Fine scale waterbody data improve prediction of waterbird occurrence despite coarse species data

Abstract

While modelling habitat suitability and species distribution, ecologists must deal with issues related to the spatial resolution of species occurrence and environmental data. Indeed, given that the spatial resolution of species and environmental datasets range from centimeters to hundreds of kilometers, it underlines the importance of choosing the optimal combination of resolutions to achieve the highest possible modelling prediction accuracy. We evaluated how the spatial resolution of land cover/waterbody datasets (meters to 1 km) affect waterbird habitat suitability models based on atlas data (grid cell of 12 × 11 km). We hypothesized that the area, perimeter and number of waterbodies computed from high resolution datasets would explain distributions of waterbirds better because coarse resolution datasets omit small waterbodies affecting species occurrence. Specifically, we investigated which spatial resolution of waterbodies better explain the distribution of seven waterbirds nesting on ponds/lakes with areas ranging from 0.1 ha to hundreds of hectares. Our results show that the area and perimeter of waterbodies derived from high resolution datasets (raster data with 30 m resolution, vector data corresponding with map scale 1:10 000) explain the distribution of the waterbirds better than those calculated using less accurate datasets despite the coarse grain of the species data. Taking into account the spatial extent (global vs regional) of the datasets, we found the Global Inland Waterbody Dataset to be the most suitable for modelling distribution of waterbirds. In general, we recommend using land cover data of a resolution sufficient to capture the smallest patches of the habitat suitable for a given species' presence for both fine and coarse grain habitat suitability and distribution modelling.

Vegetation structure derived from airborne laser scanning to assess species distribution and habitat suitability: The way forward

“…We propose 10–20 m as the optimal resolution for calculating these variables, which is a compromise between the grains usually used in studies on species distributions and habitat suitability and the typical density of national point clouds. For example, SDM studies typically use an analysis grain (i.e., the spatial unit in which the species occurrence is modelled) between 10 m and 10 km (Mertes & Jetz, 2018; Moudrý & Šímová, 2012); however, environmental data should be available at an even more detailed resolution, sufficient to capture the smallest habitat patches suitable for a given species (Gottschalk et al, 2011; Koma, Seijmonsbergen, et al, 2021; Šímová et al, 2019). In addition, the resolution of the proposed standard structural variables needs to reflect the density of point clouds.…”

Habitats as predictors in species distribution models: Shall we use continuous or binary data?

“…In contrast, atlas squares with binary presence of water almost always contain a substantial area of water bodies, possibly enough to support a persistent breeding population of a waterbird species, leading to the good performance of the binary water predictor. In line with this, Šímová et al (2019) showed that the area of water bodies derived from high‐resolution (30 m) datasets explain distributions of waterbirds better than predictors derived from coarser water datasets (including CORINE Land Cover). This may be a reason why Tuanmu and Jetz (2014) found the Global Consensus Land cover (1 km resolution) performed worse for water species than for species that from other environments.…”

“…In addition, we propose that the relative merit of continuous versus binary predictors depends on the interplay between spatial resolution of the habitat data (Domisch et al 2015, Friedrichs‐Manthey et al 2020), spatial grain at which habitats are aggregated for modelling (response grain; Seoane et al 2004, Venier et al 2004, Convertino et al 2011, Moudrý and Šímová 2012, Tuanmu and Jetz 2014, Šímová et al 2019), as well as the home range size of the species, and its degree of specialisation to the habitat (Jedlikowski et al 2016, Mertes et al 2020). For highly specialised species, the ratio between the home range size and the grain size of the response variable may be particularly important (Jedlikowski et al 2016), as it determines whether the species can gather resources from multiple grid cells, or whether it is confined to a single cell.…”

Scale mismatches between predictor and response variables in species distribution modelling: A review of practices for appropriate grain selection

“…Several studies have examined the importance of fine-grain habitat features for the analysis of speciesenvironment relationships using a relatively coarse-grained response variable (Figure 1(c); Table 3). In this type of study, authors typically use predictor variables of various origins, collected, for example, by remote sensing (Leitão and Santos 2019), fieldwork, or crowd-sourcing ( Šímová et al, 2019;Thomas et al, 2002;Venier et al, 2004). Others have coarsened the grain of the original predictors to examine the grain-dependency of species-environment relationships (e.g.…”

“…Studies that manipulate the resolution of predictor variables, so that the resulting predictor was finer than the response variable (Figure 1(c); Table 3), are mostly concerned with the importance of fine-scale habitat features for analyzing species–environment relationships (e.g. Gottschalk et al, 2011; Šímová et al, 2019). They combine response variables at a coarse resolution with predictor variables at a fine resolution.…”