The ecological forecast horizon, and examples of its uses and determinants

Abstract

Forecasts of ecological dynamics in changing environments are increasingly important, and are available for a plethora of variables, such as species abundance and distribution, community structure and ecosystem processes. There is, however, a general absence of knowledge about how far into the future, or other dimensions (space, temperature, phylogenetic distance), useful ecological forecasts can be made, and about how features of ecological systems relate to these distances. The ecological forecast horizon is the dimensional distance for which useful forecasts can be made. Five case studies illustrate the influence of various sources of uncertainty (e.g. parameter uncertainty, environmental variation, demographic stochasticity and evolution), level of ecological organisation (e.g. population or community), and organismal properties (e.g. body size or number of trophic links) on temporal, spatial and phylogenetic forecast horizons. Insights from these case studies demonstrate that the ecological forecast horizon is a flexible and powerful tool for researching and communicating ecological predictability. It also has potential for motivating and guiding agenda setting for ecological forecasting research and development.

A comparison of macroecological and stacked species distribution models to predict future global terrestrial vertebrate richness

“…applied the same approach to S-SDMs of birds in the New World and found that model algorithm, followed by GCM and their interaction, accounted for the largest amounts of variance in the forecasts. In addition to these sources of variance, the spatial and temporal scale as well as the taxonomic level considered influence the forecast horizon(Harris et al, 2018;Petchey et al, 2015). Their and our findings suggest that considering different modelling approaches and algorithms is often more important than accounting for the variation resulting from different GCMs.…”

“…This conclusion is particularly important if time or computational resources are limited. In addition to these sources of variance, the spatial and temporal scale as well as the taxonomic level considered influence the forecast horizon (Harris et al, 2018;Petchey et al, 2015). Thus, future studies should assess whether our findings still hold when looking at different spatial and temporal resolutions and different taxonomic levels.…”

Predicting species abundances in a grassland biodiversity experiment: Trade‐offs between model complexity and generality

“…Intervals, axis labels, and indices are as described in the legend to Figure 1. See Table S5b abundances were strongly regulated by density dependence and temporal autocorrelation, such that observed species abundances were largely explained by their abundance in the previous time step (Petchey et al, 2015). Note, however, that intraspecific density dependence and autocorrelation alone were not sufficient to guarantee accurate out-of-sample predictions.…”

Community‐ and ecosystem‐level effects of multiple environmental change drivers: Beyond null model testing

“…One useful ingredient of the existing null model approach is the fact that it makes clear predictions. Assessing the capacity and limitations of predictive capacity is indeed a central goal in ecology and environmental science (Houlahan, McKinney, Anderson, & McGill, 2017;Petchey et al, 2015). Another example of how the proposed theory could deviate from observed joint effects is through the influence of time.…”