Floral visitors of Trillium camschatcense (Trilliaceae) in fragmented forests

Abstract

Plant–pollinator interactions are of particular interest in conservation biology because of their ecological and agricultural importance (Bachmann & Nabhan 1996; AllenWardell et al . 1998; Kearns et al . 1998). Pollinator deficits may result in shortfalls in fruit or seed production, and ultimately, plant extinction (Rathcke & Jules 1993; Olesen & Jain 1994; Kearns & Inouye 1997). Many recent studies have shown that seed production is reduced in small, fragmented plant populations due to pollen limitation (e.g. Jennersten 1988; Aizen & Feinsinger 1994a). However, the effects of habitat fragmentation on pollinator assemblages have been rarely examined directly (Didham et al . 1996; Kearns et al . 1998). Further understanding of the pollinator dynamics in a fragmented landscape, for example, what kinds of pollinators are likely to be lost from fragmented habitats, is required to obtain useful implications for conservation of plant–pollinator interactions. In the Tokachi region of eastern Hokkaido, Japan, intensive agricultural development since the 1880s has resulted in a highly fragmented landscape with a large number of small forest remnants. Trillium camschatcense Ker Gawler (Trilliaceae), a representative woodland spring herb in this region, once often formed large ( > 5 ha), continuous populations, but is now mostly found in the understory of small remnant forests surrounded by pastureland or agricultural fields. In a previous study (Tomimatsu & Ohara 2002), we demonstrated that small, fragmented T. camschatcense populations generally produced fewer seeds than large, continuous populations and that the low stigmatic pollen load was largely responsible for the reduced seed production. The aim of the present study was to describe floral visitors of T. camschatcense in fragmented forests, where the availability of pollinators was hypothesized to be reduced (Tomimatsu & Ohara 2002). We observed floral visitors at five different-sized populations and examined the differences in species composition and visitation frequency among the populations. We also discuss the relationship between floral visitors and outcross pollen load, which was measured in our previous study (Tomimatsu & Ohara 2002).