Biased baselines? Differences in manual and computer-based modelling of blue-winged teal (Spatula discors) habitat suitability

Abstract

Assessing wildlife habitat suitability has been an important tool in species conservation for generations of wildlife managers, despite changes in technology and theoretical frameworks. The goal of this study was to quantify differences in habitat suitability indices (HSIs) for blue-winged teal derived from our re-creation of traditional paper-based outputs and GIS-based outputs using the same inputs and to determine how HSIs for blue-winged teal would respond to fine-scale applications in Miquelon Lake Provincial Park, Canada. The subsequent models assessed the four original variables presented by the 1985 USFWS model: wetland density, wetland area, horizontal cover, and proximity to wetlands for three components of breeding habitat including pair bonding, nesting, and brood rearing. For a subset of GIS models, vertical cover was also incorporated into our analyses to determine its applicability in the model’s assessment of aerial predation risk for blue-winged teal. The paper-based model consistently produced higher habitat suitability scores when compared to the GIS model, with only the brood-rearing model being slightly lower. When vertical cover was included in the GIS model either by itself or combined with horizontal cover, the area of high suitability habitat was reduced to 0% and ∼3%, respectively. Results revealed that the species did not show a special preference to high or moderately suitable habitats. Given the overestimation of habitat suitability at all breeding stages for blue-winged teal, the comparative utility of the actual HSI scores derived manually from such models are limited; however, the long-term ecological insights into model inputs are invaluable when integrated with emerging spatial data for computer-based habitat modelling.