Feeding Behavior in Laughing Gulls: Compensatory Site Selection by Young
Total Page：16
File Type：pdf, Size：1020Kb
Condor85:467-l73 G The CooperOmthological Society 1983 FEEDING BEHAVIOR IN LAUGHING GULLS: COMPENSATORY SITE SELECTION BY YOUNG JOANNA BURGER AND MICHAEL GOCHFELD ABSTRACT.-Feeding behavior of different age classesof Laughing Gulls (Larus atricillu)was studied in different habitats in Texas and Mexico; feeding methods were compared. Adults generally had higher capture successrates and shorter intervals between obtaining food items than young, while subadultswere inter- mediate in both measures.Where food was most readily available (i.e., garbage dumps and fish offal), differencesamong ageclasses were negligible.Age differences increased when intervals between obtaining food increased, reflecting the greater difficulty of the feeding task. Proportionately more young fed in those situations where their interfood intervals and successrates most closely approached those of the adults. Young gulls thus appeared to compensate for their generally lower feeding successby feeding at certain sites. Age differencesin foraging behavior and feed- coast of the Gulf of Mexico. We wanted to ing successhave been found for several avian learn whether age differences in ability varied species,including Little Blue Herons (Egretta with the difficulty of the task, and whether caerulea;Recher and Recher 1969) Brown habitat useand method of feeding differed with Pelicans(Pe1ecanu.s occidentalis; Orians 1969), age. We selectedLaughing Gulls becausethey Olivaceous Cormorants (Phalacrocorax oli- are abundant, feed in a variety of habitats, and vaceus;Morrison et al. 1978) Sandwich Terns use different feeding methods; we chose the (Sternasandvicensis; Dunn 1972) Royal Terns Gulf of Mexico region becauseboth adults and (S. maxima; Buckley and Buckley 1974) young are common there. Glaucous-winged Gulls (Larus glaucescens; We predictedthat (1) for difficult tasksyoung Barash et al. 1975, Searcy 1978) and Herring should take longer and/or be successfulless Gulls (L. argentatus;Verbeek 1977a, b, c, In- often than adults, (2) young should congregate golfsson and Estrella 1978). In general, birds in those areas where the task is easiest, and feed more effectively as they gain experience. where their feeding successapproaches that of One measure of feeding effectiveness is the adults, and (3) subadults should be interme- “capture successrate, ” defined as the number diate in ability, approachingadult rates in eas- of items obtained per number of capture at- ier tasksand young ratesin more difficult tasks. tempts. Age differencesin capture successhave been usually attributed to the difficulty of the STUDY SPECIES AND SITES task (such as plunge-diving which requires a The Laughing Gull, a monotypic, hooded gull, period of learning), or the difficulty in learning breeds locally on the Atlantic coast from New to recognizesuitable food items, or both. Many England to Florida and the West Indies, on the aspects of foraging behavior may vary with Gulf of Mexico to Yucatan, and on rocky islets age, such as the time required to search for off the coastof Venezuela and northern Mexico prey, pursue the prey once it is found, and (A.O.U. 1957). The main breedingarea is from capture,handle, and eat the prey. “Search time” the Carolinas through the Caribbean. The may be defined as either the total time a bird specieswinters from the Middle Atlantic states searchesfor food each day or the time required to Peru and Brazil. to find individual prey. Search time varies in We identified three age classes:adults (with relation to the ability to find the prey, or the an all-white tail), subadults(resembling adults percentageof successfulcaptures, since an un- in body plumage but with a partial narrow sub- successfulattempt lengthensthe time required terminal tail band), and young(dark-plumaged to locate another prey item. birds with a broad tail band, seeDwight 1925). Variations in age differences in foraging In our field work on the Yucatan Peninsula we within a speciesunder different conditionshave readily encountered enough young and adult not been hitherto studied. We examined the plumaged birds for sampling, but only a small foraging behavior of Laughing Gulls (Larus proportion of gulls in subadult plumage. atricillu)under different conditions along the We studiedLaughing Gulls in Galveston Bay 14671 468 JOANNA BURGER AND MICHAEL GOCHFELD and at the Texas City landfill in Texas, near interfood interval (using a stopwatch) for 20 Progresso, 32 km north of MCrida (Yucatan, randomly chosenindividuals of each ageclass. Mexico), and near Campeche, Champoton, When the gulls were feeding by more than one Seybaplaya, and Isla Aguada (Campeche, method, we recorded the interfood interval for Mexico). We remained at each study site for 20 randomly chosen individuals of each age at least one day. class for each foraging method. After timing the interfood interval of one individual we METHODS switchedto another, alternating ageclasses. To We examined the foraging behavior of Laugh- avoid timing the same individuals repeatedly, ing Gulls along the Gulf of Mexico near Gal- we sampled birds in all areas of the flock. We veston, Texas and in Campeche and Yucatan, compared the size of food items with the size Mexico in January 1979. To minimize con- of the bill and found that at each site, gulls fed founding variables suchas seasonand weather, on relatively the same size item; agedifferences we limited our study to three weeks, sufficient within any foraging site thus were not attrib- to study a variety of feeding situations. For utable to differencesin food particle size. How- four days we observed Laughing Gulls feeding ever, since volume and energy values differ in Galveston Bay and on garbageat the Texas among foods, we did not compare foods among City landfill (where a bulldozer intermittently study sites or feeding methods. distributed the garbage over the dump). For In addition to interfood intervals, we also the remainder of the study period we studied recorded capture successrates where possible the gulls on the Mexican Coast. All observa- (successrate = number of items obtained per tions were made on calm, clear, sunny days in number of attempted captures). At all study order to minimize differences due to weather sites, we recorded the number of Laughing (see Dunn 1973). We made observations from Gulls and the age composition every 10 min about 07:OOto 19:OOand did not find age dif- throughout the observation period, for each ferences in time of feeding. Since we did not method being studied. We made two transects mark individuals, our study does not provide along the Yucatan coastfrom Progress0to Car- information on age differences in the propor- men to count all Laughing Gulls to determine tion of each day that individuals fed (but see age distribution. Cooke and Ross 1972, Buckley and Buckley Most analyses were performed using non- 1974). parametric contingency tables or correlation We distinguishedseveral methods of feeding procedures (Siegel 1956). Where appropriate, (after Ashmole and Ashmole 1967, Ashmole we performed other analyseson interfood in- 197 1, Simmons 1972): surface-feeding on a tervals (using log-transformed data) and on garbagedump, aerial snatching,plunge-diving, percentages(using arcsine transformation; So- aerial dipping, and contact-dipping. “Plunge- kal and Rohlf 1969). diving” (where the gulls descendquickly to the water head first, and submergethemselves only RESULTS partially while capturing fish) was observed at In Texas, adults comprised 95% of the foraging a stream 3 km south of Progresso,at a river Laughing Gulls, whereas along the entire Yu- mouth 30 km west of Champoton, in the Gulf catan coast the ratio of age classeswas: 60% of Mexico 16 km southwestof Champoton, at adult, 5% subadult, and 35% young. The gulls’ Seybaplaya, and behind ferry boats at Galves- choice of feeding habitats ranged from natural ton, Texas and Isla Aguada (35 km east of (harbors, streams,rivers, ocean) to man-influ- Carmen, in Campeche). “Dipping” refers to enced(plunging for fish behind ferries, and over seabirdsdipping only their bills into the water a streamwhere a culvert had concentratedfish), to pick up small food items from the surface. and to man-made situations (garbage dump, We observed aerial dipping (dipping while offal). At Seybaplaya, Laughing Gulls fed by a flying) at Seybaplaya and Champoton, and variety of methods within sight of one another: contact dipping (dipping while swimming) at aerial dipping, contact dipping and plunge- Seybaplaya and Progress0Harbor. At Cham- diving. Although young comprised only 30% poton we observed gulls feeding by aerial dip- of the gulls present at Seybaplaya, a higher ping at a 2-m long dead fish partly exposed in percentagethan expected fed by aerial snatch- shallow water. At Seybaplaya, a fishing village ing of offal and dipping for invertebrates, and 55 km south of Campeche City, fishermen a lower percentageplunge-dived for fish (Table butchering sharks tossed offal to waiting 1). Successrate varied according to feeding Laughing Gulls. method and age (Table 1). Young were as suc- We defined the “interfood interval” as the cessful as adults when snatching offal in the period between obtaining successive food air, lesssuccessful than adults when aerial dip- items. At each feeding location we timed the ping, and much lesssuccessful than adultswhen FEEDING BEHAVIOR IN LAUGHING GULLS 469 TABLE 1. Effect of age on location of feeding and successrate (df = 2 for x2 tests). n Adult Subadult Y0Wlg x ’ P Percent occurrence OverallAerial snatching (offal) 200 6635 45 2: 89.6 0.001 Aerial dipping (invertebrates) 160 55 5 40 7.6 0.02 Plunge-diving (fish) 100 II 3 20 10.8 0.005 Successrate Aerial snatching 150 62 60 52 0.4 NS Aerial dipping 150 96 75 60 6.8 0.05 Plunge-diving 100 83 60 24 12.9 0.005 plunge-diving for fish. Considering occurrence that the proportion of young was negatively and successrates, young fed mostly where their correlated with the index (Spearman’s rank r, successrate was closestto that of adults, and = -0.68, yt = 14, P < 0.007).