Skip to main content

Spider species collected in Macaronesian islands using COBRA sampling protocols

Cite this dataset

Malumbres-Olarte, Jagoba (2022). Spider species collected in Macaronesian islands using COBRA sampling protocols [Dataset]. Dryad.


Aim: Habitat diversity has been linked to the diversity and structure of island communities, however, little is known about patterns and processes within habitats. Here we aim to determine the contributions of habitat type and inferred dispersal frequency to the differences in taxonomic structure between assemblages in the same island habitat.

Location: The Macaronesian archipelagos (Azores, Madeira, the Canary Islands and Cabo Verde).

Taxon: Spiders (Araneae).

Methods: We established forest and dry habitat sites (each with five plots) on two islands per archipelago. We collected spiders using standardised sampling protocols. We tested the differences in beta diversity separately for each habitat and for each inferred category of ballooning (an aerial dispersal strategy) frequency across geographic scales through nested non-parametric PerMANOVA. We then tested whether ballooning and habitat influenced heterogeneity in species composition (dispersion in beta diversity) in the two habitat types. We analysed the effects of habitat and ballooning on species abundance distribution (SAD) and rarity by fitting Gambin models and evaluating the contribution of ballooning categories to SAD.

Results: Communities of the same archipelago and habitat were taxonomically more similar, and beta diversity increased with geographic scale, being greater in dry habitats. There was greater species replacement among assemblages in dry habitats than in forests, with greater differences for rare ballooners. There were no differences in SAD between habitats although dry habitat sites seemed to harbour more species with low abundances (rare species) than forests.

Main conclusions: Habitat type does not only condition the differences between spider assemblages of the same habitat but also the scale at which they occur. These differences may be determined by the heterogeneity in the physical structure of each habitat as well as how much this structure facilitates aerial dispersal (ballooning), and should be considered in theories/hypotheses on island community assembly as well as in conservation strategies.


We collected spider specimens using the COBRA sampling protocol, a standardised and optimised protocol that combines different sampling methods to obtain the maximum possible number of species for a given amount of effort and produces comparable data⁠. The sampling methods of the COBRA protocol for dry habitat were pitfall trapping (12 samples, each one grouping four individual pitfall traps), nocturnal and diurnal sweeping (four samples each), and active ground search (four samples). In forest plots the COBRA protocol consisted of pitfall trapping (12 samples, as in the dry plots), diurnal and nocturnal sweeping (two plus two samples), diurnal and nocturnal foliage beating (two plus two samples), and nocturnal aerial active search (four samples). Therefore, we obtained a total of 24 samples per plot in both dry and forest habitats. In La Gomera and Tenerife dry habitats the great abundance of shrubs allowed the split of the four sweeping samples in two sweeping and two beating samples. Sampling occurred between 2012-2017 – Terceira forest (2012), Tenerife forest (2013), Pico forest (2016), Madeira forest (2016), La Gomera forest (2016), and all dry plots in Madeira, Canaries and Cabo Verde in 2017, always at the time of the year with the greatest levels of diversity (April to November).


Canary Islands Government ACIISI, Award: SE-12/02

Canary Islands Government ACIISI, Award: SE-12/03

Canary Islands Government ACIISI, Award: SE-12/04