Portal de Periódicos da CAPES

Prevalence of potentially pathogenic culturable bacteria on eggshells and in cloacae of female Pied Flycatchers in a temperate habitat in central Spain

2011; Association of Field Ornithologists; Volume: 82; Issue: 2 Linguagem: Inglês

10.1111/j.1557-9263.2011.00324.x

1557-9263

Rafael Ruiz‐de‐Castañeda, Ana I. Vela, Elisa Lobato, Víctor Briones Dieste, Juan Moreno,

Animal Nutrition and Physiology

ABSTRACT Bacteria grow on avian eggshells and thus can potentially cause diseases in developing embryos. Little is known about culturable bacteria colonizing avian eggshells in free-living birds, with most studies restricted to poultry. Our objective was to examine the culturable bacterial array growing on eggshells during incubation that could negatively affect hatching success of Pied Flycatchers (Ficedula hypoleuca) in a temperate montane habitat in central Spain. Cloacal culturable bacteria of females were also analyzed because bacteria can be vertically transmitted from females to eggs. We used fecal samples as surrogates of cloacal samples due to the small size of sampled birds. We found that eggshells and female cloacae of Pied Flycatchers harbored 24 and 40 bacterial families and species, respectively, but only a few in each clutch and each cloaca. Rod-shaped gram-negative bacteria and bacteria in the family Pseudomonadaceae were the most common bacteria on eggshells during early and late incubation and in female cloacae. Although based on small sample sizes, we found that females with rod-shaped gram-negative bacteria in their cloacae laid eggs that also had these bacteria, providing possible evidence for vertical transmission. We found no evidence for vertical transmission of Pseudomonadaceae, suggesting a possible environmental source for these bacteria. The prevalence of bacterial morphological types and major taxonomical categories on eggshells did not vary from early to late stages of incubation, providing support for the hypothesis that incubation may have bacteriostatic effects on bacterial proliferation on eggshells. Despite being primary egg invaders in poultry, we detected no effects of culturable Pseudomonadaceae or Pseudomonas luteola on hatching success. Our study represents the first to examine the culturable bacteria growing on the eggshells of a wild bird in a temperate habitat and additional studies based on culture-independent techniques are required to confirm our results. Las bacterias crecen sobre las cáscaras de huevos de las aves, y por tanto podrían causar enfermedades en los embriones en desarrollo. Poco se conoce acerca de la colonizacion de bacterias cultivables en las cáscaras de huevo de aves silvestres, y la mayoría de los estudios están restringidos a las aves de corral. Nuestro objetivo fue examinar las bacterias cultivables que crecen en la cáscara de huevos durante la incubación que podrían afectar negativamente el éxito de eclosión de Ficedula hypoleuca en un hábitat templado de montaña en el centro de España. Las bacterias cultivables de la cloaca de las hembras también se analizaron ya que las bacterias pueden ser transmitidas verticalmente de las hembras a los huevos. Utilizamos muestras de materia fecal como sustitutos a las muestras cloacales debido al pequeño tamaño de las aves muestreadas. Encontramos que la cloaca y las cáscaras de huevo de F. hypoleuca tenían 24 y 40 familias y especies de bacterias, respectivamente, pero sólo unos pocos en cada puesta de huevos y en cada cloaca. Bacterias gram-negativas en forma de bacilo y las bacterias de la familia Pseudomonadaceae fueron las bacterias más comunes sobre las cáscaras de huevo durante las etapas tempranas y tardías de la incubación, y en las cloacas de las hembras. Aunque se basa en pequeños tamaños de muestra, encontramos que las hembras con bacterias gram-negativas en forma de bacilo en sus cloacas pusieron huevos que también tenían estas bacterias, proporcionando posible evidencia de una transmisión vertical. No encontramos pruebas de transmisión vertical de Pseudomonadaceae, lo que sugiere un posible origen ambiental para estas bacterias. La prevalencia de tipos morfológicos y de las principales categorías taxonómicas de bacterias sobre las cáscaras de huevo no variaron desde el principio hasta las finales de la incubación, apoyando la hipótesis de que la incubación puede tener efectos bacteriostáticos sobre la proliferación de las bacterias en la cáscara del huevo. A pesar de ser invasores primarios de huevos en las aves de corral, no detectamos efectos de Pseudomonadaceae o de Pseudomonas luteola cultivables sobre el éxito de eclosión. Nuestro estudio representa el primero en examinar las bacterias cultivables que crecen en las cáscaras de huevo de un ave silvestre en un hábitat templado, sin embargo otros estudios basados en técnicas independientes del cultivo son necesarios para confirmar nuestros resultados. Information regarding culturable and non-culturable potentially pathogenic bacteria present on the eggshells of wild birds is scarce, with most available information restricted to poultry and focusing on the economical and medical implications of these pathogens (Humphrey 1994, Tauxe 1997, Rasekh et al. 2009). Two recent studies, both focusing on Pearly-eyed Thrashers (Magarops fuscatus), provided information about bacteria growing on eggshells in the wild (Cook et al. 2005, Shawkey et al. 2009). Based on traditional culture-dependent (Cook et al. 2005) and culture-independent (Shawkey et al. 2009) methods, both studies revealed that incubation may inhibit bacterial growth and diversification on eggshells. Cook et al. (2005, but see Shawkey et al. 2009) also found that incubation may selectively enhance the proliferation of certain types of beneficial bacteria (rod-shaped gram-positive) capable of out-competing pathogenic bacteria also growing on eggshells. In general, Cook et al. (2005) identified culturable bacteria to major group level, with rod-shaped gram-positive, enterics gram-negative, fermenters gram-negative, cocci gram-positive, and enterics gram-negative being the most common bacteria on eggshells at laying. Shawkey et al. (2009) produced a broader taxonomical description at the level of family, genus, and species. Despite the importance of these two studies for understanding the microbiology of avian eggs, their descriptive and functional outcomes may be limited to tropical habitats where temperature and moisture levels are uniformly high throughout the year, a fact of key relevance for bacterial proliferation. To our knowledge, no information is available from temperate areas regarding the prevalence of culturable bacteria colonizing eggshells at different stages of incubation and their association to those present in female cloacae. One objective of our study was to provide a description of the bacteria growing on eggshells and in female cloacae by focusing specifically on a portion of the culturable bacteria, enumerating bacteria that could potentially impact directly or indirectly hatchability in a population of Pied Flycatchers (Ficedula hypoleuca) breeding in a temperate montane habitat in central Spain. Additional objectives were to identify a possible association between culturable bacteria on eggshells and in female cloacae due to possible vertical transmission, detect possible changes in the prevalence of culturable bacteria growing on eggshells from early to late stages of incubation, and evaluate the possible effects of the most prevalent bacteria on hatching success. Study species and study area Pied Flycatchers are small (12–13 g), migratory, cavity-nesting passerines of European woodlands that readily nest in nest-boxes. Our study was conducted during spring 2008 in a deciduous forest of Pyrenean oak (Quercus pyrenaica) at 1200 m asl near Valsaín, central Spain (40° 54'N, 04°01'W). The structure of the habitat was homogeneous, with most oaks being of similar age. Egg laying by Pied Flycatchers typically begins in late May, and clutch sizes range from 2 to 8 (Sanz 1997). Females normally lay one egg per day during the laying period, and start incubating before clutch completion (generally the day the fourth egg is laid). The incubation period lasts about 12 to 15 days and only females incubate. Field sampling Nest boxes were visited daily beginning early in the breeding season and dates of clutch initiation (laying date) were recorded. We sampled 82 and 62 clutches during early (mean number of days after clutch completion = 0.56 ± 0.09 [SE]) and late (mean number of days after early incubation sample = 10.24 ± 0.10) stages of incubation, respectively. Because last-laid eggs (fifth or sixth eggs) were swabbed soon after laying, we assumed that initial samples partly reflect bacteria present in maternal cloacae, allowing us to test for vertical transmission (see details below). Only sub-samples of 56 (early incubation) and 29 (late incubation) clutches were analyzed due to logistic limitations (time limitations and cost). We selected clutch samples for which we could also obtain female cloacal samples (see below). The remaining clutch samples for which we did not have associated cloacal samples were selected randomly. To minimize possible contamination by external bacteria, we used sterile sampling techniques. Eggs were handled with latex gloves washed with 97% ethanol. We either held our breath when handling eggs or wore face-masks. Surfaces of all eggs in a clutch were entirely swabbed once for about 8 s per egg using one sterile cotton swab (Sterile R, Meus s.r.l., Piove di Sacco 35028 [PD], Italy) previously moistened with sterile phosphate-buffered saline (4 g NaCl, 0.1 g KCl, 0.72 g Na2HPO4, and 0.12 g K2HPO4), obtaining one sample per clutch. Eggs were taken one at a time from nests and placed in a bowl with clean paper tissue and sequentially placed back in nests after sampling, preventing us from sampling the same egg more than once. At intermediate stages of incubation (mean number of days after early incubation sample = 6.6 ± 1.5), 65 females were captured while incubating. We assumed that the periods between sampling cloacae and eggshells during early incubation were short enough to permit comparison of female and eggshell samples. Capturing females prior to this stage entails a high risk of nest desertion in our population and was therefore impractical. This prevented us from sampling female cloacae during the laying period. Cloacae samples were obtained from 35 of 65 captured females, but we only considered 26 samples in our study due to an excessive delay in analysis ( >7 d) for the rest. Feces produced by females during handling or present on the immediate outer membrane of the cloacal orifice were sampled with a flexible sterile cotton swab (cotton head of about 2–3 mm thick; Sterile R, Nuova Aptaca s.r.l., Canelli 14053 [AT], Italy). Swabs were uniformly impregnated with feces for approximately 1 to 3 s. Swabs could not be inserted into cloacae due to the small size of the birds. We assumed that bacteria in cloacae were similar to those in feces and other exudates ejected from the oviduct as reported for domestic chickens (Gallus gallus; Barrow 1994, Humphey 1994). Although this assumption, to our knowledge, has not been tested in any species of wild bird, fecal bacteria may represent cloacal bacteria if feces are collected immediately after ejection from the cloacae as in our study. Also, feces are usually associated with intestinal materials, but this is not strictly the case for birds where exudates from the urinary and genital systems converge with the digestive tract in the cloaca. Thus, we avoid the term fecal because bacteria collected may include those originating in other tracts (e.g., Gardnerella vaginallis detected in our study). Hereafter, therefore, we will refer to cloacal bacteria rather than fecal bacteria. Immediately after sampling, swabs were transferred into Transport Media Amies (Sterile R, Meus s.r.l.), placed in a portable cooler, and transported to the lab where they were placed in refrigerators (4° C) and stored until analyzed. All samples were analyzed within 7 d. We tested for the possible effect of time between sampling and analysis on the prevalence of the most representative culturable bacterial morphological types (cocci gram-positive, rod-shaped gram-negative, and rod-shaped gram-positive) and families on eggshells during early and late incubation (Pseudomonadaceae, Enterococcaceae, and Enterobacteriaceae) and in female cloacae (Burkholderaiceae, Pseudomonadaceae, Enterobacteriaceae, and Enterococcaceae), but found no effects for either eggshells (all P > 0.20) or female cloacae (all P > 0.16). Isolation and identification of prevalent culturable bacteria on eggshells and cloacae Swabs were placed in 3.5 ml of sterile phosphate-buffered saline, and 0.5 ml of swab supernatant was transferred to sterile vials containing 2 ml of cryopreservant media (320 ml H2O, 80 ml glycerol, 20 g skimmed milk, and 10 g tryptone) and preserved at –40°C. After the field season, 20 μl of the cryo-preserved samples was plated onto Columbia agar +5% sheep blood and MacConkey agar (Biomerieux s.a., Madrid 28034, Spain) using the streak-plate technique. We divided the circular surface of plates into four quadrants. The top-left quadrant was considered the first and the successive second, third, and fourth quadrants were considered in a clockwise direction. We first inoculated the first quadrant and then streaked the remaining quadrants in succession. Thus, the concentration of the sample was gradually reduced from the first to the fourth quadrant. Columbia agar +5% sheep blood is a highly nutritious general purpose medium for the isolation and cultivation of non-fastidious and fastidious microorganisms from a variety of clinical and nonclinical materials. The sheep-blood supplement allows the culture of most culturable pathogenic bacteria, regardless of their metabolic requirements (http://www.biomerieuxdiagnostics.com). MacConkey agar with crystal violet is a selective isolation and differentiation medium used for detection of culturable Enterobacteriaceae in samples of various origins (http://www.biomerieuxdiagnostics.com). Bacteria were incubated both aerobically and anaerobically in Columbia agar +5% sheep blood, and aerobically in MacConkey agar at 37 ± 1°C for 24 to 48 h. Anaerobic atmospheres for growing culturable fastidious bacteria were generated using GENbag anaer incubation systems (Biomerieux s.a.). We collected samples in the field in aerobic conditions, but strict anaerobes could be present in samples in the form of spores that would only grow under anaerobic conditions in the lab (i.e., Clostridium; González-Braojos et al., unpubl. data). After incubation, plates with no bacterial growth were not considered further. For analysis, we only considered colonies growing in the third and fourth quadrants where bacteria were completely isolated. Plates were analyzed simultaneously and comparatively for each of the three culture conditions described previously (aerobically and anaerobically in Columbia agar +5% sheep blood, and aerobically in MacConkey agar). Based on colony morphology, we selected all distinct colonies growing in each medium and streaked them onto Columbia agar +5% sheep blood and incubated aerobically or anaerobically at 37 ± 1°C for 24 to 48 h. Morphologically typical Enterobacteriaceae colonies were isolated from MacConkey agar. Other colonies were isolated from Columbia agar + 5% sheep blood. Colonies present on more than one plate were only isolated from one plate. Pure cultures were microscopically characterized in terms of morphology (cocci or rod-shaped) and gram staining (gram-positive or gram-negative). Finally, bacteria were biochemically identified at the genus and species levels using ad hoc multi-substrate identification systems (API® Strips, BioMérieux): rapid ID 32 Strep, API 32 Staph, API 20 NE, API 20 E, and API Coryne. Interpretations of the API® Strips were made in all cases by the same observer (Ruiz-de-Castañeda). Taxonomic classification of bacteria at the level of family was made using the Euzéby Data Base (http://www.bacterio.cict.fr/classification.html). Statistical analyses Statistical analyses were performed using Statistica 7.0 (StatSoft 2001). Bacterial morphological types (cocci gram-positive, rod-shaped gram-negative, and rod-shaped gram-positive bacteria) and taxonomical categories (family, genus, and species) were considered as binomial variables referring to the presence–absence of each bacterial morphological type and taxonomical category on eggshells during early or late incubation and in female cloacae. This allowed us to calculate the prevalence of each bacterial morphological type and taxonomic category on eggshells and in female cloacae. Prevalence for each bacterial morphological type and taxonomy category was defined as the proportion of samples showing bacterial colonies on the third and fourth quadrants of plates for the specific type or category. We controlled for the possible effects of clutch size on prevalence of most common bacterial morphological types (cocci gram-positive, rod-shaped gram-negative, and rod-shaped gram-positive bacteria) and families (Pseudomonadaceae, Enterococcaceae, and Enterobacteriaceae), but no significant effects were detected (all P > 0.19) for either early or late incubation. We used two-tailed binomial tests (α= 0.05) to detect possible associations between the presence or absence of certain bacteria in cloacae from laying females and eggshells during early and late incubation. We considered the number of pairs where females and their clutches had the same bacterial morphological type and the number of pairs where females and their clutches did not carry the same bacterial morphological type. Due to the low prevalence of most bacterial families identified on eggshells and in cloacae as well as small sample sizes and unpaired data, we analyzed the possible vertical transmission of bacteria for all bacterial morphological types and only the most prevalent bacterial family (Pseudomonadeceae) and species (Pseudomonas luteola) found on eggshells and in female cloacae. Siegel (1956) provides the one-tailed probabilities for the occurrence of various values as small as x for various N (from 5 to 25). We obtained the two-tailed probabilities by doubling the one-tailed probabilities. When N was 15%), genera, and species (prevalence >10%) due to the occasional presence of many bacterial species on eggshells. We analyzed possible effects of Pseudomonadaceae and Pseudomonas luteola on hatching success using two-tailed Fisher's exact tests (α= 0.05). Hatching success in our population during our study was high (mean = 94 ± 2 [SE]%). All eggs hatched in 57 nests (hatching success = 100%); at least one egg did not hatch in 19 nests (hatching success 80%. Thus, due to the low variability of hatching success in our population, we considered this variable as a binomial dependent variable, i.e., either total hatching success or partial hatching success when at least one egg did not hatch. Cell morphology and gram stain Bacterial growth was obtained for 28 of 56 (50%) and 22 of 29 (75.9%) clutches during early and late incubation, respectively, and for all cloacal samples. We identified three bacterial morphological types, including cocci gram-positive, rod-shaped gram-negative, and rod-shaped gram-positive. We did not detect cocci gram-negative bacteria. Eggshell bacterial morphological types during incubation Cocci gram-positive, rod-shaped gram-positive, and rod-shaped gram-negative bacteria were present on eggshells during both early and late incubation. The number of detected cocci gram-positive, rod-shaped gram-positive, and rod-shaped gram-negative bacteria on eggshells did not differ between early and late incubation (Fisher's exact test, all P > 0.07, N= 50). Bacterial families We identified 24 culturable bacterial families. The prevalence of most (20, or 83.3%) families on eggshells or female cloacae was 15% on eggshells or female cloacae. Of these families, Enterococcaceae and Pseudomonadaceae were detected on eggshells during early and late incubation as well as in female cloacae. Overall, Pseudomonadaceae was the most common family. Enterobacteriaceae was detected in female cloacae and on eggshells during early incubation, but not late incubation. Finally, Burkholderiaceae was only detected in cloacal samples (Table 1). Frequencies of Enterobacteriaceae, Enterococcaceae, and Pseudomonadaceae did not change throughout incubation (Fisher's exact test, all P > 0.50, N= 50). We found no association between the prevalence of Pseudomonadaceae in cloacae and on eggshells during either early (binomial test, P= 1.0, N= 9) or late (binomial test, P= 0.20, N= 16) incubation. Bacterial genera We identified 30 genera of bacteria on eggshells and in cloacae. Bacteria on eggshells belonged to 24 genera, with 10 genera present on eggshells during both early and late incubation. Only Acinetobacter, Enterococcus, Ochrobacter, and Pseudomonas had prevalences >10% during early or late incubation (Table 1). The abundance of these genera did not vary during incubation (Fisher's exact test, all P > 0.62, N= 50). The six remaining genera present during both early and late incubation were not analyzed due to low prevalence ( 10%. The nine remaining genera were occasionally detected in cloacae and had prevalences 10% on eggshells during early and late incubation and in female cloacae, respectively. In all cases, Pseudomonas luteola was the most common bacterium (Table 1). We found 10 species of bacteria on eggshells during both early and late incubation, but only Pseudomonas luteola had a prevalence >10% during early and late incubation. Pseudomonas luteola was the most common bacteria. No changes were detected in the abundance of Pseudomonas luteola on eggshells from early to late incubation (Fisher's exact test, all P > 0.74, N= 50). Two species of Enterobacteriaceae (Enterobacter amnigenus and Rahnella aquatilis) were independently detected from two different clutches during early incubation, but not during late incubation. We found seven and five species of bacteria in both female cloacae and on eggshells during early incubation and late incubation, respectively. Again, Pseudomonas luteola was the most common bacterium. However, we found no association between the presence of these bacteria in female cloacae and their presence on eggshells during either early (binomial test, P= 0.18, N= 9) or late (binomial test, P= 0.39, N= 16) incubation. Although sample sizes were too small for statistical analysis, we found that two of three and two of four clutches, respectively, did not harbor Pseudomomas luteola when not present in female cloacae. Acinetobacter lwoffii, Aerococcus viridans, Leuconostoc spp., and Pseudomonas luteola were detected on eggshells during both early and late incubation as well as in female cloacae. Association between cloacal and eggshell bacterial morphological types Rod-shaped gram-negative bacteria were the most abundant bacterial morphological type on the eggshells during early and late incubation and in female cloacae. We found a significant association between the presence of rod-shaped gram-negative bacteria in female cloacae and their presence on eggshells during early incubation (binomial test, P < 0.012, N= 11), with bacteria present in both female cloacae and eggshells in 10 of 12 cases. We also found a significant association between the presence of rod-shaped gram-negative bacteria in female cloacae and their presence on eggshells during late incubation (binomial test, P < 0.004, N= 19), with bacteria present both in female cloacae and on eggshells in 17 of 20 cases. We found no association between the absence of cocci gram-positive bacteria in cloacae and their absence on eggshells during early incubation (binomial test, P= 0.73, N= 8). However, the absence of cocci gram-positive bacteria in female cloacae was significantly associated with their absence on eggshells during late incubation (12 of 20 cases; binomial test, P= 0.036, N= 15). Rod-shaped gram-positive bacteria were not in female cloacae, but were present on eggshells during early incubation. Thus, the absence of rod-shaped gram-positive bacteria in female cloacae was significantly associated with their absence on eggshells during early incubation (10 of 12 cases; binomial test, P= 0.002, N= 10). The absence of rod-shaped gram-positive bacteria in female cloacae was significantly associated with their absence on eggshells during late incubation (13 of 20 cases; binomial test, P= 0.05, N= 17) Hatching success We detected no significant effects of the prevalence of Pseudomonadaceae and Pseudomonas luteola on eggshells on hatching success either during early (Fisher's exact test, P > 0.14, N= 21) or late incubation (Fisher's exact test, P > 0.57, N= 25). Culturable bacterial array on eggshells and in female cloacae Based on culture-dependent methods and focusing on the most prevalent culturable bacteria that may cause disease, we found that eggshells and female cloacae of Pied Flycatchers harbored a number of bacterial species, but only a few on each clutch and in each cloaca. Only rod-shaped gram-negative bacteria and Pseudomonadaceae were well represented, constituting most of the bacteria identified on eggshells and in female cloacae. One possible explanation for this result may be the diverse origins of culturable bacteria detected in our study that could colonize eggshells and cloacae. Bacteria detected on eggshells may be in part associated with the fecal bacteria of laying females (Barrow 1994) that, in turn, may depend on ecological factors such as diet and habitat (Blanco et al. 2006, Klomp et al. 2008), micro-environmental factors associated with the anatomical structure of the digestive tract (Berg 1996, Stevens and Hume 1998), and bacteria from reproductive tracts (Lombardo et al. 1999, Stewart and Rambo 2000, Hupton et al. 2003). In addition, bacteria on eggshells may have an environmental origin. Competitive interactions among bacteria on eggshells and in cloacae may also influence the relative abundance of the species detected, ultimately favoring the presence of only a few dominant groups (Shawkey et al. 2009) such as rod-shaped gram-negative bacteria and Pseudomonadaceae (in terms of bacterial morphological type and family, respectively), but maintaining an actively changing pool of culturable bacteria associated with the permanent flow of new maternal or environmental colonizers. The most common family of bacteria on eggshells and in cloacae of female Pied Flycatchers in our population was Pseudomonadaceae (genus Pseudomonas), possibly because the genus Pseudomonas is arguably the most diverse and ecologically significant group of bacteria on the planet (Stanier et al. 1966, Spiers et al. 2000). Pseudomonas bacteria have also been recently described as the most prevalent in the nests of hole-nesting passerines (Goodenough and Stallwood 2010). The dominance of Pseudomonadeace in our study must be carefully interpreted within the context of the culture-dependent approach we used, so we can only confirm that this was the most prevalent family among the culturable bacteria considered in our study. Enterobacteriaceae is a diverse group of bacteria that includes many of the most studied animal culturable enteropathogens, such as Salmonella spp. and Escherichia coli (Humphery 1994, Janda and Abbott 2005, Benskin et al. 2009). Shawkey et al. (2009) described a wide array of Enterobacteriaceae colonizing eggshells in a tropical songbird. We found few Enterobacteriaceae on eggshells and only two species (Enterobacter amnigenus and Rahnella aquatilis) were detected, possibly due to a loss of viability when facing environmental conditions after eggs are laid (Baker 1990). Strict anaerobes, such as Clostridium, are part of the normal microbiota of the alimentary tract of chickens (Barrow 1994). Although this may also be the case for wild birds, information about these bacteria is mainly restricted to reports of intestinal anaerobes with lethal effects on hosts (Asaoka et al. 2004, Hagen and Bildfell 2007). However, we detected no strict anaerobes in cloacal samples. Strict anaerobes may be present in cloacal samples of Pied Flycatchers (González-Braojos et al. unpubl. data), but may be uncommon and could rapidly lose their viability. However, in future studies of strict anaerobes, investigators should avoid enrichment and use culture-independent techniques. Association between cloacal and eggshell bacteria We obtained only limited indirect evidence of vertical transmission. Although based on small sample sizes, we found that females harboring rod-shaped gram-negative bacteria in their cloacae laid eggs that also harbored these bacteria, and that females not harboring either cocci gram-positive bacteria or rod-shaped gram-positive bacteria in their cloacae laid eggs that did not harbor these bacteria. This may be due to the fact that rod-shaped gram-negative bacteria were the most abundant bacterial morphological type in cloacae and on eggshells. Pseudomonadaceae were found on both eggshells and in female cloacae, suggesting possible vertical transmission. However, no female Pied Flycatchers harboring Pseudomonadaceae in their cloacae laid eggs that also harbored these bacteria. Thus, we cannot rule out the possibility that environmental Pseudomonadaceae may colonize eggshells (Goodenough and Stallwood 2010). Enterobacteriaceae were detected in female cloacae and on eggshells, but only during early incubation. If some Enterobacteriaceae are vertically transmitted, their limited presence on eggshells could be due to a loss of viability when facing environmental conditions after eggs are laid (Baker 1990). In addition, by culturing in MacKonkey agar, we would have failed to detect non-viable Enterobacteriaceae growing on eggshells soon after laying, providing an underestimation of their possible total prevalence (viable and non-viable) on eggshells. In general, our results are inconclusive concerning the possibility of vertical transmission of bacteria from females to their eggs. Eggshell bacteria during incubation The prevalence of bacterial morphological types and major taxonomical categories on eggshells did not vary between early to late stages of incubation. These results provide at least some correlative support for the hypothesis that incubation has possible bacteriostatic effects on bacterial proliferation on eggshells (Cook et al. 2005, Shawkey et al. 2009). However, we cannot exclude the possibility that alternative mechanisms independent of incubation may explain this possible stable pattern of bacterial composition during incubation. Hatching success Pseudomonadaceae has been described as the most common egg-spoilage microorganism in poultry (Dockstater 1952, Florian and Trussell 1957) due to their potential to degrade the cuticle of eggshells and increase the number of unplugged pores, exposing them to bacterial penetration (Baggott and Graeme-Cook 2002). However, we detected no effect of Pseudomonadaceae and Pseudomonas luteola on hatching success. Bacterial infection may generally take place under particularly humid conditions and when warm eggs cool down to environmental temperature after laying, producing a negative pressure that tends to pull bacteria into an egg (Bruce and Drysdale 1994). In addition, defensive adaptations against bacteria may be effective at preventing bacterial infection of embryos in temperate habitats (Peralta-Sánchez et al. 2010) as well as in humid tropical habitats more favorable for bacterial proliferation (Cook et al. 2003, 2005). By incubating, females may reduce moisture on eggshells, preventing bacterial proliferation and the potential infection of embryos (Cook et al. 2005). Limitations and future prospects Our microbiological approach was restricted to potentially pathogenic culturable bacteria on eggshells and in female cloacae and thus our results and conclusions must be strictly considered within this scope. For logistical reasons, bacterial samples we collected in the field were not processed on the same day as sampling and were cryopreserved for variable time intervals. We found no significant effect of this delay on the prevalence of major bacterial groups, but such delays and cryopreservation could affect the viability of some bacteria. Thus, in future studies, bacterial samples should be analyzed soon after collection in the field to obtain more precise information about the original prevalence of bacterial groups on eggshells and in cloacae. Our study represents the first to examine the culturable bacteria growing on eggshells in a wild bird in a temperate habitat and additional studies based on culture-independent techniques are required to confirm our results. We thank J. Donés (Director de Montes de Valsaín) for permission to work in the study locality. "Junta de Castilla y León" authorized the banding and handling of birds. We thank J. J. Soler and his research group for helpful discussions and guidance in field methodological procedures, S. Merino, J. Martínez-de la Puente, S. del Cerro, J. Rivero, and S. González-Braojos for their help in the field, and members of the DICM-VISAVET group for their guidance in laboratory analysis. This study was funded by projects CGL2010-19233-C03-02 (to J. Moreno) from Ministerio de Educación y Ciencia, Spain. E. Lobato and R. Ruiz-de-Castañeda were supported by grants MEC and JAE-CSIC, respectively. This study is a contribution to the research developed at El Ventorrillo field station and complied with current Spanish laws.