Summary 1.Resolving the controversy over the prevalence of generalization in plant-pollinator interactions requires field studies characterizing the pollination effectiveness of all a plant's floral visitors. Herein, the pollination effectiveness of all visitors to two species of barrel cactus ( Ferocactus ) was quantified. 2. Flowers of both species were pollinated almost exclusively by cactus-specialist bees: 99% ( F. cylindraceus (Engelm.) Orcutt) and 94% ( F. wislizeni (Engelm.) Britt. and Rose) of all seeds produced in this study resulted from cactus bee visits. 3. For F. cylindraceus , the cactus-specialist Diadasia rinconis was the most abundant visitor. For F. wislizeni , three cactus-specialists (including D. rinconis ) plus generalists in the family Halictidae (which did not act as pollinators) each accounted for a quarter of all visits. 4. Diadasia rinconis visits to F. wislizeni flowers were more effective (per-visit) than visits by the other two cactus-specialists. 5. Pollen-collecting and nectar-collecting visits were equally effective. Nectar-collecting visits were the most abundant. 6. Apart from the non-pollinating halictids, floral visitors surprisingly did not include commonly co-occurring generalist bees. 7. These data suggest that, just as apparently specialized flowers may be visited by a diverse assemblage of generalists, so apparently generalized flowers may be visited predominantly by specialists, and that these specialists may perform virtually all of the pollination.

No abstract

No abstract

No abstract

Increases in the incidence and severity of drought threaten the viability of rare plants in arid regions. The endangered Nichol's Turk's head cactus (Echinocactus horizonthalonius Lemaire var. nicholii L. Benson) occurs only in four small, isolated populations in the Sonoran Desert of North America. Since 1995 we have monitored a population in southeastern Arizona (USA). Here we report 23 years of observations on abundance, growth, mortality, flowering and recruitment. Abundance of plants decreased from 132 in 1996 to 40 in 2017, with 100 individuals recruited and 203 dying during the study. Individual plants grew slowly, increasing annually by an average of 0.22 cm (95% confidence interval, 0.18–0.26 cm) in diameter and 0.27 cm (0.20–0.33 cm) in height. Growth was slowest when drought was most severe and slowed as plants reached maximum size. Annual mortality increased markedly across the study period and did not vary with plant size. Annual probability of flowering increased as plants increased in diameter but not in height, and varied with precipitation and drought but not with mean annual temperature. Recruitment was higher when average temperature was higher and the number of recruits per capita increased across the study period. The annual rate of change in abundance averaged −6%, but shifted markedly from −1% during 1995–2008 to −11% during 2008–2017. Our results indicate that the population's decline was not a consequence of failed recruitment but of increased mortality, which we discuss in the context of climate and herbivory.