rn~ li -1.-r T 5 ./ a - E, 1, , EN 1% i ' 5 F W 1 = % * 4I of the FLORIDA STATE MUSEUM Biological Sciences Volume 26 1981 Number 4 COMMUNAL SOCIAL BIOLOGY OF THE SOUTHERN SAN BLAS JAY JOHN WILLIAM HARDY THOMAS A. WEBBER AND RALPH J. RAITT 1 ..9. 3 . s B* M .s~ M IE . 1 .5= a,% ./ :'#A...-91.45 0,1- im... , = il, 8 4 .*» == ' ,#SM;v 5 -5 ..P 9 9 : *=,S =? ,(= :* .. S ..5.5 11 & . .... AS 11% r T.'9 8 : * E 5 .SI2 Ah.4/S m %'*ah = - 5 / m.= 5 '6 . 5 *1 . sI Y *5 *. 58 .- =.9- ,=& I F Bf . 5..:' UNIVERSITY OF FLORIDA GAINESVILLE Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM, BIOLOGICAL SCIENCES, are published at irregular intervals. Volumes contain varying numbers of pages and are not necessarily completed in any one calendar year. OLIVER L. AUSTIN, JR, Editor RHODA J. BRYANT, Managing Editor Consultants for this issue: RUSSELL P. BALDA LEWIS W. ORING Communications concerning purchase or exchange of the publications and ali manuscripts should be addressed to: Managing Editor, Bulletin; Florida State Museum; University of Florida; Gainesville, Florida 32611. Copyright © 1981 by the Florida State Museum of the University of Florida. This public document was promulgated at an annual cost of $5025.00 or $5.03 per copy. It makes available to libraries, scholars, and all interested persons the results of researches in the natural sciences, emphasizing the circum-Caribbean region. Publication date: December 18, 1981 Price: $5.00 COMMUNAL SOCIAL BIOLOGY OF THE SOUTHERN SAN BLAS JAY JOHN WILLIAM HARDY, THOMAS A. WEBBER, AND RALPH J. RAITTI SyNopsts : We studied the communally-breeding Southern San Blas Jay , Cydnocorax (Cis·silopha) s'. sanblasianuw. from 1974 to 1978 near Acapulco, Guerrero, Mexico. In this area it inhabits only palm plantations and the small remnants of native woodland scattered among them. Most nests are built in the crowns of coconut palms. This jay lives year-round in nonmigratory groups of fairly constant composition, which on our study areas ranged in size from 13 to 26 birds one year old and older. In each group 62-82% of the members were at least 3 years old. Each group contained 6-10 breeding pairs that were probably permanently· monogamous, plus a,smaller number of nonbreeders. With the exception of replacement nests and the rare occurrence·of second broods, female breeders laid in only one nest per breeding season, and only one female laid in each nest. Breeding pairs were assisted in feeding anddefending their young by both the nonbreeders and breeders of their communal group. Some nests were attended by as many as 13 birds. Helpers. whether breeders or nonbreeders, preferred to feed fledglings rather than nestlings. Nests begun earlier in the breeding season attracted more helpers than did later ones, because of the accumulation of fledglings as the season progressed. The feeding rate per nest increased with brood size, but there was no correlation between the feeding rate per nestling and brood size. The positive correlation between feeding rate per nest and brood size is probably a result not of greater rates of feeding per attendant, but of larger numbers of attendants. each of which feeds at a rate independent of the number of nestlings. All birds of known parentage helped at their parents' nests, though none helped its parents exclusively. Most birds began breeding at 3 years of age or older. but 1- and 2-year-olds were breeders more commonly than is at present evident in the other three species of Cissilopha. Younger breeders tended to nest later in the breeding season than older breeders. Each communal group inhabited an exclusive home range, whose boundaries with other groups were maintained by mutual avoidance rather than by active defense. In one large group, whose members were dispersed over a comparatively large home range, breeders with active nests confined their activities to a small fraction of their group's home range and ex- erted a quasi-territorial dominance over other group members who entered these areas. Such individual 'core areas were not evident in smaller groups. The commonest clutch sizes were 3 and 4. Incubation lasted 17-18 days, and the nestling period was about 18-20 days. Breeding pairs were usually single-brooded. Between 50 and 68 % o f nests that had clutches produced at least one fledgling, and 22-38 % of all eggs resulted in fledglings. From 0 to 50 % of young that fledged survived to at least 1 year of age. The annual survival rate of yearling and older birds was about 75 % in most years. Change5 in group membership, by both young and adults, appear to be of infrequent but regular oc- 'John William Hardy is Curator in Ornithology. Florida State Museum. Universit,· of Florida, Gainesville FL 32611: Thomas A. Webber is a graduate st,ident in the Department of Zoology. Universitv of Florida; and Ralph J- Raitt is a Pro- fessor in the Department of Biology. New Mexico State University. Las Cruces NM 88003 HARDY, JOHN WILLIAM, THOMAS A. WEBBER, and RALPH J. RAITT. 1981. Com- munal Social Biology of the Southern San Blas Jay. Bull. Florida State Mus., Biol. Sci. 26(4):203-264. 204 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 currence. Females seemed to change groups more often than did males, Most of the im- migrants to our study groups were of unknown origin. The movements of dependent juvenile and their attendant5 are described. Southern San Bias Ja,·s are omnivorous. They seemed to forage about equally in palms, herbaceous growth, hedgerows, and jungle patches. During the breeding season they did not usually forage in flocks. There was little or no mixing of birds from different communal groups during the nonbreeding season, At that time of year they typically foraged in flocks and were less vocal than in the, breeding season. We speculate on the way in which each Ci%silopha species' characteristic grbup size, number of breeders per group, and age at first breeding, may be related.to one another and to the productivity of the habitat. TABLE OF CONTENTS INTRODUCTION. 205 ACKNOWLEDGMENTS ... 206 MATERIALS AND METHODS. 206 RESULTS ... 207 NESTS.AND PRINCIPAL HABITAT . 207 ~CCURRENCE IN OTHER HABITATS . 210 COMMUNAL STRUCTURE. 212 Group Composition 212 Nest Dispersion , 214 The Roles of Helpers. 214 Temporal Patterns of Nest Attendance and Social Status of Nest Attendants . .. .. 218 Variation in Rate of Nest Attendance 224 Kinship As a Factor in Attendance 226 Age at First Breeding . 228 Breeding Time of Adults vs. Subadults 229 PAIR BOND PERSISTENCE . 230 HOME RANGE 230 INTRAGROUP CORE AREAS . 230 NATAUTY AND MORTALITY 235 Overall Nesting Success 235 Breeding Success of Subadults vs. Adults 235 Causes of Nestling Loss. 236 Post-fledgling Survival and Mortality . 236 Emigration and Immigration 238 Juvenile Movements: Home Range and Core Area Fidelity 241 FOOD AND FEEDING HABITS . 243 NONBREEDINC SEASON . 244 DISCUSSION . 245 LITERATURE CITED 250 APPENDICES 252 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 205 INTRODUCTION Four allopatric species of Middle American jays, known as the black- and-blue jays, constitute the su_~genus Cissilopha of the widespread neotropical genus Cyanocorax (see Hardy 1969 for the taxonomic revision that we follow here, which submerges Cissilopha in Cyanocorax). These four species are the Yucatan Jay, C. gucatanicus, of southeastern Mexico and adjacent Belize; the Beechey Jay, C. beecheii, of central western Mexico in Sonora and Sinaloa; the Bushy-crested Jay, C. melanocyaneus, of Guatemala, Honduras, and El Salvador; and the San Blas Jay, C. sanblasianus, of southwestern Mexico. There are two races of the San Blas Jay. The Nelson San Blas Jay (C, s, nelsoni) ranges from the state of Sinaloa south to northern Guerrero, Mexico. The Southern San Blas Jay (C. s. sanblasianus) occurs from about 100 km northwest to about 70 km southeast of Acapulco, Guerrero, These two races are separated by about 200 km, and they differ in certain aspects of their vocalizations, softpart color, and plumage (Hardy and Raitt 1977). We have studied the breeding biology of all four species of black-and- blue jays, including both races of the San Blas Jay (Raitt and Hardy 1976, 1979; Hardy 1976; Raitt et al., ms.). Each of these forms exhibits a varia- tion on the theme of communal social behavior, in which birds live year- round in groups of basically stable composition, and breeding pairs receive assistance from other group members in feeding and defending their young. Prior to our work, the behavior of these four species in the wild was virtually unknown. These studies help to fill a large Middle American geographical gap in our knowledge of communal New World jays. Our objective in studying these four species was to determine at least the outlines of their social organization, especially their communal breeding behavior, their demography, and the nature of their habitats. The ethology and systematics of the subgenus were of secondary impor- tance to us, We decided not to attempt a long-term intensive investigation, such as that being conducted on the Florida Scrub Jay (Aphelocoma c. coerulescens) by Woolfenden (Woolfenden and Fitzpatrick 1978 and references therein), of a single species of neotropical jay. We could not conduct such a study year-round from our base in the United States. Also the study areas themselves lead a precarious existence at the mercy of neotropical agriculture and forestry (one of our study sites was razed in the midst of our work), so that we were wary of investing effort in the sort of study that depends on having a site secure from interference for a decade or more. Finally, we felt the urgency of getting some information about each species before it becomes extinct or so rare as to be impossible 206 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 to study (this is most likely in the case of the Beechey Jay; see Raitt and Hardy 1979). In short, we decided we would rather have limited informa- tion on four species than somewhat more information on only one. This paper presents our findings on the fifth and final form to be studied in thesubgenus, the Southern San Blas Jay. Because of the relative ease with which it could be studied, it yielded richer information than did any of the other black-and-blue jays, showing it to have the most elaborate and peculiar communal breeding behavior of any New World jay studied so far. ACKNOWLEDGMENTS Erik Bitterbaum, Alan B. Bolten, Howard Gates, Gary L Grabowski, Harold F. Mayfield, James R. Raitt, James R. Reitzel, Sarah Sloane, and John T, Vollertsen all par- ticipated in the field work. We gratefully acknowledge their contributions, which in many cases extended well beyond routine assistance. L.W. Oring, G.E. Woolfenden, and R.S. Balda refereed the manuscript and made many helpful suggestions. J.A. Cox gave assistance with statistical analysis. We thank them very much. We are also grateful to the Direcciob General de la Fauna Silvestre, Republica de Mdxico, for permission to collect and mark birds, to Rogers MeVaugh, University of Michigan, for identification of certain plants, and to Richard Franz for identification of stomach contents of jays. Financial support was provided by National Science Foundation Grants BMS 74-11107 and DEB 76-99735. MATERIALS AND METHODS Our principal study site was 13 km northwest of Pie de la Cuesta, Guerrero, about 2 km north of the Laguna de Coyuea. Our other study site was 6 km southeast of Acapulco Interna- tional Airport, about 2 km inlaild from Playa Encantada. We captured, aged, sexed, color-marked, and released jays of four communal groups at the main study site and one group at the other site hereafter called the southern site. Birds were caught in mist nets or by hand. We determined the birds' ages by criteria of Hardy (1973) and Hardy and Raitt (1977). By using a combination of plumage characters and eye color, one can distinguish in the field be- tween juveniles, yearlings, 2-year-olds, and "adults." Through their second year Southern San Blas Jays have brown eyes. The irides then gradually change through various shades of green until by 3 years of age they are pure yellow and remain so for life. We have found few excep- tions to this rule. Since 1977 we have encountered two 2-year-olds with greenish.yellow eyes that at a distance were difficult to distinguish from those of adults. Some 3-year-olds retained a greenish,tinge to the eye, which could be distinguished from pure yellow only with the bird in hand. On the whole, age classes remained distinguishable with remarkable reliability. Sex was determined by a combination of weight, presence of a cloacal protuberance or brood patch, and reproductive role. The birds were marked with colored plastic tarsal flags ('Saflag") and usually with anodized aluminum or plastic leg bands in the same color combinations. Individual birds are referred to by their color markers, the left leg first. Thus PIG means pink on the left, green on the right and Pi/G-L means pink over green on the left. Colors and their symbolsin this paper are: 0, orange; Bl, blue; Bk, black; W, white; Pi, pink; P, purple; G, green. An "X" denotes absence of a band. A suffix numeraI indicates that the combination was used previously on a bird now presumed to be dead. In each year of the study some birds (usually females) re- 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 207 mained unmarked in each group, but usually could be distinguished by a combination of their age, association with a particular site or marked bird, degree of hostility toward humans, and peculiarities of voice. In some cases these characteristics were deemed reliable enough to iden- tify unmarked birds from year to year. We studied the Southern San Bias Jay in five breeding seasons from 1974 through 1978. Field work in 1974 was. conducted from 25 to 30 June by three persons (approximately 16 man-days). Some birds in group 1 were flagged, two nests were found and studied, and the general nature of the main study site and the jays was assessed, including verificatibn of the detectability of age classes in the field. Field work in 1975 was conducted from 21 June to 28 July by four to six persons (approx- imately 160 man-days), all at the main study site except for two man-days at thesouthern site. Field work in 1976 was conducted from 23 June to 28 July by two to seven persons in ap- proximately 160 man-days (18 days at the southern site). In 1977 study was conducted from 10 June to 27 July by four to eight persons (appprox- imately 240 man-days). In 1978 eight days (18-25 June) were spent by two people censusing the two best-studied communal groups, ascertaining the survival of flagged birds, the size and composition of the groups, and pair bond persistence, with some effort devoted to determining the number of nests, their distribution, timing, and success. From 1975 through 1977 attention was given to netting birds; following flagged birds to detect group membership, individual roles, home ranges, and core areas (subdivisions of home ranges); and to studying nests and fledglings Figure 1 is a map of the study site, showing the distribution of vegetation, roids, lanes, fences, and other features. RESULTS NESTS AND PRINCIPAL HABITAT In the present study, 52 nests were found or were known circumstan- tially to have existed in the communal groups that we have designated 1, 2, and 3. Of these, 48 were in coconut palms on a bract or cluster of coconuts at the base of the fronds. Two were in the crowns of lime trees and two on bracts of small native palms (species not determined). All nests were within the palm groves and none within the patches of native woodland nearby. Nests ranged in height of placement from 2 m (lime tree) to 18 m. Most of the nests were 9 to 15 m up, the smallest and the very tallest palms rarely being selected as nest trees. Mean height of 39 nests that we were able to measure was 11.8 m. Certain portions of each group's home range were consistently pre- ferred for nesting from year to year. Figure 2 shows all nests given a num- ber in the main study area from 1974 through 1978. The jays nested in those parts of the groves that satisfied three requirements: (1) the canopy of the palm crowns was not too dense; (2) there were'no acacia thickets but lush herbaceous growth instead; and (3) there was native woodland nearby. The frequency with which any fronds of adjacent palms touched (expressed as a percent of the total possible) was used as an index of canopy closure (Table 1). The palms were planted in straight rows and 208 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 1234567891011121314151617181920 'irilli.. lilli -~ 3*t-*.,% u i- *~ s »~r:- i.#%4&4)n. 5744V244¤~ 9»99,-0 1 {ff*' i 1**15**f*~~ C D (/1- 3 >1:-- ~. , 52«21 .1--1 - .-fiE~~ :]62«»*i-f--1{jj»12* 1 ;0E»t~:SY'- .... 74*fpso.»5029':7,7:. '*~TF *su :~452/74110 --« , G 4,1~-2,3'2~-~ S=Sts '«*3*i «]1*7,: 1("_i~jtl~__j i L 1 G H-14,-·'4+.0- Hdit#t:*LY-4,-,· . ».:,-k#F BARE -15·55«055'' 4 -- - ....»«, 14%*At@65~6.**444~ 7%- - - ' ~ HILL ~~2~5~,- .-f ..:, f.14*t»4~\«~ -,Sin@fr5*'* d.*36*5?28%-5:8%*EM#JE--~~,SS:'5}~35*. *A w- 834%522****52*35<1114PLOWED FIEL[)~* "K ·*33~"ff~{f~1~9~* 'v,[LL 1~»==6»=..=,~~»104 HIL-=*f< K, 33«3~»«AL»mI %$--%4« 3~ J ~ - 5 -FlfKS '.:.~ 5':r·55 I 5*. -' ,/U - 'N ,-7.9 Pn M ·Sffit« S-¢ssisafs·SR««f« 1- .i,».*~***~~~~~11~N ff.S]S)Gff*St}§*{f .6,42'IG'~~,22»5*~ ' lt]7.%447<---~23***35*3%* 56*,ts . .,4 :3« N.. -045%0 R.--~' ''4*-Sib-- i IfVE-2 ,~U~i.LANE,5,e.. - ., R «.'56.'~4 -0,112·-T ':·''~ ~~:- 1. --- ' ,/ ' s ~ - r-- -- ·.:- --.-·5:- )3- 41, -c<0 -1 Ii---------T,-- ----- -:CYft-----1~:YS: L.1 :SSS** s 1234567891011121314151617181920 >00< OPEN PALM GROVE SHRUBS AND BRUSHY THICKETS 0 5 MANGO TREES 'lilli. MEDIUM PALM GROVE fatEm NATIVE BROADLEAVED TREES I HOUSE 411111'. DENSE PALM GROVE ----1 BARBED WIRE FENCE OSHED OR SHELTER 67·f HERBACEOUS GROWTH - PLOWED FIELD OR CORN FIELD -POWER LINE FIGURE 1. Mapofthe main Southern San Bias Jay study site, showing landmarks and vegeta- tion. Areas bounded by coordinates A-B/12-13 and R-S/12-16 have palms and ground level vegetation of herbaceous growth, shrubs, and brushy thickets. columns, so that each tree (unless on the edge of the grove) had four nearest neighbors. Thus, if a given palm touched, say, three of its four neighbors, its closure index was 75 %. The closure index for a sample quadrat is the mean of the closure indices of all the trees in the quadrat. Eight square quadrats, each of one hectare, were censused in 1977. Three of these were selected because they contained jay nests (quadrats 1,4,7); the other five represented habitat variations not used for nesting. The in- dex of closure for those having nests is 7.2 % or less. Of the five having no jay nests in the five years, three (2, 3, 8) had IO % or greater closure in- 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 209 , 1 9 m Y f 07 -a r---, 1,\1 ~zi(521 ~~ ~MLE~~~1 1!11 - Ctit, lEo E MARSH E 738 BARE . AAHILL | HILL ', WELL .: E/ E .0 / M,-- Eli 11@ \ PATH -- *7:/ i BRICK 1 YARD 1. /183333391;SET METERS 6 so ido FIGURE 2. Locations of all known active nests at the main study site, 1974-78. Double circles = 1974; circles = 1975; squares = 1976; hexagons = 1977; triangles = 1978. Nest l's exact location not ascertained. Approximate positions of nest 51 and 52 indicated by question- marked designations. dices; one (5) had a low index of 0.17 %; and one (6) had a 5.3% closure, was frequented by jays, and probably had nests occasionally. The relatively open-canopied groves may be selected as a means of reducing predation by snakes and squirels. In groves where the trees are far enough apart to prevent a climbing predator from traveling from crown to crown, these animals must return to the ground and ascend each tree. This difficulty may make open groves unattractive to such predators. We rarely saw squirrels in the nest areas, but found them frequently in the groves not used for nesting and having relatively closed canopies, where they moved easily between adjacent trees with intermingling fronds. Rich herbaceous growth, both for foraging and for the protection it affords fledglings, also is a feature of good nesting groves. Groves with the ground cleared and plowed for corn planting had nests only where 210 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 TABLE 1.- Coconut grove quadrat comparisons. Quadrat No. of No. of No. of Index of Jay Other nurnber trees possible touches closure use features etouches 1 156 594 43 7.2 Nests Ground forbs and hedgerows 2 196 698 133 19.0 Rare Sparse ground forage cover 3 360 1478 163 11.0 Forage Sparse ground cover 4 191 984 34 3.4 Nests Heavy forb cover 5 144 600 1 0.17 NOJay Brushy acacia use understory 6 144 600 32 5.3 Forage, Similar to probably 4 above nests previously 7 152 591 31 5.2 Nests Sparse ground cover, many hedgerows. 8 195 861 87 10.1 No Nests Heavy forb forage bordering on preferred habitat. The failure of quadrat 5 to have jay nests is possibly related to the brushy acacia growth at ground level. We found no nests where these acacias and shrubs grew in dense thickets. The jays favored nesting sites near remnants of native broadleaved woodland, either as islands of at least a few hundred me or as hedgerows in and around a grove. Each communal group seemed to have its own more or less exclusive woodland foraging ground. A particularly striking example of the importance of such tracts was seen in the case of group 4. The hill on the north side of that group's home range (Figs. 1 coordinates 18-19/J-K; 2, 14) was covered with a dense low jungle in 1976.- By the time we arrived in 1977, the hill was completely denuded for planting corn. This jungle patch had been the major tract of native vegeta- tion availabl6 to group 4 (see Fig. 14). The group had completely disap- peared by 1977. BY July 1978 the hillside jungle had begun to grow back and birds of group 2 were foraging there, but the palm grove portion of group 4's former home range was never reoccupied by jays of any group. Figure 3 shows several views of the coconut palm groves the jays in- habited. OCCURRENCE IN OTHER HABITATS Northwest of Acapulco we found Southern San Blas Jays nesting only 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 211 FIGURE 3. Vegetation of the main study site. (A) Spotting scope is directed tow·ard nest site (white arrow) in ecronut palm. (B) Mango foodtree with overstory of coconilt palms. and herbaceous ground cover. (C) Palm at the south edge of the "jungle"-a patch of native broadleaved evergreen woodland - in the group 2 hoInc range. 212 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 in palm groves. At the southern site, southeast of the city, a large remnant of coastal plain broadleaved scrub woodland is intermingled with the coconut groves and open corn fields, where, in addition to the adjacent palm groves, we found the jays nesting in the edges of the scrub. This habitat, shown in Figure 4, was very patchy; we found two patches large enough to constitute true woodland and ascertained that the jays were not nesting in their interior. The relationship of the San Blas Jay to the original vegetation is considered further in the Discussion. COMMUNAL STRUCTURE GROUP COMPOSITION. - Appendix I summarizes basic information on all the jays studied in groups 1 and 2. Only some of the group 1 birds and none in group 2 were flagged in 1974. Thus the 1974 column of Appendix I d6es not fully show group l's composition for that year and is given to show persistence of some individuals throughout later years. The same is true for the 1975 column with respect to group 2. Data on the member- ship of group 1 are complete for 1975-78 and for group 2, 1976-78. Table 2 summarizes the age composition of groups 1 and 2 for the years in which adequate data are available. Southern San Blas Jays live in groups that are unusually large for com- munal breeders (see Brown 1978). The sizes of the two groups (1 and 2) that we studied most intensively usually were about 14 and 24 respec- tively (Table 2). The three groups (3,4, and 5) that we studied in less detail never had fewer than 10 birds each. From 1976 to 1977 the size 6f group 1 remained unchanged. Group 2 increased in size from 22 to 26 birds by immigration of GPi from group 3 and possibly by recruitment of some unmarked birds from group 4 (although only 22 of these were observed. regularly; the remaining four TABLE 2. -Proportion of age classes (marked and unmarked birds) in groups 1 and 2 of Southern San Bias Jays. 1975 1976 1977 1978 Group 1 Adults, 3 yr + 8 11 10 9 2 yr 3 1 2 2 lyr 2 3 3 1 13 15 15 12 Group 2 Adults, 3 yr + 18 19 19 2* 1 3 2 lyr 3 4 4 22 26 25 1981 H A R D Y E T A L.: S O U TH E R N SAN BLAS JAYS 213 FIGURE 4. Patchy second-growth scrub woodland of the southern study site. (A) and (B) Men are indicating approximate locations ot jay nests. (l) Thicket edge; the tallest middle tree held a jay nest. (D) Worker attaching a marker tag to indicate a jay nest tree. 214 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 were seen only once or twice each and their status was never determined). Groups 3 and 4 were never counted accurately. Our impression of group 3 was that it was slightly smaller than in the previous year. Group 4 had disappeared by 1977, as discussed previously under Nests and Principal Habitat. In 1978, group 1 had 13 members, and group 2 had 26 members. Other groups were not censused in 1978. In most species of communally breeding birds, each communal group usually cares for only one or two nests (Brown 1978). The Cissilopha jays conform to this pattern (see Raitt and Hardy 1976 on Yucatan Jays, 1979 on Beechey Jays) with the exception of the Southern San Blas Jay. Groups 1 and 2 had from 6 to 10 active nests at one time. We were never able to count accurately the number of nests in groups 3,4, and 5. We found that most members of the Southern San Blas Jay groups were breeders, including some birds 1 and 2 years old. While it is not unusual in some other communal species for 1- and 2-year-olds to breed (Brown 1978), it is rare in the other Cissilopha jays. NEST DISPERSION. - Table 3 shows that group 1 nests were in a single cluster and that group 2 nests were in two subclusters, with one linking nest (33) and two outlying ones (30, 38). Table 4 shows the sizes of the nest cluster areas and the percentage of the total home ranges that they comprised. The nest clusters in the two groups in 1977 were of approximately equal density. Table 5 shows the mean distance from each nest to all other nests in its group that existed at the same time (not counting overlaps of up to 3 days). In group 1 the mean distance from each nest to all four neighbors was 117 m. To ascertain the comparable figure for group 2 we calculated the mean distance from each nest to its four nearest neighbors, which yielded a figure of 127 m. We found no evidence that breeding pairs having one or more subadult members nested farther from their nearest neighbors (Table 5, nests 39, 33, and 38) than did older pairs. The significance of nest cluster density and the placement of subadults' nests are considered in the Discussion. THE ROLES OF HELPERS. - We found 10 nests being built. We were able to study construction at half of these; building was mostly by the female of the breeding pair, assisted or accompanied by her mate. In only one of these five cases was any bird other than the pair involved in any way: at nest 32 a female (Pi/Bl-R) accompanied the breeding pair for several days during nest building but did not add material to the nest. Other nearby birds apparently did not visit the nest in the construction stage. We found no instance in which an incubating female (always the female breeder) was fed on the nest by any bird other than.her mate. 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 215 TABLE 3.-Shortest distances between neighbor nests inside and outside their communal groups. Distance.from each nest to its Distance Nearest intergroup Distancez nearest intragroup neighbor (m)' in meters neighbors in meters Group 1,1975 N-3 -- 5 44 N-4 - 3 50 N.5 - 3 44 N-4 - 6 (group 2) 135 N-9 - 5 72 Group 1, 1976 N-12-14 57 N-13-12 64 N-14-15 21 N-13-17 (group 2) 100 N-15-- 14 21 N-16--14, 15, 20 100 Group 1, 1977 N-26 -- 36 60 N-27 - 36 120 N-27- 28 (group 2) 160 N-34 - 39 85 N-36 -- 26 60 N-39-26 75 Group 2. 1975 (only three nests found) N-6-7 88 N-7-6 88 N-6-4 (group 1) 135 N-8 - 6 95 Group 2.1976 N-17 - 18 88 N-18-17 88 N-21 -- 18 180 N-21 - 19 (group 3) 250 N-22-23 50 N-23 - 22 50 N-24-23 52 Group 2,1977 (only first nesting attempts included) N-28-29 50 N-28-27 (group 1) 160 N-29 - 28 50 N-30 - 29 140 N-30 - 29 (group 3) 150 N-31--33 70 N-32-40 65 N-33--31 70 N-35-31 80 N-37-29 35 N-38-32 150 1 Mean distance between nearest neighbor nests in group 1: 62 m. group 2. 80.3 m 2 This figure is given onle for nests closest to borders of groups in which nests were known. 216 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 TABLE 4.- Sizes of home ranges and nest clusters (mz) of communal groups of San Blas Jays. Groups Home Ranges Nest Clusters Nest Cluster/Home Range 1, 1976 90,500 20,500 22.6% 1, 1977 120,500 24,500 20.3% 2, 1976 130,749 13,049* 10.0% 2, 1977 250,000 68,000* 27.2% 'totals of 2 sub-clusters TABLE 5.-Mean distance (m) from each nest to all other nests in its group. Group 1 Nests 26 27 34 36 392 X 26' - 160 160 60 75 113.0 27 160 - 125 120 120*3 135.0 34 160 125 - 170 85 135.0 36 60 120 170 - 90 110.0 39 75 120 85 90 - 92.5 117.0 Group 2 Nests 28293031323335373840X 28 - 504 180 230 370 170 300 70 320* 410 117.5 - 29 50 - 140 200 320 140 260 30 280* 370 90.0 - 30 1&Q 11Q - 260 300 220 280 140 200* 360 170.0 - 31 230 200 260 - 170 70 80 170 210 200 122.5 - 321 370 320 300 170 - 220 90 290 150 60 117.5 - 331 170 140 220 70 220 - 140 110 240 260 115.0 - 35 300 260 280 S -22 140 - 230 170 120 108.0 37 70 30 140 140 290 110 230 - 250 340 87.5 - 381 320* 280* 200* 200 150 240 170 250 180 175.0 - 40 410 370 360 360 60 260 120 340 180 - 155.0 i.il-- - - 126.0 1 Nests 26, 32, 33, and 38 each had one subadult breeder. 2 Although technically this was a second nesting, see text p. 26 for account of unique circumstances that justify its inclusion here. 1 Asterisked numbers were not used in calculating means because of nonsimultaneity. ' Underlined numbers designate four nearest simultaneous neighbors. 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 217 Helpers began to appear at the nest once the young hatched. Except for incubating or brooding females, most breeders were helpers simultaneously at one or more nests in their group, with some birds serv- ing as attendants at all nests in their group having nestlings. Using only data from 1977, when we achieved the most complete record of nest at- tendance, we can see some manifestations of the communality of groups 1 and 2. In Figure 5 a line connects each two nests that had attendants in common, regardless of their identity or status. The number of lines as a fraction of the number of possible lines depicts the communal unity (C. U.) of the group. If a line has one arrow on it, the attendants were helpers first at the nest from which the arrow originates and then later at the nest to which the arrow is directed. Similarly, two arrows on a line in- dicate reciprocity in this regard. The fraction of reciprocities recorded of the total number poskible (given a biological "need" in the form of nest- lings requiring feeding) is a communality index (C.I.). The C.I. of 0.83 (5 of 6 lines between nests that produced nestlings having two arrows) for group 1 in 1977 was about as high as it could be, because nest 39 failed prior to hatching and nest 34 failed when the nestlings still required little food. The subdivided nature of the group 2 nest clusters seems to account for that group's lower C.I., which at 0.29 (9 of 31 lines double-arrowed) is still indicative of communality. Initial study of this group suggested that it might not be unified in that no network of attendants seemed to connect the subgroup nest clusters. Long-term study revealed that while the fre- quency of appearance of birds at each other's nests was low when nests were far apart, such attendance did occur, as the lines and arrows show. A line and one arrow need be based upon only a single feeding visit from a bird primarily associated with another nest, and it might be argued that this is slim evidence for communality. This record of social relationship becomes more impressive, however, when one realizes that in the five seasons of study, the inter-nest-cluster C.I. and C.U. were zero. A comparison of the roles of parents and nonparents (pure helpers) in maintaining unity and communality shows little difference between the two (Figs. 5 and 6). Male breeders had relatively little to do at their own nests while their mates laid and incubated, and spent some of this time helping at neighboring nests. Thus most social connections by breeders were maintained by the males. Breeding females were helpers prior to egglaying, or after their young were fledged or their nests abandoned. The C.U. of group 1, based only on parental contacts, was 0.80 (8 of 10), the same as with pure helper representation. The C.I. of only 0.17 (1 of 6) is expectable when one considers the roles of breeding vs. nonbreeding at- tendants: the earlier-nesting, successful parents may be wholly involved with fledged young and have less time to reciprocate by visiting later nests to feed nestlings. Some helpers may have involvement with earlier broods 218 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 N 43 46 44 41 27 31 ?33 3 26 37 32C. U. = 0.80 C. 1. = 0.83 28 29 · 38 45 0 Fledged I or more young C.U.=056 300 Nestlings,failed C 1.0.29 0 Failed prior to hatching -4 Fed nestlings ---0 + Fed fledglings, ~ nestlings ?0 Nestling alive ot end * Helper only when nest was moved east 50m Meters 0 100 FIGURE 5. Communal contacts of groups 1 and 2 based on all nestling or fledgling attendants in 1977. Numbers near circles are nest number designations. See text, page 15 for definition of C.U. and C.I. and discussion and explanation of arrow directions. as well, but not to the same degree. Group 2 had only 12 of 36 possible social connections by parents only, for a C.U. of 0.33 and 3 of 9 (0.33) social connections were reciprocal. Apparently this lower C.U. was caused by two factors: the long distances between many nests and the greater number of nests having virtually identical timing. Birds feeding their own nestlings seldom visited other nests. Breeders of nests 33 and 37 were prominent assistants at nest 29 while building their own nests. The male breeder of nest 40 attended at nest 29 after nest 40 was abandoned. Breeders at nests 32 and 41 became helpers at nest 38 after both their at- tempts failed. TEMPORAL PATTERNS OF NEST ArrENDANCE AND SOCIAL STATUS OF NEST ATTENDANTS. - In Southern San Blas Jays at least five factors interact to produce a complex sociality: (1) the groups are unusually large for a com- munal species, (2) the number of breeding pairs per group is high, (3) the timing of the various nests within a group is staggered (Figs. 7 and 8), (4) 1981 HARDYETAL.:SOUTHERNSANBLASJAYS 219 most birds are both parents and helpers, and (5) successful early nests reduce attendance at later nests and perhaps their success. Early nests were virtually certain to attract assistance from non- breeding helpers and from late breeders who were just starting to select nest sites or to build. Later nests often had few or no helpers: parents that nested earlier were renesting if they failed or attending their own fledg- lings if they succeeded. Nonbreeding helpers also tended to move around with the juveniles and not visit later nests. Figures 9 and 10 illustrate the above points. These figures as well as Figures 11 and 12 are constructed from observa- tion periods that were virtually or completely uninterrupted, contained few or no questionable identifications of individual birds, and represented, when possible, days throughout the nesting period or a se- quence of days in the week before fledging. With two exceptions, the N ~-9==*$00 C. U.=0.80 Cl = 0.17 C.U.=038 Cl.=033 Meters 0 100 FIGURE 6. Communal contacts of groups 1 and 2, 1977, based only on attendants that were breeders in the same season. See page 15 for definition and discussion of C.U. and C.I. See Fig. 11 for other symbols. 220 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 GROUP I 1 Nest'x' ......./\\\\\\\\\\\\\\\\\ N-13 , failed 1 ?- 1 N-16 failed 1 --- 1 N-14 ~ N-12 R/FLT/VKKAJA/~~\\\\\\\\\\\\\\V N-15 GROUP 2 ....L\\\\\\\\\\\\\\\\ N-18 .....1..../\\\\\\\\\\\\\\\\\\\ N-21 ................Al\\\\\\\\\\\\\\ - 1/' N-23 1/*\A-/--\\\\\\\\\\\\\\\\ 1 N-22 ,& , foiled ~V'vyrri................./\\ 1 N-17 , , HYYVYYY\.~\\\\\\ N-24 10 6 20 25 30 5 10 IS 20 25 June July ~ No activity or unknown ~ Nestbullding ~ Fledgling/Juvenile period~ Nestling period ~ Incubation period (No. of lines = no. of young surviving) FIcuRE 7. Timing of stages in the development of nests in groups 1 and 2 at the main study site, 1976. Note staggered timing of nests in both groups. periods selected were an hour or more long. Some birds recorded as helpers in Appendices I and II did not happen to visit a nest during the observation periods represented in Figures 9-12, hence the discrepancy between the figures and appendices. In 1977, nest 27 was the first active nest in group 1. The many atten- dants included yearling nonbreeders such as BIBl and Pi/Bl-L, as well as all male parents at later nests: OPi, OBl, PiG, and PP. The young from nest 27 fledged on 18 June, though only one survived the first few days. Nest 26 (Fig. 9) was the second to fledge young, on 4 July. There- fore, nest 26 also had a sizeable group of attendants, although note that 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 221 Group I 1\\\\\\\\\. ! N 27 . N 26 \\\\\\\\\\\ N 34 27 July r~VZ79ZX[T......................../'llill'llillilliv N 36 1 ~ 1 N 39 Predatorized ~ in incubation 10 15 20 25 30 5 10 15 20 25 Group 2 L\\\\\\\N Nest 29,45 N-451•(found) 2nd Nestof pair · i N-28 - Associates i N-30 1 I.lili I N-40,43 f Predatorized 1 N-33 \Predolorized (I doy fledgling) [7VG'VEAn~/~ill\\\\\\\\\\\\:ill\\\1 N-31 7 \\\\\\ill'ill\1 1 -/V\A/V\A/\All'~ \\\\\\\\\\\\\\\1 N-37 '-/V\A/\~111 1"""""""""""4\\\\\\\\\\1 N-38 "/v/ ANI~ 1 N-32,41 N-41 7 Deserted N-35,44 N-44 10 15 20 25 30 5 10 15 20 25 June July [~K7\1 Bles,bunding Kri3] Nestling care Incubation ~ Fledgling/Juvenile period (No. of lines=no. of young surviving) FIGURE 8. Timing of stages in the development of nests of groups 1 and 2 at the main study site, 1977. Note staggering of nests in group 1 but near-simultaneity of some nests in group 2. compared to the small parental contribution to feeding at nest 27, nest 26's nucleus pair contributed roughly half the feeding visits per hour. Nest 26 fledged three young and simultaneously we introduced three or- phan fledglings from outside the study area. Unfortunately for the com- parison of the number of helpers at all nests of group 1 in 1977, nest 34 (Fig. 10A) failed after about one week of nestling life, too early to allow -7 210min. 215 170 77 108 95 25 26 100- - - - Pi? 1-IPP Pi W - 1~ - PIG 81 81-2 w10 Pi90- Pi W - -681-L - OBI - BIBI-2 80- - % F ee di ng s of Y ou ng : P ar en ts a nd H el pe rs Pi G \ BIBI-2 -5 « -7 -Piw /15\ PiW- r- 81811-/ OP'/70- - Pi/81-L \-/- 60- - Pi/81-L -4 ~F Pi/Bl-L- -9 -50- -3 %8181· bPi/Bl-L 40- 12 P-i/81 Pi / 81-L OBI 81812 L - 30- 7 -25927-1 120- Vt'N 0 PP/0 0 10 11 12 13 15 16 17 18 (26min) PM onlyJune Days Day of fledging of 2 young Nest 27: 2103 Young (Much disturbance) -7 210 min, 151 170 194 160 60 175 95 100- - - - 00 ? ? Pi W ¤081 P/O-SCr -lxxod - - 90- Pi W Pi G ~|Pi G ~-_GG - Pi/Bl P/0- P *Mod -6 Pi G - Pi/81 A.|P'/811 PIG - 4 --~~1GG % F ee di ng s of Y ou ng : P ar en ts a nd H el pe rs - L/r-*L ' Pi G 80- .- - I J Pi/81 * 081 081 L xxodll OB, -5 70- - -4 ~ ( P~ 34~/ PP Pim\1 60 - - Oal1- 1 lip4~ «sj f ~ «I f * .P,6 - 3 1~ 2 20- *~ ~*t R « 4 0 0 **6 2 e° AN 29 30 1 2 3 6 (Fledging Day)June July Days Nest 26: 3 young FIGURE 9. (Percentage contribution of attendants and rates of feeding visits at nests 27 and 26, 1977. Note a larger number of attendants at both nests but higher contribution of parents at nest 26. 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 223 -7 1 60- .EZE.in 131 -221 94 195 120 - 11 1 un g: P ar en ts a nd H el p6 rs %5- 1 1 1/70-65- lip-/f 7/711'111 -% 1«S>2*0202*I -4 =jips 1 1 Ill li ESS S S 4 -315-2 ~<20- I~---·~~ JI / 4 S \PA XX /'0- 19( p2 / / / 5 ~ T 27 28 29 30 1 2 Pg at neil June July totol time Days 1156 min.) First week Early second week Nest 34: 20r 3 young Fooc for sitter Helsers fed -7 50 min 180 65 )nly 152 190 I 65 go thnugh Pg 168 180 175 - 3 73 T ,65,90/f- 3 / 0.05). The feeding rate per nestling did not increase with brood size (r,= - 0.16, P > 0.05). The feeding rate per nest is not correlated with number of frequent attendants (rs- 0.321). The brood size is not correlated with the number of frequent attendants Crs=.-0.083). In summary, the amount of food received by nestlings and the work of bringing it by attendants form a complex relationship in which the availability of more attendants reduces the work load of in- dividual attendants, and the number of feeding visits required to serve the young is directly related to the size of the brood. Still, nests attended only by parents and a few helpers apparently received just as many feeding visits as those with many attendants. We do not have enough data to provide information on the relation- ship between the number of attendants and nesting success. KINSHIP AS A FACrOR IN ATTENDANCE. - We were able to follow only six birds of known parentage from fledging through to their roles as at- tendants at nests. All these birds helped their parents if their parents had TABLE 6.- Summary of nest attendance data for San Blas Jays, groups 1 and 2, 1976, 1977. 1981 H A R D Y ET A L.:S O U TH E R N SAN BLAS JAYS 227 Frequent Ratel j Group, Attendants Total Obs. Span of Rate of Rate/ Frequent Year Nest Brood Size (Total No.) Time in Hrs. Obs. Feeding Nestingi Attendant 9-21 July 1 12 2 4(7) 16.8 Last 11 days of 3.4/hr. 1.7 0.85 1976 nestling period 13 3 5(11) 4.3 3 days, 4.4/hr 1.5 0.88 beginning of last week of nestling period 15 2 2(7) 7.45 18-22 July 2.7/hr. 1.35 1.35 2 18 2 ?(8) 1976 21 2 2(7) 1.8 4 July 5.3/hr 2.65 - 23 3 (12) 22 2 3 or 4(8?) 6.5 Last 7 days 2.9 1.45 0.97 1 27 2 7(8) 27.35 Last 9 days 3.7/hr. 1.85 0.53 1977 26 3 5(11) 26.05 Entire 4.4/hr. 1.47 0.88 34 2-3 2(4) 16.90 First week 1.7/hr, (Destroyed) 36 2-3(early) 2(7) 27.20 Entire 1.8/hr. 1.80 0.9 1, mid-late 2 28 3 3(7?) 11.1 Final week 4.5/hr. 1.50 1.50 1977 29 2 5(14) 22.73 Final week 3.2/hr. 1.60 0.64 30 2 3(10) 3.08 Final week 2:9/hr. 1.45 0.97 31 2 3(7) 6.70 First week 1.04/hr. 0.52 33 ?>1 3(4) 9.50 First 11 days 2.0/hr. 37 3 2(3) 17.30 First 2 weeks 3.5/hr. 1.17 1.75 (almost entire) 38 ?> 1 2(6?) 14.5 From 15th to 15th day 2.4/hr. - 1.20 40 ?>1 2(3) 3.9 2 days at end of 1.02/hr. - 223.16 ca first week 1 Where data are not supplied in the last two columns, it is because they were too few or from early in the nesting period and thus do not reflect accurately'the rates of attendance. 228 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 young to feed, although help was delivered to the first nest that produced young, regardless of the identity of the breeders there. , OBl was an offspring of OPi and PiO in 1975. In 1976 he was known to have been a helper at three nests, the first that of OPi, his father. and Bl/G, his stepmother, PiO his mother having disappeared. In 1977, as a 2-year-old, 0B1 was paired to BlBk at nest 26. That nest and its eggs were taken over by OBl's father OPi and OPi's mate, Bl/G-L. OBI then was mate, Bl/G-L (G/0-L having been killed by a predator). OBl then was recorded as a helper at nest 26, thus being a helper to his own young for his father, the grandfather of the young! In 1978, OBl was apparently breeding with BlBk again, although the nest was not found. WBk was the sole offspring of WO and WB1 at nest 18 in 1976. In 1977 he was a nonbreeding helper at his parents' nest (28) and rarely at nearby nest 29. WBk was apparently still a helper at his parents' nest in 1978, but the young had fledged before we arrived. Pi/G-R was an offspring of OP and Pi/Bl-R at nest 22 in 1976. Pi/G-R helped at three nests in 1977. It helped first at nest 31, which produced young early in the breeding season. Later it helped at nest 33, its father's nest, and at nest 38, both of which had nestlings of about the same age. All three offspring of B10 and 0/G-L at nest 13 in 1976, B1B1-2, Pi/Bl- L, and PiPi-2, were helpers in group 1 in 1977. Nest 27, owned by GG and P/0-R, was the first nest known to have nestlings in group 1 in 1977. All three sibling yearlings helped there. Blfll-2 and Pi/Bl-L were also ac- tive helpers at nest 26 (owned by OPi), the second nest in the group to pro- duce nestlings. Thus at nest 26 they were feeding nestlings, which, though in a nest "owned" by OPi were actually the offspring of OBl. B1B1-2 and Pi/Bl-L were also active feeders of the fledglings of nests 26 and 27 and the orphans we introduced. In 1978, only Pi/Bl-L of the three sibs remained in group 1 and was seen as a helper with the putative fledglings of OPi and Bl/G-L. Some young, such as WBk, may even show some nearly ex- elusive attachment to their parents, but nearby nests with earlier young usually attracted attendance by such birds. birds. AGE AT FIRST BREEDING. - Of seven jays marked as either fledglings or yearlings and followed at least to their first nesting attempts (see Appen- dix I), two males PiG and OBl, and one female, BlPi, attempted to nest as 2-year-olds; three, PP and Bl/G-R, both males, and BlW, a female, nested first as 3-year-olds. GG, a male, nested first as a 4-year-old. Of all 14 yearlings whose activities were well known. only 2 (both females) attemp- ted to nest; of all 10 well-known 2-year-olds, only 4 (3 males, 1 female) at- tempted to nest. We conclude that although most individuals first at- tempt to breed in their third year, and rarely the fourth, attempts by yearlings and 2-year-olds to breed were of regular occurrence in the 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 229 groups studied. The possible significance of breeding by subadult birds (less than 3 years 01* is treated in the Discussion. BREEDING TIME OF ADULTS VS. SUBADULTS. - Nests involving subadults as breeders were usually late, among the last to be initiated in the com- munal groups studied (Fig. 13). The one exception was nest 26 in group 1, at which OBl was the 2-year-old male breeder. This nest was second only to nest 27 in the temporal sequence. It may be significant that 081 matured early in plumage and softpart color, being without crest feathers and having yellow irides only faintly tinted with green in the second year. Because we terminated study each year before the fate of late nests could be known with certainty, we lack data on the relative success of late and early breeding attempts. 37 8 Pi/0-R El: MALE 08k FEMALE 35 B ~:i C} 23 YEARS E3 2 YEARS 30 8 Bhpi 31~ 0/p,-R ~Bk/0-R P/G-R 40~]10/81-R r-h ppi 38 ~G/0-R ~ 1 YEAR 1-r~GP 29 0 Z 33 ~ Brpi 1977 GROUP 2 28 ~3 WBI WO 32 988xg 27 0 P/O-R 268= 34EIZ 3687)? 1977 GROUP I 22 Q Pi/81-R 23 ~Pi/0-R 1976 GROUP 2 GP 21 (2&-R 18 ~ V'0 17 82: WBI 'SE~== 13 E] BIO 1976 GROUP I 41 0/G-L 12 El 0856-L GBI gBX 3~PiG BkBk 1975 GROUPI 563 OPi Pio XX 11 13151719212325272931~ 3 5 7 911131517192123252729 ~ 3 5 7 9111315 MAY JUNE JULY FIGURE 13. Approximate dates of egg-laying for three age classes of breedersin first breeding attempts of the season. Pairs in which yearlings or 2-year-olds are members tend to nest later. 230 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 PAIR BOND PERSISTENCE In 17 cases both members of the pair survived from one year to the next. (Three of these cases involved unmarked adults: OW's mate from 1976 through 1978 and PP's mate from 1976 through 1977 were un- marked.) In only one known case were the birds not mated to each other in the second year. OP's 1976 mate, Pi/Bl-R, was seen briefly in June 1977 and then disappeared. The bird may have been sick and was associating with another pair (XX, a 2-year-old, and BlW) that was building nest 32. However, OP already had another mate, BlPi, at this time. We conclude that these jays normally mate for life or until one leaves the communal group. HOME RANGE Figures 14 and 15 show the almost exclusive home ranges (as well as nest locations) of San Blas Jay communal groups in 1976 and 1977. These home ranges were maintained by mutual avoidance rather than by active defense. On many occasions birds of neighboring groups came close to each other at boundaries between ranges, sometimes into the same trees, without any signs of intergroup hostility. We noted no vocalizations or postural displays specialized for territory defense. On the rare occasions when trespassing occurred the intruder moved swiftly and inconspicu- ously and avoided confrontation with residents. In 1975, group 2 birds, such as OW and WO, occasionally flew at palm-crown height with at most one stop across the nest area of group 1 to the west jungle patch (Fig. 1). This behavior apparently stopped in 1976 after the west jungle patch was largely cut down. An adult stranger netted in the heart of group l's range in June 1975 and never found there again was finally found to be a member of group 5 from 1975 through 1977. INTRAGROUP CORE AREAS Within communal group 2 we regularly saw strong evidence that breeding pairs verged on territoriality toward members of their own group. Especially when their nests had contents, male breeders restricted their movements to small "core areas" around their nests and exerted various degrees of dominance over intruders. Figure 16 shows the core areas of males G/0-R, OP, XX ad, and OBk in 1977. Group 1 rarely showed such intragroup hostility, probably because of its small size, smaller home range, and the resulting greater familiarity among its members. Male breeders of group 2 restricted their visits away from their core areas almost exclusively to gathering food hastily and returning as soon as possible. When inside their core areas, they typically made short flights 1981 HARDY ET AL.: SOUTHERN SAN BLAS JAYS 231 i , N i / /1.-»/ - / \ /\ 1 - ®12 2414 16 ® ® 2 \ 20 I8 'IO I / ./ / \ /2\ / 119 ~ 0 ~ Hill ~ /i25 1 1 3 51/ i 1 1 / 4 /1 1 /1 / 1 \ ,1./ i i-/ I- --____-,Meters ~ - \ 0 100 I V FIGURE 14. Distribution of known nests and group home ranges at the main study site, 1976. for communal groups 1-5. interrupted by long pauses, but when venturing outside, they made long uninterrupted flights to the objective and spent relatively little time there (Table 7). Of 72 recorded pauses in his core area, G/0-R preened in 7, did sentry duty in 24, foraged in 32, and scolded in 19. In his 10 visits to the central mango out of his core area, G/0-R only gathered food. Once while birds of the nest 33 subgroup were mobbing the observer at the mango tree because their fledgling was nearby, G/0-R fed in the same tree but took no part in the mobbing. Moments later G/0-R retreated to his core area and at its boundary immediately began to scold the observer. Of the 71 recorded pauses of OP in his core area, he preened in 3, did sentry duty in 13, foraged in 63, and scolded in 6. Outside his core area, OP was never seen to perform any of these behaviors except forag- ing. Birds other than the breeding pair rarely entered the core area when the residents' nest held eggs. After the young were about one week or more old, the traffic of nonbreeding helpers became more frequent and 232 BULLETIN FLORIDA STATE MUSEUM Vol. 26, No. 4 -------- 43 . I I.- I ' , ''-46 44 4127 0 0? * U0-\ 31 35 , 36 0 33 0 * i39 26 4 0 2 321 37 ~ 28 l~ 0 290 38 i' A: .-.,---~~i>~0=----+24*_-F v, 30 0?7 I n <(BARE) ~ ; 0 i flC /~ ~ /'~ ,~ 47 ---' / 42f / 0? / 1 1 { 0 ? 3 )'1 f31 r . 1. ]: Wilson Bull 91: 1-15. Skutch, A., in A.C. Bent. 1958. Life histories of North American blackbirds, orioles, tanagers. and allies. U.S. Natl. Mus. Bull. 211:335-350. Stacey. P.B., and C.E. Bock. 1978. Social plasticity in the Acorn Woodpecker. Science 202:1298-1300. Wetmore. A. 1927. The birds of Porto Rico and the Virgin Islands (Psittaciformes to Passeri- formes). New York Acad. Sci. 9:409-598. Woolfenclen. G.E., and J.W. Fitzpatrick. 1978. The inheritance of territory in group- breeding birds. Bioscience 28:104-108. APPENDIX I 252 B U LLE TIN FLO R ID A S TATE M U SEU M Vol. 26, N o. 4 COMPOSITION AND HISTORY OF SAN BLAS JAY GROUPS 1 AND 21 Color Year Age Sex ROLES When Codei Marked Marked (Crit.) 1974 1975 1976 1977 1978 GROUP 1 1 00 74 Ad a A-1 P?-Nt H-13, 15 Found (Role) H-Nx, 3,5,9 Dead June 3 2 PiPi 74 Ad 9 P?-1 (BP) 3 BkBk 74 Ad p?-l P-9, H-5? (Role) 4 WW 74 Ad 9? A-1 P?-Nx A-BlP (Role 75) H-Nx, 13, 15 5 BIN 74 Ad ? A-8 6 XX - Ad 9 P-2 (Role) 7 XX - Ad ? A-3 8 XX - 2 ? A-3 9 GG 74 1 1 ? H-5 H-12, 13, 14? P-27 A-fl? (Role) H-26 10 PP 74 1 0 A-fl H-5,9 P-15 P-34 P-Nx (Role) H-13 H-26, 27, 36 11 PiG 74 1 EY ? P-3 P-Nx? P-36 P-50 (Role) H-4,5,9 P of 0/Pi-L? H-26,27 H-13, 15 12 XX - 1 ? A-3 13 XW 74 sib~ 0 ? n Nx 14 XPi 74 sib 0 ? fl Nx H-3,5,9 15 XG 74 sib | 0 ? fl Nx 16 XO 74 sib-1 0 ? n Nx 17 OPi 75 Ad a P-5 P-12 P2-264 P-48 (Role) H-9 H-13 H-36 1981 H A R D Y E T A L.: S O U TH E R N SAN BLAS JAYS 253 18 BIO 75 Ad o· P-4 P-13 (Role) H-3,5,9 H-Nx n 19 GBI 75 Ad 9 P-4 (Role) H-5,9? 20 PiO 75 Ad 9 P-5 (Role) H-3,9 21 XX - Ad 9 P-9 (Role) 22 GW 75 1 1 H-5,9 23 XX - 1 9 P-3 (Role) 24 GO 75 0 ? 58 25 OBL 75 0 1 5 n H-12, 13, 15 Pi-264 P-Nx? (Role) P-39 with BlBk? 26 Bl/G-L 76 Ad 9 P-12 P2-264 P-48 fl (Role) 27 XX 76 Ad 9 P-15 P-34 (Role) H-36 28 0/G-L 76 Ad 9 P-13 Predatorized (Role) H-12 June 29 BlBk 76 Ad 9 P?-Nx Pi-26 P-Nx? with (Role) P-39 OBl? 30 PiP 76 1 ? H-12, 13, 15 H-BlP fl 31 PiW 76 1 ? H-Nx, 12, 13 H-26, 27, A-fl 34?, 36 32 PiBK 75 0 9 Group 3 fl P-14? (Role) sitter, dead 33 BIP 76 0 ? Nx fl 34 PiPi-2 76 sib ~ ~ 13 fl H-270? 35 B1B1-2 76 sib ? 13 0 H-27, 34, 36 36 Pi/Bl-L 76 sib 0 07 13 n H-26,27 A-n APPENDIX I (continued) COMPOSITION AND HISTORY OF SAN BLAS JAY GROUPS 1 AND 21 Color Year Age Sex ROLES When Codet Marked Marked (Crit.) 1975 1976 1977 1978 37 0/Pi-L 76 0 ? Nx fl 38 Bl/Pi-L 76 sib 0 ? 12 fl 39 G/Bl-L 76 sib 0 ? 12 fl 40 XX - Ad 9 P-36 (Role) 41 P/0-R 77 Ad 9 P-27 P-50 (Role) H-26 PiBk-2 77 sibl 0 ? 26 n B U LLE TIN FLO R ID A S TATE M U SEU M Vol. 26, N o. 4 8& 56 :t as O/G-L-2 77 sib~ 0 ? 26 n B1O-2 77 sib 0 ? 26 fl G/O-L 77 sibl 0 ? 27 fl P/Pi-L 77 sib J O 7 27 fl 81/0-L 77 0 36 n 2 ? A-fl - 1 ? A-fl GROUP 2 1975 1976 1977 1978 1 OG 75 Ad ? ? H-15?, 21 2 WO 75 Ad G H-8? P-18, H-21 P-28 P-51 fl (Role) H-29 3 OW 75 Ad a ? P-17, H-18, P-29, 45, H-28, P- 52 fl (Role) 21 H-30 fl 4 PPi 75 Ad 9 ? P- 21 P-31 , 46 P-57 (Role) 5 OP 75 2 7 ? P-22, H-23 P-33 P-56 (Role) H-29,30 6 PO 75 Ad 9 H-8? H-28 fl H-30? (BP) seen once 7 OBk 75 Ad cy ? H-24? P-37, H-29 P-55 1981 H AR D Y E T A L.: SO U TH ER N SAN BLAS JAYS 255 (Role) 30 fl 8 CPi 75 Ad 9 Group 3 ? P-35,44 P-58 (Role) 9 BkO 75 Ad ? H-8 H-18 10 CP 75 Ad 9 ? P-23 P-40, 43 P-59 (Role) H-24 11 BlW 75 1 ? H-8 H-18 P-32, 41 H-59 H-28? 30, 31, 38 12 PiX 75 1 ? 13 PC 75 0 ? 9£1 14 BlG 75 0 ? 6fl 15 PiBK 76 Ad 9 H-22, 23 (BP) 16 WBJ 76 Ad 9 P-18 P-28, H-29 P-51 11 (Role) 17 Pi/Bl-R 76 Ad 9 P-22, H-23 fl ? seen with (Role) H-24 P's of 32 twice 18 G/O-R 76 Ad a P-24, H-23 P-38, H-30 P-59, H-58 (Role) 31 n 19 0/Pi-R 76 Ad cy P-21 P-31 P-57 (Role) 20 Pi/O-R 76 Ad 9 P-23, H-22 P-37,46, H-29,30 fl P-55 (Role) 21 0/81-R 76 Ad a H-22,23 P-40, 43 (Role) H-29 22 XX - Ad 9 P-17 P-29, P-45, H-28 P-52 n (Role) H-18, 21? 23? 23 XX - Ad 9 P-24,21? 237 (Role) H-18 24 BlPi 76 1 9 H-18, 23? P-33, H-28,29 P-56 (Role) 25 81/0-R 76 1 ? H-22,23 APPENDIX I (continued) COMPOSITION AND HISTORY OF SAN BLAS JAY GROUPS 1 AND 21 256 B U LLE TIN FLO R ID A S TATE M U SEU M Vol. 26, N o. 4 Color Year Age Sex ROLES When Codel Marked Marked (Crit.) 1975 1976 1977 1978 26 Bl/G-R 76 1 e H-22,23 H-30,33,29 fl P-53 (Role) 27 WBk 76 sib 0 ? 18 n H-28,29 H-51, 52 28 WPi 76 sib~ O ? 18 n 29 Pi/G-R 76 sib 0 ? 2211 H-29, 30, 31, 33, 38 H-57 30 G/Pi-R 76 sib~ 0 ? 22 fl H-31 seen once 31 0/6-R 76 0 ? 21 fl 32 PiBk 76 sibl 0 ? 23 fl 33 81/Pi-R 76 sib~ 0 23 fl 34 G/Bl-R 76 sib 0 ? 23 fl 35 XX Ad a P-35,44 P-58 (Role) 36 P/G-R 77 Ad o· P-30, H-29,37,38? P-54 (Role) 37 Bk/Pi-R 77 Ad 9 P-30, H-29 fl P-54 (Role) 38 XX - 2 0 P-32,41 (Role) H-31? 38, 40 39 Bk/0-R 77 1 9 P-38 (Role) 40 Pi/Pi-R 77 sib 0 ? 28 fl 41 Pi/Bk-R 77 sib ? 28 n 42 81/81-R 77 sib~ ~ ? 28 fl 43 0/0-R 77 sib ? 29 fl 44 G/G-R 77 sib 0 ? 29 R 45 0/P-R 77 sibl 0 ? 30 fl 46 G/P-R 77 sib-1 0 ? 30 fl 47 0/Bk-R 77 0 ? 31 fl (Predatorized) 1981 H A R D Y E T A L.: SO U TH ER N SAN BLAS JAYS 257 48 XX - Ad 9 P-53 (Role) 49 XX - 1 ? with OW, XX, WO, WBI 50 XX - 1 ? with OW, XX, WO, WB1 51 XX - 1 ? with G/O-R, CP, and BlW 52 XX - 1 ? with 0/Pi-R, and PPi 'Code: color flag tarsal markers. A slant mark between two colors indicates they were on the same leg, designated by a suffix L"or "R." Asuffix numeral indicates that a bird dead or long missing was previously marked with this combination. X means no band. Sex (crit.): birds were sexed by role, or brood patch (BP). Roles: P= parent:H= helper: A= attendant, whether helper or parent not known; fl = fledgling: numerals = number