INTRODUCTION
The study of biting midges of the genus Culicoides Latreille (Diptera: Ceratopogonidae) is of great importance not only because females are obligate blood feeders attacking a huge diversity of vertebrates (Downes, Reference Downes1958; Kettle, Reference Kettle1995; Marquardt et al. Reference Marquardt, Demaree and Grieve2000), but also because they are vectors of a large number of transmissible agents. Some of these pathogens, including viruses (Braverman et al. Reference Braverman, Messaddeq, Lemble and Kremer1996; Mellor et al. Reference Mellor, Boorman and Baylis2000) and other parasites such as protozoa and filarial worms (Fallis and Wood, Reference Fallis and Wood1957; Atkinson et al. Reference Atkinson, Greiner and Forrester1983; Shelley and Coscarón, Reference Shelley and Coscarón2001; Garvin and Greiner, Reference Garvin and Greiner2003; Mullens et al. Reference Mullens, Cardona, McClellan, Szijj and Owen2006), have economic and veterinary importance.
Females of biting midges, the only sex that requires blood, are infected by blood parasites when they obtain a meal from an infected host. With the exceptions of few non-biting species and autogenous species that require a bloodmeal only after laying their first egg batch, most Culicoides females need to obtain blood for their first ovarian development (Downes, Reference Downes1958). Many studies on Culicoides have been conducted to identify the mechanisms affecting their host selection processes and feeding patterns. However, in the wild, there is scant information about ecological relationships between Culicoides and their hosts, especially in the case of wild birds. The main reason for the scarcity of this kind of study is probably the absence of an effective method of capture. Usually, biting midges are captured using different gadgets such as light traps, CO2 traps placed close to the animals or directly vacuuming them from the animals' bodies (i.e. Bennett, Reference Bennett1960; Braverman et al. Reference Braverman, Boorman, Kremer and Delecolle1976; Zimmerman and Turner, Reference Zimmerman and Turner1983; Mushi et al. Reference Mushi, Chabo, Isa, Binta, Kapaata and Bathuseng1999; Yu et al. Reference Yu, Wang and Yeh2000; Mullens et al. Reference Mullens, Owen, Heft and Sobeck2005). However, these methods are difficult to use in avian nests, especially for the study of midges attacking birds of species breeding in nests placed in cavities.
Arthropod–host interactions involve fascinating behavioural processes and chemosensory mechanisms and chemicals that allow vectors to express host-selection behaviours resulting in non-random biting (Mukabana et al. Reference Mukabana, Takken, Coe and Knols2002; Tomás et al. Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b). Visual as well as antennal and maxillary receptors may be involved in host-location (Bowen, Reference Bowen1991). Culicoides have receptors sensitive to a diversity of host derived products such as lactic acid, 1-octen-3-ol and CO2 (Bhasin et al. Reference Bhasin, Moredue (Luntz) and Mordue2000a; Grant and Kline, Reference Grant and Kline2003), which produce attractive effects (Blackwell et al. Reference Blackwell, Dyer, Mordue (Luntz), Wadhams and Mordue1996; Gibson and Torr, Reference Gibson and Torr1999; Marquardt et al. Reference Marquardt, Demaree and Grieve2000; Mordue, Reference Mordue (Luntz)2003; Mands et al. Reference Mands, Kline and Blackwell2004). Also, the presence of volatile pheromones produced by parous midge females may attract other females, as reported by Blackwell et al. (Reference Blackwell, Dyer, Mordue (Luntz), Wadhams and Mordue1994) in their study of an autogenous species, the biting midge C. impunctatus. In addition, as may occur under natural conditions, host-derived volatile components may interact with parous female pheromones, either attracting or repelling females as a function of the relative doses of each chemical (Blackwell et al. Reference Blackwell, Dyer, Mordue (Luntz), Wadhams and Mordue1996).
Also, host infection status may be a key factor affecting host location by vectors, because infection could affect host metabolism and therefore host-derived attractants (Torres-Estrada and Rodríguez, Reference Torres-Estrada and Rodríguez2003; Lacroix et al. Reference Lacroix, Mukabana, Gouagna and Koella2005). In humans, individuals with high intensities of infection by malaria are more susceptible to the attack by vectors (Lacroix et al. Reference Lacroix, Mukabana, Gouagna and Koella2005). However, this may not be the case for birds (Tomás et al. Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b) where higher abundances of biting midges were found in nests of female blue tits with experimentally reduced intensities of infection by medication with an anti-malarial drug, an effective method to reduce the intensity of infection by the Culicoides transmitted malaria-like Haemoproteus (Merino et al. Reference Merino, Moreno, Sanz and Arriero2000; Tomás et al. Reference Tomás, Merino, Martínez, Moreno and Sanz2005; Martínez-de la Puente et al. Reference Martínez-de la Puente, Merino, Tomás, Moreno, Morales, Lobato and García-Fraile2007). Biting midges may prefer to feed on less infected birds because blood parasites may reduce their survival (Valkiūnas and Iezhova, Reference Valkiūnas and Iezhova2004). On the other hand, the infection status could also affect host susceptibility to vector attacks through other ways such as reducing host antimosquito behaviours (Torres-Estrada and Rodríguez, Reference Torres-Estrada and Rodríguez2003). It is known that hosts use a diversity of insect-repelling strategies to avoid the attack of biting midges including anti-insect behaviours (Edman et al. Reference Edman, Webber and Schmid1974; Mooring et al. Reference Mooring, Fitzpatrick, Fraser, Benjamin, Reisig and Nishihira2003; Darbro and Harrington, Reference Darbro and Harrington2007) or the use of plants with insecticide properties (Bucher, Reference Bucher1988; Clark, Reference Clark, Loye and Zuk1991; Lafuma et al. Reference Lafuma, Lambrechts and Raymond2001). Humans, due to the sanitary and economical importance of Culicoides (Mellor et al. Reference Mellor, Boorman and Baylis2000; Ratnayake et al. Reference Ratnayake, Dale, Sipe and Daniels2006), also use different insecticides to control midge populations. There is evidence of lower abundances of Culicoides in fumigated farms as compared to non-fumigated ones (Sarto i Monteys and Saiz-Ardanaz, Reference Sarto i Monteys and Saiz-Ardanaz2003; also see Satta et al. Reference Satta, Goffredo, Sanna, Vento, Cubeddu and Mascherpa2004) that may reduce the costs associated with the activity of biting insects. In the case of birds, some species introduce in their nests plants with insect-repellent properties that could reduce the abundance of ectoparasites in avian nests (Bucher, Reference Bucher1988; Clark, Reference Clark, Loye and Zuk1991). Both naturally derived and synthesized components have been tested for their repellent effect on biting midges (Braverman and Chizov-Ginzburg, Reference Braverman and Chizov-Ginzburg1997). In wild populations, birds may also benefit from the use of insecticides if they reduce biting midge densities. In the case of blue tits Cyanistes caeruleus, it has been suggested that the use of green plants could be a mechanism of protection against parasites (Cowie and Hinsley, Reference Cowie and Hinsley1988; Banbura et al. Reference Banbura, Blondel, de Wilde-Lambrechts and Perret1994; Petit et al. Reference Petit, Hossaert-McKey, Perret, Blondel and Lambrechts2002). However, the effect of plant-derived repellents could be different among parasite species because there are both attraction and repellency effects of a particular compound among Culicoides species (Braverman et al. Reference Braverman, Chizov-Ginzburg and Mullens1999). To reveal the potential effect of insecticides on Culicoides infracommunities in avian nests, studies in wild populations should be performed. In this respect, we found in a previous study that the use of an insecticide treatment was not effective in reducing the abundance of Culicoides in blue tit nests, although a differential specific susceptibility of Culicoides species to the insecticide treatment could affect these results (Tomás et al. Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008a).
Understanding the interactions between biting midges and birds is especially interesting for the case of hole-nesting species because some of these birds predate on insect pests of gardens and forests. Also it is important to note that Culicoides is a worldwide distributed genus, with about 1254 described species (Beckenbach and Borkent, Reference Beckenbach and Borkent2003), present in most terrestrial habitats (Kettle, Reference Kettle1995; Marquardt et al. Reference Marquardt, Demaree and Grieve2000). Our aim in this study was to identify the role of different factors affecting the composition and abundance of Culicoides species in a wild population of a hole-nesting bird, the blue tit. In addition, we investigated the abundance of parous and engorged Culicoides females, because parous females are potential haemoparasite vectors and engorged females have fed recently on a host. In addition, we investigated the abundance of nulliparous females because, although they have not fed, they are potential haemoparasite vectors to the same degree as life stages after feeding.
MATERIALS AND METHODS
Study area
This study was carried out in a population of blue tits Cyanistes caeruleus breeding in nest-boxes during the spring of 2005 in a Pyrenean Oak Quercus pyrenaica deciduous forest located in Valsaín, Central Spain (Segovia, 40°53′74N, 4°01′W, 1200 m a.s.l.).
Treatments
When nestlings were 3 days old, nests were randomly assigned to fumigation and medication treatments, thereby generating groups of medicated and control pairs breeding in fumigated and control nests (14 medicated×fumigated nests, 14 medication control×fumigated nests, 15 medicated×fumigation control nests and 16 medication control×fumigation control nests). The medication consisted in a subcutaneous injection of 0·1 ml of the anti-malarial drug Primaquine (Sigma, St Louis, MO, USA) diluted in saline solution (concentration 1 mg·ml−1) when nestlings were 3 days old. Control pairs were injected with the same volume of saline solution. Treatment with Primaquine causes a reduction in the intensity of infection by blood parasites in the study population (Merino et al. Reference Merino, Moreno, Sanz and Arriero2000; Tomás et al. Reference Tomás, Merino, Martínez, Moreno and Sanz2005; Martínez-de la Puente et al. Reference Martínez-de la Puente, Merino, Tomás, Moreno, Morales, Lobato and García-Fraile2007). The fumigation treatment was carried out at 3 different times (at the nestling ages of 3, 7 and 11 days) with an insecticide solution (Stockade©, Fort Dodge Veterinaria, S.A., Vall de Bianya, Girona, Spain) comprising 0·5% Permethrin and 1% Piperonyl butoxide. Nestlings were extracted from nests prior to fumigation and left again in the nest immediately after treatment. This treatment has been previously used to reduce ectoparasite populations in nests without detection of any deleterious effect for nestlings (Tomás et al. Reference Tomás, Merino, Moreno and Morales2007b). The same methodology was employed in control nests using water instead of insecticide.
Culicoides collection and identification
During 2 days after the last fumigation, Culicoides were captured using the method described and tested by Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008a). This method consisted in the placement inside the nest-boxes of plastic Petri dishes (8·5 cm diameter; 56·7 cm2) layered with 0·5 ml of commercially available body gel-oil (Johnson's© baby chamomilla, Johnson and Johnson, Dusseldorf, Germany). This gel-oil is made up of paraffinum liquidum, hexyl laurate, ethylene/propylene/styrene copolymer, cyclopentasiloxane, butylene/ethylene/styrene copolymer, chamomilla recutita, bisabolol and perfume [FPT1353]. The effect of the fumigation treatment in non-medicated pairs was previously reported by Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008a) in the context of a methodological study to determine the efficacy of such a sticky medium to collect biting midges. On day 13, brood sizes for each nest were recorded and Petri dishes removed and stored in a freezer until their examination.
In the laboratory, biting midges were removed from dishes using xylene and maintained in absolute ethanol until their identification. All Culicoides species were initially sorted depending on their wing pattern under an Olympus SZH stereomicroscope (10×–64× magnification). However, given their minute size (usually no longer than 3 mm), for more accurate diagnosis, it was necessary to dissect many of the midges and make microscopic slide preparations of their body parts. For fixing them we used Tendeiro solution (distilled water: 35 ml; chloral hydrate: 40 g; glacial acetic acid: 18 ml; polyvinyl alcohol: 7 g). To identify them to specific level we used Kremer's (Reference Kremer1966) and Delécolle's (Reference Delécolle1985) morphological keys. Culicoides were sexed and the parity of females determined as follows: nulliparous (those that have never fed on blood), parous (those showing a burgundy pigment in the subcutaneous cells of the abdomen indicating a previously digested bloodmeal; see Dyce (Reference Dyce1969) or engorged females (those with a bloodmeal still not completely digested in their abdomen). We assume that engorged females fed on blood from birds (nestlings or adults) from the nest-box where they were captured.
Statistical analysis
Total abundance of Culicoides and each specific abundance were logarithmically (log10) transformed to normalize distributions. General regression models (GRM) (Statistica version 6.0, StatSoft, Inc. 2001) applying the forward stepwise solution, were used to investigate the relationships between the total abundance of Culicoides, the abundance of total nulliparous Culicoides females and the abundance of each species, including in the model the 2 treatments (fumigation and medication treatments) and their interaction as factors and brood size and phenology (a potential confounding variable estimated as hatching date of each brood) as covariables. Results were also confirmed using backward stepwise solutions. Residuals of the models were tested for normality. Variables reflecting total abundances included the total number of nulliparous, parous and engorged females per nest. In addition, when residuals of the models did not follow a normal distribution, non-parametric analyses were conducted. Simple correlations and Mann-Whitney U-tests were used to test for the effect of each brood size, seasonality and fumigation and medication treatments on the species richness, the abundance of total parous females and the abundance of total engorged females (both not normally distributed variables, even after log transformation). Analyses for Culicoides species were restricted to the 3 more abundant species, C. simulator, C. kibunensis and C. festivipennis (see Table 1).
Table 1. Abundance of each female stage (nulliparous, parous and engorged) for each Culicoides species captured in blue tits nests during the breeding season of 2005
(The percentage of infected nests is shown in parentheses. As several stages were present in the same nests the sum of percentages is higher than 100. However the column total shows the percentage of nests infected by each species.)
RESULTS
A total of 1531 female biting midges of 7 different species were captured in 57 nests. Only 2 males (one C. kibunensis and one C. festivipennis) were captured. In 2 additional nests we did not capture any biting midge (Table 1). In addition, 41 biting midges (2·6% of the total) could not be identified because of the absence of wings or other anatomical structures. However, unidentified individuals were also considered in total abundances. In each nest, we captured an average of 26·6 (s.d. 39·1, range 0–208) biting midges from 3·1 (s.d. 1·5, range 0–6) different species.
The abundance of total Culicoides females captured in avian nests was strongly and positively associated with brood size (Table 2; Fig. 1). The same positive significant association was found for the abundance of C. simulator (Table 2). A significant positive association was also found between the abundance of C. kibunensis and both brood size and phenology (Table 2). In addition, we found a significant effect of the interaction between medication and fumigation treatments on the abundance of C. festivipennis females (Table 2) with a higher abundance in non-fumigated nests occupied by control pairs (non-medicated) than in the rests of the groups (LSD test, all P<0·011). As residuals of this model did not follow a normal distribution, the effect of the treatment on the abundance of C. festivipennis was also tested using non-parametric statistics, including 4 treatments, medicated×fumigated nests, control×fumigated nests, medicated×control nests and control×control nests and obtaining the same conclusion (Kruskal-Wallis test: H3,59=11·12, P=0·01), that is that the medicated×fumigation control nests showed a higher abundance of C. festivipennis than the other groups.
Fig. 1. Relationship between the total abundance of Culicoides females (nulliparous, parous and engorged) and brood size in blue tit nests during the spring of 2005 (adjusted R2=0·11, P<0·01). Regression line is shown.
Table 2. GRM results after applying the forward stepwise solution for the relationship between the total abundance of Culicoides and the abundance of each species with the medication and fumigation treatments and their interaction, brood size and phenology
(Adjusted R2 values are shown.)
The abundance of nulliparous females was significantly associated with brood size (model: adjusted R2=0·13, P<0·01; brood size: F1,57=9·40, P<0·01). In addition, no significant association was found between the abundance of parous females and brood size, phenology or treatments (Table 3). In addition, the abundance of engorged females was significantly higher in nests with larger broods (Table 3). Also, the abundance of engorged females was significantly lower in fumigated nests than non-fumigated nests (Table 3). In addition, no significant association was found between the abundance of engorged females and phenology or medication treatment (Table 3). Finally, a significant and positive association was found between species richness and brood size (Table 3). In addition, although we found that fumigation significantly reduced the species richness in nests (Table 3), no significant association was found between species richness and either phenology or medication treatment (Table 3).
Table 3. Relationship between the abundance of total parous females, total engorged females and Culicoides species richness and brood size, hatching date and medication and fumigation treatments
(Significant relationships at P<0·05 are marked in bold. Spearman R (rs) and adjusted Z (Z) values are shown.)
DISCUSSION
Here we report the Culicoides infracommunity composition and examine different factors determining their abundance in wild blue tit nests. All the Culicoides species found in this study have been previously cited for the Iberian Peninsula (Delécolle, Reference Delécolle and Carles-Tolrá Hjorth-Andersen2002), and 3 of them, C. festivipennis, C. kibunensis (quoted as C. cubitalis) and C. truncorum (quoted as C. sylvarum) have been previously captured on wild avian hosts (buzzards Buteo buteo nests; Votỳpka et al. Reference Votỳpka, Oborník, Volf, Svobodová and Lukeš2002; Podlipaev et al. Reference Podlipaev, Votỳpka, Jirků, Svobodová and Lukeš2004). We found that 3 Culicoides species had prevalences above 60% and that a very low proportion of nests were free of biting midges. Because vector abundances may determine the prevalence of blood parasites in their hosts (Sol et al. Reference Sol, Jovani and Torres2000; Yu et al. Reference Yu, Wang and Yeh2000), our results are in accordance with a previous study in the same host population reporting a high prevalence of infection by haemoparasites (Merino et al. Reference Merino, Moreno, Sanz and Arriero2000).
Many, if not all, biting insects have evolved a complex sensory system designed to detect and locate hosts with different receptors including chemo- and visual-receptors (Gibson and Torr, Reference Gibson and Torr1999; Grant and Kline, Reference Grant and Kline2003). Blood sucking insects use host-derived odours as cues to detect their hosts (Gibson and Torr, Reference Gibson and Torr1999; Mordue, Reference Mordue (Luntz)2003). As shown in electrophysiological studies on several Culicoides species, these products are effective in stimulating biting midge receptors (Bhasin et al. Reference Bhasin, Moredue (Luntz) and Mordue2000a; Grant and Kline, Reference Grant and Kline2003; Sollai et al. Reference Sollai, Solari, Masala, Crnjar and Liscia2007), and their attractive effect on Culicoides species has been reported both when they are present on their own (Blackwell et al. Reference Blackwell, Dyer, Mordue (Luntz), Wadhams and Mordue1996; Braverman et al. Reference Braverman, Wegis and Mullens2000; but see Bhasin et al. Reference Bhasin, Moredue (Luntz) and Mordue2000b) and in interaction with other host products (such as CO2) (Gibson and Torr, Reference Gibson and Torr1999; Bhasin et al. Reference Bhasin, Moredue (Luntz) and Mordue2000b; but see Braverman et al. Reference Braverman, Wegis and Mullens2000). In the case of birds, some of the kairomones responsible for inducing feeding could be the compounds produced by uropygial glands, as previously reported for other blood-feeding arthropods (see Russell and Hunter, Reference Russell and Hunter2005 and references therein). For that reason, if more nestlings are capable of producing a higher amount of these products we could expect the pattern obtained here, with higher abundances of Culicoides in nests with larger broods. Accordingly, the abundance of Culicoides in avian nests increased with nestling age (a correlate of nestling size) (Tomás et al. Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008a). In a previous study Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b) reported the effect of other variables (nest size, nestling condition, female infection status, the abundance of other ectoparasites and parental provisioning rates) also affecting the total abundance of Culicoides in avian nests, although they did not find a significant effect of nestling brood mass (a correlate of brood size) on total Culicoides abundance. However, the different experimental designs used here could explain discrepancies between the results of the two studies. For example, we captured biting midges with Petri dishes during a period of 2 days, while Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b) captured Culicoides using a piece of plastic tape during one day. In addition, a considerably lower number of Culicoides was captured during 2005 (1531 Culicoides females) than in 2004 (more than 2300 Culicoides) when the study by Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b) was carried out. Differences in the species composition of Culicoides between both studies may also affect results, but unfortunately this information is not available for the study by Tomás et al. (Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b). More experimental studies modifying brood size or the concentration of host attractants should be done in avian nests to reveal the actual importance of these cues to host detection by ornithophilic midges.
We also found a significant association between the medication treatment and the abundance of biting midges in avian nests. Although, we did not measure the efficacy of the medication treatment to reduce the parasite load, we can assume an effect of the treatment in reducing the intensity of infection by Haemoproteus, the most common blood parasite affecting this population (see for example, Merino et al. Reference Merino, Moreno, Sanz and Arriero2000; Tomás et al. Reference Tomás, Merino, Martínez, Moreno and Sanz2005; Martínez-de la Puente et al. Reference Martínez-de la Puente, Merino, Tomás, Moreno, Morales, Lobato and García-Fraile2007). The results reported here on the effect of the medication treatment increasing the abundance of C. festivipennis in avian nests support a previous study conducted in the same population during 2004 (Tomás et al. Reference Tomás, Merino, Martínez-de la Puente, Moreno, Morales and Lobato2008b) where authors found a higher abundance of biting midges in nests occupied by medicated female birds. Also, our results support a previous study where Darbro et al. (Reference Darbro, Dhondt, Vermeylen and Harrington2007) found that Culex mosquitoes are less likely to feed upon birds infected with Mycoplasma gallisepticum maintained in captivity. Different possibilities could explain the higher abundance of this insect species in nests attended by medicated birds. One possibility could be that C. festivipennis is able to discriminate between heavily or lightly infected hosts and feed preferentially on those with lower intensities of infection as a defensive strategy due to the harm that blood parasites produce to biting midges (Desser and Yang, Reference Desser and Yang1973; Valkiūnas and Iezhova, Reference Valkiūnas and Iezhova2004). Alternatively, it could be possible that the increase in bird provisioning rates associated with the medication treatment (Merino et al. Reference Merino, Moreno, Sanz and Arriero2000; Tomás et al. Reference Tomás, Merino, Moreno, Morales and Martínez-de la Puente2007a) help this insect species in host location. The possibility that the medication treatment modifies, at least in part, the release of host odours that are used as cues for host location by midges should be also considered. In this respect, if C. festivipennis was more sensitive to those changes, we could expect an absence of any significant effect of bird medication on the abundance of other Culicoides species. More studies to identify the effect of the medication treatment on the abundance of biting midges in avian nests are needed.
On the other hand, although the total abundance of Culicoides was not affected by the fumigation treatment, we clearly found an effect of the insecticide on the abundance of engorged females, suggesting that the insecticide reduced the efficiency of blood feeding by midges, probably through their avoidance of the nesting material saturated with insecticide where nestlings were located. Another possibility could be that the treatment killed a certain proportion of the midge population that rendered a reduction in abundance of engorged midges. In fact, this could also be the reason, at least in part, for the lower species richness found in fumigated nests with respect to control nests. In previous studies where we used the same insecticide, a significant reduction in the abundance of other ectoparasites (fleas, mites and blowflies) was found (Tomás et al. Reference Tomás, Merino, Moreno and Morales2007b; Lobato et al. Reference Lobato, Merino, Moreno, Morales, Tomás, Martínez-de la Puente, Osorno, Kuchar and Möstl2008), suggesting that the higher mobility of biting midges with respect to other nest ectoparasites (mites, fleas and blowflies) could explain the differential efficiency of the treatment between nest-dwelling and flying ectoparasites.
Finally, we found that phenology, estimated as hatching date, is an important factor affecting the abundance of some vector species probably due to its association with meteorological conditions. There are many reports on the effects of both meteorological factors and seasonality on Culicoides biology in terms of development, adult survival, distribution, abundance and activity rates (Bishop et al. Reference Bishop, McKenzie, Barchia and Harris1996; Gerry and Mullens, Reference Gerry and Mullens2000; Mellor et al. Reference Mellor, Boorman and Baylis2000; Wittmann et al. Reference Wittmann, Mellor and Baylis2001; Garvin and Greiner, Reference Garvin and Greiner2003; Sarto i Monteys and Saiz-Ardanaz, Reference Sarto i Monteys and Saiz-Ardanaz2003; Lysyk and Danyk, Reference Lysyk and Danyk2007; Martínez-de la Puente et al. Reference Martínez-de la Puente, Merino, Lobato, Rivero-de Aguilar, del Cerro, Ruiz-de-Castañeda and Moreno2009). The relationship between the abundance of C. kibunensis and host phenology suggests that early dates in the host breeding season were less favourable for the development of this species. Overall, our results represent the first evidence for different factors affecting the Culicoides infracommunity in the nests of wild birds.
We thank Javier Donés (Director of ‘Montes de Valsaín) for permission to work in the study area. The Junta de Castilla y León authorized the ringing and handling of birds. This study was funded by projects CGL2006-14129-C02-01 (to S. M.) and CGL2007-61251 (to J. Moreno) from Ministerio de Educación y Ciencia. J. M.-P. is supported by a grant from ‘El Ventorrillo’ field station. J. M. was supported by an FPI and a post-doctoral grant from Ministerio de Ciencia y Tecnología. E. L. was supported by an FPU grant from MEC. G. T. was supported by FPI grant from the Comunidad de Madrid and a post-doctoral grant from MEC. Authors thank N. Pages Martínez, F. Muñoz Muñoz, and J.-O. Moreno Vidal for their help in the lab. A.V. Bordería kindly supplied the Petri dishes. This study is a contribution to the research developed at ‘El Ventorrillo’ field station. Two anonymous referees considerably improved a previous version of the manuscript with their constructive comments.
