Relationship between insect density and plant damage for the golden loosestrife beetle, Galerucella pusilla, on purple loosestrife (Lythrum salicaria)

Assessing cause-and-effect relationships of weed biological control programs requires linking herbivore density with plant abundance. Spatial and temporal fluctuations in herbivore density, feeding niche and behavior, and anti-predator behavioral responses often make it difficult to count individuals directly. An alternative to direct counts is to estimate herbivore density indirectly from feeding damage. To apply the indirect approach, one must first establish the functional form of the density–damage relationship. We manipulated the density of adult and larval Galerucella pusilla on the target host plant, purple loosestrife (Lythrum salicaria), and estimated damage to shoots in greenhouse and field-cage experiments to quantify the relationship between beetle density and plant damage. Insect densities were initially fixed for each treatment for adults (4, 8, 12, 16, and 20 beetles per plant) and larvae (4, 8, 12, 16, 20, 24, 28, 32, and 64 larvae per plant) and subsequently allowed to vary due to mortality and emigration. Realized larval density (daily means averaged over a 14-day period) ranged from 3 to 48 larvae per plant across levels of the density treatment and adult density (daily means averaged over a 22-day period) ranged from 4 to 14 adults per plant. The relationship between damage and density was linear for adults on ln-transformed scales and linear for larvae on untransformed scales over the range of densities examined. Larvae damaged a mean leaf area of 0.23 cm2 per individual per day (in the period between emerging from the egg stage and entering the pupal stage) and adults damaged a mean of 0.78 cm2 per individual per day. Measures of percent damage using two rigorous methods (counting pixels of scanned images using a computer and counting cells on graph paper overlaid on the leaf image) were highly correlated with quick visual estimates of percent damage by an observer guided by precalculated leaf damage charts (r = 0.84). These results indicate that plant damage is an increasing function of insect density and that visual estimates of leaf area damaged can be used to estimate G. pusilla beetle density for field populations below carrying capacity.