Niche modelling identifies low rainfall, but not soil type, as an important habitat requirement of the fossorial Australasian trapdoor spider genus Cantuaria (Hogg, 1902)

• Published in: Austral ecology Vol. 46; no. 7; pp. 1070 - 1083
• Main Authors: Smith, Victoria R., Vink, Cor J., Fountain, Emily D., Cruickshank, Robert H., Paterson, Adrian M.
• Format: Journal Article
• Language: English
• Published: Richmond Blackwell Publishing Ltd 01.11.2021
• Subjects: Climate change; Climatic conditions; Conservation; Data analysis; Data collection; Datasets; Gene flow; Habitat loss; Habitat selection; Habitats; idiopidae; Modelling; niche modelling; Niches; Populations; Rainfall; Random variables; Soil conditions; Soils; Spatial analysis; Spatial data; spider; Spiders; Vegetation; Vegetation type
• ISSN: 1442-9985; 1442-9993
• DOI: 10.1111/aec.13039

Habitat loss is one of the greatest drivers for extinction worldwide. Understanding a taxon’s habitat requirements is crucial in being able to protect it from decline. The Australasian trapdoor spider genus Cantuaria (Idiopidae) is widespread and diverse within the islands of New Zealand, but little is known of its ecology. We studied the habitat requirements of Cantuaria, using niche modelling, to inform conservation efforts and predict how the genus may fare under future climate conditions. Two data sets were employed. A fine‐scale data set used hand‐collected data from the vicinity of individual burrows for localised soil, vegetation, and human disturbance data directly inside and adjacent to Cantuaria populations. A spatial GIS data set at the landscape level mapped presence/absence of populations, as well as general environmental and soil conditions, and vegetation type. The two data sets were analysed separately, using linear models for the fine‐scale data set (17 presences, 17 absences). General linear models were used to analyse the spatial data set (72 presences, 17 absences), with variables selected using a random forest analysis in R. In the fine‐scale data set, no significant variables were found. In the spatial data set, rainfall explained most (71.4%) of the variation, according to the random forest analysis, followed by soil group (33.71%) and elevation (27.12%). Some variation was explained by more than one factor. The two models with the lowest AICC numbers included rainfall (P < 0.01) and elevation (P = 0.06), or rainfall only (P < 0.01), as variables. We conclude that low rainfall is desirable for populations of most Cantuaria spp., although some have been found where rainfall is up to 3000 mm/year. Climate change may cause challenges for Cantuaria conservation, particularly in areas with increased rainfall, such as Southland, Tasman, the West Coast, and Central Otago. Elevation and soil type may also be important factors in Cantuaria habitat selection, but further research with gene flow analysis, captive studies, more structured sampling, and/or a greater sample size may reveal more about the effects of elevation and soil type on Cantuaria populations.