbivory by detecting specific environmental cues before damage occurs. But can plants ‘anticipate help?’ Building on the optimal defence and information transfer models of induced plant defence, we argue they can. We find literature support for key assumptions of the ‘anticipating help’ hypothesis, which proposes plants can (1) detect cues that signal reliable protection from enemies of herbivores (bodyguards), and (2) downregulate direct anti-herbivore resistance when bodyguards compensate. In an original a priori test of the assumptions of cue detection and downregulation of direct resistance, we use a meta-analysis of sequential herbivory experiments. We found that plants express induced susceptibility (dampened direct resistance) towards leaf-chewing herbivores only after induction by myrmecophilous sap-feeding herbivores, a putative cue for reliable ant-mediated protection against chewing herbivores. More generally, we expect ‘anticipating help’ behaviour in plants when local environmental cues predict reliable anti-herbivore protection from bodyguards that compensate for dampened direct resistance at a reduced fitness cost. The ‘anticipating help’ hypothesis can explain several enigmatic issues, such as induced susceptibility, associational resistance of plants, and how indirect resistance may benefit plant fitness under a wider range of conditions than previously recognised.

Benthic–pelagic coupling, the reciprocal exchange of materials between benthic and pelagic habitats, has traditionally emphasised pelagic influences on benthic systems. Yet, the role of benthic biological processes in shaping pelagic microbial dynamics remains underexplored. We investigated how surfgrass and mussels regulate nitrogen and dissolved organic matter (DOM) cycling and their cascading effects on heterotrophic bacteria in Oregon tide pools. We quantified biogeochemical fluxes and bacterial responses before and after foundation species removal during contrasting upwelling regimes. Mussel-dominated pools released high concentrations of ammonium and nitrate, while surfgrass pools transformed DOM that fueled bacterial growth; upwelling intensified these benthic–pelagic linkages. Removing foundation species dampened nutrient release in mussel pools and reduced DOM-fueled bacterial growth in surfgrass pools, ultimately decoupling benthic productivity from pelagic microbial growth. Our results demonstrate the critical role of foundation species to pelagic microbial processes and underscore the vulnerability of coastal microbial dynamics to their global decline.

The routine generation of species distribution data at scale remains a challenge. We used aquatic environmental DNA metabarcoding to sample vertebrate species across the 30,000 km² Gaoligongshan region along the China–Myanmar border. In just 56 calendar days (33 researcher-field-days + 69 researcher-lab-days), we detected 389 vertebrate species, of which 35 are Red-Listed. We introduce the ‘eDNA-aware’ OccPlus occupancy model, which accounts for false-negative and false-positive error in the field and lab. OccPlus leverages the taxonomic breadth of eDNA datasets by using ordination to estimate species occupancies. We recover known biogeographic patterns and find that native terrestrial and fish species have higher occupancies inside protected areas while domesticated species and non-native fishes have higher occupancies outside them. Our study demonstrates how eDNA metabarcoding can obtain high-quality, granular, timely, trustworthy and efficient species distribution data to facilitate nature conservation and restoration.

Assessing the physiological costs of sociality remains challenging due to complex interactions between environmental and social factors. To overcome this challenge, we integrated game theory with empirical research to examine how interspecific competition affects energy expenditure in the cooperatively breeding Asian burying beetle Nicrophorus nepalensis. Our nested tug-of-war model made two predictions: beetles facing interspecific competition would experience (1) lower physiological costs due to reduced investment in intragroup conflict, despite increased cooperative effort; and (2) decreasing physiological cost differences between dominance ranks. Using thermal imaging, we analysed body temperature as a proximal indicator of thermogenic effort during carcass work. Results supported our model's two key predictions: beetles competing with blowflies exhibited lower body temperatures, with less pronounced temperature differences between social ranks. Thus, reduced social conflict outweighs the energetic costs of increased cooperation in the face of external threats, providing insights into how environmental conditions shape energy allocation in social species.

Humans interact with wild animals through lethal and non-lethal activities. While theory predicts that these interactions should alter animal behaviour, the relative magnitude of impact is not well understood. We conducted a systematic review and meta-analysis of the past 30 years of research, focusing on changes in foraging, vigilance and movement behaviours in wild animals. We found that lethal human activities (e.g., hunting) cause significant behavioural changes, with targeted species increasing vigilance and reducing foraging in affected areas. Active non-lethal activities (e.g., tourism) elicited similar but weaker patterns, with many species showing little to no change in their behavioural responses. In contrast, passive non-lethal interactions (e.g., roads) produced highly variable responses. Overall, human-induced fear elicits responses in wild animals that broadly align with predictions from the risk allocation hypothesis. However, the magnitude and direction of animal responses depend on the type of human activity and the ecological context. The most pronounced behavioural changes occur where humans pose a direct lethal threat. Gaps in the literature, uneven data within and across species, and limited information on the history or context of interactions currently limit our ability to better predict when and why animals change their behaviour in response to humans.

Understanding animal space use is central to ecology and conservation, and movement-based habitat selection models provide powerful tools for identifying preferred environments. However, in heterogeneous landscapes, movement constraints limit habitat accessibility, decoupling realised space use from underlying preferences and complicating efforts to scale individual-level movement processes to landscape distributions. We present a steady-state distribution (SSD) framework that integrates locally estimated habitat selection and movement into sparse Markovian transition matrices to predict emergent space use. Across simulations and empirical case studies, SSD consistently outperformed occurrence-based models in predicting space use from individual- to population-level scales. Moreover, SSD proved to be more efficient—and sometimes even more accurate—than more assumptive agent-based model simulations (ABMs). Our approach bridges a key gap between mechanistic movement models and spatial prediction, providing a scalable and computationally efficient approach for translating individual movement behaviour into landscape-scale distribution patterns relevant for ecological inference and conservation planning.