Comparing two methods for estimating floral resource availability for insect pollinators in semi-natural habitats

Abstract

SummaryPollinator and flowering plant interactions play a critical role in maintaining most terrestrial ecosystems, including agroecosystems. Although estimates of floral resource availability are essential to understand plant–pollinator relationships, no generally accepted methodology exists to date. We compared two methods for sampling floral resources in a single meadow. About every three days, we recorded species lists of insect-pollinated plants with abundance categories assigned to each species (hereafter referred to as scanning) and we counted the flowering shoots in 36 2 × 2 m quadrats (hereafter quadrat sampling). These methods were compared with respect to (i) the number of species detected, (ii) estimated floral resource abundance, and (iii) temporal changes in flowering. With scanning, we found more potential nectar-plant species and species were found earlier than with quadrat sampling. With the latter, abundant species were found with higher probability than the scarce. Flower abundances were correlated between the two methods. We predicted that a cover of 6.3 ± 3.6% should be used for an appropriate estimate of flower abundance in our study site, although the optimal cover probably varies across different habitats. Furthermore, flower abundance changed 6% per day compared to the flowering peak. Overall, scanning seems to be more appropriate for detecting presence and the timing of species, while quadrats may provide higher resolution for abundance estimates. Increased sampling coverage and frequency may enhance research accuracy and using scanning and quadrat sampling simultaneously may help to optimize research effort. We encourage further development of sampling protocols.