Early warning indicators of population collapse in a seasonal environment

Abstract

Recent studies have demonstrated that generic statistical signals derived from time series of population abundance and fitness‐related traits of individuals can provide reliable indicators of impending shifts in population dynamics. However, how the seasonal timing of environmental stressors influences these early warning indicators is not well understood. The goal of this study was to experimentally assess whether the timing of stressors influences the production, detection and sensitivity of abundance‐ and trait‐based early warning indicators derived from declining populations. In a multi‐generation, season‐specific habitat loss experiment, we exposed replicate populations of Drosophila melanogaster to one of two rates of chronic habitat loss (10% or 20% per generation) in either the breeding or the non‐breeding period. We counted population abundance at the beginning of each season, and measured body mass and activity levels in a sample of individuals at the end of each generation. When habitat was lost during the breeding period, declining populations produced signals consistent with those documented in previous studies. Inclusion of trait‐based indicators generally improved the detection of impending population collapse. However, when habitat was lost during the non‐breeding period, the predictive capacity of these indicators was comparatively diminished. Our results have important implications for interpreting signals in the wild because they suggest that the production and detection of early warning indicators depends on the season in which stressors occur, and that this is likely related to the capacity of populations to respond numerically the following season.

Community structure determines the predictability of population collapse

“…In addition to this, if however, an environmental perturbation impacted phenotypic traits directly, than the potential for traits to be used as an indicator increases significantly (Baruah et al 2019). Such instances could be when there was a decline in food availability that could impact phenotypic traits such as body size, or a size-based harvesting regime that could also lead to a shift in body size (Clements and Ozgul 2016, Burant et al 2021). However, our environmental perturbation did not have direct impacts on mean traits directly, and hence observable change in mean phenotypic dynamics were negligible, leading to poor performance of trait-based signals measured at the species level.…”

“…Consequently, in order to improve forecasts of population collapse, warning signals that incorporate information from fitness-related traits (known as trait-based EWS), such as body size, have been proposed (Clements and Ozgul 2016, Clements et al 2017, Baruah et al 2019, 2021). Although such EWSs have been shown to forecast rapid shifts in a variety of ecological systems in response to continuous (Drake and Griffen 2010, Suweis and D’Odorico 2014) or seasonal environmental perturbation (Burant et al 2019, 2021), relatively few studies have evaluated the performance of EWS in predicting population level collapses in a multispecies context (Boerlijst et al 2013, Dakos and Bascompte 2014, Dakos 2017). Whilst it is known that EWS could be useful in forecasting community level transitions (Carpenter et al 2011, Spanbauer et al 2016), the impact of community structure and species interaction on the utility of EWS is relatively still unknown.…”

Timeline to collapse

“…This individual-to-population approach allows us to use individual stress responses as early indicators of change in population conditions, and to measure the impacts of the stressors in multiple dimensions simultaneously. Such an approach expands on recent work in the field of EWS, which consider abundance based EWS and shifts in the mean body size of the population simultaneously, leading to an increase of the overall predictive power (Clements et al 2017, Burant et al 2021. However, these approaches ignore the fact that such signals are not necessarily expected to change concurrently, 5 but rather may occur sequentially as individuals' plasticity buffer them against negative environmental conditions.…”

“…Moreover, even in cases of the assumed "ramped" disturbance, the mutable nature of biological systems may create situations where the sequence of signals may be different (e.g. body traits shift occurs first, triggering then behavioural shift (Burant et al, 2021). Fully applying the framework requires studying species that show plastic and quantifiable behaviours and morphological traits, where gathering data is easy at the individual level, and thus it may not be fully applicable to sessile organisms (e.g.…”

Community structure determines the predictability of population collapse

“…Consequently, to improve forecasts of population collapse, warning signals that incorporate information from fitness‐related traits (known as trait‐based EWS), such as body size, have been proposed (Baruah et al, 2019, 2021; Clements et al, 2017; Clements & Ozgul, 2016). Although such EWS have been shown to forecast rapid shifts in a variety of ecological systems in response to continuous (Drake & Griffen, 2010; Suweis & D'Odorico, 2014) or seasonal environmental perturbation (Burant et al, 2019, 2021), relatively few studies have evaluated the performance of EWS in predicting population‐level collapses in a multispecies context (Boerlijst et al, 2013; Dakos, 2017; Dakos & Bascompte, 2014). Although it is known that EWS could be useful in forecasting community‐level transitions (Carpenter et al, 2011; Spanbauer et al, 2016), the impact of community structure and species interaction on the utility of EWS is relatively still unknown.…”