|Document Type:||Journal Article|
|Title:||Model-based inference for estimating distribution changes in marine fishes|
|Author:||James T. Thorson, M. L. Pinsky, Eric Ward|
|Journal:||Global Change Biology|
|Keywords:||climate change, species distribution model, range shifts, California Current, center of gravity, abundance weighted average, ,|
Changing climate is already impacting the spatial distribution of many taxa, including bees, plants, birds, butterflies, and fishes. A common approach to detecting range shifts in response to climate change is to first quantify changes in the center of the population’s distribution (the center of gravity, COG). However, a conventional “abundance-weighted average” (AWA) estimator of COG confounds species range shifts with changes in the spatial distribution of available survey data. The AWA estimator also does not estimate the standard error of COG in individual years, and therefore cannot be used to distinguish between range shifts and random error in estimating COG. We therefore propose an alternative species density function (SDF) model, which estimates shifts in species distribution over time with associated uncertainty while potentially integrating information from multiple surveys. We show that both the center of the SDF and its standard error are simple to compute, and illustrate the method by analyzing data from two surveys spanning 1977-2013 for 19 marine fishes along the U.S. West Coast. We find that the SDF estimator shows significant northward shifts for 6 of 19 species (with southward shifts for only 2). The SDF illustrates that pelagic species (e.g., Pacific hake and spiny dogfish) have greater interannual variability in COG than bottom-associated species. We also compare the AWA and SDF estimators when applied to data from a single survey ranging from 1977-2004, and find substantial differences between estimators that are likely caused by shifts in sampling distribution (which affect the AWA but not the SDF estimator). We caution that the common AWA estimator can yield inappropriate inference whenever sampling designs have shifted over time, and conclude by suggesting further improvements in model-based approaches to analyzing climate impacts.
|Theme:||Ecosystem approach to improve management of marine resources|
Understand how climate influences ecosystem variability.
Assess ecosystem status and trends.