Experimental evidence that both parties benefit in a facultative plant-spider mutualism

Spiders are known to influence plant fitness, and vice versa. Yet, it has not been appreciated that these facultative relationships can be mutualistic. I examined the interaction between Phryganoporus candidus, a subsocial Australian spider, and the extrafloral nectary-bearing shrub Acacia ligulata to explore variability in mutualistic interactions over a three-year period. Spiders enhanced seed production by reducing seed predation by heteropterans, wasps, and weevils. Because spider colonies occupy only a fraction of a plant's volume, average benefits ranged from 0.4 to 6% increases in whole-plant seed production. These benefits were strongest in years of low seed production, suggesting that spiders may buffer plants against female reproductive failure. To evaluate benefits for spiders, I established experimental spider colonies on three common hosts. Spider performance (persistence and prey capture rates) on live A. ligulata and live hopbush Dodonaea viscosa exceeded that on dead acacia, suggesting that live hosts are more beneficial than dead hosts. Stable-isotope analyses demonstrated that colonies living on the three hosts differed substantially in diet, providing a possible mechanism for the observed differential suitability of hosts. However, the analyses were unable to establish conclusively that A. ligulata extrafloral nectar was an important reward for spiders. Variability in the A. ligulata P. candidus system suggests that plant-spider associations, like other facultative protection relationships, likely vary along a continuum from antagonism to mutualism.