Linyphiid spiders and Azteca instabilis ants: A temporal association influenced by resource availability, plant structure and a disease

Background/Question/Methods   Linyphiidae spiders are involved in the control of herbivore populations in ecosystems and agroecosystems. In addition, it is becoming evident that some linyphiids are associated with ants that might be acting as ecosystem engineers. In Germany, Schuch et al. (2008) reported that linyphiids inhabiting the grassland floor were positively associated to Lasius niger ants. In coffee agroecosystems, Vandermeer et al. (2002) reported that linyphiids were associated to the ant Azteca instabilis and Marín (unpublished data) found that the morphospecies Linyphiidae1 was more abundant in coffee plants patrolled by A. instabilis. In this study we elucidate some mechanisms responsible for this association by addressing the following questions. Does Linyphiidae1 find higher insect resource availability in the presence of A. instabilis than in its absence? Does vegetation structure affect the abundance of Linyphiidae1? In June-August of 2009, Linyphiidae1 abundance was monitored weekly in 60 coffee plants with presence/absence of A. instabilis ants. Additionally, insect resource availability was quantified in presence/absence of A. instabilis by placing horizontal sticky traps (7.5 x 11.5 cm) between the branches of coffee plants and by sampling Linyphiidae1's webs. Coffee plant height and number of branches per plant were also recorded.

Results/Conclusions   Results indicate that the association between Linyphiidae1 and A. instabilis varied over time. At the beginning and end of the sampling period Linyphiidae1 was not associated with A. instabilis while in the other seven sampling dates Linyphiidae1 was strongly associated with the ant. Both sticky traps and web revision reported that insect availability was significantly higher in the presence of A. instabilis than in its absence (t=2.38, p<0.027 and t=3.50, p<0.001 respectively). The analysis of Linyphiidae1 webs indicated that Coccus viridis, the coffee scale, was the most common insect followed by hymenopterans and dipterans. In addition, coffee plant size also influenced Linyphiidae1 abundance. The temporal association between Linyphiidae1 and A. instabilis seems to be coupled with the dynamics that occurs amongst A. instabilis, its mutualist, C. viridis, and L. lecanii, a fungus that attacks C.viridis and disrupts the mutualism. Results support that the low abundances of Linyphiidae1 in the presence of A. instabilis recorded at the end of the sampling period might be a consequence of L. lecanii attack on C.viridis, since this insect was the main prey found in Linyphiidae1 webs. We conclude that coffee plant size; insect availability and L. lecanii are factors involved in the association Linyphiidae1-A.instabilis.