Monday, August 2, 2010

PS 19-154: Annual grass fungal endosymbionts decrease the vulnerability of symbiotic-free neighbors to Sypha maydis

Marina Omacini1, Maximiliano Mallarini2, Cecilia Casas2, and Claudio M. Ghersa2. (1) IFEVA/CONICET, Faculty of Agronomy, Buenos Aires University, Argentine, (2) FEVA/CONICET, Faculty of Agronomy, Buenos Aires University, Argentine

Background/Question/Methods

Symbiotic associations between plants and fungus are of major importance as drivers of ecological processes at different organization levels and scales. Asexual fungal endophytes of Neotyphodium (Clavicipitaceae) grow in aboveground tissues of many cool-season grasses. These endosymbionts may alter host plant chemistry and ecophysiology, usually increasing host´s resistance to herbivores. Despite of the increasing awareness that grass‑endophyte symbiosis may affect whole neighborhoods, little is known about its impact on biotic interactions between nearby symbiotic-free plants and other organisms.We performed an outdoor experiment to investigate if symbiotic-free Bromus unioloides plants gain protection from aphid herbivory owing to their proximity to Lolium multiflorum plants defended by endophytes. We created vegetation patches (50cm x 50cm) with four different endophyte infection levels (I:5%, II: 32%, III: 60%, IV: 95%) by sowing different proportions of seeds with and without the symbiont. Six months after planting date, three pots with Bromus unioloides (focal plants) were located into each of grass patches and were infested, one week later, with 40 aphids of Sypha maydis (Aphididae). We measured the abundance of S. maydis live and parasitized and of other naturally colonizing aphids in both grasses.

Results/Conclusions

Endophyte infection level of the neighboring vegetation significantly affected the establishment of S. maydis on B. unioloides plants (P<0.0001). Aphid abundance was four times higher in focal plants immersed within patches of low endophyte infection (average values were 167 and 206 aphids/plant for I and II infection levels, respectively) than within patches of high infection (43 and 46 aphids/plant for III and IV infection levels, respectively). The abundance of S. maydis colonizing L. multiflorum plants was low (less than 9 aphids/plant) regardless of the treatments (P=0.80). Neither parasitized aphids nor other aphid species were detected during the experiment. L. multiflorum reproductive (P=0.023) and aerial vegetative (P=0.014) biomass at the end growing season were positively related with endophyte infection levels. B. unioloides biomass did not differ among treatments (P=0.72). Mechanisms for this associational protection are still unclear. However since treatments did not affect top-down controls nor B. unioloides plant performance, we speculate that volatiles may be involved modulating grass vulnerability to generalist aphid herbivores in endophyte infected L. multiflorum patches.