Few studies have quantified parental care patterns and diets of Mississippi Kites (Ictinia mississippiensis). Using video-recording systems, we quantified and analyzed nesting Mississippi Kite parental care and food delivery in eastern Arkansas. During the incubation period, male kites appeared to spend more time (
= 311.8 min/day, SE = 30.27) on the nest than females ( = 249.6 min/day, SE = 30.62), though the difference was not statistically significant. Females spent significantly more time on the nest during the brooding period than did males. Female kites also spent significantly more time shading nestlings than males. The amount of time that females stayed on the nest overnight during the incubation ( = 886 min/night) and brooding periods ( = 815 min/night) was longer than that of males ( = 806 and 739 min/night, respectively). Male kites delivered more food items (58%) than females (42%) and also delivered significantly more vertebrate food items than did females. Insects were the most common type of food items (80%) delivered to nests. Vertebrates made up 5% of the food items, and the remaining 15% could not be identified. Our data indicated differential sex-division in parental contributions by male and female Mississippi Kites, but a similar contribution in terms of overall parental effort.
Sex-specific parental care varies among species of raptors, but in general the female's presence at the nest is typically greater than that of the male. Newton (1978) suggested three types of division of labor that occur between the sexes of raptors during incubation: (1) males do not contribute to incubation, (2) males temporarily relieve the female, and (3) the sexes share equally in incubation. For many raptor species, the male provides little help with brooding, but delivers most of the food for the young (e.g., Newton 1978, Collopy 1984, Holthuijzen 1990, Good et al. 2001, Meyer et al. 2004). Attendance at the nests by both adults of some raptors decreases as the nestlings age (e.g., Collopy 1984, Dykstra et al. 2003). Also, females are more likely to leave the nest for longer periods of time to forage as the nestlings get older (Newton 1978, Collopy 1984, Dykstra et al. 2003).
Parental care of the Mississippi Kite (Ictinia mississippiensis) and other species of kites has been the focus of only a few studies and these lacked continuous observation periods or were based on small sample sizes. Based on limited data, adult Mississippi Kites in different geographical locations have been reported to show different patterns of parental care. In New Mexico at a single nest, the female brooded more than the male and the male fed the young more frequently (Airth-Kindree 1988). However, Shaw (1985) found that females incubated more than males in Texas (n = 5 nests); males brooded more than females; and females made significantly more food deliveries than males. Most other studies have reported that male kites made more food deliveries to the nest than females (Glinski and Ohmart 1983, Airth-Kindree 1988, Botelho et al. 1993).
Data on food items delivered to nests have been reported in several studies of Mississippi Kites, but Parker (1999) suggested that only one of these studies provided unbiased quantitative data. Quantification of diet for raptors often consists of analysis of pellets (Haywood et al. 1993) and/or prey remains (Sergio and Boto 1999), or observation of nest activities for long periods of time from a blind (Dykstra et al. 2003). Data based on prey remains can be biased because of loss of prey remains to scavengers (Rogers et al. 2005). Biased representation of food items in pellets is well known (e.g., Bednarz and Dinsmore 1985, Simmons et al. 1991). Direct observations may be influenced by observer fatigue (Rogers et al. 2005) and by the method used (Margalida et al. 2005) or the length of observation period may not sample nest activities and food items effectively. Also, the process of building of a blind in the nesting area can disturb the nesting adults and may bias results due to its presence near the nest. Furthermore, to optimize observations of the activities in the nest, the blind should be elevated high enough to allow observation into the nest bowl. In Arkansas, Mississippi Kites nest high in the canopy of super-emergent trees (St. Pierre 2006, Bader 2007). Hence, it is difficult to build a blind at a height above the ground that would allow for optimal and safe observation of nest activities. Therefore, we employed time-lapse video recording systems to observe activities at Mississippi Kite nests.
Video camera systems have been used to document nesting activities and analyze diets of birds for over two decades (e.g., Ouchley et al. 1994, Delaney et al. 1999, Booms and Fuller 2003, Lewis et al. 2004a). These systems have provided an accurate assessment of nest activities, prey delivery rates, and prey identification, especially as technology has improved (Lewis et al. 2004b, Rogers et al. 2005). There are numerous advantages of using video cameras, including the ability to observe multiple nests simultaneously, the close view of the nest, the reduced disturbance in the nesting area, and the permanent record of data that can be evaluated multiple times and by different observers.
We used time-lapse video recording systems at Mississippi Kite nests to quantify parental care patterns, identify food items delivered to nests, determine rates of food delivery, and explore how these aspects of parental care are related to sex. In addition, we investigated the relationship between reproductive success and food delivery rates. Specifically, we addressed the hypothesis that Mississippi Kites that have limited access to food resources have higher rates of nest failures (see Wiehn and Korpimäki 1997, González et al. 2006). Thus, we predicted that successful Mississippi Kite nests should have significantly higher food delivery rates than unsuccessful nests.
Study Area
The White River National Wildlife Refuge (WRNWR) is located in southeastern Arkansas. The WRNWR is approximately 62 800 ha in size and is divided into two units by Arkansas Highway 1. The South Unit of WRNWR, where the majority of this research was conducted, is the larger of the two units (approximately 41 400 ha).
The WRNWR consists mostly of bottomland hardwood forest, with some upland forest, fallow fields, agricultural fields, moist-soil impoundments, and 356 natural and man-made lakes. Dominant tree species of WRNWR include Nuttall oak (Quercus nuttallii), overcup oak (Q. lyrata), bald cypress (Taxodium distichum), sugarberry (Celtis laevigata), hickory species (Carya spp.), and green ash (Fraxinus pennsylvanica).
Methods
Nest-searching and Monitoring
Nest-searching for Mississippi Kite nests began in April 2004 and 2005. Searches were conducted with an outboard motorboat, all-terrain vehicle (ATV), kayak, or on foot. Details of our nest-searching techniques are described in Bader and Bednarz (2009).
Mississippi Kite nests were monitored and classified as occupied if an adult kite was observed on the nest site on two or more separate days. Nests were monitored every 3–4 d with a spotting scope or binoculars from a distance of approximately 50–100 m to minimize disturbance. Observed activities of adults at nests were used to determine nesting phase (incubation or brood-rearing) and to estimate the dates of hatching and fledging.
Video-recording of Kite Nests
We used three high-resolution infrared video-recording systems designed by Fuhrman Diversified Inc., (Fieldcam: Field Television System: LDTLV/Box/Versacam/IR60, Seabrook, Texas, U.S.A.) to monitor nests. These systems were designed for use in the field 24 hr/d. The recording systems consisted of a Sony VHS time-lapse recorder (SVT-LC300, New York, New York, U.S.A.) and the camera was a Sony Color Ultra-High Resolution Versacam (Bader and Bednarz 2009). We also used two supercircuits color mono-power infrared video cameras (PC177IR-1color, Liberty Hill, Texas, U.S.A.) and two supercircuits VHS time-lapse recorders (NCL3300, Liberty Hill, Texas, U.S.A.) to monitor kite nests. Videotapes were viewed through a television monitor using a Sony time-lapse videocassette recorder (SVT-RA168, New York, New York, U.S.A.).
We placed each video camera at a separate Mississippi Kite nest at which adult birds had been incubating for at least 7 d. We did not disturb nests during the first 7 d after laying to reduce the likelihood of the kites abandoning their nest. If possible, the camera was placed at or above the nest to provide a complete view of the activities that took place in the nest. The Supercircuits camera systems were placed in the nest tree approximately 40 cm above the nest.
Cameras recorded each nest 24 hr/d until the young fledged or the nest failed. The systems used VHS tapes and recorded 8 frames per sec, allowing a single T-160 videotape to last for 72 hr. Two or three 12-V deep-cycle marine batteries placed at the base of the nest tree powered each camera system for 3 d. Every third day, we changed batteries, replaced the videotape, and viewed the monitor to ensure the camera remained focused on the nest. Once the nest failed or the young fledged, the camera system was removed and mounted at another occupied nest.
Analysis of Video Data
The mean incubation bout length in min, mean number of min that each adult spent incubating during daylight, the mean brooding bout length, and the mean number of min each adult spent brooding during daylight were calculated from recorded video data. The proportion of brooding done by each adult each week after the egg(s) hatched was also determined at each nest studied. The length of time each kite spent on the nest overnight during the incubation and brooding periods was recorded from the last incubation switch until the kite left the nest the following morning. The mean shading bout length and the mean number of min each adult spent shading the nestling(s) each day was also quantified.
Prey items delivered to the nests were identified to taxonomic order or finer taxa if possible. All items that could not be identified at least to order were considered unknown. The daily mean number of food items delivered to the nests by each adult was determined. The prey item most commonly delivered was also determined weekly and for the entire nestling period. The gross number and number per order of food items delivered to the nests by each adult were calculated for the entire nestling period.
Sex of male and female kites was assigned by observing behaviors of each video-recorded adult and learning to identify each individual by differences in plumage and unique characteristics. Male kites were typically a lighter gray and also had a sleeker and more “well-groomed” appearance; whereas, female kites were typically a darker gray and were not as well-groomed and usually had small white spots visible between the misplaced feathers on their back. These subtle plumage characteristics can be readily discerned from individual kites visiting nests through the intimate knowledge that we obtained by making repeated observations of the birds close-up in the video monitor. Further, DNA analysis on five adults confirmed that we were able to distinguish between sexes accurately using the video analyses based on these plumage characteristics.
Capturing Mississippi Kites
We used a mist-net system with a live Great Horned Owl (Bubo virginianus) or Red-shouldered Hawk (Buteo lineatus) near nests with nestlings (7–30 d of age) to capture adults (Barber et al. 1998). The mist-net system consisted of two 2.6 × 6 m mist nets (72-mm mesh) connected to a pulley system on telescoping metal poles. This allowed the top of the net to be elevated ∼7 m aboveground with the bottom of the lower net ∼1.5 m off the ground. The total dimensions of the system using two standard mist nets were 5.2 m tall × 6 m wide. We climbed to each nest and removed the nestlings at 3–4 wk of age. Adults and nestlings were banded with a U.S. Geological Survey (USGS) aluminum band and two or three plastic color bands. We recorded several linear measurements and mass, and collected a blood sample from each kite.
Sex Determination by DNA Analysis
A 0.05-cc sample of blood was drawn from the brachial vein of each kite captured (Faaborg et al. 1995). The blood sample was placed in Longmire Solution (Longmire et al. 1988) and stored until analyzed in the lab. Following the protocol outlined by Donohue and Dufty (2006), we extracted DNA from the blood, isolated the CHD-Z and CHD-W bands (Griffiths et al. 1998), and determined the sex of the kites.
A DNeasy Blood and Tissue Kit (Qiagen Inc., Valencia, California, U.S.A.) was used to extract the DNA from the blood samples. The same primers as described by Donohue and Dufty (2006), 2550F and 2718R, were used. The thermal cycling settings were as in Donohue and Dufty (2006). We used a 1% agarose gel containing 1.2 uL of ethidium bromide at a concentration of 10 mg/mL. The electrophoresis was run at 80 V for 45 min, then viewed and photographed under UV light.
Statistical Analysis
We used a mixed-model analysis of variance (ANOVA) in SAS statistical software (SAS Institute 1999) to compare nest attendance patterns and food delivery rates of adult Mississippi Kites. The analyses were performed on the raw data when distributions were normal, and square-root-transformed data or log-transformed data as needed to normalize data.
Results
Video-recording of Kite Nests
We located 21 Mississippi Kite nests during the 2004 field season and recorded data at seven of the nests using two systems designed by Fuhrman Diversified and one supercircuits system. The first camera was set up on 21 May 2004 and the last camera was taken down on 28 August 2004. During this time period, 176 d of data were recorded (
= 25 d/nest). A total of 75 d and ca. 1800 hr of video data from four of the seven video-recorded nests were analyzed for nest attendance patterns. The video data from the remaining three nests were not analyzed because of poor view of activities in the nests.
During the 2005 field season, we located 18 Mississippi Kite nests and recorded data at nine of the nests using three cameras designed by Fuhrman Diversified and two supercircuits cameras. The first camera was set up on 2 June 2005 and the last camera was taken down on 8 August 2005. A total of 225 d of data were collected ( = 25 d/nest). A total of 147 d and 3528 hr of video data from seven of the nine video-recorded nests were also analyzed for nest attendance patterns. Data were not analyzed from the remaining two nests because of poor view of the nest or short period of time (<3 d) that data were collected.
Eleven nests were used to quantify the nest attendance patterns. Only one of these nests had more than one nestling during the observed period. This nest contained two nestlings which were ca. 7 days old at the start of the video observation. When the nestlings were ca. 17 d old (day 10 of video observation), the smaller nestling was killed by its sibling. Thus, except for 10 d of video data (total = 222 d of data analyzed), all data and analysis were based on nests with a brood size of one nestling.
Nest-attendance Patterns
Analysis of video indicated that males incubated longer than did females during daylight hours and males also had longer bout lengths than did females (Table 1). Females brooded significantly (P = 0.0382, F = 5.42, df = 1; Table 1) more than did males during daylight hours. Female kites brooded during the first week after hatching at all nests. Male kites brooded some of the time during the first week (Fig. 1) at all but one of the nests. The female kite's attendance began to decline at the nests after the first week, but increased during the fourth week of the nestling period, then declined to 0 during the fifth week (Fig. 1). The male kite's attendance also diminished after the first week (Fig. 1). Male kites at six of the eight nests did not brood during the third week of the nesting period, and one and no males brooded during the fourth and fifth weeks of the nestling period, respectively (Fig. 1). The mean bout length (Table 1) for females during the brooding period was also greater than that of the males.
Figure 1
Mean percent time adult female and male Mississippi Kites spent on nests during the nestling stage by week in the White River National Wildlife Refuge, Arkansas, 2004–05. Percentage was calculated based on the times when an adult was present on the nest. Week refers to age of nestling (e.g., week 1 = nestling 1–7 d of age). No adults brooded during week 5.
Table 1
The means, standard errors (SE), and ranges of time spent on the nest per 24-hr period by male and female adult Mississippi Kites during the incubation and brooding periods in the White River National Wildlife Refuge, Arkansas, 2004–05.
Female kites spent significantly (P = 0.016, F = 8.38, df = 1) more time shading the nestling(s) than the males and had significantly (P = 0.022, F = 7.13, df = 1) longer shading bout times (Table 1). Although not statistically significant (Table 1), the mean time female kites spent on nests overnight during the incubation period and brooding period (P = 0.413, F = 1.1, df = 1, and P = 0.295, F = 1.3, df = 1, respectively) was greater than that of the males in our small sample (Table 1). Exchanges by adults did not occur at night during either the incubation or brood-rearing periods.
The longer overnight stays were a result of the female coming to the nest in the early afternoon (13:00–14:00 H) and staying until the next morning. If there was a rain event early in the afternoon, the female would brood during the rain and stay for the rest of the day and into the night. If it rained in the morning, the female would stay until the rain stopped, adding several hours to the overnight nest attendance bout.
Adult kites did not brood on some nights starting the third week of the nestling period at two of the nine nests monitored. Brooding overnight stopped completely at one nest during the third week of the nestling period because the female was depredated and the male did not brood at night. For the remaining nests, brooding overnight stopped completely during the fourth week of the nestling period. One nestling, ca. 5 wk old, stayed away from the nest (roosted in branches) two nights prior to fledging, but returned to the nest the following mornings at about sunrise to receive food from the adults.
Diet Analysis
Food items delivered by adult Mississippi Kites to the nestlings were quantified at seven (127 d, 1524 hr) of the 16 nests observed by video cameras in 2004 and 2005. Data were not quantified at nine nests because the camera setup was not adequate for a clear view of the entire nest bowl, video did not have adequate resolution to identify most food items, or some nests were only recorded during the incubation period and failed before food delivery data could be recorded. Four orders of insects, one order of amphibians, one order of reptiles, one order of mammals, and two orders of birds were documented (Table 2). In addition, five more birds were documented, but could not be identified (Table 2). Insects made up 80% of the food items delivered to nests. The most common insects delivered (52.1%) to the nests were cicadas (Homoptera: Cicadidae). Cicadas were the most common food item fed to nestlings by all adult males and they were the most common food item fed by five of the seven adult females. Dragonflies and damselflies (Odonata) were the second most common food item (26.1%) fed to the nestlings (Table 2). Dragonflies and damselflies were the most common food items fed by two adult females and were the second most common food item fed by the remainder of the adult kites.
Table 2
Prey items fed to nestling Mississippi Kites by adults as observed with video cameras in the White River National Wildlife Refuge, Arkansas, 2004–05.
Other insects delivered to the nest include; grasshoppers and katydids (Orthoptera, 1.7%), and butterflies and moths (Lepidoptera, 0.07%; Table 2). Vertebrates made up only 5.3% of the food items delivered to the nest, the most common of which were Anura (4.8%), including eastern cricket frogs (Acris crepitans), southern leopard frogs (Rana sphenocephala), and unidentified species of toads (Bufonidae). Other vertebrates delivered to the nest included skinks (Squamata), and birds (Aves). One Mississippi Kite nestling was a food item because it was fed to its sibling by its mother after it had been killed by that sibling. The only mammal delivered to a nest was a small bat (Chiroptera). We could not identify 418 items (14.7%; Table 2) because they were either too small to see clearly or the view of the item was blocked by the adult. Several unknown food items were thought to be beetles (Coleoptera), pieces of dragonflies, or items that had been regurgitated by the adult.
Adult male kites made more food deliveries (1646, 58%) in both 2004 and 2005 than did adult female kites (1203, 42%). The mean number of food items delivered daily by male kites was 13.5 (n = 144 d at nine nests, SE = 1.26) and ranged from 3 to 43 items. The mean number of food items delivered daily by female kites was 10.7 (n = 129 d at nine nests, SE = 1.26) and ranged from 1 to 41 items. Food delivery rates of female and male kites did not differ (P = 0.309, F = 1.2, df = 1). The adult male kites made significantly (P < 0.001, F = 22.0, df = 1) more daily deliveries of vertebrate food items than did females. The mean daily delivery rate of vertebrate food items by males was 1.06/d (SE = 0.14), compared to a mean daily delivery rate of vertebrate food items by females of 0.11/d (SE = 0.15). Male kites delivered 94% of the frogs (n = 118), 89% of the toads (n = 8), 88% of the skinks (n = 7), 100% of the birds (n = 7), and the only bat (Chiroptera). Only one male kite delivered food items (n = 5) to an incubating female. The video data documented adult males and females delivering two food items simultaneously to a single nestling on eight occasions. These food items consisted of two dragonflies (n = 4), a dragonfly and a cicada (n = 2), a frog and a dragonfly (n = 1), and two cicadas (n = 1). Males only fed the nestling(s) directly when the female was absent from the nest or when both the male and female were feeding the nestling(s), at all other times, the male would pass the item to the female.
The most commonly delivered items changed between week 2 and 3 of the nestling period. Odonates were the most common items during the first 2 wk of the nestling period (Table 3). The most commonly delivered items switched to cicadas the third week of the nestling period and continued to be the most common items throughout the rest of the nestling period (Table 3).
Table 3
Numbers and percentages (%) of the two primary food groups delivered to Mississippi Kite nestlings by adults in the White River National Wildlife Refuge, Arkansas, 2004–05. Week refers to age of nestling (e.g., week 1 = nestling 1–7 d of age).
Although the difference was not statistically significant (P = 0.503, F = 0.5, df = 1) food delivery rates were slightly higher at unsuccessful nests ( = 24.5/d, SE = 2.34) than at successful nests ( = 22.5/d, SE = 1.36).
Sex Determination by DNA Analysis
Analysis of DNA to determine sex of captured adult and nestling kites was conducted on 14 blood samples (6 adults and 8 nestlings; Bader 2007). In total, five male and one female adult kites and four male and four female nestling kites were sexed. Five of the six adults (83%) were identified correctly based on a brief one-time, in-hand assessment of coloration. The bird that was identified incorrectly was a male, mistaken for a female. The misidentified kite was not at a video-sampled nest. Therefore, it did not compromise any of the sex-related parental care data. The other five adult kites that were sexed were video-recorded and used in the parental care analysis.
Discussion
Identification of sex by plumage while having only one member of a breeding pair in hand was 83% accurate (5 of 6 adults). The variation in light gray colorations that distinguish males from females were not always conspicuous when viewing one member of the pair separately (Parker 1999). However, the use of behavioral cues and plumage to identify the sex of breeding adults was more accurate when viewing the videos, because both adults were often on the nest simultaneously, thus colors and characteristics of both individuals could be compared directly. Distinguishable colors in the plumage of adults and distinct feather characteristics of the females made determination of the sex of the adults on video recordings accurate. Color-band combinations could rarely be seen in the videos because of camera placement. Therefore, bands were rarely used to identify individual kites.
Only two studies reported incubation bout lengths (Evans 1981, Shaw 1985). Evans (1981) suggested that both female and male kites in southern Illinois incubated for 120 to 240 min (n = 12 nests) and that female bout lengths “appeared slightly longer.” Shaw (1985) reported that females incubated 53% of the time, males incubated 43% of the time, and the nest was unattended 4% of the time in west Texas (n = 5 nests). Shaw reported that the mean incubation bout lengths of female and male kites were 105 and 85 min, respectively. In Arkansas, we found that incubation bout lengths were similar between male and female kites (Table 1).
Newton (1978) suggested three types of division of labor that occur between the sexes of raptors during incubation: the males not incubating, the males temporarily relieving the females, and equal sharing by sexes. Mississippi Kites in this study seem to fit the third type with the male doing 51% of the incubation and the female doing 49% of the incubation. The mean amount of time males spent during daylight hours on the nest during the incubation period (312 min) was longer than that of females (250 min). After the eggs hatched, the mean amount of daily time males spent on the nest decreased to 80 min, and the females' mean amount of daily time spent on the nest decreased to 158 min. This decrease in the amount of time that the males spent on the nest during the nestling period reflects a role switch from incubating eggs to primarily hunting and food delivery. Moreover, males also made more food deliveries (58%) than females throughout the nestling period. The same is true for most male raptors (Newton 1978). For example, male Golden Eagles (Aquila chrysaetos) delivered 83% of prey items and 95% of prey biomass in Idaho (Collopy 1984). Collopy (1984) stated that the males' role during the nestling period was to provide food, while the female did the majority of the brooding. Female kites did the majority of the brooding throughout the nestling period with the exception of one nest, where the female was killed the third week of the nestling period. For urban-nesting kites in Texas, Shaw (1985) reported that male kites brooded 63% of the time and females brooded 34% of the time during the early part of the nestling period. Furthermore, she recorded that female kites (56%) delivered significantly more food per week than males (44%; n = 5 nests, 327 feedings). We documented that male kites in Arkansas brooded less and made more food deliveries than females. This study and Shaw's (1985) indicate that the roles of adult kites may vary depending on urban or rural setting, or by geographical location.
Evans (1981) reported that kites in Illinois brooded overnight until the nestlings fledged. We documented that brooding overnight became irregular during the third week of the nestling period at some nests. Overnight brooding at a majority of the nests sampled ceased during the fourth week of the nestling period. The absence of adults at the nests during the night may be an antipredator defense. Video documentation (Bader and Bednarz 2009) of predator attacks at night show that adult kites do little or nothing at all to defend their young and leave the nest until daylight the next morning. The absence of an adult in the nest may result in less scent around the nest, which predators such as rat snakes (Elaphe spp.) use to locate nests (Halpern 1992). The warm nights (typically >21°C) in southern Arkansas may also be another reason why adults here do not brood during the night. By the third and fourth week of the nestling period, nestlings can thermoregulate independently. Furthermore, the fact that one nestling stayed away from the nest at night, then returned in the morning, may also be seen as an adaptation to avoid predators. The scent that attracts predators to nests, which may include scent of food remains in the nests, is probably the strongest by the end of the nestling period and may be more likely to attract predators at that time. Martin et al. (2000) documented that predation increased later in the nesting period at passerine nests. Therefore, the nestling may be safer on the limbs of the tree away from the nest than it is in the nest.
The food items most commonly delivered to the nestlings were insects, whereas vertebrates composed only 5% of the identified food items. Other studies have reported similar diets. Glinski and Ohmart (1983) documented that 85% of food deliveries were insects in Arizona. Shaw (1985) reported that adult kites delivered 95% insects and 5% vertebrates to nests in Texas. Cicadas were the most commonly delivered food items in this study during both years. Cicadas were also the most commonly delivered food item in New Mexico (Airth-Kindree 1988, Botelho et al. 1993), in Texas (Shaw 1985), and in Arizona (Glinski and Ohmart 1983). Overall, dragonflies were the second most frequently delivered items in this study, but were the most common food items delivered the first 2 wk of the nestling period. Odonates were not documented as part of the Mississippi Kite's diet in Arizona (Glinski and Ohmart 1983) or New Mexico (Airth-Kindree 1988, Botelho et al. 1993), and were rarely consumed in Texas (Shaw 1985). However, dragonflies were one of the primary food items for kites in Illinois (Evans 1981). Odonates and cicadas were both present at the onset of kite hatching (mid-June). Odonates may have been taken most commonly during the first 2 wk because of their local abundance around nest sites, ease of capture, and smaller size. We observed that the smaller size of the Odonates made it easier for adults to tear apart and feed to young nestlings (T. Bader and J. Bednarz unpubl. data). The larger-bodied cicadas took more time and effort for adults to tear apart and feed to young. By the third week of the nestling period, nestlings were able to handle cicadas and feed themselves, making it more efficient for adults to capture and deliver cicadas to the nests and then continue hunting.
Food delivery rates at successful and unsuccessful nests did not differ significantly. Although we expected that adults successful at producing young would have higher food delivery rates, in fact, delivery rates were slightly higher at unsuccessful nests. Therefore, the data did not support our hypothesis that Mississippi Kites that provided less food to their nestlings at WRNWR would have higher rates of nest failures. Our findings do not necessarily refute the food limitation hypothesis. Lack (1954), Newton (1980), and others emphasized that food often limits reproductive success and recruitment of individuals into a breeding population. In support of this hypothesis, several studies have shown that supplemental feedings alter parental-care strategies and also increase nestling survival (e.g., Yom-Tov 1974, Soler and Soler 1996, Ward and Kennedy 1996, Wiehn and Korpimäki 1997, Dewey and Kennedy 2001, González et al. 2006). Although we documented no significant difference in food delivery rates between successful and unsuccessful nests, we suggest that food may be limited for the entire Mississippi Kite population in the WRNWR, resulting in smaller brood size and low reproductive success compared to other kite populations (Bader and Bednarz 2009).
Acknowledgments
This project was primarily funded by the U.S. Fish and Wildlife Service (USFWS) and Arkansas Game and Fish Commission (AGFC) through a State Wildlife Grant. We would like to thank Karen Rowe (AGFC) for all her knowledge, support, and assistance on this project. Further, we thank the USFWS for help on the logistics on this project. We also thank Richard Hines (USFWS) and the staff at the White River National Wildlife Refuge for all the help they provided on this project and also for looking for kites. We especially are grateful to the 2004 and 2005 research technician, Waylon Edwards, for his endless amount of hard work and dedication to this project. We greatly appreciate the assistance in the field and in the lab provided by Jeremy Brown, T.J. Benson, Dick Baxter, Travis Edwards, Jim Rowe, Erika Bader, and Amy St. Pierre. Finally, we thank Cheryl Dykstra, Antoni Margalida, Travis Booms, Jennifer Coulson, Jim Parker, and other anonymous reviewers for their helpful comments on earlier versions of this manuscript.
