In this study, the prevalence of Haemosporidians in birds of Northern Michigan was investigated with an emphasis on the parasite Plasmodium in passerine birds. Birds caught and banded by mist-netters were sampled for blood and recorded by species, net number, sex, age, and swab number. The total prevalence of blood parasites in the samples was 23.8%. Birds caught at the CEMA banding site had the highest prevalence of malarial infection, and the birds of the NOMA banding site had the lowest. Mann-Whitney Tests resulted in two-tailed p-values of 0.001, 0.762, and 0.003 for comparisons of prevalence between NOMA and CEMA, CEMA and SOMA, and NOMA and SOMA, respectively. Kruskal-Wallis Tests regarding bird age, location, and families resulted in p-values of 0.259, 0.004, and 0.009, respectively. Tests for sex showed no significant differences between prevalence of infection in male and female hosts. Parasites in nonhuman animals offer insight in understanding and treating human parasites, and are thus frequent subjects of scientific study. Avian species of malaria were the first models for studying the biology of the human Plasmodium (Slater 2005). In fact, they are so similar that for a long time, there was doubt as to whether the parasites of bird and human malaria were really different (Manwell 1935). Studies by Waters and colleagues (1993) showed that Plasmodium
falciparum , the most deadly infective agent of human malaria, was more closely related to avian parasites than any other species. Not only are the causative agents similar, but the symptoms of human malaria are also similar to those of bird malarias (Kudo 1971).  Previous studies have also attempted to compare the parasite-host relationship of avian and human malarias, plus evaluate the evolutionary dynamics of a system of many hosts and several multihost parasites (Gandon 2004). Co-evolution between the malaria parasite and its bird hosts have also been of interest. For example, studies of the Hawaiian Honeycreeper suggest that avian parasites do not hinder their hosts' reproductive success possibly because of evolution (Kilpatrick 2006). In addition, another Haemosporidian Leucocytozoon  in ducks has been studied for its effects on brood size (O'Roke 1934).  However, avian Plasmodium in Northern Michigan has not been studied in much depth. Its prevalence in the birds should be between 10 and 30% (Feldman et al. 1995). The current study attempted to examine the possibilities of a trend in the incidence of various species of Plasmodium in groupings of birds according to species, age, sex, or size. Furthermore, prevalence of blood parasites in passerine birds were compared to those found in Mergus merganser based on previous findings of Haematozoa   in Michigan (DeJong et al., 2001). 	Data collection was carried out from July 10 to August 3, 2006 in Emmett County, Michigan. Birds were caught using mist nets, which were arranged in a loop formation with ten nets per loop. There were a total of three sites where birds were captured: NOMA (North Maple River), CEMA (Central Maple River), and SOMA (South Maple River). Data of every bird captured were recorded into organized data sheets, including information such as species name, age, sex, and net number. If blood was taken from a bird, the blood smear number was also recorded. A total of 189 blood samples were taken from a variety of birds; 159 of these were obtained through mist-netting, and the rest were obtained with help of Harvey Blankespoor. 	Blood samples were taken from birds' feet using sterile lancets, and then slides pre-cleaned with acid ethanol were held against the small droplet of blood. The blood collected on these slides was smeared with the edge of another slide, then fixed in methanol to be stained later in the lab. Each fixed slide was immersed in Coplin jars containing Giemsa stain for 15 seconds, placed in distilled water for 15 seconds, and finally, rinsed quickly in two more jars of fresh distilled water.  Dried slides were viewed using the 100x power oil immersion lens. Observation of malaria within the blood cells was carried out by manually scanning the surface of the slide for five minutes (roughly 1000 red blood cells). Prevalence of infection was determined by dividing the number of infected passerine birds by the total number of passerine birds examined. Overall prevalence was determined by dividing the number of infected birds by the number of total birds examined. Mann-Whitney and Kruskal-Wallis tests were used to calculate differences of the distribution of avian blood parasites and the intensity of infection between age, gender, and location of the species sampled. These helped to determine any relationship between age or gender and a particular infection and these analyses were done in SPSS 14.0 for Windows. The total prevalence of blood parasites in the samples was 23.8%. Table 1 summarizes the data of the 14 families of birds that were examined. The family Parulidae was most represented in the sample, but the Anatidae had the highest overall prevalence of infection and prevalence within the family. The family Parulidae also had the highest prevalence of infection within the mist-netted birds (Table 2, Figure 1). At species level, the Baltimore Oriole and Common Grackle had the highest prevalence of parasitic infection, then the Song Sparrow was next highest (Table 3, Figure 2).  Most of the birds caught were in their hatch year age, which also had the highest prevalence excluding the group of five birds which were of unknown or indeterminate ages; no birds of the second year category were found to be infected (Figure 3).  Birds caught at the CEMA banding site had the highest prevalence of malarial infection, and the birds of the NOMA banding site had the lowest (Figure 4). Mann-Whitney Tests resulted in two-tailed p-values of 0.001, 0.762, and 0.003 for comparisons of prevalence between NOMA and CEMA, CEMA and SOMA, and NOMA and SOMA, respectively (Tables 4, 5, and 6).  	Kruskal-Wallis Tests regarding bird age, location, and families resulted in p-values of 0.259, 0.004, and 0.009, respectively (Tables 7, 8 and 9). Tests for sex showed no significant differences between male and female. Table 1. Prevalence of parasitic infection  Table 2. Family, sample size, number infected, and prevalence of infection only in mist-netted birds *The overall prevalence based on the total mist-netted birds sampled N=159. Table 3. Family, species, scientific name, sample size, avian parasite, and prevalence of birds sampled from mist-netting * The Black-Capped Chickadee was the only bird sampled in the family Paridae. The Baltimore Oriole and Common Grackle were the only sampled species in the family Icteridae Figure 1. Prevalence of blood parasites in passerine bird families Correction: Fringillidae should be Emberizidae. Figure 2. Prevalence of parasitic infection in host species grouped by families. From left to right: in the Parulidae family, species are Ovenbird, Northern Waterthrush, Black and White Warbler, Nashville Warbler, and American Redstart; Turdidae include Hermit Thrush and Veery; Emberizidae include the White-throated Sparrow and Song Sparrow; Black-capped Chickadee is the sole representative of Paridae; and the Icteridae include the Baltimore Oriole and Common Grackle. Figure 3. Prevalence of parasitic infection by host age Figure 4. Prevalence of malarial infection by banding site Table 4. Result of Mann-Whitney Test for the prevalence of infection between central and north Maple River Table 5. Results of Mann-Whitney Test for the prevalence of infection between central and south Maple River Table 6. Result of Mann-Whitney Test for the prevalence of infection between north and south Maple River Table 7. Kruskal-Wallis Test by Age Table 8. Kruskal-Wallis Test by Location Table 9. Kruskal-Wallis Test by Family 	The data from this study show that prevalence of avian malaria in Northern Michigan was statistically significant. The prevalence of infection in all of the samples collected was 23.8% and the prevalence in only passerine birds was 10.7%, which supports previous research that blood smears demonstrate infection rates between 10 and 30% (Feldman et al. 1995). 
aemoproteus and Leucocytozoon  are host-specific and are found within related species of a particular family (Fallis et al. 1974), but species of the genus Plasmodium are not as family-specific (Bennett et al. 1982). Results from the current study showed that Plasmodium was represented in five families. Birds of the family Anatidae had the highest prevalence of haemosporidian infection, but this is because of the ubiquitous nature of Leucocytozoon  in that family of birds. The family Corvidae also had a high prevalence of 100%. Prevalence among species also varied dramatically. One factor that may explain the variance is habitat preference of infected bird species. Mosquitoes are found much more often near standing water, which suggests that birds nesting and breeding near standing water should be more susceptible to avian malaria. The Ovenbirds of the family Parulidae are common in all woods with dry floor; the Northern Waterthrushes are mainly in conifer bogs, wooden swamps, and along streams in heavily shaded ravines; the Black-and-White Warblers are common in conifer bogs; the Nashville Warblers are common in moist conifer woods and bogs or open, dry woods; the American Redstarts are common in deciduous woods. The Hermit Thrush of the family Turdidae are common in semi-open woods, and the Veeries are common in densely shaded, often moist woods. The White-throated Sparrows of the family Emberizidae are common in the shrubby edges of conifer bogs and heavily shaded coniferous-deciduous woods, and the Song Sparrows are common in brushy areas. Black-Capped Chickadees of the Paridae family are commonly found in wooded areas. The Baltimore Oriole of the family Icteridae is common at the edges of open deciduous woods, and the Common Grackle is common near lakes and streams, particularly where there are pines and other conifers. (Pettingill 1974). Overall, some of the infected bird species prefer habitats that are potentially mosquito-dense while others do not.  	There was a propensity for prevalence of infection to decrease with bird age except for birds in their second year (Figure 3). These data were inconsistent with previous findings which showed juveniles to be more susceptible to infection than adult birds (Atkinson and Van Riper 1991). Birds should be highly vulnerable to insect bites during the two weeks until they leave the nest because of their bare, feather-less skin, allowing an opportunity for them to be infected by Plasmodium. The small sample size of this study most likely limited the accuracy of the data; with a larger, more thorough sample, one would be able to better interpret the relationship of host age and haematozoan infection. Prevalence of infection was also high in the group of birds with unknown age, since only two out of only five birds were infected.  	The prevalence of infection at the three banding locations differed from one another, and was statistically significant. This could be a result of birds having varying degrees of exposure to the mosquito vector. Birds caught in mist nets near water should show a higher prevalence of infection than those caught elsewhere. Possibly, the CEMA and SOMA locations had more mosquito-rich areas in it, and as a result reflected a higher prevalence of bird malaria in the data.  	Tests for different prevalence in sex showed no significance, which is a result of inadequate data. Many of the birds caught in the mist nets could not be characterized for sex, because they were too young and juveniles are difficult to assess gender even with the expertise of collaborators of this project.  	Prevalence of Haemosporidian infection throughout the birds of Northern Michigan varied significantly by host family, species, age, and location. There are many possible factors that could contribute to this: bird behavior, habitat, concentration of mosquito vectors, climate, or co-evolution (Atkinson and Van Riper 1991). Further studies that look into this would be useful in comparison to distribution patterns of Plasmodium species that infect humans. No conclusions could be made about the effect of Haemosporidians on the bird populations of Northern Michigan, but continued pursuit of this question could supplement data and make more accurate analyses possible. Keely Dinse deserves the most credit in this project for allowing the authors to join her in collecting birds and for giving them all of the needed bird identification data. I would like to acknowledge Harvey Blankespoor for giving his assistance in identifying the parasites, Mike Akresh for helping data collection and data analysis, Dave Gonthier for data collection and assistance in parasite identification, Paula Furey for her assistance in using the camera microscope. Lastly, the University of Michigan Biological Station deserves credit for making this project possible.  The Michigan Corpus of Upper-level Student Papers (MICUSP) is owned by the Regents of the University of Michigan (UM), who hold the copyright. The corpus has been developed by researchers at the UM English Language Institute. The corpus files are freely available for study, research and teaching. However, if any portion of this material is to be used for commercial purposes, such as for textbooks or tests, permission must be obtained in advance and a license fee may be required. For further information about copyright permissions, please contact micusp-help@umich.edu. The recommended citation for MICUSP is: Michigan Corpus of Upper-level Student Papers. (2009). Ann Arbor, MI: The Regents of the University of Michigan.