Alastair Pickering

1. Passive acoustic monitoring (PAM) is increasingly used for ecological research, biodiversity monitoring, assessment, and reporting. Automated species classifiers make it feasible to process large audio datasets but generate numerous detections that often need validation before use in downstream analyses or formal outputs. 2. Method development in PAM has focused on classifier building and downstream models that account for imperfect detection, yet the practical step between these - post-classification validation - remains weakly supported and is often implemented through ad hoc workflows. This increases manual handling, creates scope for transcription or consolidation errors, limits transparency and makes it difficult to document what was reviewed. 3. We introduce PAMalytics, an open-source, no-code, local browser-based application to support post-classification validation as a standardised workflow stage. PAMalytics ingests detections from any classifier, allows users to define how detections are sampled for review, and presents selected detections alongside their spectrograms with audio playback in one unified interface. Sampling strategy and review decisions are tracked alongside reviewer identity improving traceability and reproducibility across the validation workflow. 4. Case studies with Conservation International Cambodia and Imperial College London demonstrate PAMalytics in two validation settings. In Cambodia, gibbon predictions from a large, uneven dataset were sampled within sites, with likely classifier errors prioritised for validation. At Imperial, Amazon bird detections were sampled across each species' classifier-confidence range before biodiversity metrics were derived. In both cases, PAMalytics reduced manual handling and validation time. By turning an ad hoc step into an accessible, structured workflow for conservation practitioners, PAMalytics fills a practical gap in the PAM bioacoustics pipeline and strengthens the link between automated detections and evidence used in biodiversity monitoring and reporting.

Climate change is expected to alter forest community composition and functioning, with consequences for the ecosystem services forests provide. However, most macroecological projections focus on individual species distributions and offer limited insight into whether entire communities will remain functionally compatible with future climatic conditions. Here we quantify the risk that present-day forest communities will become functionally misaligned with projected climates using a trait-based approach. We analysed forest inventory data from more than 42,000 mature plots across the United States and Canada. For each plot we estimated community-weighted means for 24 functional traits describing leaf economics, hydraulic function, wood structure, abiotic tolerances and symbiotic strategies. We modelled relationships between community functional composition and environmental conditions, and used these relationships to estimate the trait profiles most compatible with projected late-century climates (2080–2100). Trait–environment misalignment (TEM) risk was quantified as the multivariate distance between current community trait composition and the trait profile associated with the projected future climate at each location, accounting for covariance among traits and intraspecific trait variation. Projected climatic conditions favour trait combinations associated with greater hydraulic capacity and reduced cold and shade tolerance. However, the magnitude of functional misalignment varies strongly across space. The highest TEM risk occurs in high-latitude and montane conifer forests across western and central North America, whereas many mid-latitude broadleaf and mixed forests show lower risk because projected climatic changes reinforce existing drought-adapted functional strategies. Critically, high species richness was the strongest predictor of reduced risk, reinforcing the importance of biodiversity in buffering against adverse outcomes. Our results suggest that many forests are projected to experience climatic conditions associated with functional strategies that differ from those characterising the current community. By identifying where the largest functional adjustments are implied, this trait-based framework provides a scalable way to pinpoint forests most likely to experience suboptimal climate conditions and to prioritise monitoring and climate-adapted management.

Community-based conservation (CBC) initiatives aim to reconcile biodiversity protection with local livelihoods, yet their effectiveness in protecting wildlife remains uncertain, often hinging on local management ^1,2 . We evaluated a globally significant CBC model in Kenya’s Greater Maasai Mara Ecosystem (GME), where conservancies, run jointly by Maasai landowners and the tourism sector, employ rotational cattle grazing to support both wildlife and pastoralism ^3,4 . Using a ∼1200 km ² grid of 180 camera traps across gradients of livestock pressure in Maasai Mara National Reserve and three conservancies in 2018, we collected and analysed over 2 million images with a customised AI-powered pipeline. We found a positive impact of observed cattle pressure on mammal community occupancy and species richness, except for at the highest levels of cattle grazing. However, sheep and goat grazing and proximity to infrastructure had a negative impact. These results provide evidence that wildlife and pastoralism can coexist under community-led stewardship ⁵ , but only with active management and targeted control of emerging threats. AI tools such as our image classifier may contribute to more adaptive community-led management of these areas ⁶ . As conservation policy shifts beyond formal protected areas, our findings support CBC as a scalable model for conserving biodiversity within working landscapes, offering a pathway to meet global targets while maintaining local livelihoods ⁷ .

Animal vocalizations encode rich biological information—such as age, sex, behavioural context and emotional state—making bioacoustic analysis a promising non‐invasive method for assessing welfare and population demography. However, traditional bioacoustic approaches, which rely on manually defined acoustic features, are time‐consuming, require specialized expertise and may introduce subjective bias. These constraints reduce the feasibility of analysing increasingly large datasets generated by passive acoustic monitoring (PAM). Transfer learning with Convolutional Neural Networks (CNNs) offers a scalable alternative by enabling automatic acoustic feature extraction without predefined criteria. Here, we applied four pre‐trained CNNs—two general purpose models (VGGish and YAMNet) and two avian bioacoustic models (Perch and BirdNET)—to African forest elephant ( Loxodonta cyclotis ) recordings. We used a dimensionality reduction algorithm (UMAP) to represent the extracted acoustic features in two dimensions and evaluated these representations across three key tasks: (1) call‐type classification (rumble, roar and trumpet), (2) rumble sub‐type identification and (3) behavioural and demographic analysis. A Random Forest classifier trained on these features achieved near‐perfect accuracy for rumbles, with Perch attaining the highest average accuracy (0.85) across all call types. Clustering the reduced features identified biologically meaningful rumble sub‐types—such as adult female calls linked to logistics—and provided clearer groupings than manual classification. Statistical analyses showed that factors including age and behavioural context significantly influenced call variation ( P < 0.001), with additional comparisons revealing clear differences among contexts (e.g. nursing, competition, separation), sexes and multiple age classes. Perch and BirdNET consistently outperformed general purpose models when dealing with complex or ambiguous calls. These findings demonstrate that transfer learning enables scalable, reproducible bioacoustic workflows capable of detecting biologically meaningful acoustic variation. Integrating this approach into PAM pipelines can enhance the non‐invasive assessment of population dynamics, behaviour and welfare in acoustically active species.