FEEDING OF WESTERN AND SEMIPALMATED SANDPIPERS IN PERUVIAN WINTER QUARTERS

MYRTr•E J. ASHMOLE

LARGEnumbers of shorebirdsthat breed in North America spend the winter in the regionof the Bay of Paracas,150 miles southof Lima, Peru. During a short period in October and November 1965 I had the oppor- tunity to observesome aspects of the feedingbehavior of mixed flocks of two very similar species,the SemipalmatedSandpiper (Ereunetes pusillus) and the Western Sandpiper(E. mauri), which has a considerablylonger bill. They were feedingin a smallharbor-like inlet with a narrowentrance, situatedat La Puntilia, Pisco,about 2 miles north of the Bay of Paracas. The tidescaused important changes in this habitat, and th'esevaried con- siderably from day to day; the lowest tides virtually emptied the inlet, while the highest of the low tides left the inlet bottom at least one third coveredwith water. At low tide, three zonescould be distinguished;water, wet mud (shiny), and dry mud (dull). Thesezones are presumablyequiva- lent respectivelyto zonesD and E, zone B, and zone A in Figure 3 in Recher'sstudy of the feedingof shorebirds(Ecology, 47: 393-407, 1966). The extent of the three zoneschanged continually as the water flowed in and out of the inlet, and as wet mud dried on the outgoingtide and dry mud becamewet mud on the incomingtide. The numbersof birds in the inlet varied from day to day; in late Oc- tober and the first half of November there were usually between 20 and 100 SemipalmatedSandpipers (average 33) and between3 and 40 Western Sandpipers(average 8). The Western Sandpiperswere frequentlyin two or three loosegroups within the flock of SemipalmatedSandpipers, and the two speciesfed over the same general area, frequently only a meter or so apart. I noticedonly occasionalinter- or intraspecificaggression. My objective was to determine the extent to which these two species differ in their exploitationof the feedinghabitat. As closeas possibleto the time of low tide, I made a seriesof countsof the flock in quick suc- cession,noting first how many of each specieswere in water, th'en how many were on wet mud, and finally how many were on dry mud. Birds at the boundariesof the defined substrateswere ignored,for instancebirds standingon wet mud but feedingin water. These countswere repeated severaltimes on eachof 6 days. The totals for each day (Table 1) show that on most days the highestpercentage of both speciesfed on wet mud. On all days the Western Sandpipertended to be relatively more abundant in water, and the SemipalmatedSandpiper to be more abundanton dry mud. The consistencyof this trend indicates a real difference in zone

131 The Auk, 87: 131-135. January 1970 132 MYRTLE J. AS•rOLE [Auk, Vol. 87

TABLE 1 NUMBERSO1 • WESTERNSANDPIPERS (.ERENEUTES MAURI) ANDSEMIPALI•iATED SANDPIPERS (.E. PUSILLUS FEEDINGIN DIFFERENTZONES ON A TIDAL MUDFLAT

Number Total numbers(and %) in eachzone of Date Species counts Water Wet mud Dry mud 4 November E. mauri 6 5 (17%) 24 (80%) 1 (3%) E. pusillus 6 13 (5%) 127 (49%) 117 (46%) 5 November E. mauri 5 8 (57%) 3 (21%) 3 (21%) E. pusillus 5 12 (25%) 20 (42%) 16 (33%) 10 November E. mauri 5 8 (15%) 35 (66%) 10 (19%) E. pusillus 5 9 (7%) 26 (20%) 94 (73%) 18 November E. mauri 5 26 (22%) 69 (60%) 21 (18%) E. pusillus 5 29 (7%) 289 (72%) 81 (20%) 19 November E. mauri 3 24 (24%) 66 (65%) 11 (11%) E. pusillus 3 19 (11%) 128 (72%) 31 (17%) TOTALS g. maurz 24 71 (23%) 197 (63%) 46 (15%) E. pusillus 24 82 (8%) 590 (58%) 339 (34%) utilization betweenthe species,even thoughthe successivecounts on each day cannot be consideredto be completelyindependent of each other be- cause they were made over fairly short periods and involved many of the same birds. Although Recher (1966) did not observethe Western and SemipalmatedSandpipers together in the samearea, his histogramsshow that in the separateareas where he watchedthem the SemipalmatedSand- piper spentmore of its time on dry mud than did the WesternSandpiper. I also determined the proportion of feeding time spent in probing, as opposedto pecking,by each of the specieswithin each of the three zones, ignoring birds close to the boundaries (Table 2). Probing I define as feeding by jabbing continually with at least the tip of the bill always below the surface, except for brief momentswhen the bird changedits probing site; pecking is feeding by pecking the surface, each peck being a clearly distinct feeding movement. I obtained data by watching in- dividual feedingbirds, chosenat random, in each of the zonesthat had feedingbirds presentduring the observationperiods. I recordedthe total time of the observationon one stopwatch and the amount of time spent probing as opposedto peckingon a secondwatch. Each observationwas continueduntil the bird was disturbed or stopped feeding, which always occurred within 3 minutes. The average length of the observation was just under 1 minute and only four observationsextended beyond 2 minutes. I watched the birds only while feeding and made no attempt to determine what proportionof their total time they spent in feeding. In water, the WesternSandpiper spent all its feedingtime probing,while the shorter-billedSemipalmated Sandpiper probed for only half of the time. On wet mud the Western Sandpiper probed three fourths of the Jan. 1970] Western and Semipalmated Sandpipers 133

TABLE 2 AI•OUNT OF FEEDII•G TII•E SPENT IN PROBING AS OPPOSED TO PECKING

Zone

Species Water Wet mud Dry mud Number of days of E. mauri 3 5 3 observation E. pusillus 3 5 3 Number of E. mauri 8 15 7 observations E. pusillus 8 16 8 Total time watched E.•nauri 294 1071 363 (seconds) E. pusillus 368 1499 390 Time spent probing E.•nauri 293 804 2 (seconds) E. pusillus 204 432 0 Per cent of feeding E. •nauri 100 75 1 time spent in probing E. pusillus 55 29 0 time and the SemipalmatedSandpiper only one third of the time. Even on dry mud the WesternSandpiper probed occasionally. The individual observationsshowed slight variation, but only once did I see a Semi- palmated Sandpiper probe in wet mud for a greater proportion of its feedingtime than a Western Sandpiperon the sameday. Althoughbirds feedingat the water'sedge were excludedfrom quantitativeconsideration becauseof the difficulty of definingthis feedingzone, I watchedSemi- palmatedSandpipers there on one day feed largely by probing. Combiningthe informationin Tables 1 and 2 showsthat the Western Sandpiperspent about 70 per cent of its feedingtime in probing,com- paredwith 21 per cent for the SemipalmatedSandpiper. The rest of the feedingtime it spentin pecking.As might be expected,although the two speciesfeed in such closeassociation, they evidently exploit the habitat in somewhatdifferent ways; the Western Sandpiper,with its longer bill, probedbelow the surfacefor much of its feedingtime and concentrated on shallowwater and wet mud, while the SemipalmatedSandpiper fed to a greaterextent by pecking,spending most of its time on the wet mud and the dry mud. Apart from the two sandpipers,six other speciesof shorebirds(Cha- radrii) fed regularly in the inlet during the end of Octoberand the first half of November. Short-billedDowitchers (Limnodromus griseus) were present in larger numbers (average 89 per day) than any other species. They generallyfed in tight groupsin shallowwater, constantlyprobing, often submergingthe wholebill, but alsocame out onto wet mud to probe. Willets (Catoptrophorussemipalmatus) were present in only small num- bers (average4), and fed isolatedfrom one another. They tended to be in deeperwater than the dowitchers,although sometimesa singlebird fed 134 MYRTLE J. Asu•oLE [Auk, Vol. 87 in sh'allowwater with a group of dowitchersor peckedamong the stones and seaweedsat the water'sedge. Sanderlings(Crocethia alba) were fairly numerous(average 18) and fed almostentirely by probingat the water's edge among dowitchers, Western Sandpipers,and occasionallySemi- palmated Sandpipers.(Outside the inlet, Sanderlingsfed in the more typical way--running downafter the recedingwaves, feeding rapidly, and fleeingup the strandagain.) A smallnumber (average3) of SpottedSand- pipers (Actitis macularia) fed by peckingmainly on wet mud near the waterline. They weresolitary feeders and werehostile to membersof their own and other species.Turnstones (Arenaria interpres) were also present (average20) and had more varied feedinghabits than the other species. They searchedin the kitchen garbagethat was thrown over the harbor wall severaltimes a day, peckedon the mud not in very closeassociation with other species,and excavatedholes in the mud with their beaks,while other small speciespecked in the upturnedmud. Numerous (average39) SemipalmatedPlovers (Charadrius semipalmatus)fed fairly well spaced out on wet and dry mud, often closeto the sandpipers.They were mainly after pink "worms" for which there was considerablecompetition. The plovers,Willets, Turnstones,and the Grey Gulls (Larus modestus),but not the sandpipersoften chasedeach other for a worm for severalminutes at a time. When feedingon wet mud and at the water's edge, the two sandpipers came into closecontact with Short-billedDowitchers and Sanderlings,but competitionwith other specieswas minimal in all three feedingzones. Recher (1966)noted the Least Sandpiper(Erolia minutilla) feedingin associationwith the Western Sandpiperon mud flats on the Californian coast,although it also frequentedmarsh areas there. On the New Jersey shore he saw the Least Sandpiperin the same general area as the Semi- palmatedSandpiper, but thereit fed mainly in the tidal marshand avoided the mudflats where the Semipalmated Sandpiper was feeding. In Peru the Least Sandpiperis rarely seenin the samearea as the other two sand- pipers; it occursmore frequently in fresh water and brackish'marshy habitats and not on tidal mudflats. Thus the Western Sandpipermay be found feedingalongside either the Least Sandpiperor the Semipalmated Sandpiper,although where the two latter very similar speciesoccur in the samearea, they tend to feed in separatehabitats.

SUMMARY Westernand SemipalmatedSandpipers in winter quartersin Peru tended to exploitthe mudflat habitat in differentways. The longer-billedWestern Sandpiperspent about 70 per cent of its feeding time probing, chiefly on wet mud and in the water, while the SemipalmatedSandpiper spent Jan. 1970] Westernand SemipalmatedSandpipers 135 only 21 per cent of its feedingtime in probing and fed mainly on wet mud and dry mud. A third species,the Least Sandpiper,was only rarely seen in the region and was not seen feeding in associationwith the other two. Amongother Charadriiformesfeeding regularly in the area, only Short-billedDowitchers and Sanderlingsfed in closeassociation with the sandpipers. c/o PeabodyMuseum of Natural History, Yale University,New Haven, Connecticut 06520.