Temporally invariant structure despite changes in the identity of interactions in a pollination network

CHACOFF, NATACHA; LUCIANO CAGNOLO; DIEGO VÁZQUEZ; ERICA STEVANI; BENIGNO PADRÓN; IMENA DORADO

Palma de Mallorca, Islas Baleares, España

Simposio; The Evolutionary Ecology of Plant-Animal Interactions: From genes to communities; 2008

Spanish association of terrestrial Ecology

The understanding of organization and its determinants are methodological and conceptual challenges in the study of plant pollinator networks (Olesen et al. 2007). Nestedness and thus interaction asymmetry are among the most widely observed patterns in mutualistic networks. Species abundance, morphology and phylogeny have been proposed as possible causes of the structures observed in networks. However, flower and visitor abundance may exhibit great year to year variation (Herrera 1995, 1996). Despite this variation, most network studies have considered (some times making coevolutionary inferences) only a snapshot of a single season. The objective of this study was to evaluate how parameters of the network and the interactions change between years.We investigated year-to-year variation in a pollination network from the Monte desert of Argentina. Flower visiting animals were observed on different plant species in two consecutive flowering seasons (2006 and 2007) between September and December. We sampled the whole community as comprehensively as possible. A total of 545 five minute observation periods were conducted for all sites and plant species in 2006 and another 349 observation periods in 2007. With these data we constructed separate networks for each year. We report the number of species and network descriptors including size, connectance, linkage density and nestedness, and explored how persistent individual interactions were between years.We observed 1619 interactions in 2006 and 1854 interactions in 2007. The number of species of visitors were quite similar in both years (60 and 62), while the number of plant species raised in 2007 (20 for 2006 and 34 for 2007). Connectance and linkage density were both very similar between years (0.1 and 0.09 for connectance and 4.85 and 4.80 for linkage density). Networks were highly nested and temperature was also similar between years 5.37 and 6.86 for 2006 and 2007. Almost 40% (26 species) of flower visitors and 60% (16 species) of plants were included in the networks in both years, but only 30% of the interactions among those species occurred in both years.In this study we found that while network structure did not vary substantially between consecutive years, there was considerable variation in the identity of visitors and flowering plants and their interactions. This result indicates that despite on the apparent constancy of network structure, ecological and evolutionary inferences based on single-year studies need to be taken with caution and might actually be wrong.