Published online: 31 Jan 2011.

This article was downloaded by: [Cornell University Library] On: 13 February 214, At: 7:38 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 172954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK International Journal of Acarology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/taca2 Biology of House Finch feather mites, Proctophyllodes pinnatus (Acari: Proctophyllodidae), parallels variation in preen gland secretions Meena Haribal a, Heather Proctor b, André A. Dhondt a & Eloy Rodriguez c a Laboratory of Ornithology, Cornell University, Ithaca, NY, 1485, USA b Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada E- mail: c Plant Biology, Cornell University, Ithaca, NY, 14853, USA Published online: 31 Jan 211. To cite this article: Meena Haribal, Heather Proctor, André A. Dhondt & Eloy Rodriguez (211) Biology of House Finch feather mites, Proctophyllodes pinnatus (Acari: Proctophyllodidae), parallels variation in preen gland secretions, International Journal of Acarology, 37:1, 75-9, DOI: 1.18/1647954.21.495952 To link to this article: http://dx.doi.org/1.18/1647954.21.495952 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

International Journal of Acarology Vol. 37, No. 1, February 211, 75 9 TACA BIOLOGY OF HOUSE FINCH FEATHER MITES, PROCTOPHYLLODES PINNATUS (ACARI: PROCTOPHYLLODIDAE), PARALLELS VARIATION IN PREEN GLAND SECRETIONS Internat. J. Acarol. Meena Haribal 1, Heather Proctor 2, André A. Dhondt 1 and Eloy Rodriguez 3 1. Laboratory of Ornithology, Cornell University, Ithaca, NY 1485, USA (e-mail: mmh3@cornell.edu and aad4@cornell.edu); 2. Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada (e-mail: hproctor@ualberta.ca); 3. Plant Biology, Cornell University, Ithaca, NY 14853, USA (e-mail: er3@cornell.edu). (Received 2 August 29; accepted 2 May 21) ABSTRACT Except for Dubinin s classic works in the 195s, there is very little information on the natural history or population dynamics of feather mites (Astigmata: Acariformes, Pterolichoidea). We studied variation in numbers of the different life stages of the feather mite Proctophyllodes pinnatus (Nitzsch) (Analgoidea: Proctophyllodidae) on captive House Finches Carpodacus mexicanus (Müller) (Passeriformes: Fringillidae) from December 23 to November 24. Simultaneously, we also studied how preen gland secretion varied in those birds. We monitored 2 House Finches (1 individuals of each sex) for the presence of mites on their wing feathers. There was seasonal variation in the abundance and prevalence of mites on different individual birds. Most birds did not show any mites from December to April. Mite numbers started to increase in July and peaked in August September. In September, we observed more mites of early stages than in any other months. We also observed higher proportions of adults in August and October and fewer adults than expected in September, suggesting that more than one generation was involved. By November, very few mites were present on the wing feathers. The variation in mite numbers paralleled variation in the composition and quantity of preen gland secretions produced by the host House Finches. In most of the captive birds, secretion peaked in the month following peak mite-abundance. During the period when mite abundance was highest, secretions contained a higher diversity of chemicals. Prevalence of mites on wild House Finches showed similar trends to those observed in captive birds. Relative abundances of life stages on wild birds collected in July were similar to those observed for captive House Finches in the same month. Key words Astigmata, Acariformes, Carpodacus mexicanus, Proctophyllodes pinnatus, ectosymbionts, seasonal variation, natural history, evolution, preen gland secretions. INTRODUCTION Feather mites (Astigmata: Pterolichoidea, Analgoidea) have been collected from all extant orders of birds with the exception of Rheiformes (Proctor, 23; Proctor and Owens, 2; Mironov and Proctor, 28). Most vane-dwelling feather mites are considered to consume components of preen gland secretions that are present on the feathers as well as small particles (e.g. spores, pollen) caught on the feathers (Dubinin, 1951; Mironov and Dabert, 1999; Blanco et al., 21; Proctor, 23). Although mites are present on most species of birds, their natural history and ecology have rarely been studied in detail, with most effort being directed towards taxonomy. The main exception is the massive body of work by the Russian Current address: Boyce Thompson Institute, Tower Road, Ithaca, NY 14853, USA ISSN 164-7954 print/issn 1945-3892 online 211 Taylor & Francis; printed 31 January 211 DOI: 1.18/1647954.21.495952 http://www.informaworld.com

76 Haribal et al. 211 feather mite biologist V. B. Dubinin, who published three massive volumes in the 195s (1951, 1953, 1956) that documented the astonishing range of morphology, behavior and ecology displayed by these mites. Since Dubinin, a number of studies have looked at relationships between mite load and host condition (Blanco et al., 1997; Jovani and Blanco, 2; Wiles et al., 2; Blanco and Frias, 21; Blanco et al., 21), host behavior and mite numbers (Blanco and Frias, 21) and molting status and mite distribution on feathers (Atyeo and Windingstad, 1979; Wiles et al., 2). Seasonal phenologies of feather mites on hosts have been reported for a few species. Blanco and Frias (21) estimated densities of mites (which were not identified beyond feather mites ) on Barn Swallows (Hirundinidae: Hirundo rustica Linnaeus) at different times of the year and found that total mite numbers are highest in August; however, they do not differentiate among the various life history stages of mites. Mironov (2) studied various life stages of Monojoubertia microphylla (Robin) (Proctophyllodidae) found on Common Chaffinches Fringilla coelebs Linnaeus (Fringillidae). He reported that there was seasonal variation in mite numbers and life stages and also between male and female birds, particularly during the nesting season. Neither study involved counting mites on same birds. Based mainly on correlative studies, the majority of feather mites appear to be benign commensals (Blanco et al., 1999; Dowling et al., 21), although there is some evidence that in certain circumstances they can act as parasites (Thompson et al., 1997; Harper, 1999; Figuerola et al., 23) or as mutualists (Blanco et al., 1997; Brown et al., 26). Most of these studies collected data for any particular host individual only once, and therefore do not provide information about temporal variation in mite abundance and host characteristics. From first principles we might predict annual variation in numbers of vane-dwelling mites to correlate with ambient temperature (Dubinin, 1951; McClure, 1989), host breeding period (e.g., Mironov, 2), or changes in the quantity or composition of preen gland secretions. There is some evidence of seasonal variation in preen gland size and secretory output. For example, glands are larger during breeding season in Pycnonotus cafer (Linnaeus) (Pycnonotidae) (Bhattacharyya, 197), and in some other species there is seasonal change in internal morphology of the glands (Jacob, 1978; Jacob and Ziswiler, 1982). Galvan and Sanz (26) and Galvan et al. (28) found a positive correlation between preen gland size and the mite load on birds. In most phenological studies, mites (and glands, if examined) are measured once in their study, where the study periods ranged from two to several months. It is very rare that the same individual bird was looked at more than once. An exception is Brown et al. (26) in which they recaptured the same individuals at least three times, but they did not indicate if the mite load varied between these recaptures on the birds and there is no information on the period of time between the recaptures. It is important to study the mites on the same individuals for a length of time to understand mite phenology as mite age structure on individuals may vary considerably. We chose House Finches, Carpodacus mexicanus (Müller) (Passeriformes: Emberizidae), as model subjects for this study because as a part of another study we had access to wild-caught and captive birds throughout the year. Mites have been observed on House Finches throughout the year (McClure, 1989; Dhondt et al., unpublished), but no mites were identified to species by these researchers. Two species of feather mites of the genus Proctophyllodes, P. pinnatus (Nitzsch) and P. vegetans Trouessart (Proctophyllodidae), have been recorded from House Finches (Atyeo and Braasch, 1966; Hartup et al., 24). In our preliminary investigation we found a species keying to P. pinnatus to be the only ectosymbiont on the captive House Finches: subsequently we did find some specimens keying to P. vegetans on a few of the wild birds as well as even rarer specimens of a Trouessartia sp. (Trouessartiidae). In this paper, we present observations of variation in abundance and distribution of different life stages of Proctophyllodes on feathers of C. mexicanus and for the first time present correlations between feather mite loads and preen gland secretions as measured repeatedly on the same individual birds. We collected several samples of preen gland secretions from captive House Finches throughout a 12 month period and also observed the mites on their flight feathers to address the following questions: how do the different active life stages of mites vary with season and sex of birds? How do secretions and mite numbers vary over the year? Also, because of the rarity of such information, we describe some behavioral observations of individual Proctophyllodes on feathers. MATERIALS AND METHODS Sources of birds Mites and preen secretions were sampled repeatedly from 2 banded captive House Finches, 1 individuals of each sex caught as HY (hatch year) birds in Ithaca, New York, but less than 2 years old that were held in an outdoor aviary in Ithaca New York, USA. Four of these birds died during the experimental period at various times. All captive birds were housed at ambient temperatures and daylengths. They were fed ad lib with a commercial pelleted diet, Roudybush, Inc., Cameron Park, CA, USA. All procedures and protocols were approved by the Cornell University Institutional Animal Care and Use Committee (Protocol #-9).

Vol. 37, No. 1 Internat. J. Acarol. 77 Dhondt s research group who were studying the dynamics of Mycoplasma gallisepticum in House Finches, also collected data on mite prevalence on free living birds during their field surveys. We have used data collected from January 24 to December 24 to assess prevalence and abundance of mites on wild birds. Collection and analyses of preen gland secretions We collected preen gland secretions every month from December 23 to November 24, except during May when the birds were breeding. Secretions were collected from individual birds by gently pressing a sterile cotton swab over the gland, collecting secretions from the opening 3 4 times until no more secretion squeezed out. The swabs were transferred to individual glass vials and stored tightly capped at 4 C until analyzed. Although the cage contained some additional juvenile birds after May, we did not separate juvenile birds from experimental birds nor collect secretions from these birds. We analyzed the samples by Gas Chromatography coupled with Mass Spectrometer (GC-MS) on an HP 689 coupled to a MSD 5972A mass detector. The details of the methods were as described in Haribal et al. (25). Observations of mite behavior We collected mites from wild caught and captive House Finches. The feathers bearing mites were placed in a Petri dish that was examined under the microscope. We observed behaviors of mites at various magnifications under a dissecting stereomicroscope, Stemi 2 Zeiss, with Fostec light set up. We occasionally took short video clips of less than a minute each as well as still photographs with a Nikon Coolpix 57 digital camera held to the microscope eyepiece. Appropriate movies were then extracted for still shots using Quick- Time pro 7 and exported as TIFF files, which were further processed with D-interlace Video filter and converted to gray scale in Adobe Photoshop CS. In total we spent more than 2 hours observing mites and collected a total of about 3 min of video shots and 2 still photographs. We also maintained the mites in the lab at 4 C and at 38 ± 2 C to see how temperature affected their survival in captivity. Feathers with 5 3 mites were stored in a vial and placed in a fridge or an incubator maintained at 4 or 38 ± 2 C in dark conditions. A piece (approximately 1 3cm 2 ) of Whatmann filter paper No. 1 was introduced in to the vial and was moistened with three or four drops of MQ (Millipore quality) water to maintain humidity in the vial. No additional food was offered other than whatever was present on the feathers. Three replicates for each condition were set up. Mites were observed daily during the next few days to see if they survived. Observations of mites on birds Mites were monitored on the captive birds. During the entire study period captive birds did not have any contact with outside birds. They were, however, in the same room as an additional group of House Finches that were maintained in a different cage separated by 1 cm, but not monitored for mites. When secretions were sampled from captive birds, we also recorded the presence of mites on the wing feathers. We monitored for mites on captive birds once a month from December 23 to November 24 except for January and May, because in January it was too cold in the housing facility to conduct work for a long period and in May because the birds were breeding. To sample birds, we first caught all birds and placed each individually in a brown paper bag. Birds were then removed one by one to collect samples of secretions and for observation of mites. Other researchers have observed that vanedwelling feather mites on passerines are distributed symmetrically on both wings (Behnke et al., 1995, 1999; Jovani and Serrano, 24); therefore, we monitored mites on the left wing only by holding the wing open to the light to see if mites were present on the flight feathers. In December 23 only one of the birds had observable mites, but by the beginning of July 24 six birds had mites. Between July and November 24, we collected the apparently most heavily miteladen wing feather (which was either a primary or a secondary that had most number of mites) from each mite infested bird. The mite-laden feathers were observed at various magnifications under a dissecting stereomicroscope, Stemi 2 Zeiss, with Fostec light set up. We noted the presence of various life stages: eggs, larvae, protonymphs, tritonymphs and adults. As it was difficult to consistently differentiate among some stages, we grouped them into the following four groups: I: eggs and larvae (eggs were lumped in this group because only 1 eggs were observed overall); II: nymphs; III: adults and IV: pairs in copula (adult male and female) or precopula (adult male and tritonymph). Numerous exemplars of the mites were identified by one of us (HP) as closest to Proctophyllodes pinnatus (Nitzsch) using Atyeo and Braasch (1966). This species of mite has been reported from numerous genera of fringillid birds in Europe and North America. In their monograph, Atyeo and Braasch (1966) state that the location of the type specimen of P. pinnatus is unknown. Given the broad geographic and host ranges, and the fact that illustrations and measurements in Atyeo and Braasch (1966) were not based on the original type, it is quite possible that the mites we collected from captive House Finches are not P. pinnatus in the strict sense, but rather represent an undescribed species closely related to it. Voucher

78 Haribal et al. 211 specimens are deposited in the acarological collection of the E. H. Strickland Entomological Museum of the Department of Biological Sciences at the University of Alberta, and some exemplars with Bird Population Studies at Cornell Laboratory of Ornithology. In subsequent observations MH identified the mites. Although we used only a single feather to study mites, we assumed that mites were of similar life history stages and distributed in similar proportions on other feathers. Our subsequent observations on other wild birds suggest this may not be true at all times of the year. In 26 28 some wild-caught House Finches were observed for mites using a field microscope, and all the mites on the secondary and primary feathers were recorded by taking photographs using the Dinolite field microscope, an LCD microscope with continuous focus 5 2 magnification (AM413T-FVW) attached to a computer screen via USB connection. The software Dinocapture interface allowed us to take digital photographs and store them. The photographs of left wing feathers of 13 birds were used to calculate the distribution of different life stages of the mites on the feather surface. We divided the feather vane area into three parts and counted mites of different life stages and locations of mites in these areas, whether near or away from rachis or on barbules. Statistical analyses Abundance of life stages of mites was analyzed by months stages table (5 4) and Chi square tests. Occurrence of mite life stages on the different parts of the feather was also analyzed by location stages table (6 4) and the null hypothesis that different life stages of mites equally and randomly occupy the feather surface was tested by a Chi square test. We used a one-way ANOVA to test for seasonal variation of prevalence and abundance in wild caught birds. Chemical composition of preen secretions was analyzed using Total Ion Current (TIC) obtained for individual peaks. TIC is directly proportional to the concentration of the individual chemical contents in the secretions at that retention time. For analyses of seasonal variation in amounts of all components, we summed the TIC values integrated for each peak (as area under each peak obtained by integrations of the chromatogram) of individual birds in a given month and averaged it for that month over 15 2 individuals and analyzed variation by GLM (Generalized Linear Model). We could not measure absolute quantities of secretions in individual birds, as we are not sure if we removed all the secretion present in the gland by our methodology. Nevertheless, because most samples (more than 95%) were collected by MH, which minimized inter-person variation in sampling, we can consider that variation in the amount of secretions obtained from this similar collection effort does reflect changes in amounts of the secretions produced by the birds over the seasons. In addition, we could also visually observe that much more secretion oozed out when the gland was pressed during summer than in the winter. We compared GLM results using Duncan s multiple range test for variation in amounts of collected secretions between months. RESULTS Behavioral and life-history observations of Proctophyllodes pinnatus We investigated longevity of the mites in the laboratory under low and high temperatures. Many mites died by the day after collection, and most of the mites that were maintained at 38 ± 2 C died in 5 6 days with a few individuals living for more than a week. For example, 3 mites of different life stages were placed in an incubator on 11 June 24. Only eight of these were still alive on 15 June 24 and by 2 June 24 all were dead. Survival rates of the mites varied between different sets. The mites that were kept at 4 C could be kept alive for longer periods of 8 12 days. We followed this procedure routinely to keep mites alive to be used in for later experiments. In one separate instance, in a vial collected 24 days earlier kept at 4 C, we found a few them were still alive. On the feathers, different life stages of P. pinnatus tended to occupy different parts of the feathers (Figs. 1 3 and Table 1). Eggs were laid on the ventral surface of the feather, parallel to the rami of the barbs, in the basal third of the feather, half way between the rachis and the edge of the inner vane (Fig. 3a). Eggs are sausage-shaped, smooth, translucent, and shiny. They were smaller than the distance between two rami of the two adjacent barbs. Oviposition behavior was observed once. A female who was carrying an egg moved around the whole feather, visiting the edge of the feather and stopping at several spots. After approximately 5 min she deposited the egg about 1.5 cm from the basal third area on the barbs about more than half distance away from the rachis towards the edge of the vein (Fig. 1, Table 1). During the search, the egg, which was about 2/3 the length of the female s body, was more than 2/3 outside her body. Several clusters of 2 4 eggs were observed, but it is unclear whether this was due to the same female sequentially ovipositing at the same site, or if the eggs were placed there by different females. Larvae are translucent and difficult to see, but were mostly observed in the depression on the ventral side along the main rachis on inner side of the vane (Fig. 3b). They were observed significantly more often than expected along the distal third of the feather

Vol. 37, No. 1 Internat. J. Acarol. 79 Proximal end Basal third Medial third Distal third Region where eggs are laid Inner vane Fig. 1. A typical flight feather of House Finch depicting terminology used in the text. Numbers 16 14 12 1 8 6 4 2 Eggs Larvae Nymphs Adults Basal third r Medial third r Distal third r Basal third b Medial third b Distal end Distal third b Fig. 2. Bar chart showing distribution of different stages of mites on the feathers obtained by pooling data from the Dinolite digital photos of all primaries and secondaries on the left wing of 13 wild caught birds. r = along the rachis, b = on the barbs. where the depressions are small enough to hold them snugly (Table 1). The rachis forms narrower depressions with the barbs as the rachis is thinner in this area than in the other two sections of the feather. Nymphs were also observed along the rachis but mostly in the medial thirds, but occasionally in basal and distal thirds (Figs. 2 and 3b, Table 1). They were often active but moved mostly along the rachis back

8 Haribal et al. 211 Ramus a b Female Barbs and barbules Eggs Nymph Larva c Male e d Male-tritonymph precopulatory guarding Fig. 3. a e. Life history stages of Proctophyllodes pinnatus mites on House Finches a. eggs; b. nymph and larva; c. female; d. male and e. male-tritonymph precopulatory guarding.

Vol. 37, No. 1 Internat. J. Acarol. 81 Table 1. Results of chi-square test on 6 4 table for occupancy of mite life stages on the feather. Life stages Basal third r Medial third r Distal third r Basal third b Medial third b Distal third b Eggs Obs 1 Exp. 3.95 3.2.16 2.38.31 CChi. 3.95 3.2 618.16 2.38.31 Larvae Obs 48 138 Exp. 73.47 59.47 2.92 44.31 5.83 CChi. 8.83 13.68 2.92 44.31 5.83 Nymphs Obs 87 26 3 4 Exp. 47.4 38.37 1.88 28.59 3.76 CChi. 33.9 3.99 1.88 22.9.2 Adults Obs 117 4 149 16 Exp. 127.18 12.96 5.5 76.71 1.9 CChi..82 38.5 5.5 68.11 3.46 Notes: The data is based on results from seven birds, whose all flight feathers of left wing were observed using the Dinolite field microscope. Not all feathers had mites. The feather vane is approximately divided into three equal lengthened portions, basal, medial and diatal (excluding shaft) and r = along the rachis and b = on the barbules. Obs = observed, E = expected, CChi = cell chi-square. Bold numbers show cells with very high positive chi square values and bold in italic show cells with negative values. and forth. Especially, the tritonymphs can be observed moving on the barbs (Fig. 3b). Tritronymphs were often seen in precopula with adult males, or resting on the barbules along with adults. Adults were seen either along the rachis or in between the barbs, parallel to the rami of barbs when resting (Fig. 3c e). The location of the adults seemed to depend on the total number of mites on the feather. If the density was low they were mostly found along the rachis, but if it was high then they were found on the barbs. Also, they were found mostly on the inner side of the vein. Very occasionally, mostly when the density was high, we observed them lying along the rachis on the outer side of the vein. Male-female pairs were observed fairly often, including males in precopula with tritonymphs (Fig. 3e). Generally, male-female pairs moved in the direction that the male was pointing, but a few times, we observed females pulling the male in the direction she was facing. Both males and females are capable of moving on many kinds of surfaces including smooth glass, plastic containers, rubber container caps and filter paper. They move sideways as well as forward and backward. Mites of all active life-stages were often observed moving their mouthparts as if feeding; however, for the most part we were unable to distinguish what, if anything was being eaten. In two cases, an adult P. pinnatus were observed consuming unidentified amorphous organic matter. Life history and abundance of mites on birds Over the year, mite abundance varied considerably on captive House Finches. Between December and April, when flight feathers were observed against light, most of the captive birds did not show any mites except for an individual bird each in December and April. In June, three birds had a small number of mites. In July six birds had at least some mites. Under the microscope, we observed mites of all stages on the feathers that were collected between July and November. The number of mites differed significantly from July to November (Fig. 4, Table 2) and the proportion of mites of different life stages varied significantly over the 5 months (c 2 = 157.5, df = 8 and P <.1). All mite stages were present in July (Fig. 4). In August, adults were present in higher numbers than expected (Table 2). In September and October all three life history stages (I, II and III) were observed but mites of earlier life stages (I) were in significantly higher proportion than expected. In November, only two individuals had visible mites on the wings (which were either adults or tritonymphs). Eggs and larvae were no longer observed. The number of pairs in copula/precopula also varied significantly between July and November (c 2 = 298.1, df = 4, P <.1) In November,

82 Haribal et al. 211 I II III IV Mean no. of mite life 9 8 7 6 5 4 3 2 1 Fig. 4. Mean numbers of mite life stages on birds in different months. I = eggs and larvae; II = nymphs; III = adults are plotted along primary Y axis (in different shades of grey) and IV = male-tritonymph precopulatory gaurding (as dots) plotted along secondary Y axis. only two pairs were observed (Fig. 4). Two of the birds did not show any mites throughout our study period. Overall, numbers of all life-history stages increased from July to September on most birds. In August, as many as 15 out of 2 birds showed mites. Overall number of mites reached higher numbers in August and September, but for some bird hosts, a peak in mite numbers was reached earlier than in the others (Fig. 5). The mite numbers on individual birds did not vary significantly among individuals (F = 1.38, df = 19, P =.16), but differed over 5 months (F =4.83, df=4 and P <.1 for 5 months). When males and females were analyzed separately to see if there was any difference in mite abundance over 5 months period, there was an overall difference in mite occurrence (c 2 =15, df = 4, P <.1), but most of the deviation was for the month of July, which had comparatively a smaller sample size of birds with mites (six birds) so the observed significance may have been accidental (Fig. 5). The average number of mites of each life stage on wild birds in July 26 was similar to that observed in captive birds in July 24, although the number of early life stages (group I) was lower in July 26 (Fig. 6). Examination of data collected by Dhondt et al. from field sampling (January December 24) of House Finches in Ithaca area in New York reveals seasonal variation in mite prevalence and abundance (Fig. 7). This indicated that the mite abundance peaks in post breeding season and differed significantly between months (F = 2.8, df = 4, 1, P <.1) Jul Aug Sep Oct Nov 9 8 7 6 5 4 3 2 1 Mean no. of mating pairs (Table 3). This observation corroborates what we observed for captive birds. In wild birds, the mite abundance was at its lowest from March to June. Secretions The chemical components and quantities of the secretions varied throughout the year. For example, the secretions in December were made of up two major and a minor homologous series of long chain monoester components [homologous compounds have the same functional groups, but differ in mass by series of 14 or 28 (CH 2 or C 2 H 4 ) mass units], while secretions in August had one major homologous series monoester and had at least more than five more homologous series that were structural isomers of the major components (Fig. 8). The number of isomers was highest in the mixture when there was more secretion (Fig. 8). The quantification of secretions was made based on the sum of total ion currents for all components averaged over 15 2 birds in each month. As the monthly average came from a sample size of 15 to 2 birds, we believe that the differences we observed in secretion concentrations and relative proportions of the components were natural. Detailed analyses of chemistry of the components and variation are being published elsewhere. Also, the amounts of secretions showed two stepwise increases. First there was an increase in secretion amount in April and another increase in amount of secretions in August (Fig. 9). Each level differed significantly from the other. Similarly, the components also differed significantly. The comparison of total number of mites and total amount of secretion based

Vol. 37, No. 1 Internat. J. Acarol. 83 Table 2. Results of chi-square test on 5 3 table mite life stages observed on the feather and months. Month Life stages I II III Total Jul Obs 24 47 97 168 Exp 43.75 27.99 96.26 CChi 8.92 12.91.1 Aug Obs 18 191 71 181 Exp 281.53 18.1 619.37 CChi 36.62.66 13.26 Sep Obs 342 135 455 932 Exp 242.73 155.28 534. CChi 4.6 2.65 11.69 Oct Obs 29 19 367 685 Exp 178.4 114.13 392.47 CChi 5.25.23 1.65 Nov Obs 1 32 33 Exp 8.59 5.5 18.91 CChi 8.59 3.68 9.7 Total 755 483 1661 2899 I = eggs and larvae; II = nymphs and III = adults. Obs = observed, Exp = expected, CChi = cell chi-square. Bold numbers show cells with very high positive chi square values and bold in italic show cells with negative values. on total TICs suggest that they paralleled in variation suggesting, that when there was an increase in amount of secretion there was a subsequent increase in mite abundance; however, we do not know whether the relationship is coincidental or causal, and if the latter, whether the increase in secretion caused an increase in mite abundance or the reverse way (Fig. 9). Since we collected information on mite infestation and secretion chemistry for individual birds during the 7 6 5 4 3 2 1 Jul 253 285 29 249 284 137 274 146 139 26 261 Mites on individual birds period when mite numbers greatly increased, and because the month in which mite numbers peaked varied between individuals, it was possible to determine how mite numbers and secretion quantity co-varied. Five of the 2 experimental birds were missing some data or died, so were not used in the analyses. In 1 of 15 birds, the quantity of secretion peaked a month after mite abundance reached its peak, e.g. if the mite abundance reached its peak in August on an individual, increased secretion was observed in September and then decreased in subsequent months in most cases. In two birds, maximum mite abundance and increase in quantity of secretions occurred in the same month. Only in 3 out of 15 birds, we did not find the general relationship (Fig. 1). Overall, therefore, maximum amount of secretion reached its highest amounts after the mites reached their highest abundance (Wilcoxon s paired rank test W+=24, W = 96, N = 15, P <.5 or 95% CI and T-test t = 2.434, df = 14, P <.5). DISCUSSION Both in our captive house finches and in wild birds there was clear seasonal variation in the presence and abundance of mites. Proctophyllodes pinnatus seems to reproduce from July to November, a period that coincides with the post breeding and molt period of its hosts. Based on the observations of mites of wild-caught birds by McClure (1989) and our results (Fig. 6), there is one major breeding season for mites during the post-breeding season and molt of the hosts and a minor peak in mite abundance in winter. We do not know if it is the same species having two peaks or Aug Sep Oct Nov F Fig. 5. Bar chart depicting distribution of mites on different individuals between July and November. F = females, M = males. The numbers on X-axis represent band numbers of the individual birds. M 151 299 276 278 148 294 136 277 267

84 Haribal et al. 211 8 7 6 5 4 3 2 1 Wild in July Captive in July I II III IV Fig. 6. Comparison of proportion of mites of different life stages on all primary and secondary feathers of left wing of wild caught birds observed as digital photographs using Dinolite (field microscope) and the captive birds in month of July. if two different species of mites are involved in two different peaks as researchers have not identified the species of mites that were present on the birds during this period. In our captive House Finches, only one bird was observed with mites in December and one in April. From July onwards many more birds had mites. Of the 2 captive birds, one bird (261) that hosted most mites died in October. On two of the birds, we did not find mites in any months. The number of mites varied from a few mites on some birds in any given month to a few hundred on some individuals. This suggests that physiological or behavioral variation in individual birds influences mite loads. Although Hartup et al. (24) observed higher Proctophyllodes loads on wild male House Finches from Wisconsin, we did not find that host sex was a significant correlate of mite load in our captive birds. There was variation in number of adults and juvenile mite stages over the 5 months in which captive birds had mites. This suggests that there is probably more than one generation involved as there were higher proportions of eggs and larvae (category I) in July and in September and higher proportions of adults (category III) in August and in November. For example, in August the early life stages were lower and the adult stages were higher than expected [Cell c 2 36.62 (lower) and 13.26 (higher)], while the reverse was observed in September. Wild birds N = 8 Captive birds N = 7 We do not have any idea as to how long it takes for P. pinnatus to complete its life cycle from egg to adult. The apparent disappearance of mites from birds in November may be explained by a likely hypothesis that on the hosts, mites might hibernate in some parts other than flight feathers during the non-breeding season either as eggs, juvenile forms or adults. Dubinin (1951) observed that the mites hibernate in various stages on different parts of the feathers. We only checked for mites on primary and secondary feathers, though in a few cases we did look at tertials. Presence of specific life stages at specific locations on the feather suggest that there are some additional biological factors involved that make mites occupy specific sites. Juvenile stages (nymphs and larvae) mostly reside along the rachis, between the rachis and barb joints, because they may not have enough strength to hang on to the more flexible barbules when the birds take flight, and/or may not be structurally adapted to reside on barbules. As we did not observe juvenile stages moving more than a few mm from the rachis and as we observed their mouth parts moving this suggests that they find their sustenance near the rachis. However, when they are in bright light of the microscope they may not move to the barbules because of bright light. It is possible that when birds are resting at night it might be safer for the younger stages to move around the plumage.

Vol. 37, No. 1 Internat. J. Acarol. 85 4 Birds with heavy mite infestation (%) 3 2 1 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Fig. 7. Graph depicting percent of wild House Finches with heavy mite infestation throughout 24 in Ithaca NY area. Table 3. Variation in mite abundance based on scores tested for significance using two-way ANOVA for birds with different mites score by month. Source of variation Sum of squares (SS) df Mean of squares (MS) F P-value F crit Mite scores 6769.88 4 1692.47 2.79 2.46E-9 2.6 Months 1 1 2.7 Error 3254.97 4 81.37 Total 1,24.86 54 Another intriguing aspect of mite biology was oviposition sites. In almost all species of feather mites, a female matures and lays a single egg at a time (Dubinin, 1951). Therefore, it is interesting that we frequently observed clusters of eggs not only in House Finches but also in other species such as House Sparrows. It was also unclear why eggs are laid on the barbules, away from the rachis in the basal third region where other stages of the mites were rarely observed. Could there be chemical cues involved in the selection of oviposition sites? Once the egg has hatched the larvae need to move to rachis from the distal third area. We found about 35% of larvae in the medial third portion of rachis. Maybe some of these were just moving to distal end of the rachis after hatching. Perez (1997) also observed taxon-specific oviposition sites in feather mites associated with parrots, in this case, members of the same genus of mites tended to lay eggs in the same place. Dubinin (1951) has also observed that oviposition sites are specific for some species and that some species of feather mites are viviparous at some time of the year and that developed larvae emerge directly from the

86 Haribal et al. 211 C 32 C 33 Aug C 32 1 6 C 33 Total Ion Current (TIC) 1 5 Dec 15 2 25 3 R 2 R 1 O R 3 m C 32 C 33 C 33 n C 33 O O 2.5 C 32 2.5 C 33 21.5 21.5 min m, n, o are real positive numbers R1, R2, R3 = (CH 2 ) n + H Fig. 8. GC chromatograms of typical August and December secretions indicating clear differences in the composition of components. In set graphs are expanded regions of the chromatograms from 2.5 to 21.5 min showing different homologous isomers in August and December. The chemical structure represents a typical monoester where m, n, o and R (1 3) represent presence of possible substitutions and chain lengths of acyl and alcohol moieties. female. We did not see any evidence of viviparity. In wild-caught House Finches we did find a few specimens of other mite species, Proctophyllodes vegetans and a Trouessartia sp. (Trouessartiidae). It would be very interesting to understand whether they interact with each other. For example, Atyeo and Windingstad (1979) found six species of mites on a single flight feather of Sandhill Crane and they believed that the

Vol. 37, No. 1 Internat. J. Acarol. 87 Mean Sec/bird Mean no. of mites/bird Sum of TIC X1 6,, 5,, 4,, 3,, 2,, 1,, Jan Feb Mar Apr Fig. 9. Graph depicting, mean (±SE) total preen gland secretions (as sum of TIC of all peaks) per bird, averaged over 15 2 birds (primary Y-axis) and mite abundance (secondary Y-axis) as mean (±SE). Letters a and b indicate that the two groups are significantly different from each other (Duncan s Test P.5). mites occupied specific locations. So it would be interesting to understand ecology and interactions of more than once species of mite on the same host. Bush and Malenke (28) observed clear interaction between two species of bird lice, the presence of one species of body louse affecting the abundance of a wing-dwelling louse. The mite life stages on wild caught birds in July and that of captive birds in July were very similar. Life stages in the wild and in the captive birds may be similar but variable among individuals. Preen gland secretions of the House Finches varied in quality and quantity throughout the year. On average in a given month the GC chromatogram pattern of the secretion was very characteristic. In House Finches the quantity of secretions began to increase with the arrival of the breeding season, reaching a peak in the post-breeding season. The secretion amounts and complexity varied. Increase in the amount of secretions may have been caused by environmental cues such as temperature, weather or/and physiological condition as some of the evidences suggest (Haribal et al., 25, 26, 29). In Jun Jul Aug Sep Oct Nov Dec 7 6 5 4 3 2 1 Mites response to increased secretions, mite populations might increase, as more secretion means more food, thus providing favorable conditions for mites to breed. Birds apply secretions 2 3 times a day (Gill, 1995) and these contain a mixture of numerous compounds that differ in volatility and solubility. Thus, some compounds may evaporate from the feathers faster than others. This could lead to an accumulation of higher boiling chemical materials on the feathers over a period of time. If feather mites remove the build-up of nonvolatile waxes and other non-waxy components, and other organisms such as fungi and bacteria which may colonize these build-ups, they would be beneficial as suggested by Blanco et al. (21). However, if the population of mites increases so much so that they deplete compounds useful to the bird faster than they can be replaced, then the mites might become harmful to the birds, which might result in the evolution of some mechanism that would check the build-up of mites. This scenario might explain why (1) certain compounds, such as homologous series of some esters are present in secretions year round, while others are found only in late summer and fall; (2) that

88 Haribal et al. 211 Change in TIC after mites peaked (%) 8 7 6 5 4 3 2 1 1 42 A 51 64 2 16 17 Jul Aug Sep Fig. 1. Percent change in amount of secretion of individual birds after the mite numbers peaked. The letters above the individual bars indicate the months in which the mites peaked, numbers indicate the maximum numbers counted on the feather and the difference in amounts of secretion (as sum of TIC) for the month after the peaks of mites and the month mites peaked. The bar in dark with an asterisk is the individual with very high infestation and died later in October. Amount of secretion was significantly higher in the month after the mite infestation reached a peak in individual birds. (95% CI T-test. df = 14, P <.5). when the total amount of secretion increases the number of components also increases during that period; (3) why the amount and complexity of secretions in individual birds is subsequent to the peak in mite abundance. A peak in mite abundance occurred at different times in different individuals. In most birds, the amount of secretion reached a maximum after mite numbers reached a peak. Two possibilities are that the increase in amounts of secretion and components is a response to the occurrence of mites, or that birds have evolved to produce higher amounts of secretions in this period regardless of presence or absence of mites. Our study of House Finch secretions and Proctophyllodes pinnatus suggests that there is interaction between the two organisms potentially mediated by secretions; however, with only observational data we cannot tease apart the cause and effect. Manipulative experiments are required in order to determine if high mite loads are the cause or consequence of increased preen secretions. There are numerous other ectosymbionts that 14 4 * B C D E F G H I J K L M N O 64 Individuals 65 64 12 99 74 78 Oct are present on the feathers that might affect the hosts, and also other mite species and other organisms infest birds on different body parts. A complete understanding of the natural history of these organisms and dynamics of secretions and host behavior related to these is necessary. We are currently, experimentally modifying mite loads on birds and measure the resulting secretion responses to understand the interaction and are conducting long-term studies to observe the effects of mites on birds. ACKNOWLEDGEMENTS This research was funded by a Cornell University Biogeocomplexity and Biodiversity Initiative award to AAD and ER and by NSF (DEB# 54336 and 62275). We wish to thank Melanie Driscoll, Chris Jenelle, Tom Muscato, Elliot Swarthrout, Keila Dhondt and Mari Kimura in help with collection of samples. We also wish to thank anonymous reviewers for their comments and suggestions on the previous versions of the manuscript.

Vol. 37, No. 1 Internat. J. Acarol. 89 REFERENCES Atyeo, W. T. and N. L. Braasch. 1966. The feather mite genus Proctophyllodes (Sarcoptiformes: Proctophyllodidae). Bull. Univ. Nebr. State Mus. 5: 1 354. Atyeo, W. T. and R. M. Windingstad. 1979. Feather mites of the Greater Sandhill Crane Grus canadensis-tabida. J. Parasitol. 65(4): 65 658. Behnke, J., P. McGregor, J. Cameron, I. Hartley, M. Shepherd, F. Gilbert, C. Barnard, J. Hurst, S. Gray and R. Wiles. 1999. Semi-quantitative assessment of wing feather mite (Acarina) infestations on passerine birds from Portugal: Evaluation of the criteria for accurate quantification of mite burdens. J. Zool. (Lond.). 248(3): 337 347. Behnke, J. M., P. K. McGregor, M. Shepherd, R. Wiles, C. Barnard, F. S. Gilbert and J. L. Hurst. 1995. Identity, prevalence and intensity of infestation with wing feather mites on birds (Passeriformes) from the Setubal Peninsula of Portugal. Exp. Appl. Acarol. 19(8): 443 458. Bhattacharyya, S. P. 197. A comparative study on the histology and histo chemistry of uropygial glands. Indian Sci. Congr. Assoc. Proc. 57(4): 371 372. Blanco, G. and O. Frias. 21. Symbiotic feather mites synchronize dispersal and population growth with host sociality and migratory disposition. Ecography 24(2): 113 12. Blanco, G., J. Seoane and J. de la Puente. 1999. Showiness, non-parasitic symbionts, and nutritional condition in a passerine bird. Ann. Zool. Fenn. 36(2): 83 91. Blanco, G., J. L. Tella and J. Potti. 1997. Feather mites on group-living Red-billed Choughs: A non-parasitic interaction? J. Avian Biol. 28(3): 197 26. Blanco, G., J. L. Tella, J. Potti and A. Baz. 21. Feather mites on birds: Costs of parasitism or conditional outcomes? J. Avian Biol. 32: 271 274. Brown, C. R., K. R. Brazeal, S. A. Strickler and M. B. Brown. 26. Feather mites are positively associated with daily survival in cliff swallows. Can. J. Zool. 84(9): 137 1314. Bush, S. E. and J. R. Malenke. 28. Host defence mediates interspecific competition in ectoparasites. J. Anim. Ecol. 77(3): 558 564. Dowling, D. K., D. S. Richardson and J. Komdeur. 21. No effects of a feather mite on body condition, survivorship, or grooming behavior in the Seychelles warbler, Acrocephalus sechellensis. Behav. Ecol. Sociobiol. 5(3): 257 262. Dubinin, V. B. 1951. Fauna USSR. Paukoobrasnye (Arachnida), 6(S). Feather mites (Analgesoidea). Ch. I. An introduction to their study. Academy of Sciences, USSR, Moskow-Leningrad. 363 pp. Dubinin, V. B. 1953. Fauna USSR. Paukoobrasnye (Arachnida). Feather mites (Analgesoidea). Ch. II. Families Epidermoptidae and Freyanidae. 412 pp. Dubinin, V. B 1956. Fauna USSR. Paukoobrasnye (Arachnida). Feather mites (Analgesoidea). 6(7). Ch. III. Family Pterolichidae. 814 pp. Figuerola, J., J. Domenech and J. C. Senar. 23. Plumage colour is related to ectosymbiont load during moult in the serin, Serinus serinus: An experimental study. Anim. Behav. 65(3): 551 557. Galvan, I., E. Barba, R. Piculo, J. L. Canto, V. Cortes, J. S. Monros, F. Atienzar and H. Proctor. 28. Feather mites and birds: An interaction mediated by uropygial gland size? J. Evol. Biol. 21: 133 144. Galvan, I. and J. J. Sanz. 26. Feather mite abundance increases with uropygial gland size and plumage yellowness in Great Tits Parus major. Ibis. 148(4): 687 697. Gill, F. B. 1995. Ornithology. 2nd ed. W.H. Freeman, New York. Haribal, M., A. Dhondt and H. Proctor. 26. Parallel seasonal variation in preen gland secretion and occurrence of feather mites. J. Ornithol. 147(5, Suppl. 1): 36. Haribal, M., A. Dhondt and E. Rodriguez. 29. Diversity in chemical compositions of preen gland secretions of tropical birds. Biochem. Syst. Ecol. 37(1): 8 9. Haribal, M., A. A. Dhondt, D. Rosane and E. Rodriguez. 25. Preen gland secretions of passerines and their chemistry: Different pathways to same goal? why? Chemoecology 15(4): 251 26. Harper, D. G. C. 1999. Feather mites, pectoral muscle condition, wing length and plumage coloration of passerines. Anim. Behav. 58(3): 553 562. Hartup, B. K., B. Stott-Messick, M. Guzy and D. H. Ley. 24. Health survey of House Finches (Carpodacus mexicanus) from Wisconsin. Avian Dis. 48: 84 9. Jacob, J. 1978. Uropygial gland secretions and feather waxes. Chem. Zool. 1: 165 211. Jacob, J. and V. Ziswiler. 1982. The uropygial gland. pp. 199 322. In: Farner, D. S., J. S. King and K. Parker (Eds.). Avian Biology. Vol. VI. Academic Press, New York. Jovani, R. and G. Blanco. 2. Resemblance within flocks and individual differences in feather mite abundance on long-tailed tits, Aegithalos caudatus (L.). Ecoscience 7(4): 428 432. Jovani, R. and D. Serrano. 24. Fine-tuned distribution of feather mites (Astigmata) on the wing of