We found a high bird richness in the three study areas evaluated being similar among them. These findings are consistent with research carried out in the Neotropics, where there is a great heterogeneity of habitats and climates, even in small patches of vegetation a high richness of species is found51.
Overall, 863 individuals belonging to 90 different species were caught in mist nets. We obtained 678 blood samples for Haemosporidian diagnosis resulting in 66 infections (prevalence: 9.73%) for the three sites evaluated. The number of infections and the prevalence were similar to those reported for the Neotropics. More explicitly, in areas near to these three dams, Martínez Alvarado52 found a prevalence of 14.8% and Pulgarín et al.53 reported a prevalence of 14.3%. However, clear differences were found with respect to the parasite genera studied. Haemoproteus and Plasmodium presented a similar number of infections, 30 and 27, which represented a prevalence of 4.42% and 3.98%, respectively. The literature reports that both genera present a heterogeneous distribution, a huge variable prevalence (0-100%), and a great diversity of lineages56–57, a pattern also found here. On the other hand, we found that the number of infections of Leucocytozoon was low (n = 9, prevalence 1.33%) with 5 out of the 9 Leucocytozoon infections found in migratory species including Catharus minimus (n = 1) and Catharus ustulatus (n = 4). Lotta-Arevalo56 and Matta and Rodriguez57 found that the prevalence of Leucocytozoon varied from 10–25% in low altitude locations and is in general, reported in migratory birds. In addition, it is possible that the low prevalence of Leucocytozoon is due to the low availability of appropriate or susceptible vectors for transmission and maintence of the parasite cycle.
Although we found a similar number of infections and prevalence of haemosporidian among the evaluated sites, differences were found in some species of bids. For example, in the case of A. aurantiirostris, high prevalences were found in Porce II and Porce III (80% and 62.5%, respectively) while none of the 11 individuals sampled in Playas were infected. Species of this family (Passerelidae, formerly Emberizidae) usually show high prevalences of infection (> 40%) in the Neotropics54,62. However, our results suggest that local differences may exist in the prevalence of infection, which could be due to different factors. First, it is possible that the availability of food, associated with greater plant cover, favors the response of the species to infection, where better resources improve the body condition of birds, making individuals less susceptible to infection58. In support of this possibility, in the Playas dam, there is a high plant diversity59 that may potentially explain the absence of parasite infection in A. aurantiirostris in this locality. Several studies indicate that the availability of food for birds increases the innate and adaptive response to immune challenges produced by emerging diseases60. Additionally, the ecology of the vector can affect the patterns of infection, since insect vectors could be favored by open areas showing a greater diversity in areas with heterogeneous anthropized coverage, including agricultural landscapes61. Thus, the vector community is an important predictor, for example, explaining the variation in the prevalence of Plasmodium in birds28, so future efforts should implement the characterization of avian haemosporidian vectors in the Neotropics.
We found significant relationships between the number of infections, the estimated richness of Haemoproteus and Plasmodium, and the variables reflecting the avian community, that is, bird dominance. However, we failed to identify any significant relationship between haemosporidian prevalence and the community variables considered (host richness and dominance).
The relationships between the exponential and logarithmic terms of host dominance and parasite variables were not unexpected, due to the high number of species recorded (270) and under the hypothesis that a high host diversity may provide different niches for different parasite lineages62,64, it is possible that amplification patterns are favored. This could be due to the availability of host-saturated environments under favorable environmental conditions for the development of the parasite life cycle68,70. Species vary in their diluting and amplifying capacity based on their abundance, susceptibility, and transmission potential, and thus certain species can disproportionately affect disease risk2,71. This makes it necessary not only to evaluate the patterns associated with the diversity of a host community but also to analyze independently whether each host is competent or not for the parasite and whether their abundances generate a dilution or amplification effect in the community, as has already been evaluated with Passer domesticus in Spain66.
In our case, species such as Manacus manacus, Machaeropterus striolatus, Myiarchus tuberculifer and Arremon aurantiirostris, which present the highest number of infections in almost all the sites sampled, are the ones that generate the amplification effect in these communities. Huspeni and Lafferty68 found that, during restoration or succession processes, host species present higher prevalences and richness of parasites compared to more conserved sites. The studied dams present around 30 years of restoration processes as dam protection zones. In addition, the species of birds are usually found in poorly conserved areas and under restoration processes69. However, as new non-competent (diluting) species become available and as forests become older and the species assemblage stabilizes, it could be expected that the abundance of competent (amplifying) species will be reduced and, therefore, exposure to the parasite will be reduced2.
Although some parasite interactions have been characterized in detail28,31,33, little is known about the dynamics or even the patterns of parasites at the community level. For example, Ricklefs et al.67, evaluated relationships between hosts and prevalence of malaria parasites, finding that this presented a U-shaped relationship (quadratic regression) with host sample size, however, they assumed abundances of individuals captured in mist nets as a relative proportion of abundances in communities, unlike our study, where census-reported abundances were taken into account to estimate the dominance index, which may represent a proxy for community-level abundances. The low representation of studies at the avian community level and haemosporidian infection including census data within the analysis represents a novelty in the characterization of these interactions, which highlights the need to further evaluate the different actors in these interactions, such as mosquitoes and infection in other vertebrate groups, which although not taken into account within these analyses, could be an important factor in explaining the form of these relationships.