Teaching Issues and Experiments in Ecology

THE ECOLOGICAL QUESTION: Does cowbird brood parasitism reduce songbird numbers? WHAT STUDENTS DO: Students use a spreadsheet to compare numbers of individuals of up to 40 bird species in sites with and without cowbirds. SKILLS: Using a spreadsheet to make figures. ASSESSABLE OUTCOMES: Figure from spreadsheet data, written interpretation of data, written analysis of the published paper. The brown-headed cowbird (Molothrus ater) in North America is an obligate or generalist brood parasite of over 200 songbirds. Numerous studies show reproductive losses for individual host species (e.g. Strausberger and Ashley 1997). It has long been assumed that cowbird parasitism can result in changes in songbird communities, but surprisingly the De Groot and Smith (2001) study is the first to assess whether this brood parasite can change overall bird community composition. They state their hypothesis as: brown-headed cowbirds change the composition of songbird communities by depressing the number of suitable host individuals. This work has management implications in regard to cowbird removal programs designed to protect endangered birds. De Groot and Smith worked in northern Michigan where there is an extensive cowbird removal program to protect the Kirtland's Warbler (Dendroica kirtlandii). This bird only nests in young jack pine forests in northern Michigan. In 1971, a census recorded only 201 singing males, and both limited/poor quality of habitat and cowbird parasitism were proposed as responsible for their decline. A cowbird removal program began in 1972. In this study, De Groot and Smith compare songbird communities in cowbird removal sites and control areas. The researchers divided the songbird communities into Suitable Hosts (birds that accept cowbird eggs and feed their young) and Unsuitable hosts (birds that reject cowbird eggs). They selected 10 Removal sites adjacent to active cowbird traps in operation for 5-11 yrs. Control sites were > 5 km from cowbird traps or any area that had not experienced cowbird removal within the last 5 yr. They matched general habitat characteristics for Removal and Control sites. In 1997 they added Control sites < 10 km from cowbird traps. Site areas were a half circle radius of 1 km. Birds were identified by sight or sound on transect lines (four 1 km lines each 60 degree apart). Between dawn and 1000 on all sites birds were counted twice during summer in 1996 and three times in 1997. Cowbirds were assessed with additional playback of female chatters to improve likelihood of cowbird detection rates, …