Departing from an ideal: An asymmetric, bimodal and non-Equatorial latitudinal gradient of marine diversity in Western Atlantic burrowing shrimps (Decapoda: Axiidea and Gebiidea)
Data files
Sep 13, 2021 version files 36.55 KB
Abstract
Aim: Despite the generality of the latitudinal gradient of species diversity (LDG) phenomenon, there is growing evidence showing deviations from an idealized pattern, i.e., a single peak of species richness symmetrically centered in the Equator, but the underlying causes remains little studied. We here evaluate the existence of departures from the idealized LDG in a group of marine crustaceans and the explanatory role of environmental variables.
Location: Coastal shelf (< 200 m depth) along the Western Atlantic coast (46ºN to 47ºS)
Taxon: Burrowing shrimps (Decapoda: Axiidea and Gebiidea).
Methods: We assessed the shape of the LDG in 100 burrowing shrimp species using the reported latitudinal ranges of distribution. Species richness was calculated in one degree-latitude bands using a range-through approach. The shape of the LDG was statistically evaluated in terms of latitudinal symmetry, number of modes and location. We evaluated the importance of 10 environmental variables (proxies of different hypotheses categories) predicting LDG using a random forest model.
Results: Burrowing shrimps exhibit an increase in diversity towards the tropics, but departures to the idealized LDG were evident. The LDG is asymmetric between hemispheres and bimodal (two peaks within the tropics), and this trend cannot be explained in terms of sampling artifacts. A random forest model explains 92% of species richness, but the only significant variables were bottom seawater temperature, bottom seawater salinity, bottom seawater temperature range, and tidal range. These predictors were non-linearly related to species richness, and only bottom seawater temperature, bottom seawater salinity showed a significant phylogenetic signal.
Main conclusions: The LDG of burrowing shrimps is driven ecophysiological restrictions of species, reflecting the role of evolutionary (i.e., ‘time for species accumulation’ and/or ‘diversification dynamics’) and ecological processes (i.e., ‘ecological limits’). Departures from the idealized LDG could be explained by the non-linear responses of species richness to environmental conditions, the spatial structure of these environmental conditions, and a varying degree of phylogenetic conservatism of key ecophysiological constraints.
Methods
We gathered a database documenting the northernmost and southernmost limit of the distribution of 100 burrowing shrimp species inhabiting the coastal shelf (< 200 m depth) along the WA coast, from Canada (46ºN) to Patagonia (47ºS). This represents ca. 15 % and 83 % of all valid species of Axiidea and Gebiidea described for the world and western Atlantic, respectively (World Register of Marine Species, www.marinespecies.org, accessed on July 17th 2020). The information was collected from an exhaustive literature review and fieldwork carried out from 2007 to present along the coast of Central America and South America (Hernáez, 2014, 2018a; Hernáez, Gamboa-González, & De Grave, 2015; Hernaez, Miranda, & Tavares, 2020; Hernáez & Vargas, 2013; Hernáez, Villegas-Jiménez, Villalobos-Rojas, & Wehrtmann, 2012; Hernáez & Wehrtmann, 2007; Hernáez, Windsor, Paula, & Santana, 2020). The taxonomic status of each species was validated using the World Register of Marine Species. The reported latitudinal ranges are provided in the Dataset S1.
We evaluated the importance of 10 environmental variables predicting species richness (Table 1). The data was obtained from BioOracle 2.0 (Assis et al., 2018), GMED (http://gmed.auckland.ac.nz) (0.05º resolution each), and Aquamaps database v2.0 (Kaschner et al., 2008) (0.5º resolution). Values were averaged over one-degree latitudinal bins, using only pixels on the coastal shelf (<200 m depth). The information of environmental variables is provided in the Dataset S2.
See references in the article.
Usage notes
Check the README file and annotated instructions in the Rscript to reproduce all analyses presented in the article.