Dairy cattle management factors that influence on-farm density of European starlings in Ohio, 2007–2009

•Dairy cattle management factors were assessed for possible association with the number of starlings per milking cow (starling density).•Zero-inflated negative binomial model was the best fit model.•The risk factors identified with the count portion of the model were mainly related to feeding management of the cows.•The odds of zero starling counts were associated with manure management and distance of farms from roosting sites.

Potential dairy farm management and environmental factors that attract European starlings (Sturnus vulgaris) to dairy farms were explored. During the period from 2007 to 2009, 150 dairy farms were each visited twice (once during the summer and again in the fall) and the number of starlings was recorded. Risk factors were assessed for possible association with the number of starlings per milking cow (starling density), using a zero-inflated negative binomial model. Starling density was higher on farms visited in 2007 compared to those visited in 2008 or 2009. The interaction term between feeding method and feeding site was significantly associated with starling density on farm; generally, feeding outdoors was associated with increased starling density. The odds of a zero starling count (compared to a count greater than zero) was higher on farms that removed manure from barns weekly or less frequently than weekly compared to those that removed manure daily or after every milking. The odds of a zero starling count decreased with increasing distance of a farm from the closest night roost. Identifying on farm risk factors that expose farms to starlings will help farmers develop strategies that minimize the number of birds on their farms and thereby reduce physical damage to the farms as well as the potential for pathogen transmission from birds to cattle and humans.