Philornis parasitism: impact on nestlings and risk factors involved

Parasitic botfly larvae (Philornis ssp., Diptera: Muscidae) are found in nests of several bird taxa, although prevalence and nestling tolerance vary considerably among species. Here we describe patterns of botfly infestation in blue-black grassquit (Volatinia jacarina) nestlings. We identified the most typically affected nestling body parts and assessed parasite prevalence, impact on nestling survival, changes in nestling body shape and mass index. Additionally, we tested whether climatic conditions, nest morphology and habitat characteristics are associated with larvae abundance. Blue-black grassquits had low breeding success (15%), but most failures resulted from predation by vertebrate predators. We estimated that only 1% of nestlings died due to botfly infestation, and the number of larvae in nestling body did not affect nest success. Infected chicks exhibited a higher body mass to tarsus length ratio, and higher tarsus asymmetry. Previous studies indicate that adult grassquits with a higher body mass index have lower dominance status and mating success. Thus, we argue that although botflies had a small impact on offspring survival, they may reduce fitness in adulthood. There was no evidence that environmental conditions and nest morphology are linked to the number of larvae on nestlings. Territories with higher food supply had lower infestation rates. Possibly, food-rich habitats allow parents to invest more time in offspring care (brooding nestlings), thus protecting them from fly attacks. The present study brings to light new perspectives concerning bird-botfly interactionFunding provided by: Coordenação de Aperfeiçoamento de Pessoal de Nível SuperiorCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100002322Award Number: 471945/2013-7Funding provided by: Conselho Nacional de Desenvolvimento Científico e TecnológicoCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100003593Award Number: 1789/2015Funding provided by: Animal Behavior SocietyCrossref Funder Registry ID: http://dx.doi.org/10.13039/100009702Award Number:Study site and species This study was carried out within the University of Brasília campus, in central Brazil (15°45'S; 47°52'W), in an area of 20 ha. The vegetation is classified as Cerrado sensu stricto (tropical savanna) with high plant diversity (Assunção and Felfili 2004, Aguilar et al. 2008). We monitored breeding activities of blue-black grassquits between November to April, across two breeding seasons (2017-2018; 2018-2019). Throughout the breeding season, males blue-black grassquits defend territories and attract mates by performing multimodal displays (Manica et al. 2016, Manica et al. 2017). Both sexes build the nest and provide parental care. The small cup-shaped nests are placed in the forks of shrubs (more frequently) or in dense grass undergrowth (Carvalho et al. 2007, Aguilar et al. 2008). Predation is considered the main factor leading to lack of breeding success (Almeida and Macedo 2001, Carvalho et al. 2007, Aguilar et al. 2008, Dias et al. 2010). Cheating behaviors, such as extra pair paternity and intraspecific brood parasitism, occur in blue-black grassquits (Carvalho et al. 2006, Manica et al. 2016). Little is known about botfly infection in this species, but literature reports indicate that they are infected by the subcutaneous P. glaucinis and P. trinitensis (Teixeira 1999). Data collection We searched for nests by slowly walking across the field site at least twice each week, inspecting the vegetation and watching for birds carrying nesting materials. New and active nests were checked at intervals of up to three days, until chicks fledged or the nest was lost to predation. Incubation and nestling periods lasted up to 10 days each (Carvalho et al. 2007). If eggs were present in a nest on a given day and eggs hatched on the following day, we assumed that hatching had occurred on the second day. For chicks hatching after a checking interval of two-three days, or nests found in the nestling period (20 of 180), we estimated hatching day by comparisons with chicks of known age. The disappearance of eggs before hatching or of nestlings before seven days of age, associated with a damaged nest, was attributed to nest predation. Nestling death due to botfly larvae infestation was assumed when nestlings were found dead in the nest and larvae were found in the nest or nestling. Nest desertion was assumed when parents no longer cared for the eggs, which remained in the nest > 10 days. Whenever eggs were deserted, we collected and opened them to check for development. After this inspection, eggs were classified as "infertile" if no embryo was found, or "death in development" if we found a dead embryo. Nest success was assumed when nestlings disappeared from the nest at or later than 7 days post-hatch and no signs of predation were detected (i.e. the nest remained intact in the vegetation). Nestling body condition was recorded up to three times for each nestling, and included mass (taken with a 10g, 0.1g resolution Pesola spring scale), both left and right tarsus length (digital caliper Mitutoyo 500-196-30B, 0.01 mm resolution), and number of larvae and their location on the nestling´s body (we did not identify larvae to the species level). A body mass index was calculated by dividing nestling weight by average tarsus length. In adult grassquits, it has been found that body mass index correlates negatively with intestinal parasite load (Costa and Macedo 2005, Aguilar et al. 2008) and social dominance (Santos et al. 2009). Nest body asymmetry was calculated based on the absolute difference between left and right tarsus lengths. The location of larvae was mapped onto the nestling´s body areas: head-neck, wings, legs, and main body (see figure 1); large larvae could occupy more than one area. We did not band nestlings at the nest since this could influence natural predation rate by increasing the contrast between nestlings and nest background material. Climatic data were obtained from the open database provided by the Brazilian meteorological institute (Instituto Nacional de Meteorologia - INMET), which provided a regional sampling location less than 10 km from the study site. For each nest, we averaged the daily rainfall and temperature for a period encompassing 14 days, from seven days before and after hatching. When nests were no longer active, they were collected for measurements associated with their architecture, after which they were deposited in the museum collection Coleção Ornitológica Marcelo Bagno, at Universidade de Brasília. We used two variables to characterize nest architecture: nest wall density and nest wall openness. Nest wall density was calculated as the ratio of nest weight to wall volume (mg/mm3). Nest weight was obtained with a high precision balance (Shimadzu BL320H, 1mg resolution) after the nest was air dried at 75ºC for 24h. Nest wall volume was estimated as the difference between external and internal nest wall volumes. Volume was estimated using the semi-ellipsoid volume formula V = 2/3 × π × a × b × c, where a and b are perpendicular measurements of nest outer/inner diameter and c is nest height/depth. Nest wall openness was estimated as the average of four measurements of nest wall openness taken on different sides of the nest. A detailed description of the method is provided in Biagolini-Jr and Macedo (2019). In brief, photos of the nest were taken with a white styrofoam ball (50mm in diameter) placed inside the nest chamber. Then, with an image editor software (GIMP version 2.10) we cropped the styrofoam ball image section and converted it to a black and white scale. Using the R package bwimage (Biagolini-Jr 2019), we estimated nest wall openness as the proportion of white pixels relative to total number of image pixels. We assessed food availability and vegetation density in five spots at distances of 3 m from the nest within two weeks after nestlings fledged. We did not demarcate parental territories, but assume that the sampled spots had a high probability of falling within territories (Aguilar et al. 2008). We estimated the abundance of seed resources by averaging the number of seed inflorescences counted in 50 x 50 cm grids placed at each of the five spots (Manica et al. 2014). Vegetation density was estimated in five plots of 30x100 cm, by adapting the Zehm et al. (2003) method. In summary, a photograph was taken of the vegetation against a panel of 100 x100 cm white cloth placed perpendicularly to the ground on the largest side of the plot. The photograph was converted to a pure black and white image (GIMP version 2.10). Vegetation density was estimated as the proportion of black pixels relative to total number of pixels (Biagolini-Jr and Macedo 2019). Host nest density was estimated as the number of grassquit nests, within a radius of 50 meters of the focal nest being considered, with clutches that hatched in the period of 10 days before and after nest hatching date at the focal nest. We chose this range because it encompasses both the nestling period and the botfly pupation period from a possible previously infected nest (Saravia-Pietropaolo et al. 2018). We used the R package geosphere Version 1.5 (Hijmans et al. 2019) to calculate distances between nests