Multiple-scale resource selection of an undescribed urban invertebrate (Onychophora: Peripatopsidae) in Dunedin, New Zealand

Abstract

Urbanisation is expanding at an unprecedented rate, but ecological studies within human dominated environments are in the minority. Furthermore, within the field of urban ecology there is an insufficient focus on invertebrates. Among the threatened invertebrates found in cities are Onychophora, a “vulnerable” phylum with restricted habitat ranges. Habitat disturbance is the primary threat to Onychophora, yet very little is known about the ecology of the phylum internationally, and even less is known about their response to urbanisation. Urban development has resulted in population declines of Onychophora, and possibly species extinctions. However, onychophoran populations have been found in cities, and while this is noteworthy, their urban habitats have been only qualitatively described. A comprehensive, quantitative understanding of the urban features that promote the presence of Onychophora may become important for managing these vulnerable invertebrates.

Identification of important ecological variables is a priority for understanding conservation needs. In this study the habitat selection of an undescribed species of Onychophora was investigated in the city of Dunedin, New Zealand between the months of April and July, 2012. City reserves and forest fragments, as well as the surrounding urban matrix, were sampled to evaluate habitat selection. Resource selection functions were developed at multiple scales using an information theoretic approach. The role of a variety of potential predictor variables was modelled using logistic regression (micro-and macro-scales) and maximum entropy (landscape-scale).

Resource selection modeling revealed twelve key environmental variables within forest and park habitat patches, and the surrounding urban matrix. Important variables within patches were: the presence of tree fuchsia (Fuchsia excorticata), the presence of small arachnids, an abundance of long and wide decaying cover objects, and moist shady conditions. At the landscape scale the important variables were: forest and park perimeter-toarea ratios, residential garden conditions, and the presence of amenity habitat. The importance and implications of each variable are discussed in the context of Onychophora ecology, and urban conservation.

The results emphasise the need to avoid broad generalisations about Onychophora habitat selection, the need to manage onychophoran populations based on the habitat requirements of individual species, and the importance of habitats that are typically assumed to be inadequate. The research also underscores the importance of the matrix in modeling urban habitat selection, and management of urban Onychophora. City parks and urban gardens comprise a continuum of overlapping resources, necessitating communication and collaboration between municipal management and private landowners.