UC Berkeley

Life history variation is a general feature of arthropod systems, but is rarely included in models of field or laboratory data. Most studies assume that local processes occur identically across individuals, ignoring any genetic or phenotypic variation in life history traits. In this study, we tested whether field populations of Pacific spider mites (Tetranychus pacificus) on grapevines (Vitis vinifera) display significant intraspecific life history variation associated with host plant cultivar. To address this question we collected individuals from sympatric vineyard populations where either Zinfandel or Chardonnay were grown. We then conducted a "common garden experiment" of mites on bean plants (Phaseolus lunatus) in the laboratory. Assay populations were sampled non-destructively with digital photography to quantify development times, survival, and reproductive rates. Two classes of models were fit to the data: standard generalized linear mixed models and a time-to-event model, common in survival analysis, that allowed for interval-censored data and hierarchical random effects. We found a significant effect of cultivar on development time in both GLMM and time-to-event analyses, a slight cultivar effect on juvenile survival, and no effect on reproductive rate. There were shorter development times and a trend towards higher juvenile survival in populations from Zinfandel vineyards compared to those from Chardonnay vineyards. Lines of the same species, originating from field populations on different host plant cultivars, expressed different development times and slightly different survival rates when reared on a common host plant in a common environment.