Estimating the influence of population density and dispersal behavior on the ability to detect and monitor Agrilus planipennis (Coleoptera: Buprestidae) populations.

Emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), a phloem-feeding pest of ash (Fraxinus spp.) trees native to Asia, was first discovered in North America in 2002. Since then, A. planipennis has been found in 15 states and two Canadian provinces and has killed tens of millions of ash trees. Understanding the probability of detecting and accurately delineating low density populations of A. planipennis is a key component of effective management strategies. Here we approach this issue by 1) quantifying the efficiency of sampling nongirdled ash trees to detect new infestations of A. planipennis under varying population densities and 2) evaluating the likelihood of accurately determining the localized spread of discrete A. planipennis infestations. To estimate the probability a sampled tree would be detected as infested across a gradient of A. planipennis densities, we used A. planipennis larval density estimates collected during intensive surveys conducted in three recently infested sites with known origins. Results indicated the probability of detecting low density populations by sampling nongirdled trees was very low, even when detection tools were assumed to have three-fold higher detection probabilities than nongirdled trees. Using these results and an A. planipennis spread model, we explored the expected accuracy with which the spatial extent of an A. planipennis population could be determined. Model simulations indicated a poor ability to delineate the extent of the distribution of localized A. planipennis populations, particularly when a small proportion of the population was assumed to have a higher propensity for dispersal.

Dispersal of Agrilus planipennis (Coleoptera: Buprestidae) from discrete epicenters in two outlier sites.

Emerald ash borer, Agrilus planipennis (Fairmaire) (Coleoptera: Buprestidae), a phloem-feeding beetle native to Asia, has become one of the most destructive forest pests in North America. Since it was first identified in 2002 in southeast Michigan and Windsor, Ontario, dozens of isolated A. planipennis populations have been discovered throughout Michigan and Ontario, and in 12 other states and the province of Quebec. We assessed realized A. planipennis dispersal at two discrete outlier sites that originated 1 yr and 3 yr earlier from infested nursery trees. We systematically sampled ash trees within an 800 m radius of the origin of each infestation to locate galleries constructed by the progeny of dispersing A. planipennis adults. Our sampling identified eight trees at the 1 yr site infested with a mean +/- SE of 11.6 +/- 8.4 A. planipennis larvae and 12 trees at the 3 yr site with 25.8 +/- 11.1 larvae per meter squared. Dendroentomological analysis indicated that A. planipennis populations were predominantly undergoing a 2 yr (semivoltine) life cycle at both sites. Colonized trees were found out to 638 and 540 m from the epicenters at the 1 yr and 3 yr sites, respectively. Logistic regression was used to determine whether the likelihood of A. planipennis colonization was affected by wind direction, ash phloem abundance, distance from the epicenter, or land-use type (i.e., wooded, residential, agricultural, or urban). Results show that the probability of A. planipennis colonization was significantly affected by ash phloem abundance and decreased with distance from the epicenter.