Our second issue of the year is now online!
Senior Editor Lee Hsiang Liow has selected five featured articles, find out all about them below. We also have one article from the Special Feature on Citizen Science, a joint venture across the British Ecological Society journals which held an open call for papers. Read all about the Special Feature in this editorial.
Citizen Science Special Feature
Performance of species distribution models with citizen science data (free access) To explore the consequences of spatial bias and class imbalance in presence–absence citizen science data, Steen at al. used eBird citizen science data for 102 bird species from the northeastern USA to compare spatial thinning, class balancing and majority‐only thinning. They created species distribution models (SDMs) using two parametric or semi‐parametric techniques and two machine learning techniques. When testing the predictive abilities of the SDMs using an independent and systematically collected reference dataset, they found large variation in performance depending on thinning and balancing decisions.
Methods for studying biotic interactions in phenological analyses Phenological events play a key role in modulating ecosystem services, but the complex and interlinked nature of ecosystems indicates that interactions among different taxa during phenological events can have consequences for the entire ecosystem. Currently, there is a lack of a unified criteria on the methodologies studying phenology and biotic interactions. In answer to this, de la Torre Cerro & Holloway perform an extensive integrative review of works evaluating phenology and biotic interactions.
Leveraging spatial information to forecast nonlinear ecological dynamics Empirical dynamic modelling, an equation‐free nonlinear forecasting method, is receiving growing attention, but it requires long time series to produce accurate predictions. Though most ecological time series are short, spatial replicates are often available. Here, Johnson et al. explore how utilising available spatial data can improve our ability to forecast ecological dynamics.
Unsupervised acoustic classification of individual gibbons Advances in automated signal detection have increased the scope of passive acoustic monitoring, but distinguishing between individual animals—which is necessary for density estimation—remains a major challenge. Here, Clink & Klink use an existing dataset of Bornean gibbon female calls with known identity to test the ability of three different unsupervised clustering algorithms to distinguish between individuals. They conclude that unsupervised techniques may be useful for providing additional information regarding individual identity for passive acoustic monitoring applications.
Quantifying the impact of an inference model in Bayesian phylogenetics (open access) Bilderbeek et al. present pirouette, a free and open‐source R package that assesses the inference error made by Bayesian phylogenetics for a given macroevolutionary diversification model. pirouette makes use of BEAST2, but its philosophy applies to any Bayesian phylogenetic inference tool. pirouette’s usage is described, providing full examples in which they interrogate a model for its power to describe another.
Analyzing individual history through size-classified matrix population models (free access) Matrix projection models (MPMs) are used to analyse population dynamics, but are not structured to incorporate the influences of individual histories. Historical MPMs (hMPM) were developed to incorporate these impacts, but their complexity has left them little used. Here, Shefferson et al. present lefko3, an R package that provides simple, quick methods to estimate and analyse hMPMs, as well as ahistorical MPMs. This dramatically reduces the difficulties in testing the impacts of individual history on population dynamics.
The Turtle on the Cover
This month’s cover image shows a female flatback sea turtle Natator depressus leaving a nesting beach, fitted with an accurate Fastloc‐GPS tag. In tracking studies it is often difficult to determine the sample sizes necessary to describe the spatial distribution of a population. In this issue, Shimada et al. examine existing approaches, highlight caveats and propose a new method to better assess appropriate sample sizes. Large datasets of movements of flatback turtles are analysed to illustrate this new method, which has broad applicability for the post‐hoc validation of sample sizes of tracking data across a wide range of taxa, populations and life‐history stages of animals. Code for the analysis is compiled in the open‐access R package SDLfilter. Photo Credit: © C.J. Limpus