In many species of birds, parental care is provided by both parents to maximize offspring survival, and there may be important trade-offs between maximizing food gathering and nest protection during the nesting period. The role of parental care in determining reproductive success was investigated in Wood Storks (Mycteria americana), and specifically how the trade-off between frequency and duration of foraging trips and nest protection contributed to the nesting outcome. Parental behavior of 85 pairs of Wood Storks was monitored throughout the nesting season in two breeding colonies in Palm Beach County, Florida, USA. Wood Storks gradually increased the frequency, but not the duration, of foraging trips as chicks developed. The ratio of hatchlings to fledglings was positively associated with the frequency of foraging trips during late chick development. Intra-specific aggressions resulting in nest takeovers affected 32% of the nests. Occurrence of nest takeovers were higher for later-breeding pairs, and happened primarily in the first few weeks of incubation, but was not affected by the degree of joint nest attendance of both parents. These results establish a functional link between parental effort and reproductive outcome in Wood Storks, and highlight the importance of frequent foraging trips, but not nest attendance, by parents.
Any pre- or post-breeding investment made by a parent that increases offspring survival (parental care; Trivers 1972; Westneat and Sherman 1993; Royle et al. 2012) may affect reproductive success (Eggert et al. 1998). For many species of birds, including most waterbirds (Del Hoyo et al. 1992), bi-parental care is the rule (> 90%; Silver et al. 1985; Cockburn 2006; Harrison et al. 2009). The degree of parental care received by offspring can be crucial for their survival and, as a consequence, can affect the reproductive success of the parents (Elowe and Dodge 1989; Dijkstra et al. 1990; Boland et al. 1997). For instance, quality of parental care and body condition of parents, instead of quality of eggs (egg size), has been shown to affect chick survival in Short-tailed Shearwaters (Puffinus tenuirostris; Meathrel et al. 1993).
Wood Stork (Mycteria americana) behaviors such as incubation, brooding, number of feedings or nest protection have been studied (Clark 1980; Bryan and Coulter 1991; Bryan et al. 2005), but the effect of Wood Stork behavior on reproductive success has not been investigated yet. To our knowledge, no study has quantified the direct impact of food provisioning on survival of chicks in Wood Stork. In particular, altricial and semi-altricial hatchlings cannot feed themselves, yet require a copious and steady flow of nutrients to fuel rapid growth and development (Schwagmeyer and Mock 2008). However, provisioning takes time and energy, and parents should trade off optimal levels of offspring provisioning versus nest defense (Mutzel et al. 2013), clutch size (Dijkstra et al. 1990), clutch mass (Hébert and Barclay 1988), or parental body condition (Erikstad et al. 1997). Moreno et al. (1999) showed that increasing the intake of food positively affected reproductive success, but little is known about how the trade-off between food provisioning and nest protection impacts reproductive success.
In Wood Storks, both parents provide parental care (Kahl 1962, 1971; U.S. Fish and Wildlife Service 2001). Earlier studies suggest that daily care from both parents is strictly necessary to ensure survival of eggs and offspring in this species (Clark 1980; Bryan et al. 2005). Survival of nestling Wood Storks is affected by many factors, including predation (Rodgers 1987), human disturbance (Bouton et al. 2005), intraspecific aggression that can lead to nest takeover (Bryan and Coulter 1991), contamination with toxic chemicals (Fleming et al. 1984; Burger et al. 1993), as well as weather conditions, such as storms associated with strong winds (Coulter and Bryan 1995; Bouton et al. 2005; Bryan and Robinette 2008). However, the reproductive success of Wood Stork seems to be primarily related to prey availability in the environment (Ogden 1994; Griffin et al. 2008), which affects the ability of parents to provide sufficient food to sustain the development of chicks until fledging. Wood Storks feed mainly on fish (Kahl 1962, 1971; Ogden et al. 1976) captured in 15-50 cm of water (Coulter and Bryan 1993) using “tactolocation” (Kahl and Peacock 1963; Kahl 1964; Clark 1979). This technique is extremely sensitive to variations in fish availability. Sufficient rains prior to the breeding season are required to increase wetland water levels and increase prey population growth, followed by decreased water levels during the nesting phase to concentrate prey and ensure efficient foraging when energetic needs are highest (Kushlan et al. 1975; Ogden and Nesbitt 1979; Beerens et al. 2015). In contrast, heavy rains during the breeding season can result in colony abandonment due to dispersion of prey (Frederick and Collopy 1989; Ramo and Busto 1992).
Biparental contributions to nest protection and food provisioning by Wood Storks have been confirmed in several studies (Kahl 1962; Coulter et al. 1999; Griffin et al. 2008). Wood Stork parents must budget their time efficiently to provide adequate food and at the same time protect their young in the first weeks after hatching, when chicks are unable to defend themselves or thermoregulate independently (Bryan and Coulter 1987). In this first phase, the continuous presence of a parent on the nest is necessary (Clark 1980; Bryan et al. 2005). After three weeks, chicks exhibit a behavioral change, becoming aggressive to any approach by conspecifics and other species (Kahl 1971) and are able to thermoregulate (Clark 1980). This increased independence allows both parents to forage simultaneously when food requirements of the chicks are at their peak, leaving the nest unattended (Kahl 1962). Nestling Wood Storks remain in nests for 50 to 60 days before fledging (Kahl 1971; Coulter et al. 1999) and continue to return to nests to be fed by their parents for another one to three weeks (Kahl 1971; Borkhataria et al. 2012).
The trade-off between nest attendance and food provisioning has been noted in many species of birds (Komdeur and Kats 1999; Fontaine and Martin 2006; Tilgar et al. 2010). A key assumption in this trade-off is that nest attendance is related to nest success, through several mechanisms, such as direct care for chicks and defense of nest against predators and conspecifics (Giese 1996; Schmidt and Whelan 2005). To our knowledge, no study has investigated the relationship between parental care (measured as frequency of foraging trips, their duration, and nest attendance) and reproductive success in Wood Storks, but this has been studied in other wading birds (Miller and Burger 1978); likewise, the importance of parental attendance during the nestling period has been little studied in this group. Thus, our objective was to determine if parental care (nest attendance and foraging behavior) varies during the nesting season and, if so, if such variation affects reproductive success. We predicted that the number of parental foraging trips would increase with the age of chicks, and the mean duration of foraging trips would decrease. We also predicted a positive relationship between number of foraging trips and reproductive success. Second, we assessed a possible relationship between the time spent by parents at the nest and the occurrence of takeover events. We predicted that the percentage of time spent at the nest simultaneously by both parents would reduce the likelihood of takeovers.
Methods
Study Area
We collected data on parental activities of 85 pairs of Wood Storks on 61 nests (some pair lost nests due to takeovers and abandoned nests) in two colonies in Palm Beach County, Florida, USA (Fig. 1). We observed 32 nests on two islands located within the Wakodahatchee Wetlands (26° 47′ 87″ N, 80° 14′ 34″ W), nesting on pond apple (Annona glabra), dahoon holly (Ilex cassine), and sabal palm (Sabal palmetto) (Bays et al. 2000). We also monitored 29 nests in pond apple and sabal palm in the BallenIsles Country Club, situated on a single island within a golf club (26° 83′ 01″ N, 80° 10′ 91″ W), located 46.6 km north of Wakodahatchee Wetlands (Fig. 1). It is important to note that Palm Beach County and the large adjacent protected marsh lands (including Loxahatchee National Wildlife Refuge) are a hotspot for the year-round distribution of resident Wood Storks, which may indicate that foraging habitat is generally good in this area (Picardi et al. 2020). The 2017 breeding season in Palm Beach County was characterized by lower precipitation (2.68 mm in January and February, 5.26 mm in March through May) than the long-term average (9.33 mm for December through February 1981-2010, 12.76 mm for March-May 1981-2010; data from the National Oceanic and Atmospheric Administration).
Figure 1.
Location of Wood Stork (Mycteria americana) study colonies (white circles) in Palm Beach County, Florida, USA. Protected and natural areas are presented in light gray; roads and interstates as grey and black lines, respectively. Inset shows an adult Wood Stork attending its nest with a three-weeks-old nestling.

Data Collection
We observed breeding behavior in the two colonies from 31 January-2 June 2017. Two observers conducted separate 5-h-long surveys biweekly at each colony, once in the morning (07:30 hr to 12:30 hr) and once in the afternoon (12:30 hr to 17:30 hr), as to homogenize time gaps between successive surveys at a single site (every 3.5 days, Monday morning and Thursday afternoon at BallenIsles, Tuesday morning and Friday afternoon at Wakodahatchee Wetlands). At each site, two groups of nests were observed in weekly alternation by both observers to prevent observer bias. Wood Stork behavior was observed using binoculars (12×50) and recorded using an SLR camera with a 600-mm telephoto lens. We began data collection when most pairs of Wood Storks were either building nests or beginning to incubate eggs. We identified each individual Wood Stork based on individually unique skin patterns on their head (Clark 1980; Bryan and Coulter 1991). We built a photographic database for individual identification consisting of photos of the right and left profile of each stork. Because Wood Storks lack evident sexual dimorphism, sexing partners of a pair was only possible when we witnessed copulation (Clark 1980; Fujioka and Yamagishi 1981; Bryan and Coulter 1991).
At the beginning of each survey, we recorded the status of each nest (construction, incubation, or post-hatching, and if individual parents present), took pictures of the individuals present, and recorded arrivals and departures. Identification was always made later using the photo database. The takeovers were either observed directly or inferred from the data showing the presence of different adults in the nest. A takeover can be carried out by a single individual or by a pair who are seizing an already built nest. This may result in the cessation of egg incubation or the death of the original pair's chicks present in the nest at the time of the takeover. Trips were categorized as foraging or gathering nest-building material based on whether the parents regurgitated food to nestlings upon return or brought back twigs and other woody material, respectively. Trips were initially classified as unknown when parents neither fed nor carried nest material when returning. We found that 96% of foraging trips (showing regurgitation) lasted more than 44 min. Using that information, we categorized any unknown trips longer than 44 min as foraging trips. Laying dates for each nest were estimated by back-dating from hatching dates, using an average incubation period of 28 days (Rodgers and Schwikert 1997) and were matched to observations of parental brooding behavior (especially sitting position). We counted hatchlings in each nest where visual counting was feasible. We estimated reproductive success at each nest as the proportion of chicks (ratio of fledglings to hatchlings) still alive after 8 weeks, i.e., the estimated time when young birds leave the nest for the first time (Middleton and Prigoda 2001; Bryan et al. 2005). We determined early and late pairs according to their nesting date. Most pairs (48 out of 61) initiated nesting within the first week of study, before 8 February 2017, and were considered early nesters. A second wave initiated nesting between 20 February and 21 April 2017, and were considered late nesters.
Analytical Methods - Changes in Frequency and Duration of Foraging Trips
We modeled both frequency and duration of foraging trips as non-linear functions of weeks-since-hatching using generalized linear mixed models (Mirman 2014). The overall shape of the curve was captured with inclusion of orthogonal polynomials on time up to the fourth order, with individual-within-nest random effects on all terms (for frequency only; for the duration analysis, the use of complete trips with known time of departure and arrival limited the sample size, and the estimation of random effects was not possible). Using a subset of individuals of known sex, sex differences were tested with the inclusion of an additive and multiplicative fixed effect of sex. Similarly, we included an additive and multiplicative fixed effect of the calendar date of nest initiation to test its effect on frequency and duration of foraging trips.
Effect of Frequency of Foraging Trips on Reproductive Success
The effect of frequency of foraging trips on reproductive success was assessed using a logistic regression on the ratio of fledglings to hatchlings in each nest (at 8 weeks after hatching). Fixed effects of the frequency of foraging trips during the early pre-flight stage (weeks 1-4) and during the late pre-flight stage (weeks 5-8) were included in the regression, after checking for their correlation. We have not reported results for the post-flight stage because the data were not sufficient to constitute a robust data set. In fact, at the end of the young's development (between 9 and 12 weeks), parents return very little to the nest to feed them. It happened several times that the parents did not return to the nest during the 5.5 hours of observation.
Nest Attendance and Risk of Takeover
We fit semi-parametric proportional hazards (SPPH) models to the time spent by parents at the nest prior to takeovers, expressed either as a function of calendar time or time within the nest cycle. The instantaneous risk of successful takeover of a nest at a given week was modeled as a function of the baseline hazard experienced by all individuals, and the proportion of time with at least one adult or two adults present at the nest during the week. In the semi-parametric approach, weak assumptions about the baseline hazard are made, allowing the estimation of the relative risk of takeover. Covariate effects are then estimated using a partial likelihood that does not require estimating the baseline hazard.
All statistical analyses were performed in the software R 3.3.0 (R Core Team 2017) using notably the packages “lme4” (version 1.1.13; Bates et al. 2015), “survival” (version 2.40.1; Therneau and Grambsch 2000), and “cowplot” for graphs (version 0.7.0; Wilke 2016).
Results
Wood Stork Monitoring
Between 31 January and 2 June 2017, we monitored 61 nests (29 in BallenIsles, and 32 in Wakodahatchee Wetlands), corresponding to 85 nesting attempts from individually identifiable pairs of Wood Stork (with sex identified for individuals of 71 pairs). Of these nest attempts, 27 were taken over by another pair, five were abandoned, and 53 either succeeded or were still active at the end of data collection (Fig. 2). Among surviving nests, we were able to track the fate of chicks until fledging for 29 nest attempts (see Table 1). For the other nest attemps, we were not able to track the fate of chicks until fledging due to either takeover, death of chicks, abandoned nest or the end of the observation period. We found an average of 3.00 ± 0.46 SD hatchlings per nest, and an average of 2.59 ± 0.57 SD chicks fledged per nest. We obtained a total of 70 5-h-long nest observations, and observations from the two sites were similar. For instance, comparing the two sites we obtained an average foraging rate of 0.243 h-1 for BallenIsles Country Club versus 0.239 h-1 for Wakodahatchee Wetlands. Moreover, the pairs with the highest chick survival were those with the highest mean frequency of foraging trips per hour in each site. We thus pooled both colonies to increase robustness of our results.
Figure 2.
Evolution of nest status (n = 61) during the nesting season in Wood Stork (Mycteria Americana). Color indicates the phases of each nest: Building phase (striped light gray), Incubation phase (dark gray), Post-hatching phase (light gray), Abandoned (striped dark gray). Nest takeovers (n = 27) are indicated by black crosses.

Table 1.
Reproductive success for 29 Wood Stork (Mycteria Americana) nests with known outcome in Palm Beach County, Florida, USA: Number of nests broken down by the number of chicks per nest at the time of hatching (rows) and at the time of fledging (columns).

Changes in Frequency and Duration of Foraging Trips
Adding orthogonal polynomials successively to the constant model of frequency of foraging trip significantly improved the fit until the quadratic term (χ2(7) = 14.792, P = 0.039; Table 2A), whereas adding a cubic or quartic term did not further improve the fit further (resp. χ2(1) = 1.011, P = 0.315; χ2(6) = 0.607, P = 0.436; Table 2A). Using the quadratic model as a baseline, model selection revealed that an additive or multiplicative effect of either sex (Table 2B) or initiation date (Table 2C) did not significantly improve the fit (Table 2B). The model including the effect of first- and second-order polynomials was kept for the rest of analyses (Table 3). This model showed a significant effect of the first-order orthogonal polynomial term (0.225 ± 0.06, t910 = 13.501, P < 0.001; Table 3) demonstrating a positive linear relationship between the frequency of foraging trips and the progression of chick development (Fig. 3A).
Table 2.
Model selection for frequency of foraging trips by Wood Storks (Mycteria Americana) relative to sex of individual adults and nest initiation date. K = number of parameters in the model; AIC = Akaike Information Criterion score; LL = log-likelihood; and χ2, df and P (>χ2) indicates the statistic, the associated degrees of freedom and P-value for the comparison between each model and the previous one (selected model is indicated in bold).

Adding orthogonal polynomials to the constant model of duration of foraging trip did not significantly improve the fit (all P > 0.05; Table 4A), although the first-order polynomial was close to statistical significance (F1,155 = 3.612, P = 0.059; Table 4A). We thus simplified the baseline foraging duration model as a simple linear model including a main effect of the number of weeks since hatching. Using the simple linear model as a baseline, model selection then showed that the additive or multiplicative effect of sex (Table 4B) and initiation date (Table 4C) did not significantly improve the fit. The simple linear model including no effect of sex or start time of nesting was thus kept for the rest of analyses, and indicated a weak trend of decreasing duration of foraging trips through time since hatching (-2.629 ± 1.388, t27 = -1.893, P = 0.060; Table 5; Fig. 3B).
Table 3.
Coefficients and their significance of the best model for the frequency of foraging trips during the post-hatching phase in Wood Stork (Mycteria Americana).

Effect of Frequency of Foraging Trips on Reproductive Success
We divided the post-hatching phase into three stages: early pre-flight (weeks 1-4 post-hatching), late pre-flight (weeks 5-8) and post-flight (weeks 9-12; Clark 1980). The frequency of foraging trips increased during each stage of chick development. However, the effect of the frequency of foraging trips on reproductive success was not significant during the early pre-flight stage (weeks 1-4; Z = 1.025; P = 0.306; Table 6). Frequency of foraging trips during the late pre-flight stage had a significant positive effect on reproductive success (Z = 2.929; P = 0.003; Table 6; Fig. 4). The frequency of foraging trips had no effect on the absolute number of fledglings (Z = 1.035; P = 0. 438). Pairs with the highest chick survival (P = 1) were those with the highest mean frequency of foraging trips per hour (0.25 to 0.35 h-1), whereas pairs with the lowest chick survival (P = 0.5) had the lowest mean frequency of foraging trips per hour (0.12 to 0.16 h-1). The mean frequency of foraging trips during the early and late pre-flight stages were not correlated (r = 0.216, t27 = 1.148, P = 0.261).
Figure 3.
Frequency (A) and duration (B) of foraging trips through time during the post-hatching nesting phase in Wood Stork (Mycteria Americana) (n = 157 5h-long nest observations; r2 = 0.22 and r2 = 0.02, respectively). Points represent average values for each week since hatching ± SE, and the light gray line represents the best model fit (see text for details).

Nest Attendance and Risk of Takeover
Both adults of each pair were simultaneously present at nests more than half of the time during nest building and the first week of incubation, but this proportion declined rapidly during incubation (Fig. 5A). Until the beginning of the early pre-flight stage, at least one adult was constantly at the nest. Then, the presence of even one adult gradually decreased, reaching a minimum of 10% of the time at the end of the nesting season (Fig. 5A). In our study, 32% of nests experienced a takeover. Of the 27 takeovers, the vast majority took place in Wakodahatchee Wetlands (26) vs. only 1 in BallenIsles. Most takeovers occurred between 13 February to 19 March 2017 (Fig. 5B). We found that the amount of time with presence of at least one parent at the nest did not significantly affect risk of successful takeovers (eβ = 1.030, Z = 1.289; P = 0.197; Table 7), and the amount of time with both parents present at the nest had a significant positive effect (increased risk) on successful takeovers (eβ = 1.017; Z = 2.506; P = 0.012; Table 7). Similarly, the date of initiation of incubation had a positive effect on the risk of takeovers, with later initiation dates associated with higher risk (eβ = 2.135; Z = 1.971; P = 0.048; Table 7; Fig. 5C).
Table 4.
Model selection for duration of foraging trips by Wood Storks (Mycteria Americana) relative to sex of individual adults and nest initiation date. SSR and SSE are the residual and explained sums of squares, respectively, with their associated degrees of freedom; F and P (>F) indicate the statistic and P-value for the comparison between each model and the previous one (selected model is indicated in bold).

Table 5.
Coefficients and their significance of the best model for the duration of foraging trips during the post-hatching phase in Wood Stork (Mycteria Americana).

Discussion
We found that the mean frequency of foraging trips per hour showed a gradual increase in time with the developmental stage of chicks. However, our sample of chicks in the post-flight stage was small and highly variable among different pairs, thus limiting our inference during the final stage. The duration of foraging trips was not influenced by the developmental stage of chicks. These results are in general agreement with those of previous studies of Wood Storks (Clark 1980; Bryan et al. 1995, 2005). Parents meet the increased energetic needs of chicks (Kahl 1962) by increasing the number rather than the duration of foraging trips (Bryan et al. 1995). During the post-hatching phase, the mean frequency of foraging trips per hour and their duration were not different between the sexes, and were not related to the onset date of incubation. Whether or not the duration of foraging trips is always stable as chicks grow is unclear, but may depend on the mosaic of wetland conditions in the area surrounding the colony (Coulter and Bryan 1993; Bryan et al. 1995). The dry season was not interrupted by major rainfall events that could reverse gradual water dry-down, and this resulted in good conditions for Wood Stork foraging in our study area (Kushlan 1986; Botson et al. 2016).
Table 6.
Coefficients and their significance of the logistic model for reproductive success ratio in Wood Stork (Mycteria Americana).

We found that the mean rate of foraging trips in the late pre-flight stage, but not in the early pre-flight stage, affected the proportion of chicks that fledged. Although our results showed that pairs with the highest chick survival were those with the highest mean frequency of foraging trips per hour, and that pairs with the lowest chick survival had the lowest mean frequency of foraging trips per hour, it is important to note the possible circularity of this relationship: in fact, one could argue that pairs that reached the late pre-flight stage with an already reduced brood size consequently decreased the frequency of trips due to reduced demands from the offspring, rather than the other way around. However, our results show that the relationship between reproductive success and frequency of foraging trips is independent of the absolute number of fledglings. For example, pairs with two successful chicks but low fledging success had either higher (0.23-0.28 h-1 for success ratio of 0.67, see Fig. 2) or lower (0.13-0.16 h-1 for success ratio of 0.5, see Fig. 4) foraging trip frequency than those with 100 % success ratio (0.17-0.23 h-1, see Fig. 4). If the absolute number of chicks was driving the frequency of foraging trips, we would expect the same frequency for an equal number of chicks independently from the initial brood size. This result supports an effect of foraging trip frequency in determining chick survival ratio, and not vice-versa. Because we found no effect of the frequency of foraging trips on the absolute number of fledglings, fledging success is apparently determined by other factors, such as mortality related to predation, sibling competition, or disturbance, as well as factors affecting clutch size (Burger 1982; Rodgers 1987; Bouton et al. 2005).
Figure 4.
Reproductive success ratio (fledglings/hatchlings) as a function of frequency of foraging trips in late pre-flight nesting stage in Wood Stork (Mycteria Americana) (n = 29 5h-long nest observations; r2 = 0.43). The shape of dots indicates the number of fledglings in each nest (diamond = 3; cross = 2; circle = 1) and the line indicates the logistic fit (with 95% confidence interval).

Takeover behavior appears to be widespread in Wood Storks, where it may affect more than a third of pairs in a colony (Bryan and Coulter 1991). In our study, 32% of nests experienced a takeover. The risk of a takeover was more than double for late pairs (nest initiation after 13 February). The lower occurrence of takeovers for early pairs could be explained by better intrinsic characteristics of individuals (Johnson and Kermott 1990) such as larger size, higher energy reserves, greater aggressiveness, or higher social status (i.e. dominance hierarchy rank). We also found that the time of simultaneous nest attendance by both parents did not reduce the occurrence of takeovers. On the contrary, and surprisingly, pairs with greater nest attendance times by both individuals had a greater chance of undergoing a takeover. Because Bryan and Coulter (1991) found that all takeovers occurred when a single individual was present at a nest, we expected that pairs that spent more time together at the nest would have a lower chance of getting their nest taken over, but, instead, we found that the presence of both individuals was associated with a greater risk of undergoing a takeover. Pairs may increase joint attendance when their perceived risk of attack is greater, but we have no data to test this possible explanation. In any case, increased parental attendance did not result in reduced risk of nest failure, suggesting that there may not be a clear trade-off between attendance and time spent foraging. Perhaps attendance of both parents at the nest does not provide appropriate protection against takeovers, and body condition of parents may be a determining factor as shown in the study by Meathrel et al. (1993) in Short-tailed Shearwaters.
Figure 5.
Nest attendance (A) by adult Wood Storks (Mycteria Americana) through time (n = 79 pairs). The color indicates nest attendance of two adults (dark gray), one adult (light gray) or no adults (striped light gray) (B: building phase; I: incubation phase; H: post-hatching phase). Kaplan-Meier survival curve (with confidence interval) for the risk of nest takeovers in 79 Wood Stork nests as a function of the week of the year (B) or week since start of incubation (C). Colors further distinguish between early pairs (dashed lined, top) and late pairs (dotted line, bottom).

Table 7.
Coefficients and their significance of the semi-parametric proportional hazards (SPPH) model applied to the risk of takeovers of Wood Stork (Mycteria Americana) nests. Exponentiated coefficients can be interpreted as multiplicative effects on the hazard, i.e. the instantaneous risk of takeover, holding other covariates constant. For instance, late starters have a risk more than twice as high as early starters (eβ = 2.135).

Wood Storks have long breeding cycles (∼110 days), and are often limited by food availability at the end of the dry season (Kushlan et al. 1975). Our results suggest that the frequency of foraging trips during the late pre-flight stage is positively associated with the reproductive success of Wood Storks. To our knowledge, this is the first time a link is established between parental effort and reproductive success for this species. Moreover, our findings highlighted an important mechanism of nest failure, nest takeovers, and that the risk associated with it varies throughout the season. Thus, in addition to constraints related to seasonal water level fluctuations, a late initiation of nesting can be associated with low reproductive success due to a higher probability of takeover events. It is important to point out that we have one season of data and that the reproductive success of Wood Stork is primarily related to prey availability in the environment (Ogden 1994; Griffin et al. 2008). This food availability is important for the development of chicks until fledging. Thus, a study carried out over several breeding seasons in the same location could provide more accurate breeding estimates, since foraging conditions are a dynamic system component.
Acknowledgments
We are grateful to Gary Myers and the management office of BallenIsles Country Club and to Robert Nelton and the Palm Beach County Water Utilities Department for granting access to their premises. Gary Ritchison provided valuable feedback on an earlier version of this manuscript, as did two anonymous reviewers. This work was supported by the USDA National Institute of Food and Agriculture (Hatch project 1009101).