Habitat Characteristics Influencing Cowbird Parasitism
Because of its length, variety of habitats, and stream boundaries, Illinois is host to a large number of bird species. Populations of many bird species have plummeted, however, over a relatively short period of time. This decline may be attributed to many factors including the effects of pesticides on reproduction, overhunting, and conversion of native landscapes for human activities. Within the past two decades, the impact of nest parasitism by the brown-headed cowbird has aroused interest. Although the cowbird is native, it is wrecking more havoc on certain bird populations than ever before. The cowbird does not rear its own chicks but lays eggs in the nest of other birds. Cowbird chicks often hatch earlier and grow faster outcompeting the resident chicks. Those birds hardest hit by parasitism are the neotropical migratory songbirds of the forest interior.
The dwindling of our forests has had a major impact on the species that rely upon this particular habitat. About 1820, when European settlement occurred, Illinois had 13.8 million acres of forest, which made up about 38% of the land in the state. In 1980, however, only four million acres of forest were left making up about 12% of the state. Biologists and environmentalists have been made aware of the importance of conserving our resources by observing different species that depend on this environment.
Major contributors to the destruction of forests have been agriculture and the timber industry. Fields, utility roadways, and power lines are invading the interior forests. The smaller an interior forest becomes, the smaller the microhabitat is for the species to exist. Areas of habitat, like interior forests, are losing their integrity because the exterior buffers are decreasing. This decrease in exterior forests causes competition for space among species.
One concern is the decrease in habitat for migratory songbirds. Jerry Sullivan states, in the article Sinking in the Shawnee, "…forest fragmentation, a process that exposes these birds of the forest interior to the hazards of life on the edge, results in high rates of predation and brood parasitism"(Sullivan, 1992). Eastern forests provide critical nesting habitat for Neotropical migratory songbirds. Conservation biologists have been concerned that cowbird parasitism is a major factor reducing breeding success of Neotropical migrants in North America (Link and Caldwell, 1996). Significant declines in populations of forest interior species have been noted during the past twenty years (Robinson, 1992). Among various factors contributing to these population declines, nest parasitism by brown-headed cowbirds has received the most attention in the central hardwood forests.
The brown-headed cowbird (Molothrus ater) is a small blackbird with a short sparrow-like bill. The cowbird averages eighteen centimeters in length. Their natural habitat includes farms, fields, barnyards, roadsides, wood edges, and river groves (Peterson 1980). Cowbirds usually feed in short grass habitats or with large grazing mammals (Thompson, 1997). They breed in a wide variety of habitats from prairie to forest but may select habitats with high host densities (Thompson, 1997). The female cowbird does not spend the time and energy involved in building a nest. She will locate the nest of some other species to incubate her eggs and feed her offspring. The female will wait for the host to leave its nest and will then quickly lay its egg(s). The host often accepts the other egg(s) to its clutch. The offspring of the cowbird require less time for incubation giving it an advantage to dominate feeding time (Alcock 1979).
The cost of reproduction for the songbird is being greatly affected by the parasitism of the cowbird. It is important for the songbird to maximize its reproductive success. However, if the songbird’s clutch size is increased, it will cost her in later reproductive years. The consequences of cowbird parasitism are particularly severe for a host species already reduced in numbers (Robinson, 1993).
The cowbird will lay 30 to 80 eggs per season in the nests of other species (Robinson 1997). According to Robinson (1997), the abundance of the cowbird can pose a severe problem for many of the 200 or more hosts’ species. An example of this is in the forests of Illinois where 30%-90% of the nests of most migratory songbirds is parasitized, often with more than one cowbird egg (Robinson 1997). On the other hand, less than 10% of the nests of most grassland species were parasitized across a wide variety of sites (Robinson, 1997). The orchard oriole had over 90% of its nests parasitized with an average of 2.2 cowbird eggs per nest at Lake Shelbyville, a shrubland habitat. In the Shawnee Forest, however, less than 40% of the nests were parasitized with an average of only one cowbird egg per nest (Robinson, 1997). The cowbird apparently does not prefer only forest species for host sights, but will also parasitize shrubland hosts. However, the shrubland hosts, unlike forest interior species, appear to have effective defenses against the cowbirds egg(s) where the songbird does not (Robinson 1997).
Researchers have used various methods to evaluate cowbird habitat preference and the impact of parasitism on host species. In Green Ridge State Forest in eastern Allegany County, Maryland, six habitat types were sampled: (1) forest-powerline edge, (2) forest-open canopy road edge, (3) forest closed canopy road edge, (4) forest brush edge, (5) forest-stream edge and (6) forest interior (Evans and Gates, 1997). Tests were randomly performed to see if the cowbird had a preference to a certain type of vegetation. The methods used in these particular areas included counts of cowbirds made within a four-hour period beginning at sunrise from May 20th - June 26th, 1995. Six different habitats were counted in a rotation manner in order to control any discrepancies. (Evans and Gates 1997). Data was collected to find total vegetation volume of the habitat that the cowbird preferred.
Evans and Gates (1997) made 1,980 observations of 60 bird species, including 34 known cowbird hosts. Brown-headed cowbirds were detected in all habitat types, except forest interior. The ranking of the cowbird occurrence in edge types was forest-brush at 50%, stream 46.4%, powerline 33.3%, open road 14.3%, and closed road 10.3% edges. According to Evans and Gates (1997), the cowbird ranked eleventh (2.6%) in overall species abundance, and varied from eighth most abundant in forest-brush and stream edges to fourteenth most abundant in forest-closed road and open road edges. The forest interior had the lowest bird and host species abundance on average. When in edge habitats, birds were detected most often close to the boundary between adjacent habitats, particularly if bordered by a road or stream (Evans and Gates, 1997b). The cowbird does not prefer the forest interior but does exhibit a relationship to shrubby, low-vegetative forests. Cowbirds in this area used only certain microhabitats for breeding. These certain areas may contribute to the cowbird's selection for breeding areas. Cowbirds concentrate in local areas where host densities are high (Evans and Gates 1997). It is proposed by Evans and Gates (1997) that cowbirds are attracted to distinct visible edges formed by canopy openings in the forest landscape and, secondarily, by the occurrence of a high basal area and the total vegetation volume. When the cowbird has settled near a forest edge, the activity of other bird species is likely used to further refine selection of a suitable breeding area.
There are several methods used to determine the parasitism by the cowbird on host species. One of these methods, empirical Bayes estimation of proportions, is used by Link and Hahn (1996) to demonstrate cowbird parasitism. The Bayesian analysis is a method used in studies ranging from epidemiology to population census. There have been few studies done with the Bayesian method outside medical settings. Bayesian methodology is used to estimate the size of populations from data collected. This particular analysis is used to estimate the parameters prior to distribution. The empirical Bayes application is employed to estimate proportions using information obtained in a study of parasitism by the brown-headed cowbird. The Bayes method aided in identifying species that had the greatest parasitism rates (Link and Hahn, 1996). However, like other studies, there can be a variation in the parasitized nests that are sampled, and this can cause problems in ranking the host species (Link and Hahn, 1996).
For the empirical Bayes method, a hypothetical application is used to determine factors influencing the host species. Three of five nests for Species A (60%) and 19 of 50 nests for Species B (38%) were parasitized. There would have to be an understood typical parasitism rate to conclude the percentage of parasitism on Species A and B. The Bayes method does not work according to typical logic. The percentage alone is not enough information to draw conclusions concerning parasitism. If the percent were 55% then the conclusion would be that Species B (38%) had a low rate. However, Species A (60%) would be a more approximate measure of rate. If that typical parasitism rate were 20%, then evidence would indicate that Species B had a high rate of parasitism. How these two species would rank in parasitism would depend on previous knowledge of a typical rate. The information for Species A and B is not substantial enough to conclude the ranking of parasitism. "A Bayesian analysis can be thought of as a combining of existing knowledge with new knowledge (sample data)…"(Link and Hahn, 1996). The first step to analysis in the empirical Bayes method is to estimate unknown parameters and obtain an estimate of the average parasitism rate on host species. (Link and Hahn, 1996). A heuristic explanation of Bayes estimation distinguishes the parameters of the prior from the parameters of primary interest. The parameters of the prior are hyperparameters, which give a description of the distribution of these rates. The empirical Bayes analysis gives an estimation of the unknown hyperparameters and gives an estimate of the prior distribution (Link and Hahn, 1996). The beta-binomial model gives a description of the empirical Bayes proportions. This model can assist in using new knowledge to update the data.
The application of the empirical Bayes method allows for improvement in an estimation of individual parameters. Link and Hahn concluded that wood thrush is less likely to be parasitzed by the cowbird. However, these results conflict with data reported by Robinson. Robinson had concluded that the wood thrush in Illinois had high parasitism rates (Link and Hahn, 1996). The different nesting behaviors of host species may effect the differing parasitism levels (Burhans, 1997).
The habitat has an impact on the cowbird’s increase in parasitism. Burhans studied the relationship between the habitat and the nest microhabitat of field sparrows (Spizella pusilla) and indigo buntings (Passerina cyanea) to brood parasitism by the brown-headed cowbird (Molothrus ater) (Burhans1997). The way in which cowbirds select hosts is not clear. Female cowbirds apparently focus on the behavior of the host in searching for nests. Levels of brood parasitism may be influenced by the increase of the proximity of perches (Burhans, 1997). Studies have indicated that there is a connection between the height of the nest and parasitism frequency (Burhans, 1997). The method Burhans used to support his findings were to take counts of host nests parasitized in the months of April to July. Research was done on habitats that included 1) old fields, 2) upland fields, and 3) grazed agricultural field. The female cowbird seemed to be attracted to snags, and they were often sighted on dead trees, shrubs, fenceposts, and telephone poles (Burhans, 1997). The host species with the most secluded nests experienced less parasitism by the cowbird. Nests that were not as well hidden were more likely to be impacted by parasitism.
Research that was done on the indigo bunting’s nesting was compared to the field sparrows and nonforest indigo bunting nesting. However, it was determined that habitat may be of more influence than nest seclusion. "Side concealment explained parasitism for all indigo buntings and those in nonforest habitat, but not for indigo buntings in forests alone. Parasitism increased from field sparrow, to nonforest buntings, to forest buntings, whereas side nest concealment decreased within this sequence" (Burhans, 1997). The forest indigo bunting nests were more likely to be parasitized by cowbirds. The parasitism rate approached 90% in the beginning of the breeding season (Burhans, 1997). One reason that cowbird parasitism may be high in forested areas is that the female cowbird prefers forested areas in the morning for breeding. Nest microhabitat traits such as concealment can influence parasitism, but the major contributor may be habitat and characteristics of the host species. In Burhans study, he concluded that "nest microhabitat features may influence probability of parasitism, but species and habitat characteristics may explain frequency of parasitism" (Burhans, 1997).
In a report to the USDA Forest Service, Thompson and Dijak (1997) studied the differences in movements, home range, and habitat preferences of female cowbirds in three Midwestern landscapes. In their study, they discovered that forest and shrubland habitats were preferred habitats in the morning. In the afternoon, preferred areas were grass, feedlot, and developed habitats. The cowbird’s movement and home range could be a larger area because of increased host competition in fragmented forests (Thompson and Dijak, 1997). The three study areas were in Illinois and Missouri. According to Thompson and Dijak (1997), "brown-headed cowbird numbers, and cowbird/host ratios vary greatly among these study areas". Sixty-seven percent of the nests that were parasitized in Jonesboro, Illinois were in areas with 55% forest and 36% rowcrop habitat. In Ashland, Missouri cowbird parasitism was at 59% in areas with 50% forest, 32% grassland, and 13% rowcrop. Carr Creek, Missouri had only 5% parasitism in areas with 93% forest and limited grassland at 4%. The correlation between cowbird parasitism and lack of forest area gives a good indication of how the habitat can influence the rate of parasitism.
The cowbird had definite patterns in habitat use. Of the observations that were made of the cowbird, about 60% or more were in the morning within forest or shrubland habitats. In the afternoon, more than 50% of the observations made were in grass, rowcrop, and feedlot habitats (Thompson and Dijak, 1997). At the Jonesboro site, cowbirds moved farther between feeding and roosting areas than compared to the other two sites. "Cowbirds had very large home ranges, particularly when compared to that predicted for similar sized passerines" (Thompson and Dijak, 1997). Three factors that reflect a large home range distribution is host rich forest habitats, agricultural feeding areas, and secure roosting habitats (Thompson and Dijak, 1997). In this study, it is concluded that large grassland habitats are important to support cowbird populations. There are serious conservation implications for cowbird hosts because of the dependence cowbirds have on the feeding habitat in forested landscapes (Thompson and Dijak, 1997).
These studies of the brown-headed cowbird and the impact that it has had on migratory songbirds illustrate that the habitat that each one of these species inhabits can affect the other’s success. The home range of cowbirds could become large because of increased competition for hosts resulting from greater numbers of cowbirds in more fragmented forests. The problem for the songbird is the fragmentation of the forest. This exposes the songbird to the hazards of predation and brood parasitism.
The cowbird is a migratory bird that will move south for the winter to feed (Figure 2). At one time the lack of food supply during the winter decreased the cowbird population. Today, however, modern agricultural practices offer the cowbird a better food supply increasing their population. The cowbird is now able to survive winter months giving it an advantage to increase in population. The increase in the cowbird population size influences competition. Greater competition among breeding cowbirds leads to invasion within fragmented forests.
Interior songbirds have not evolved a defense against such parasitism by the cowbird and have been termed "acceptors" of the cowbird egg. However, there are species of birds that have defenses against parasitism. These "rejecters" may include such birds as the cedar waxwing (Bombycilla cedrorumm), gray catbird (Dumetella carolinensis), American robin (Turdus migratorius), and the brown thrasher (Toxostoma rufum). A study by Rothstein was done specifically on the cedar waxwing’s defense. The cedar waxwing is able to reject the cowbird egg by lifting it out of its nest. The cedar waxwing’s most cost effective means is to eject the cowbird egg. The less time and energy spent in ridding the nest of this parasite, the more time and energy spent in raising the young. The size of the cedar waxwing’s beak makes it one of the smallest rejecter species. Unfortunately, this may cause the cedar waxwing to destroy the cowbird egg by pecking at the egg or even abandoning the nest. Rothstein’s study, "Experiments on defenses cedar waxwings use against cowbird parasitism," demonstrates the capability of some species to reject parasitism. His methods included placing artificial eggs in the host nest to determine the means used by the cedar waxwing to reject the egg. His studies were conducted in Cheboygan and Emmet Counties, Michigan, during the months of June to August in 1968 and 1969. In his research the cedar waxwing "rejection occurred at 69.0% or 40 of the 58 nests" that were artificially parasitized (Rothstein, 1976). The cedar waxwing varies in its ability to reject parasitism giving it a rate higher than most other rejecters have. Energy will be spent by the cowbird finding hosts that will accept its eggs, and will rarely lay its eggs in a nest that may be rejected.
The cost of parasitism to interior songbird species is significantly lower reproductive success. Loss of habitat coupled with a growing cowbird population presents serious challenges to many passerines. For some species, extirpation may be inevitable. Immediate measures are essential to insure that the sweet sound of the songbird is not muted by the deafening succession of exploitive land uses. The decline of songbirds in the wake of indiscriminate resource use may be a harbinger deserving our attention. These factors should include the abundance of cowbirds, the preferred habitat, and characteristics of the host species when consideration of the nest site and its features are being related to cowbird parasitism.
Researchers for this page:
Jennifer Bilyeu Elaine Haggitt Donna Short
