Condor, 82:4061116 @ The Cooper Ornithological Society 1980

SEX-SPECIFIC DIFFERENCES IN WINTER DISTRIBUTION PATTERNS OF

JAMES D. NICHOLS AND G. MICHAEL HARAMIS

ABSTRACT.-Winter band recovery distributions of North American Can- vasbacks (Aythya walisineria) suggested that males and females exhibit com- parable degrees of fidelity to general wintering areas. Of banded during the winter, the proportion of males was found to be higher in northern than in southern areas. Winter band recovery distributions of birds banded in particular areas during the summer were found to differ significantly between sexes, with females being recovered farther south. Factors that may have affected the of sex-specific wintering distributions include: (1) pos- sible reproductive benefits derived by males who winter in the north and thus reach northerly breeding areas early; (2) in body size, which may render the smaller females especially susceptible to periods of inclement weather and food shortages; and (3) interactions between sexes in which males may control food supply when food is scarce. Two lines of evidence from field data on Canvasbacks in the Chesapeake Bay suggest the existence of competition between males and females. First, Canvasbacks trapped during winter in smaller bodies of water tended to have higher pro- portions of females and weigh less than birds trapped in large open bodies of water. Second, analysis of aerial photographs of wintering rafts of Canvas- backs showed patterns of intersexual segregation, with females being found more frequently on peripheral areas of rafts.

Lack (1954, 1966) and Fretwell (1972) have ratios of North American waterfowl vary suggested that “limitation” of temperate geographically during the winter. Hoch- populations most likely occurs during baum (1944: 132) suggested that Canvas- the winter or nonbreeding season. It is backs (Aythya valisineria) may exhibit a therefore of interest to examine the distri- differential distribution of sexes on the win- bution and abundance of bird populations tering grounds. Allen (1931:354) and De- during this period of the year. In many Graff et al. (1961:79-80) noted the high pro- species of migratory birds in the Northern portion of males among Canvasbacks Hemisphere, males appear to winter farther wintering in , and suggested that north than females (see Nice 1933, Lack females were more numerous farther south. 1944, Selander 1966, Ketterson and Nolan The possibility that male and female Can- 1976 and references therein). It has been vasbacks winter in different areas is inter- suggested, (Witherby et al. 1939:219) that esting from both biological and manage- female and young diving in Great ment points of view. The continental Britain generally migrate farther south than population is thought to exhibit adult males. Evidence of such a sexual dif- the most disparate sex ratio (65-70% male) ference has been provided in North Amer- of all North American ducks. This ratio (and ica for Common Mergansers (Mergus mer- the low “reproductive potential” which it ganser; Anderson and Timkin 1972) and implies) has been labeled “a fundamental Buffleheads (Bucephala albeola; Erskine problem in the population dynamics of the 1971), in for Common Pochards Canvasback” (Trauger 1974). Canvasbacks (Aythya ferina) and Tufted Ducks (Aythya hatch approximately equal numbers of fuligula; Salomonsen 1968), Mallards (Anus males and females (Hochbaum 1944:51, platyrhynchos; Lebret 1950, Nilsson 1976), Sowls 1955:164); hence, the disparate pop- European Wigeon (Anus Penelope), Green- ulation sex ratio must be due entirely to dif- winged Teal (Anus crecca), Northern Shov- ferential mortality. Indeed, in an analysis of elers (Anus clypeata), and Pintails (Anus Canvasback banding and recovery data, we acuta; Lebret 1950). Bellrose et al. (1961) found mortality rates to be significantly summarized evidence suggesting that sex greater in females than males in each of the WINTER DISTRIBUTION PATTERNS OF CANVASBACKS 407 three populations examined (Nichols and where ri is the linear rank of observation i, n is the Haramis 1980). The existence of this differ- number of observations in the first sample, and m de- notes the number of observations in the second sam- ence emphasizes the importance of exam- ple. We then compute the resultant or vector sum of ining any sexual differences that may occur the first sample as: in distribution or general characteristics and habits. If winter mortality is as important to RI = [(k co~/ji)~+ ($ SinU)]‘ ” migratory bird populations as has been sug- gested, then the investigation of possible The null hypothesis of no difference between the two differences in the winter distribution pat- bivariate samples is then rejected for large values of terns of the two sexes may help to explain R,. Mardia (1967) has shown that for n + m > 17: the higher female mortality rates and resul- U = 2R:(m + fi - 1)lmn tant disparate sex ratios. Our primary objective was to test the gen- is approximately distributed as chi-square with two degrees of freedom. Under the hivariate normal distri- eral hypothesis that male Canvasbacks tend bution, the efficiency of Mardias’ test relative to Ho- to winter farther north than females. If this tellings’ T ’ (its parametric competitor) was found to lie hypothesis was corroborated, our secondary between 79% and 93% (Mardia 1968). Our own Monte objective was to explain the phenomenon Carlo simulation work with distributions that more or at least to obtain information on factors closely resemble band recovery patterns have resulted in approximated powers similar to those obtained us- that might have affected its evolution. ing hivariate normal distributions. “Ties” between ob- servations from the two samples occurred in some of METHODS our tests with empirical recovery distributions, and we computed approximate chi-square test statistics in the Our data were obtained from three sources: the com- manner suggested by Robson (1968). puter retrieval files of the U.S. and Wildlife Ser- Aerial photographs of Canvasback rafts, taken at an vice Bird Banding Laboratory (BBL), Laurel, Mary- altitude of 150 m, were examined under a stereobinoc- land; aerial photographs of Canvasback rafts in the ular scope to identify males and females. Negatives Chesapeake Bay; and banding and weight records of were projected with an overhead viewer to map rafts Canvasbacks trapped in the Chesapeake Bay. The BBL for tests concerning the relative distribution of the data pertained to the period 1929-77, whereas the oth- sexes. er data were collected during the winter of 1977-78. The BBL data used here included 2,663 recoveries and We investigated the winter weight-loss of Canvas- backs by trapping them in mid-January and again in over 55,000 bandings. Only birds categorized as “nor- mal, wild” were used in our analyses. Recovery rec- early March 1978 in upper Chesapeake Bay. Welded- wire diving traps baited with corn were used. ords were restricted to birds shot or found dead. All Trapped birds were held for a minimum of 6 h before birds banded during the winter (January-February) weighing to permit the drying of and the pas- were considered to be adults, whereas summer-banded sage of much of the ingested bait. This procedure re- birds (July-September) were aged as either young duced the variance of our data. Before weighing, we (first year) or adult (older than one year). inspected the gullet of each bird for corn content; only In two portions of our analysis we tested the null individuals containing little or no corn were consid- hypothesis that geographic distributions of band recov- ered in the analysis. Birds were weighed to the nearest eries from two samples came from the same popula- tion. Each observation (band recovery) is characterized 5 g. by latitude and longitude of recovery measured to the nearest lo-min coordinate. The statistical problem is RESULTS AND DISCUSSION thus one of testing whether two-sample distributions belong to the same bivariate population. Kolmogorov- TEMPORAL VARIATION IN WINTER DISTRIBU- Smirnov test statistics (e.g., Sokal and Rohlf 1969:571- TION PATTERN 575) were computed to test the null hypothesis that the empirical marginal distributions of latitude and lon- We wanted to learn whether Canvasbacks gitude (longitude was transformed via the relationship return to the same general wintering areas 1” longitude = (1” latitude) (cos A), year after year, because we believed that the presence or absence of such a tendency where A is the latitude; see Raisz 1962:147) were uni- variate normal. This null hypothesis was rejected (I ’ < would be important in interpreting any pat- 0.01) for both latitude and longitude in the first several tern of geographic variation in sex ratio that recovery distributions tested, and we concluded that might exist. We chose to address this ques- parametric tests based on the assumption of bivariate tion at the population level first by compar- normality would be inappropriate for these data. For testing pairs of empirical distributions of band ing direct (first year after banding) recovery recoveries we used the nonparametric test suggested distributions from particular winter banded by Mardia (1967, 1972: 197-201). Briefly, this test first samples with indirect (second and later involves computing the centroid or center of gravity of years after banding) recovery distributions the combined two-sample distribution. Vectors are (similar rationale was employed by Cowar- then considered from this centroid through each sam- ple point (band recovery), and the points are ranked, din 1977:6-7). For this analysis we used based on the vector directions. These directions or an- BBL data from Canvasbacks banded during gular observations are then replaced in the first sample January and February in particular states by: during specific years, and recovered during pi = 2?T r,/(n + m), i = 1, . . . n, the period from 15 December through 28 408 JAMES D. NICHOLS AND G. MICHAEL HARAMIS

TABLE 1. Results of testing the null hypothesis of no and for the single banded sample of New difference between direct versus indirect winter re- York females (Table 1). The summary test covery distributions of winter-banded Canvasbacks.” statistic for New York males was also sig-

Total nificant. We found no difference between state Year recoveries Chi-squa& df direct and indirect recovery distributions Males for any of the banded samples in other states. Hence, we conclude that birds from Maryland 1956 20 0.24 2 1957 20 4.48 2 these states exhibited the same general win- 1964 66 0.07 2 tering distribution patterns year after year. 1965 33 1.57 2 The difference between the direct and in- Total 6.36 8 direct recovery distributions of New York New York 1955 62 6.01** 2 birds could be attributable to two factors: 1) 1956 81 3.10 2 Unlike the other tested samples, Canvas- 1957 198 24.26*** 2 backs banded in New York during the years 1958 115 11.54*** 2 1965 24 4.76* 2 noted in Table 1 were exposed to large year- Total 49.67*** 10 to-year fluctuations in winter hunting pres- sure associated with varying hunting season Virginia 1969 24 0.48 2 dates in New York (see Martin and Carney 2 California 1969 82 3.05 1977). 2) Some Canvasbacks may tend to Females winter in New York during some years and Maryland 1964 21 0.66 2 in areas farther south during other years, New York 1957 52 7.20** 2 probably depending upon weather condi- tions. This possibility is consistent with the a Includes birds banded during January and February and recovered during the period from 15 December through 28 February. suggestion of Stewart et al. (1958) and b See Mardia (1967). * .05 < P 1.10; ** .Ol < P < .05; *** P < .Ol. DeGraff et al. (1961) that there is a con- siderable annual interchange of wintering Canvasbacks between New York and the February. All such banded cohorts having Chesapeake Bay area. at least 20 winter recoveries were used. We also studied annual variation in win- Records were further restricted by exclud- tering distribution by examining the pro- ing all recoveries for which exact (to lo-min portional distribution of winter recoveries block) location data were not available. among flyways. In this gross analysis we se- A total of 13 banded samples resulting in lected recoveries (15 December through 28 778 winter recoveries satisfied our selection February) of birds banded during January criteria. For each of these samples we used and February, 1929-77, and first grouped Mardias’ (1967) test to test the null hypoth- them by flyway of banding and sex. The pro- esis of no difference between the distribu- portional distribution of these recoveries tions of direct and indirect winter recover- among the four flyways and Canada is pre- ies (Table 1). Maryland and New York sented in Table 2. Birds banded in the Cen- contained several years for which selection tral flyway are omitted from this table be- criteria were met for male Canvasbacks, and cause of the small number of recoveries. A summary statistics for these states were thus high proportion of winter recoveries from obtained by summing the chi-square statis- both males and females banded during win- tics and their associated degrees of freedom. ter in the Atlantic and Pacific flyways oc- The null hypothesis was rejected for four of curred in the flyway of banding. Canvas- the five banded samples of New York males backs banded in the Mississippi flyway

TABLE 2. Proportional flyway distribution of winter recoveries from winter-banded Canvasbacks (1929-77).”

Proportion of total recoveries in recovery flyway Bnndmg T&d flyway S3‘ recoveries Atlantic Mississippi Central Pacific Canada Atlantic Male 1,390 .938 ,035 ,001 ,006 .020 Female 404 ,960 ,017 ,010 .005 ,007 Mississippi Male 100 ,210 ,760 ,020 ,000 .OlO Female 34 ,147 .794 .059 .ooo ,000 Pacific Male 246 ,000 .ooo .ooo .996 ,004 Female 99 ,000 ,000 .ooo 1.000 .ooo

a Include\ birds handed during January and February and recovered from 15 December through 28 February. WINTER DISTRIBUTION PATTERNS OF CANVASBACKS 409 seem to exhibit less fidelity to wintering TABLE 3. Sex ratios of Canvasbacks banded during areas as indicated by the relatively high pro- January and February (1929-77). portion (about 0.15 to 0.21) of recoveries oc- Pro- curring in the Atlantic flyway (Table 2). No. hrds Pi:,; These data are consistent with the conclu- Flyway State banded sion of Stewart et al. (1958) based on winter Atlantic New York 15,161 0.78 counts that major shifts of Canvasbacks may New Jersey 1,372 0.71 take place between the Mississippi Valley Delaware 3,288 0.56 Maryland 16,788 0.74 and mid-Atlantic regions. Stewart et al. Virginia 3,793 0.68 (1958) also noted that numbers of wintering North Carolina 729 0.50 birds may shift between the Detroit River South Carolina 117 0.31 and Lake St. Clair area and the Chesapeake Mississippi Michigan 1,325 0.79 Bay. Alabama 403 0.40 Finally, we were interested in whether Louisiana 1,444 0.67 male and female Canvasbacks seem to ex- Central Texas 118 0.51 hibit comparable degrees of fidelity to gen- Pacific 729 0.67 eral wintering grounds. Male and female California 5,239 0.64 test results presented in Table 1 were sim- ilar for both Maryland and New York, sug- gesting a similar pattern of fidelity in the two sexes. We also addressed this general essarily equivalent to those of the sampled question using the data of Table 2. We used populations at the time of banding. Instead 2 x 2 contingency tables to test the hypoth- we simply assumed that any inherent sex- esis that the proportions of males recovered specific difference in trappability would be within and out of the flyway of banding reflected in all of the samples, and that dif- were not different from those of females. ferences in trap sex ratios between different The null hypothesis could not be rejected areas would indicate differences in popu- for birds banded in either the Atlantic (x ’ = lation sex ratios. 2.89, 1 df) or the Mississippi (x2 = 0.17, 1 In this analysis we used 50,862 winter df) flyways. Pacific flyway bandings could banding records from BBL files summarized not be tested in this manner but exhibited by state of banding. We selected only birds virtually identical recovery distributions banded during January and February, and with all but one recovery occurring in the pooled data for all years (1929-77). Banding flyway of banding. Thus we conclude that sex ratios for states in which at least 100 the general degree of temporal variation in Canvasbacks were banded are presented in winter distribution patterns was similar for Table 3. A number of 2 x 2 contingency males and females. tests was conducted to test for differences in banded sex ratios between various pairs LATITUDINAL VARIATION IN SEX RATIO OF of states. In the Pacific flyway, Oregon WINTERING BIRDS bandings had only a slightly higher sex ratio We first approached the question of wheth- (x2 = 3.1, 1 df, P < 0.10) than the California er male Canvasbacks tend to winter farther sample. However, the two northernmost north than females by testing the specific states of the other flyways, New York and null hypothesis that sex ratios (expressed as Michigan, had significantly higher (P < proportion male) of banded samples in 0.001) sex ratios than all other states when northern states were higher than those ob- tested via a series of individual compari- tained in southern states. Inferences drawn sons. In the Atlantic flyway, where the from this test do not require the assumption banded sample was largest, the southern- that male and female Canvasbacks are most state (South Carolina) showed a sig- equally trappable. Petrides (1944) and Bell- nificantly lower (P < 0.001) sex ratio than rose et al. (1961:396400) showed that in all other flyway states. In addition, the certain ducks (Tribe Anatini) males are North Carolina banded sample had a lower more readily trappable than females. How- sex ratio than all Atlantic flyway states to its ever, Bellrose et al. (1961:398) also pre- north (P < 0.001). The latitudinal trend in sented evidence of an instance in which no sex ratio is not consistently shown within sex-specific difference in tendency to enter the mid-Atlantic states (New Jersey, Dela- traps existed for Blue-winged Teal (Anus ware, Maryland, Virginia); Delaware exhib- discors). In our analysis of latitudinal vari- ited a lower sex ratio than Virginia. We sus- ation in sex ratio of banded birds, we did pect that the differences in sex ratio not assume that these sex ratios were nec- between areas of similar latitude can be in- 410 JAMES D. NICHOLS AND G. MICHAEL HARAMIS

TABLE 4. Results of testing the null hypothesis of no difference between winter recovery distributions of male versus female Canvasbacks banded during the summer.a

Mean latitude state or Total province Age of birds recoveries MaI?‘ Female Chi-square” df P

Alaska Adult 119 41” 39” 5.73* 2 0.06 Saskatchewan Young 51 37 34” 0.82 2 0.66 Manitoba Young 72 37 34” 9.88*** 2 0.01 Minnesota Young 56 36” 34” 3.16 2 0.21 Nevada Young 55 37” ’ 37 0.01 2 0.99

Total 19.60** 10 0.03

aIncludes birds handed during the period July through September, 1950-77, and recovered during the period 15 December through 28 Fehrnary. b See Mardia (1967). * .05 < P < .lO; ** .Ol < P < .05; *** P < .Ol.

terpreted in terms of factors affecting sex- least 20 records within a single age category specificity of local geographic distribution (either young or adult). All recoveries (both patterns (see later discussion). Most of the direct and indirect) resulting from summer birds banded in these mid-Atlantic states bandings were considered. Only five states come from similar latitudes. Similarly, dif- and provinces met the sample size criterion: ferences existed among the three Gulf Coast Alaska, Saskatchewan, Manitoba, Nevada, states (Alabama, Louisiana, and Texas), but and Minnesota. Male versus female recov- they all showed lower winter sex ratios than ery distributions were compared using Mar- Michigan. Thus, we conclude from the sum- dias’ (1967) test. Results of these tests, as mary of banding data presented in Table 3 well as mean latitudes of the recovery dis- and the corresponding chi-square tests that tributions, are presented in Table 4. Winter winter-banded samples in southern states recovery distributions of males and females generally showed lower sex ratios than differed significantly for birds banded in those in northern states. On the basis of this Alaska and Manitoba. The summary test sta- sample (most major Canvasback wintering tistic computed over all areas also showed areas are included among the states ana- a significant difference between male and lyzed; Stewart et al. 1958), we suggest that female recovery distributions. In addition, southern latitudes generally contain propor- the mean latitudes of the male recovery dis- tionately more wintering females than tributions (see Table 4) were higher than northern latitudes. those of females for all five banding areas Our second approach to the general hy- (even in Nevada where the rounded means pothesis of latitudinal variation in sex ratio were equal, the mean female latitude of re- was to examine the sex-specific winter band covery was actually 0.7” less than the mean recovery distributions of birds banded dur- for males). Thus, we conclude that females ing the summer on the breeding grounds. from particular breeding areas tend to win- The null hypothesis was that males and fe- ter farther south than males from the same males banded in the same general breeding areas. areas showed equivalent winter recovery distribution patterns. The alternative hy- FACTORS AFFECTING THE EVOLUTION pothesis was that male and female recovery OF SEX-SPECIFIC WINTER DISTRIBUTION distributions differed. Specifically, we won- PATTERNS dered whether females were recovered far- Our results corroborate the general hypoth- ther south than were males from the same esis that females tend to winter farther breeding area. We used BBL data on birds south than males. We assume that this dif- banded during July, August, and Septem- ference in distribution reflects a sex-specific ber. Unfortunately, because of the low den- difference in the evolutionary costs and sity and generally poor trappability of Can- benefits associated with wintering at spe- vasbacks on the breeding grounds, sample cific latitudes. We can easily envision ben- sizes of banded birds were quite small. We efits that might accrue to birds wintering at therefore pooled data for years 1950 through southerly latitudes (see later discussion). 1977 for these tests, rather than consider Thus, there must be some cost associated groups of birds banded during specific with such behavior or we would expect nat- years. We required that the combined male ural selection to favor male Canvasbacks and female sample of winter (15 December wintering as far south as females. The ex- through 28 February) recoveries include at istence of a “cost of migration” is either im- WINTER DISTRIBUTION PATTERNS OF CANVASBACKS 411 plicitly or explicitly assumed in most hy- (1977) found that males endured fasting sig- potheses dealing with the evolution of bird nificantly longer than females. Similarly, migration (e.g., see Lack 1944, 1954, Cohen experiments on the effects of simulated 1967, Cox 1968, von Haartman 1968, Gauth- winter storms on Mourning Doves (Zenaida reaux 1978). Although migration certainly macroura) led Ivacic and Labisky (1973) to costs energy, this cost must reduce either suggest that adults (particularly males) pos- survival or fecundity expectations in order sessed a greater potential for surviving win- to be of importance in natural selection. ter storms than juveniles (particularly fe- Ketterson and Nolan (1976:686-687) dis- males). Kendeigh (1945) found that female cussed several types of mortality risks (in- House Sparrows (Passer domesticus) died creased predation, severe weather, difficul- sooner than males when starved at various ty in finding sufficient food in unfamiliar temperatures. Ketterson and Nolan (1976) stop-over locations) that “may be in rough suggested that sex-specific differences in proportion to the time (distance) spent mi- fasting endurance may have been important grating.” We have no reason to suspect that in the evolution of the latitudinal variation such migratory costs would be any greater in winter sex ratio observed in Dark-eyed (per unit of migration distance) for males Juncos. In a later experiment, they found no than females. sex-specific differences in short-term (over- Another potential cost to males wintering night) fasting ability in Dark-eyed Juncos, in southerly latitudes involves possible dis- but suggested that differences may exist advantages of arriving on courtship or over longer periods (Ketterson and Nolan breeding grounds later than males winter- 1978). Among field studies, Harrison and ing farther north. Data on temporal changes Hudson (1964) found that female European in Canvasback composition in an area Wigeon and Sheld-Duck (Tadorna tadorna) of the Chesapeake Bay led Welling and died in disproportionate numbers during a Sladen (1979) to conclude that males left the period of severe winter weather. In addi- Bay earlier in spring than did females. tion, female movement away from areas Hochbaum (1944:26-27) found that most during periods of harsh winter weather pair formation and associated courtship ac- have been noted in tivities at Delta Marsh, Manitoba, occurred (Bucephala clangula; Nilsson 1970a), after the birds ’ arrival in April and May. Green-winged Teal (Bennett and Bolen Observations at Lake Christina, a migration 1978), Pintails (Alford and Bolen 1977), and concentration area in Minnesota, indicated Common Mergansers (Anderson and Timkin the occurrence of considerable courtship 1972), possibly suggesting increased vul- activity and pair formation in April (Smith nerability of females to such conditions. 1946). Smith also observed that the very We can envision two factors that could first spring into Lake Christina were render female Canvasbacks less likely to made up predominantly of males and sug- survive winter than males. The first factor gested that males fly northward ahead of fe- involves body size and is a consequence of males and wait on the breeding grounds for the “fasting-endurance hypothesis” (Calder the arrival of potential mates. By wintering 1974:115) which has been proposed as a in the north, males may attain a competitive possible “explanation” for the geographic advantage in obtaining appropriate breed- variation encompassed by Bergmanns’ Rule ing territories and mates. However, as noted (see Lindsey 1966, Rosenzweig 1968, Boyce for Dark-eyed Juncos (Junco hyemalis) by 1979). This hypothesis basically asserts that Ketterson and Nolan (1976), it would prob- the ability to endure periods of fasting ably be possible for males wintering in the should increase with increasing body size. south to start spring migration earlier than The energetic arguments and rationale un- northern birds, thereby erasing any poten- derlying this hypothesis have been dis- tial advantage to northern birds associated cussed by Calder (1974: 110-115). with early arrival at the breeding ground. In the Common Goldeneye, Nilsson The principal “benefit” expected to be (1970a) found that females (the smaller sex) derived by birds wintering in the south is foraged and fed especially actively during an increased probability of surviving the the coldest periods of the winter. During a winter in a milder climate. Sex-specific dif- period of severe winter weather, Harrison ferences in winter survival probabilities and McLean (1947) found female European have been demonstrated and hypothesized Wigeon to be “considerably thinner and in other avian species. In cold-room.exper- more wasted than the drakes.” If the fasting- iments with White-crowned Sparrows (Zo- endurance hypothesis is true for Canvas- notrichia leucophrys), Ketterson and King backs, then males should be able to endure 412 JAMES D. NICHOLS AND G. MICHAEL HARAMIS periods of fasting better than females be- in which dominant individuals exclude sub- cause they are larger. We used weights of ordinates from the best feeding areas (Jen- adult male and female Canvasbacks trapped kins et al. 1975, Patterson 1977). In a study in Chesapeake Bay during winter 1978 as a of diving ducks wintering in South Caroli- measure of the difference in body size. At na, Alexander and Hair (1979) found that Bay Ridge, Maryland, 18-21 January, the Canvasbacks established and defended in- mean weight for adult male Canvasbacks dividual foraging sites and that adult males (X = 1,326 g * SE of 5.44, n = 247) was 96 won a large proportion of encounters with g (7.8%) greater than that for adult females females. In addition, the sex ratio in a con- (2 = 1,230 ? 5.56 g, n = 208). Later in win- centration area of a preferred food source ter at Gibson Island, Maryland, 7-10 March, was found to be much higher (74% males) the mean weight of adult males (X = 1,224 t than in other areas (48% males). Alexander 3.57 g, n = 496) was 120 g (10.9%) more and Hair (1979) concluded that male Can- than that for females (2 = 1,103 k 6.97 g, n = vasbacks seemed to be dominant over fe- 96). The group of Canvasbacks found at males during the winter. Gibson Island during March (after ice break If intersexual competition and resultant up) was essentially the same group trapped interactions are important aspects of the at Bay Ridge in January. Weights of birds wintering ecology of Canvasbacks, then we obtained during these two trapping periods would predict that some degree of local geo- can thus be used to estimate average weight graphical segregation of the sexes should loss between mid-winter and early spring. exist, in addition to the latitudinal segrega- The difference between mean weights for tion noted earlier (see Gauthreaux 1978). If males trapped during the two periods was males are dominant over females, for ex- 102 g representing a 7.7% average decrease ample, we would expect to find intersexual in weight between January and March. The differences in occupation, with difference for females was 127 g, represent- males occupying the more attractive habi- ing a 10.3% average decrease in weight over tats. In the Chesapeake Bay area, mid-win- the two months. The greater proportional ter trappings and observations of Canvas- decrease in the average weight of females backs indicate that higher proportions of supports the plausibility of the hypothesis females occur in small flocks frequenting that females may not be able to survive outlying water areas (creeks, ponds, and harsh winter periods as well as males. other small and confined bodies of water) Another factor, intersexual competition, than in main and tributaries (large, may also contribute to the lower probabili- open bodies of water), where large rafts of ties of overwinter survival in female Can- predominantly males are found (Haramis, vasbacks. Both Lack (1954, 1966) and Fret- pers. observ., Welling and Sladen 1979). well (1972) suggested that winter shortages Similar differences between the sexes in lo- of food are “limiting” to temperate bird cal habitat occupation have been noted in populations. Lack (1954:245) further sug- other diving ducks by Nilsson (1970b). We gested that the observed tendency among used trapping data from one area where a male partial migrants to remain in the few birds occupied a small body of water breeding area over winter, results from their (Castle Haven trap site, trapped 14-18 Feb- ability to compete and survive better than ruary 1978) and another area where many females because of their higher position in birds occupied a large body of water (Bay the “pecking order.” Gauthreaux (1978) has Ridge trap site, trapped 18-22 January recently suggested that dominant individu- 1978) to test the hypothesis that a greater als of migratory species generally occupy proportion of females was trapped in the the nonbreeding habitat that is closest to the smaller water area. The chi-square test sta- breeding grounds and that subordinates tistic indicated that a significantly greater must travel farther to nonbreeding areas. proportion of females (38.2% female in the The importance of intraspecific interactions small area, n = 76, versus 20.6% female in and dominance relationships to obtaining the open area, n = 1,085; x2 = 11.77, 1 df, l- resources and surviving during the winter tailed test, P < 0.01) was trapped in the has been demonstrated for some passerines smaller body of water. A similar difference by Fretwell (1968, 1969, 1972). Ketterson between the proportion of female Canvas- (1979) and Ketterson and Nolan (1976,1979) backs captured at trap sites for puddle ducks have discussed the relevance of behavioral (creeks, ponds and other small water bod- dominance to winter sex ratios Tf Dark-eyed ies) versus those for diving ducks (main es- Juncos. Sheld-Duck populations are thought tuaries adjacent to large rafts of birds) was to be limited during the winter by conflicts noted during 1978 winter trapping opera- WINTER DISTRIBUTION PATTERNS OF CANVASBACKS 413

TABLE 5. Results of tests on the distribution of male and female Canvasbacks within wintering rafts in the Chesapeake Bay area, February 1978.

Raftphoto Total Sex ratio Coefficient Nearest neighhor Edge (0-E) ID number Canvasbacks (proportion male) of segregationa chi-squar& ’ chi-quare ’ females on edge”

19 2,474 0.914 0.05 7.38*** 8.27*** 5.7 15 701 0.889 0.11 8.48*** 2.21* 5.5 1 536 0.920 -0.09 - 4.98** 7.0 3 499 0.864 0.01 0.08 3.04** 6.7 5 494 0.903 -0.04 - - - 73 475 0.905 0.17 12.95*** 33.08*** 16.7 49 249 0.827 -0.06 - - 55 227 0.899 0.17 6.69*** - 2 220 0.932 0.01 0.02 - - 53 140 0.807 0.07 0.68 2.26* 3.4 18 131 0.779 0.16 3.29** 1.55 2.9 39 92 0.913 0.07 0.51 - - 26 72 0.653 -0.06 - -

aComputed as described in Pielou (1961, 1969). Coefficient has a range of L- 1, 11. b Computed as suggested by Pielou (1961, 1969). A one-tailed test was used. Test statistics not presented for S < 0. c Computed as described in text. A one-tailed test was used. d Shows the difference between the observed number of females (0) and the expected number (E) on the edge. Positive values indicate more females on the edge than expected under a null hypothesis of no sex-specific difference in distribution. A negative value indicate7 fewer females on the edge than expected. * 0.05 < P < 0.10; ** 0.01 < P < 0.05; *** P < 0.01.

tions by Maryland Department of Natural 96), and the difference was significant (t = Resources (DNR) personnel (V. Stotts, un- 20.8,123 df, P < 0.001, Snedecor and Coch- publ. data). We compared the sex ratio of ran 1967:104-106). The small bodies of Canvasbacks obtained in Maryland DNR water containing higher proportions of fe- puddle duck traps (32.1% female, n = 56) males therefore also contained birds in with the sex ratio of the Bay Ridge birds and poorer physical condition. This finding is again found significantly more females in consistent with the hypothesis that the pres- the puddle duck traps (x2 = 4.21, 1 df, l- ence of females in such areas may have re- tailed test, P < 0.05). Thus, we conclude sulted from their subordinate relationship that the distribution patterns of male and with males. female Canvasbacks differ at the local geo- Finally, if intersexual dominance is im- graphic level. portant to wintering Canvasbacks, we In addition, Canvasbacks captured at would expect interactions to be reflected in small sites generally appeared in worse con- the sex-specific dispersion patterns within dition than those trapped in large water flocks. Field observations suggested that areas, so we compared weights of birds from male and female Canvasbacks tended to these two types of areas. Adult males and segregate to varying degrees within winter- females trapped at Castle Haven, 16-18 ing flocks or rafts and that females tended February 1978, weighed 234 g and 189 g, to occur more frequently on the periphery respectively, less than birds trapped at Bay of rafts. These observations were formulat- Ridge, 18-22 January 1978. However, Can- ed as hypotheses and were tested using the vasbacks trapped before ice in January aerial photographs of Canvasback rafts tak- would be expected to weigh more than en in the Chesapeake Bay on 9 February birds taken in February regardless of trap 1978. The 13 rafts used in this analysis location. We therefore compared weights of ranged in size from 72 to 2,474 birds; sex Canvasbacks from the February Castle Ha- ratio varied from approximately 65% to 93% ven sample with those from a large body of male (Table 5). water taken after ice (Gibson Island site For each raft, we tested the hypothesis trapped 7-10 March 1978). The mean that the sexes were spatially segregated by weight of Castle Haven adult males (f = determining nearest neighbor relationships 1,092 g, n = 31) was lower than that of Gib- for all birds in the raft, and constructing a son Island adult males (2 = 1,224 g, n = 2 x 2 contingency table based on the num- 496), and the difference was significant bers of intersexual versus intrasexual near- (t ’ = 11.3, P < 0.001, Snedecor and Coch- est neighbor relationships as suggested by ran 1967:114-116). Similarly, the mean Pielou (1961, 1969). We also computed Pie- weight of Castle Haven adult females (X = 10~s‘ (196l?,oefficient of segregation (S) as: 1,041 g, n = 29) was lower than that for Gib- 1-(observed number of intersexual nearest son Island adult females (X = 1,103 g, n = neighbor relationships) / (expected number 414 JAMES D. NICHOLS AND G. MICHAEL HARAMIS

of intersexual nearest neighbor relation- occurs during the post-breeding period with ships). Values of S near zero indicate an un- males congregating early on large lakes to segregated population, whereas positive molt and females remaining with young and values indicate segregation of males and fe- later molting individually or in small groups males. Negative values indicate a tendency in marshes (see Hochbaum 1944, Oring for intersexual aggregation or pairing. Of 1964, Olson 1965, Bergman 1973). Our cor- the 13 rafts, 9 exhibited positive S values, roborated hypothesis that females occur in and the test statistics (presented only for higher proportions in creeks and small out- rafts with S > 0) suggested a significant de- lying water areas than in large water areas gree of segregation in five of these rafts during winter is consistent with both the (Table 5). We conclude that the birds in at ideas of behavioral dominance relationships least some of the rafts tended toward inter- and of different habitat preferences. The sexual segregation. This conclusion is con- tendency of females to winter farther south sistent with a general hypothesis of behav- than males may thus result partially from a ioral interactions occurring between males preference for certain (e.g., creeks and females, although other interpretations and small water areas) that may be more are possible. abundant in the south. The observation, We tested the hypothesis about the ten- however, that Canvasbacks in the small dency of females to occur on the edges of water areas in the Chesapeake Bay tend to rafts using an ad hoc procedure. For each be in worse condition than birds trapped in raft for which the concept of edge seemed open water areas, and the tests demonstrat- applicable, we drew a band of equal width ing sex-specific dispersion patterns within (two times the approximate average dis- flocks seem to support the notion of ongoing tance between neighboring birds in that behavioral interactions that can become es- raft) around the periphery of the raft. We pecially important during periods of winter then totaled the males and females occur- stress. We conclude that intersexual com- ring in this band as well as those judged to petition in Canvasbacks probably contrib- be stragglers or outliers (birds occurring utes importantly to the latitudinal differ- outside the main raft body). We then com- ences in wintering areas of the two sexes. pared the proportions of males and females Effects of this competition for food may in- in the edge area with those occurring in the clude the historical evolution of some de- raft interior, using a 2 X 2 contingency test. gree of differential habitat preference of the The concept of edge has little meaning sexes, as well as ongoing behavior that pro- when associated with “linear” rafts or with motes segregation of the sexes. widely dispersed rafts. We therefore chose seven rafts for the edge versus interior test. ACKNOWLEDGMENTS The null hypothesis of equal proportions of We thank J. Goldsberry for taking the aerial photo- the two sexes in the edge and interior areas graphs and J. Hines for computer programming sup- port. We thank the following Maryland Wildlife was rejected (P < 0.10) for six of the seven Administration personnel for their cooperation and in- tested rafts (Table 5). All seven of the tested valuable assistance with the field aspects of the study: rafts contained more females on the edge F. Compher, S. Dawson, D. Hammett, L. Hindman, R. than expected under the null hypothesis. Maldeis, and V. Stotts. Banding and recovery records We thus conclude that females in these win- were made available by the U.S.F.W.S. Bird Banding Laboratory. We gratefully acknowledge the efforts of ter rafts tended to occur on the edges of the the individuals who have banded Canvasbacks over rafts rather than in the interiors. the past several decades. Useful comments on this The observations of Alexander and Hair manuscript were provided by W. Conley, P. Geissler, (1979) together with our test results indicate R. Kirby, F. Martin. F. Percival and K. Reinecke. We thank W. Alexander; S. Gauthreaux and F. Percival for that behavioral interactions are an impor- helpful discussions about Canvashack distribution pat- tant component of competition between the terns and dominance relationships, and V. Stotts, L. sexes for winter food. We believe that in- Hindman, and M. Perry for discussions of Canvasback tersexual competition is, at least to some wintering ecology. Finally, we acknowledge the tech- degree, an active and ongoing ecological nical assistance of P. Bowley, S. Burgoyne, L. Moyer, and R. Perry. process. However, current differences in sex ratios found in particular habitats, and LITERATURE CITED even at different latitudes, could also be the ALEXANDER, W. C., AND J. D. HAIR. 1979. Winter for- result of sex-specific habitat affinities or aging behavior and aggression of diving ducks in preferences that have evolved historically South Carolina. Proc. Southeast. Assoc. Fish in response to intersexual competition for Comm. Annu. Conf. 31:226-232. ALFORD, J. R., III, AND E. G. BOLEN. 1977. Influence resources (see Selander 1966). For example, of winter temperatures on Pintail sex ratios in Tex- segregation of the sexes in Canvasbacks also as. Southwest. Nat. 21:554-556. WINTER DISTRIBUTION PATTERNS OF CANVASBACKS 415

ALLEN, A. A. 1931. The Canvasback. Bird-Lore KETTERSON, E. D., AND V. NOLAN, JR. 1978. Over- 33:347-360. night weight loss in Dark-eyed Juncos (Junco hye- ANDERSON, B. W., AND R. L. TIMKIN. 1972. Sex and m&s). Auk 95:755-758. age ratios and weights of Common Mergansers. J. KETTERSON, E. D., ANDV. NOLAN, JR. 1979. Seasonal, Wildl. Manage. 36: 1127-l 133. annual, and geographic variation in sex ratio of BELLROSE, F. C., T. G. SCOTT, A. S. HAWKINS, AND J. wintering populations of Dark-eyed Juncos Uunco B. LOW. 1961. Sex ratios and age ratios in North hyemalis). Auk 96:532-536. American ducks. Ill. Nat. Hist. Surv. Bull. 27:391- LACK, D. 1944. The problem of partial migration. Br. 474. Birds 37: 122-130, 143-150. BENNETT, J. W., AND E. G. BOLEN. 1978. Stress re- LACK, D. 1954. The natural regulation of num- sponse in wintering Green-winged Teal. J. Wild]. bers. Clarendon Press, Oxford. Manage. 42:81-86. LACK, D. 1966. Population studies of birds. Clarendon BERGMAN, R. D. 1973. Use of southern boreal lakes Press, Oxford. by postbreeding Canvasbacks and Redheads. J, LERRET, T. 1950. The sex-ratios and the proportion of Wild]. Manage. 37: 160-170. adult drakes of Teal, Pintail, Shoveler and Wigeon BOYCE, M. S. 1979. Seasonality and patterns of natural in the Netherlands, based on field counts made selection for life histories. Am. Nat. 114:569-583. during autumn, winter and spring. Ardea 38: 1-18. CALDER, W. A. 1974. Consequences of body size for LINDSEY, C. C. 1966. Body sizes of poikilotherm ver- avian energetics, p. 86-144. In R. A. Paynter [ed.], tebrates at different latitudes. Evolution 20:456- Avian energetics. Nuttall Omithol. Club Publ. No. 465. 15. Cambridge, MA. MARDIA, K. V. 1967. A non-parametric test for the bi- COHEN, D. 1967. Optimization of seasonal migratory variate two-sample location problem. J. R. Stat. behavior. Am. Nat. 101:5-17. Sot., Ser. B 29:320-342. COWARDIN, L. M. 1977. Analysis and machine map- MARDIA, K. V. 1968. Small sample power of a non- ping of the distribution of band recoveries. U.S. parametric test for the hivariate two-sample loca- Fish Wild]. Serv., Spec. Sci. Rep., Wild]. No. 198. tion problem in the normal case. J. R. Stat. Sot., COX, G. W. 1968. The role of competition in the evo- Ser. B 30:83-92. lution of migration. Evolution 22: 180-192. MARDIA, K. V. 1972. Statistics of directional data. Ac- DEGRAFF, L. W., D. D. FOLEY, AND D. BENSON. 1961. ademic Press, London. Distribution and mortality of Canvasbacks banded MARTIN, E. M., AND S. M. CAnmY. 1977. Population in New York. N. Y. Fish Game J. 8:69-87. ecology of the Mallard. IV. A review of duck hunt- ERSKINE, A. J. 1971. Buffleheads. Can. Wild]. Serv. ing regulations, activity, and success, with special Monogr. Ser. No. 4. reference to the Mallard. U.S. Fish Wild]. Serv., FRETWELL, S. D. 1968. Habitat distribution and sur- Resour. Puhl. 130. vival in the Field Sparrow (Spizella pu.siZlu). Bird- NICE, M. M. 1933. Migratory behavior in Song Spar- Banding 39:293-306. rows. Condor 35:219-224. FRETWELL, S. D. 1969. Dominance behavior and NICHOLS, J. D., AND G. M. HARAMIS. 1980. Inferences winter habitat distribution in juncos (Junco hye- regarding survival and recovery rates of winter- m&s). Bird-Banding 40: l-25. banded Canvasbacks. J. Wild]. Manage. 44:164- FKETWELL, S. D. 1972. Populations in a seasonal en- 173. vironment. Princeton Univ. Press, Princeton, NJ. NILSSON, L. 1970a. Food-seeking activity of south GAUTHREAUX, S. A., JR. 1978. The ecological signif- Swedish diving ducks in the non-breeding season. icance of behavioral dominance, p. 17-54. In P. P. Oikos 21: 145-154. G. Bateson and P. H. Klopfer [eds.], Perspectives NILSSON, L. l970b. Difficulties in estimating the true in ethology. Vol. 3. Plenum Press, New York. sex ratio of ducks from winter counts. Int. Congr. HARRISON, J., AND M. HUDSON. 1964. Some effects of Game Biol. 8:86-88. severe weather on wildfowl in Kent in 1962-63. NILSSON, L. 1976. Sex-ratios of Swedish Mallard dur- Wildfowl Trust Annu. Rep. 15:26-32. ing the non-breeding season. Wildfowl 27:91-94. HARRISON, J., AND A. MCLEAN. 1947. The effect of OLSON, D. P. 1965. Differential vulnerability of male severe weather on Wigeon. Br. Birds 40:218. and female Canvasbacks to hunting. Trans. N. Am. HOCHBAUM, H. A. 1944. The Canvasback on a prairie Wildl. Nat. Resour. Conf. 30:121-135. marsh. American Wildlife Inst., Washington, DC. ORING, L. W. 1964. Behavior and ecology of certain IVACIC, D. L., AND R. F. LABISKY. 1973. Metabolic ducks during the postbreeding period. J. Wild]. responses of Mourning Doves to short-term food Manage. 28:223-233. and temperature stresses in winter. Wilson Bull. PATTERSON, I. J. 1977. Aggression and dominance in 85: 182-196. winter flocks of Shelduck (Tadorna tudornu (L)). JENKINS, D., M. G. MURRAY, AND P. HALL. 1975. Anim. Behav. 25:447-459. Structure and regulation of a Shelduck (Tadornu PETRIDES, G. A. 1944. Sex ratios in ducks. Auk tudornu L.) population. J. Anim. Ecol. 44:201-231. 61:564-571. KENDEIGH, S. C. 1945. Resistance to hunger in birds. PIELOU, E. C. 1961. Segregation and symmetry in J. Wild]. Manage. 9:217-226. two-species populations as studied by nearest- KETTERSON, E. D. 1979. Aggressive behavior in win- neighbour relationships. J. Ecol. 49:255-269. tering Dark-eyed Juncos: determinants of domi- PIELOU, E. C. 1969. An introduction to mathematical nance and their possible relation to geographic ecology. Wiley-Interscience, New York. variation in sex ratio. Wilson Bull. 91:371-383. KETTERSON, E. D., AND J. R. KING. 1977. Metabolic RAISZ, E. 1962. Elements of cartography. McGraw and behavioral responses to fasting in the White- Hill, New York. crowned Sparrow (Zonotrichia leucophrys gum- ROBSON, D. S. 1968. The effect of ties on critical val- b&i). Physiol. Zool. 50:115-129. ues of some two-sample rank tests. Biometrics KETTERSON, E. D., AND V. NOLAN, JR. 1976. Geo- Unit, Cornell Univ., Ithaca, NY. BU-258-M. graphic variation and its climatic correlates in the ROSENZWEIG, M. L. 1968. The strategy of body size sex ratio of eastern-wintering Dark-eyed Juncos in mammalian carnivores. Am. Mid]. Nat. 80:299- (Junco hyemulis hyemulis). Ecology 57:679-693. 315. 416 JAMES D. NICHOLS AND G. MICHAEL HARAMIS

SALOMONSEN, F. 1968. The moult migration. Wild- TRAUGER, D. L. 1974. Looking out for the Canvas- 19:5-24. back. Part 1. Ducks Unlimited 38: 12-15, 30, 36. SELANDER, R. K. 1966. Sexual dimorphism and dif- VON HAARTMAN, L. 1968. The evolution of resident ferential niche utilization in birds. Condor 68: 113- versus migratory habit in birds. Some consider- 151. ations. Ornis Fenn. 45:1-7. SMITH, J. D. 1946. The Canvasback in Minnesota. Auk WELLING, C. H., AND W. J. L. SLADEN. 1979. Can- 63:73-81. vasback sex ratios on Rhode and West Rivers, SNEDECOR, G. W., AND W. G. COCHRAN. 1967. Statis- Chesapeake Bay, 1972-78. J. Wildl. Manage. tical methods. Iowa State Univ. Press, Ames, IA. 43:811-813. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry. Free- WITHERRY, H. F., F. C. R. JOURDAIN, N. F. TICE- man, San Francisco. HURST, AND B. W. TUCKER. 1939. The handbook SOWLS, L. K. 1955. Prairie ducks. Wildlife Manage- of British birds. Vol. 3. Witherby, London. ment Inst., Washington, DC. STEWART, R. E., A. D. GEIS, AND C. D. EVANS. 1958. Migratory Bird and Habitat Research Lab., U.S. Fish Distribution of populations and hunting kill of the und Wildlife Service, Laurel, Maryland 20811. Ac- Canvasback. J. Wildl. Manage. 22:333-370. cepted for publication 29 October 1979.

Condor, 82:416 @ The Cooper Ornithological Society 1980

RECENT PUBLICATIONS

North American Birds of Prey.-Text by William to the birds of the region. Aimed at birders, it opens Mansell, paintings by Gary Low. 1980. William Mor- with an explanation of classification and naming. Fol- row & Co., Inc., New York. 176 p. $29.95. This is an lowing chapters consider the annual cycle, reproduc- attractive, large-format book about 27 of the most tion and nesting, molts and , migration, and widely distributed species of North American hawks the effects of habitat alteration on bird life. The second and owls. Since the author and the illustrator are both half of the book is largely devoted to the types of nat- Canadians, the selection is limited to species that occur ural habitats in the Monterey Bay area and the birds in Canada. The justification for yet another book on associated with each. It closes with a history of orni- this often-covered ground is to exhibit Mr. Lows’ fine thology in the area. The style is informative and sci- paintings and sketches, which interpret careful obser- entific, yet lucid and non-technical. While the book vation with a meticulous technique. In a short essay will chiefly interest those within its region, it can be on each bird, Manse11 interweaves the basic habits read with profit by birders elsewhere along the Pacific with his personal experiences and thoughts, but his coast. Photographs; bibliography; appendices; index. admiration for his subjects is marred by anthropo- morphism. Appendices summarize diets, describe fal- conry, and discuss the effects of pesticides on falcon- The Life of the Hummingbird.-Alexander F. iformes, all too briefly. A book for the coffee tables of Skutch. Illustrated by Arthur B. Singer. [1980]. Crown those who enjoy birds of prey and bird art. Publishers Inc., New York. 95 p. $15.95. Although not so marked, this is a reprint of a work first published in 1973 (and reviewed in Auk 92:397-398). It offers a Call Collect, Ask for Birdman.-James M. Vardaman. good overview of the natural history of hummingbirds, 1980. St. Martins’ Press, New York. 256 p. $10.95. Bird- based on the literature as well as the authors’ consid- ing has become highly organized and competitive for erable field experience. The writing is clear, graceful, some of its enthusiasts. One of its top practitioners here and non-technical. Singers’ many color paintings are recounts his record-breaking attempt to see 700 species attractive and well reproduced; they do not merely of North American birds within one year. He very near- present the birds but show something about their hab- ly succeeded, thanks to careful planning, luck, help its. The book is aimed at general readers, yet it can be from a multitude of other birders, a sustained effort, enjoyed by those who are already acquainted with the and a large expenditure of money. The fast-moving ad- subject. References, index. venture story tells of strategy sessions and sudden trips, of surprise finds and disappointments. A book for devotees of the sport. Photographs, appendices, charts, Journal V, 1979-1980.-The World Pheasant Asso- sighting ledger, map. ciation. 132 p. Paper. f 5.50. Source: Secretary, W.P.A., Daws Hall, Lamarsh, Bures, Suffolk, CO8 5EX, En- gland. This volume offers 11 articles about various gal- The Bird Year/A Book for Birders/with special ref- linaceous birds in nature or captivity. The studies of erence to the Monterey Bay Area.-John Davis and pheasants in the Himalayas and southeast and of Alan Baldridge. 1980. The Boxwood Press. 224 p. Pa- megapodes in New Britain yield useful census data per. $5.95. Source: The Boxwood Press, 183 Ocean and information on habits. The volume includes an an- View Blvd., Pacific Grove, CA 93950. Here is a region- nual review of the W.P.A. book reviews, and a list of al bird book on a new plan: not a catalog of species but recent periodical literature. Maps, diagrams, and pho- an introduction to avian biology, with special reference tographs, many in color.