Summary
The effect of artificial habitat in altered landscapes on species interactions and their suite of enemies is largely unknown. Water mites have been associated with reduced fitness of model damselflies. Mite parasitism was variable, but higher for Ischnura verticalis damselflies from natural, than from artificial, wetlands in the same region. There were no differences in timing of sampling, temperature during sampling, or host age or sex composition of samples between wetland types. Landscape structure might constrain mite presence or abundance at wetland sites or wetland type might be a better predictor of mites, based on factors such as prey abundance. Fewer mites on damselflies from numerous artificial wetlands means that the strength of parasite‐mediated selection is likely less than would be inferred if only natural wetlands were surveyed. Such effects of human changes in habitats on host species probably occur often.
Methodology
Ischnura verticalis is a common damselfly (Odonata, Zygoptera, Coenagrionidae) throughout Canada and the United States and prefers lentic habitats. Males, mature (pruinose) females, and younger (red‐orange) females can be distinguished readily (Walker 1953). Like many other coenagrionid damselflies, I. verticalis is a host for several species of ectoparasitic mites of the genus and subgenus Arrenurus . The parasitic larval mites attach to the damselfly larvae when it is nearing emergence. These parasites then transfer to the newly‐forming host imago, attach and secrete a feeding tube (Smith 1988). The mite(s) then engorge on the host damselfly (Smith 1988) and finally detach from the host when the host is near or above water. After returning to water, the mite completes its life cycle, which includes deutonymphal and adult stages that are predators of microcrustacea.
Water mites have been associated with reductions in survivorship, fecundity and male mating success of damselflies, using either or both experimental and observational approaches (reviewed by Forbes and Robb 2008). In some species, individuals allocate resources towards fending off parasitic attacks and such resistance is assumed to be costly (Forbes and Robb 2008). Parasites might therefore influence selection on hosts even if they do not influence populations per se (e.g. damselfly populations might be affected more by weather, Thompson 1990).
Sampling took place at 24 wetlands (in a 3446 km2 area) in the vicinity of the cities of Ottawa (ON) and amalgamated city of Gatineau (QC), Canada. Each wetland was classified as being either artificial (17 wetlands) or natural (7 wetlands). Natural wetlands were those that developed through natural geological and hydrological processes (Keddy 2000). Artificial wetlands were those that were not present historically and have been established as storm water drainage compounds, excavation sites and drainage ditches. If we could not reasonably ascribe type of wetland to a site, then that site was not included in our study. Artificial wetlands were at least 10 yr old, whereas all natural wetlands were present in aerial photographs taken in 1972 to 1975. All sites were permanent wetlands at least from 2002, varied in size from 1 to 5 ha, and contained well‐developed emergent vegetation (details available on request). Natural wetlands were generally located in the city's “Greenbelt” or in Gatineau Park (both areas are relatively free of development), whereas artificial wetlands often were located near housing and shopping centre developments. Those artificial wetlands were not formerly natural wetlands that had development in or around their margins. These wetlands were chosen as part of a larger study on the effects of artificial and natural habitats, in low versus high forested cover landscapes, on odonate diversity. The subset of wetlands chosen had to have I. verticalis present.
We netted damselflies at all wetlands over an approximate 20‐d period beginning 10 June, 2005 and ending 1 July, 2005 (this represents a very small proportion of the total flight period which extends from mid‐late May until at least late August, Forbes unpubl.). The sampling was designed to collect ca 30 individual damselflies at each wetland site, while being able to visit all wetlands over a three‐week period. Order of sampling was interspersed such that the natural wetlands were sampled at the same time as blocks of artificial wetlands. Damselflies were netted individually, following field identification. Netted individuals were sexed. Females were aged approximately as either pruinose (mature) or young red‐orange adults. We know from previous work that red‐orange females have not had time to lose their mites, if they were parasitized. The numbers of mites on each damselfly (which were invariably attached to the undersurface of the thorax and/or abdomen) were enumerated with the aid of a 10× loupe. Recording whether an individual was parasitized was not subjected to error. However, we were less precise in our mite counts for 6 of 737 individuals with >35 mites. Notwithstanding, our counts by two observers agreed to within 3 mites. Furthermore, this occasional error is not likely to have been influenced by either host sex or age or wetland type sampled.