Patch Selection by Snowy Egrets

Transcription

1 Patch Selection by Snowy Egrets TERRY L. MASTER 1,4, JOHN K. LEISER 2, KAREN A. BENNETT 3, JENNIFER K. BRETSCH 1 AND HEATHER J. WOLFE 1 1 Department of Biological Sciences, East Stroudsburg University, East Stroudsburg, PA 18301, USA 4 Internet: terry.master@po-box.esu.edu 2 Biology Department, Northampton Community College, Monroe Campus 3 Old Mill Road, Tannersville, PA 18372, USA 3 Delaware Natural Heritage Program, Division of Fish and Wildlife, Smyrna, DE 19977, USA Abstract. Models of the Snowy Egret (Egretta thula), that use the tendency of wild birds to join foraging aggregations, were used in manipulative experiments to test the ability of wild individuals to use environmental and social information indicative of patch quality, including pool depth, prey accessibility, prey density, and prey exploitation, when choosing foraging patches. Snowy Egrets, but not some other aggregation species, preferred shallow (<20 cm) compared to deep (>40 cm) pools, perhaps because they could wade throughout the shallow pools, disturbing fish prey and making them more accessible for capture. Snowy Egrets responded to an artificial increase in fish density by visiting prey-enhanced pools more than pools with most fish removed or unmanipulated pools. Snowy Egrets also responded to a decrease in prey density resulting from recent exploitation by aggregations by visiting these pools less than unmanipulated pools. These results indicate that, in addition to the presence of other individuals, Snowy Egrets used discernible pool characteristics such as depth, as well as the availability and density of prey when choosing patches in which to forage. Received 2 March 2004, accepted 25 August Key words. Egret model, Egretta thula, foraging, mixed-species aggregation, patch selection, Snowy Egret. Waterbirds 28(2): , 2005 A predator s decision regarding the patch in which to forage should ultimately be based on the net rate of prey capture from that patch (McNair 1982; Chen et al. 2001) or some measure of the predator s ability to exploit prey within that patch (Schoener 1971). Selection should favor those individuals that respond to patchy prey distributions by differentially exploiting the most profitable patches (MacArthur and Pianka 1966; Pyke et al. 1977). Differential exploitation requires an assessment of patch quality using information gained from a number of sources, including observing the physical and social environments, sampling different patches, and analyzing reward rates within patches (Tolman and Wilson 1965; Krebs 1974; Erwin 1983; Valone and Giraldeau 1993; Bateson 2002). The use of reward rates is valuable to foragers as they correspond to prey density and accessibility within the patch (Erwin 1988; Burke and Fulham 2003) and often relate to whether or not a patch was previously exploited. Avoidance of a previously exploited patch is particularly important in environments where prey renewal is relatively slow compared to the food requirements of the predator, and where the activities of the foragers themselves may alter the distribution, abundance, and availability of prey (Armstrong et al. 1987; Erwin 1988). For example, the Snowy Egret (Egretta thula) noticeably impacted prey fish density within salt marsh pools (Master 1992). The apparent density of prey items has been shown to influence the egrets foraging success and should be expected to play a role in their subsequent choice of patch (Erwin 1983; Master et al. 1993). Although their mechanism of patch choice is not entirely clear, Snowy Egrets may select patches based on prey density in a number of ways, including disturbance of the water s surface by the prey (Kushlan 1976a,b) or physical characteristics of the pools that indicate prey accessibility, such as turbidity which suggests recent disturbance by a foraging aggregation. The egrets may also possess spatial memory, having a cognitive map of the pools most recently visited (Menzel and Wyers 1981), or they may systematically investigate pools within a marsh such that encounters with exploited patches are uncommon (sensu Pyke 1978). Perhaps most importantly, egrets may use social cues such 220

2 SNOWY EGRET PATCH SELECTION 221 as the presence of a feeding aggregation within a pool as indicators of prey availability (Krebs 1974; Kushlan 1976a; Master 1992). Snowy Egrets are influential in exploitation of salt marsh pools in New Jersey where they are the commonest participants in mixedspecies feeding aggregations and where they likely serve as a focal species around which aggregations assemble (Master 1992). Here, we investigated the patch-choice decisions of Snowy Egrets and other species participating in mixed-species foraging aggregations in a salt marsh in southern New Jersey. The study was designed to examine the independent influences of predator aggregation and prey density and accessibility: (1) by evaluating the physical characteristics of the patches (i.e., pool depth) and (2) by manipulating prey fish density (a) through capture and removal or enhancement and (b) through exploitation by a natural foraging aggregation. STUDY AREA AND METHODS New Jersey offers a suitable environment for investigating patch selection in egrets; individual pools form numerous, discrete patches whose fish are exploited by wading bird aggregations. The major factor impacting prey density among pools is the foraging activity of the waders (Master 1992), while pool depth and vertical profile are the major influences on prey accessibility (Master 1992). Any influences that the waders have on prey density are erased following monthly spring tides which replenish fish numbers and pool water levels (Master et al. 1993); thus, the cycle of prey exploitation, replenishment and depth variation is repeated monthly. The study site consisted of a 2,500 ha salt marsh located in Stone Harbor, Cape May County, New Jersey ( N, W). Individual pools within the marsh have flat bottoms and vary in depth from 3-60 cm (Master 1992). On average, there are 810 individual pools km -2 within the study area ranging in surface area from 1-55,000 m 2 (Master 1992). They are isolated from daily tidal influence except during full moon periods when spring tides inundate the marsh to a depth of cm. Inundation lasts for a period of 1-2 h daily over 6-8 days, depending on monthly tidal cycles. Birds from local nesting colonies of herons, gulls and terns form foraging aggregations on the pools from sunrise until approximately h (Master 1992). The major participating species include the: Snowy Egret (Egretta thula); Great Egret (Ardea alba); Little Blue Heron (Egretta caerulea); Tricolored Heron (Egretta tricolor); Glossy Ibis (Plegadis falcinellus); Laughing Gull (Larus atricilla); Common Tern (Sterna hirundo) and Forster s Tern (S. forsteri) (Master 1992). Mummichogs (Fundulus heteroclitus) and Sheepshead Minnows (Cyprinodon variegatus) in approximately equal proportion, account for 95% of the fish found within the pools. Fish are confined to individual pools until the spring tides allow for dispersal over the surface of the marsh; monthly dispersal results in replenishment of fish numbers in pools depleted by foraging wading birds. Fish numbers within a pool increase in proportion to pool surface area (Master 1992). Mean (±SE) fish density within pools is 66 ± 43 fish m -2 (Master 1992). Both species of fish gulp air, creating disturbance of the surface of pools (Kushlan 1976a,b; Master 1992). Observations of Snowy Egrets and the other aggregation species visiting pools on the marsh were conducted from 1 June to 30 August during five summers ( and 1999). Visitation rates were derived from hourlong observation sessions conducted on pool arrays from sunrise to h, when flight activity to and from feeding and nesting sites was maximal. Individual pools used in the experiments were close enough to each other (<50 m apart) to be viewed simultaneously, thus providing a choice for birds flying overhead. During observations, the number of Snowy Egrets and all other birds visiting focal pools was recorded. To enhance visitation rates on experimental pools, model Snowy Egret aggregations composed of life-sized plastic decoys (Cabela s Sporting Goods Co.) were arrayed on the pools in each of the experiments. Each model aggregation consisted of five decoys that were placed 1 m apart with three birds in the water and two positioned along the shoreline. Snowy Egret decoys were used because they proved most attractive in previous experiments designed to investigate aggregation formation (Master 1992). The models simulated the early stage of aggregation formation when birds flying overhead are attracted to a group (Master 1992). The presence of models increased Snowy Egret visitation rate from 1.10 ± 0.25 to 1.80 ± 0.20 individuals h -1. During this period, interested birds typically flew directly down to the aggregation without circling overhead initially and very few aborted their descent to the pools. Wild birds were considered to have selected a pool when one or more landed or walked to within 3 m of it. Only one type of experimental observation session was conducted per day; no observations were conducted during spring tide periods when fish distribution was in a state of flux. Birds selecting focal pools were categorized as initial arrivals that were attracted to model flocks when other wild individuals were not present, or secondary arrivals that joined model flocks with wild individuals already in attendance. The social stimulus (model flock) for initial arrivals was identical at all pools under observation, thus any preference exhibited was likely to be influenced by other pool characteristics including physical features and/or prey conditions. Considering initial arrivals as a separate category eliminated the confounding influence of the birds cuing on one another s behavior. The social stimulus for secondary arrivals, comprised mostly of Snowy Egrets, was probably not similar for all pools during an observation session, depending upon the number and distribution of wild individuals present. Thus, differences in social stimulus among pools could not be discounted as a cause for preferences exhibited in this category. Snowy Egrets and other species were separated for initial arrivals, but were lumped together as secondary arrivals. 1. Recognition of Pool Physical Characteristics Snowy Egrets and other waders should prefer shallow pools that enable them to wade throughout the patch rather than deeper pools, where the birds would be confined to the perimeter. To determine whether waders chose pools based on water depth, we compared initial

3 222 WATERBIRDS arrivals on eight pairs of pools that were matched for surface area and perimeter length but where one pool was shallow (<20 cm depth) and the other deep (>40 cm). Secondary arrivals of all birds were considered on five of these pool pairs. Pool depths were chosen based on mean Snowy Egret leg length. Comparisons between pools were conducted using paired t-tests (Zar 1999); tests were performed with the aid of the statistical software STATISTICA (1998 ed. (c) Statsoft, Inc.). 2a. Recognition of Prey Density Fifteen sets, each of three pools, were used to study the effects of prey removal or enhancement on the initial arrival decisions of Snowy Egrets. Secondary arrivals were observed on nine of the fifteen sets. Each set of focal pools was matched for model placement, depth and other physical characteristics and included: one control pool that underwent no manipulations; one trapped pool, and; one enhanced pool. The trapped pool had a substantial portion of its prey fish removed using minnow traps the day before observation. Following removal from the trapped pool, the fish were introduced into the enhanced pool. It was assumed that this procedure nearly doubled the fish density of the enhanced pool, as pools of similar area and perimeter length tended to have similar fish densities (Master 1992). The visits by Snowy Egrets and the other birds were analyzed with one-way repeated measures ANOVAs with appropriate contrasts (Zar 1999). 2b. Recognition of Prey Depletion To determine whether foragers recognized previously depleted patches, eight sets of pools were observed for initial and secondary arrivals. Each set consisted of three pools; one that had been recently exploited by a natural foraging aggregation, one from which prey fish were trapped on the day prior to observation, and one that was unmanipulated and unvisited by an aggregation as determined by surveillance prior to the day of observation. The pools were matched as closely as possible on physical characteristics. Pools were compared using one-way repeated-measures ANOVAs with appropriate contrasts. RESULTS 1. Pool Physical Characteristics. Significantly more initially arriving Snowy Egrets visited a shallow rather than a deep pool (Paired t-test: t 7 = 2.31, P < 0.05; Table 1). Other initially arriving species did not exhibit a preference for pool depth (t 7 = 0.98, n.s.; Table 1). Secondarily arriving species showed a preference for shallow pools (t 4 = 2.66, P < 0.05; Table 1). 2a. Prey Density. Although the overall ANOVA was not significant (One-way repeated-measures ANO- VA: F 2,28 = 2.38, n.s.; Table 1), initially arriving Snowy Egrets differentially visited pools based on prey availability. That is, contrasts within the ANOVA revealed that initially arriving Snowy Egrets visited unma- Table 1. Initial and secondary arrivals by Snowy Egrets and other aggregation participants visiting focal pools. Mean ± S.E. number of Initial arrivals Secondary arrivals Pool Type Snowy Egrets Other species All species 1. Deep 1.31 ± ± ± 1.24 Shallow 2.62 ± ± ± (N = 8) (N = 5) 2a. Unmanipulated prey ± ± ± 0.52 Trapped prey ± ± ± 0.81 Enhanced prey 2.20 ± ± ± 0.42 (N = 15) (N = 9) 2b. Unmanipulated prey 5.63 ± ± ± 1.22 Trapped prey 2.13 ± ± ± 0.68 Naturally exploited prey 2.63 ± ± ± 0.77 (N = 8) (N = 8) 1 Significantly more Snowy Egrets visited shallow compared to deep pools during initial arrivals. 2 Significantly more birds visited shallow compared to deep pools during secondary arrivals. 3 Enhanced pools were visited more than other pool types combined for Snowy Egrets. 4 The unmanipulated treatment is significantly different from both of the other treatments individually for initial arrivals by Snowy Egrets. 5 Unmanipulated prey and trapped prey pools are different for each experiment since they were done at different times.

4 SNOWY EGRET PATCH SELECTION 223 nipulated and trapped pools similarly (ad hoc contrast: F 1,14 = 0.02, n.s.), but visited enhanced pools more frequently than the other pool types combined (F 1,14 = 8.41, P < 0.05). Enhancement of pools with prey fish did not affect initial visitation rates of other foragers (F 2,28 = 0.35, n.s.; Table 1). Secondary arrivals of all species were similar for the three types of pools (One-way repeated-measures ANOVA: F 2,16 = 1.09, n.s.; Table 1). 2b. Prey Depletion. The overall ANOVA was significant for initial arrivals (F 2,14 = 3.71, P < 0.05), but Snowy Egrets tended to visit the unmanipulated pool more than the trapped or exploited pools (ad hoc contrasts within the ANOVA: trapped vs. exploited pool: F 1,7 = 0.02, n.s.; average of trapped and exploited vs. unmanipulated pool: F 1,7 = 4.67, n.s.; Table 1). Other initially arriving foraging species did not show this trend (F 2,14 = 2.03, n.s.; Table 1), and no secondarily arriving species showed a preference (one-way repeatedmeasures ANOVA: F 2,14 = 0.16, n.s.; Table 1). DISCUSSION In this study, it was expected that Snowy Egrets and other wading birds would select foraging patches based on physical and social environments and on the direct availability of fish prey in salt marsh pools. Initially arriving Snowy Egrets, but not other initially arriving aggregation participants, more readily exploited shallow compared to deep pools. These results were consistent with previous studies that showed water depth to be a major factor in determining the number of foraging Snowy Egrets in coastal wetlands in Florida, North Carolina and California (Kushlan 1976a; Custer and Osborn 1978; Hom 1983; Edelson and Collopy 1990). Depth is distinguishable from the air by pool color (T.L.M. pers. obs.) and is unrelated to prey density as determined by mark-recapture studies, but in southern New Jersey does appear to be an indication of prey accessibility. That is, Snowy Egrets should be better able to disturb prey fish when wading into a pool relative to when foraging by leaning over the edge, as they often do for deep pools (Master 1989). Typical egret foraging behavior such as foot stirring, should work more effectively when the birds are capable of disturbing fish, making the prey more vulnerable to capture. That the other initially arriving aggregation participants displayed no preference for depth may be a reflection of their tendency to select pools based primarily on the presence of Snowy Egrets, or, in this instance Snowy Egret models, whose numbers were equal on both pool types. In contrast, secondarily arriving birds showed a preference for shallow pools. This may have occurred because these birds were primarily Snowy Egrets, or because the secondarily arriving birds were attracted to the model flock s numbers being enlarged by initial arrivals on the shallow pool. Secondarily arriving birds could have also used the foraging movements of the wild individuals already present as an indication of greater success in shallow pools. Any combination of these factors could have also been responsible. In addition to preferring shallow pools, initially arriving Snowy Egrets settled in patches that had enhanced prey density. This result supports the notion that foragers should choose a patch based on the potential to maximize prey consumption and is consistent with the work of Kushlan (1976a,b) who found that Snowy Egrets recognized surface water disturbances caused by fish gulping air. Although not seen in the other initially or secondarily arriving species participating in aggregations, fewer initially arriving Snowy Egrets visited pools that lacked fish activity, due to recent natural exploitation or trapping, compared to the unmanipulated pools. Both in this study, and previously (Kushlan 1976a,b), Snowy Egrets appeared to cue on pools that showed active, available, and accessible prey. Recognition of a lack of fish activity in naturally exploited pools could have been aided by alteration of pool color and clarity resulting from disturbance of bottom sediments by birds; however, surface disturbance by fish activity, or the lack thereof, is presumably the only cue available in trapped pools, as trapping caused little substrate disturbance.

5 224 WATERBIRDS The presence of foragers without disturbance of the water s surface by the fish, which is likely to decline as predators arrive or by the movement of the foragers themselves, may indicate that a patch has been exploited, or that the feeding rate by the foragers has declined. For instance, Sih (1998) suggested that greater aggregation of foragers should be observed in high resource patches, where prey intake is maximized, but it may be the case that new individuals should not join an aggregation without other evidence that the patch has not been exhausted of prey. Initially arriving Snowy Egrets primarily used physical pool characteristics and the visible appearance of prey within a pool in their decisions to land within a patch. Other initially arriving species did not appear to do so, probably because they cue on Snowy Egrets which were present in the form of an equal number of models on clusters of experimental pools, thus providing no obvious choice for the other species. Secondary arrivals of all species preferred shallow pools, but did not appear to respond to fish activity levels. This inconsistency may be attributable to variability in the number and species composition (and hence stimulus value) of birds already present at the pools. These results suggest that Snowy Egrets base their patch choices on the accessibility of prey and on the potential to maximize their foraging success. ACKNOWLEDGMENTS We thank M. Itzkowitz for helpful suggestions regarding fieldwork, experimental design and manuscript preparation and D. Klem, Jr. for assistance in the field. Support facilities were provided by Lehigh University and the Wetlands Institute of Southern New Jersey. This project was supported by grants from the Pennsylvania State System of Higher Education, the Wetlands Institute and the Department of Biological Sciences of East Stroudsburg University. LITERATURE CITED Armstrong, D. P., L. G. Clifton and G. D. Sutherland Should foragers remember where they ve been? Explorations of a simulation model based on the behavior and energetics of territorial hummingbirds. In A. C. Kamil, J. R. Krebs and H. R. Pulliam, Eds. Foraging Behavior. Plenum Press, New York. Bateson, M Context-dependent foraging choices in risk-sensitive starlings. Animal Behaviour 64: Burke, D. and B. J. Fulham An evolved spatial memory bias in a nectar-feeding bird? Animal Behaviour 66: Chen, T-C., R. F. G. Ormond and H-K. Mok Feeding and territorial behaviour in juveniles of three coexisting triggerfishes. Journal of Fish Biology 59: Custer, T. W. and R. G. Osborn Feeding-site description of three heron species near Beaufort, North Carolina. In A. Sprunt, IV, J. C. Ogden and S. Winckler, Eds. Wading Birds. National Audubon Society Report No. 7, New York. Edelson, N. A. and M. W. Collopy Foraging ecology of Wading birds using an altered landscape in central Florida. Florida Institute of Phosphate Research, Bartow, Florida. Erwin, R. M Feeding habits of nesting wading birds: spatial use and social influences. Auk 100: Erwin, R. M Predator-prey interactions, resource depression and patch revisitation. Behavioural Processes 18: Hom, C. W Foraging ecology of herons in a southern San Francisco Bay salt marsh. Colonial Waterbirds 6: Krebs, J. R Colonial nesting and social feeding as strategies for exploiting food in the Great Blue Heron (Ardea herodias). Behaviour 51: Kushlan, J. A. 1976a. Wading bird predation in a seasonally fluctuating pond. Auk 93: Kushlan, J. A. 1976b. Environmental stability and fish community diversity. Ecology 57: MacArthur, R. H. and E. R. Pianka On optimal use of a patchy environment. American Naturalist 100: McNair, J. N Optimal giving-up times and the marginal value theorem. American Naturalist 119: Master, T. L The influence of prey and habitat characteristics on predator foraging success and strategies: a look at snowy egrets (Egretta thula) and their prey in salt marsh pannes. Unpublished Ph.D. dissertation. Lehigh University, Bethlehem, Pennsylvania. Master, T. L Composition, structure and dynamics of mixed-species foraging aggregations in a southern New Jersey salt marsh. Colonial Waterbirds 15: Master, T. L., M. Frankel and M. Russell Benefits of foraging in mixed-species wader aggregations in a southern New Jersey salt marsh. Colonial Waterbirds 16: Menzel E. W. and E. J. Wyers Cognitive aspects of foraging. In A. C. Kamil and T. D. Sargent, Eds. Foraging Behavior: Ecological, Ethological and Psychological Approaches. Garland STPM Press, New York. Pyke, G. H., H. R. Pulliam and E. L. Charnov Optimal foraging: a selective review of theory and tests. Quarterly Review of Biology 52: Pyke, G. H Are animals efficient harvesters? Animal Behaviour 26: Schoener, T. W Theory of feeding strategies. Annual Review of Ecology and Systematics 2: Sih, A Game theory and predator-prey response races. In L. A. Dugatkin and H. K. Reeve, Eds. Game Theory and Animal Behavior. Oxford University Press, New York. Tolman, C. W. and G. F. Wilson Social feeding in domestic chicks. Animal Behaviour 13: Valone, T. J. and L. Giraldeau Patch estimation by group foragers: what information is used? Animal Behaviour 45: Zar, J. H Biostatistical Analysis, 4th ed. Prentice Hall, Inc. Upper Saddle River, New Jersey.