Edward C. Murphy

Density and foraging ecology of woodpeckers following a stand-replacement fire

Throughout its geographic range, the black-backed woodpecker (Picoides arcticus) is rare and appears very similar in its foraging ecology to 2 broadly sympatric congeners, the three-toed (P. tridactylus) and hairy woodpecket (P cillosus. The purposes of our study were to test for differences in foraging ecology of the black-backed, three-toed, and hairy oodpeckers following a stand-replacement fire and to evaluate the importance of such fires to the viability of populations of the black-backed woodpecker. In boreal forests of interior Alaska, endemic population densities of three-toed woodpeckers are low (<0.1/ha), and black-backed woodpeckers are extremely rare. Following the Rosie Creek fire near Fairbanks, Alaska, in June 1983, both species increased markedly. Densities of both species briefly exceeded 0.2/ha and remained high in a 67-ha plot at the edge of the burn during the following 2 years. By December 1986, densities had declined to <0.1/ ha. Black-backed woodpeckers fed primarily on charred portions of moderately to heavily burnt spruces and almost exclusively by excavating larval wood-boring beetles (Cerambycidae). Three-toed woodpeckers fed on less-burnt spruces and foraged in and immediately under the bark; bark beetle (Scolytidae) larvae predominated in their diet. In contrast to earlier studies, our results demonstrate substantive differences in foraging sites, behavior, and diet of these 2 species. Foraging ecology of male hairy woodpeckers and black-backed woodpeckers, particularly females, was similar. In all 3 species, particularly the hairy woodpecker, females fed lower on trees and were far less numerous than males in the study area, which suggested intersexual displacement from foraging sites and habitats selected by males. In summer 1985, following initial adult emergence of the 1983 cerambycid and scolvtid cohorts, woodpeckers declined markedly and were absent by late spring 1986. Our results suggest the black-backed woodpecker is extremely specialized in its foraging niche, exploiting outbreaks of wood-boring beetles in dying conifers for only 2-3 years after fires. Consequently, this species may be particularly vulnerable to local and regional extinction as fire suppression intensifies and programs of intensive salvage logging are pursued following fires.

Intergenerational conflicts in gulls

Abstract Gulls (genus Larus) are known both to adopt and to attack and kill unrelated offspring wandering through their territories. The tendency of adult gulls to accept unrelated offspring appears to counter standard Darwinian natural selection, and has been interpreted as kin selection, altruism, or reproductive error. Another alternative, not generally advanced, is that there is an intergenerational conflict in which gull chicks receiving inadequate parental care seek parental care from unrelated adults, and the adults try to prevent this from happening. In this study, chicks that abandoned their natal territories were being fed at low rates or were growing slowly compared to siblings, or were the recipients of siblicidal aggression. Of the chicks that abandoned their parents, around 20–30% became successfully adopted, and a similar percentage were killed by conspecific adults. Selection should favour the chicks winning the conflict, since if they are not adopted they will perish, whereas the cost to an adult that adopts an unrelated chick will be measured in lost time and energy, which may have little or no effect on future reproductive potential. The concept of an intergenerational conflict in which adults care for unrelated offspring may be of general applicability to those species that have precocial offspring.

High annual variability in reproductive success of kittiwakes (Rissa tridactyla L.) at a colony in western Alaska

(1) All aspects of reproductive performance of kittiwakes (Rissa tridactyla L.) fluctuated markedly in 1975-89 at a colony of about 10000 nesting pairs in the north-eastern Bering Sea, Alaska. (2) In breeding seasons following cold springs, breeding was delayed, fewer nests were built, fewer nests contained eggs, and clutch size, hatching success and the number of chicks fledging per nest were reduced. Complete reproductive failures followed the three coldest springs in the 15-year period. (3) Reproductive success (chicks fledged per nest) was highest following moderately warm springs; however, growth rates of chicks and fledging success were extremely low following the warmest spring on record. (4) Spring air temperatures were highly correlated with break-up of sea ice and these factors probably influenced seasonal warming trends in the sea water and the consequent availability of prey such as sandeels (Ammodytes hexapterus Pallas) near the colony. (5) During this 15-year period, interannual variability in spring air temperature was pronounced in comparison to that during the previous 68-year period; however, simulations using May temperature as a predictor of reproductive success suggested that high annual variability in reproduction has occurred throughout this century. (6) Numbers of adult-plumaged birds on the cliffs in mid-season were markedly lower and more variable in years when few nests were built, suggesting that relatively few experienced adults breed and consistently are present at the colony in years when overall reproductive performance is poor.

The paradox of pelagic food webs in the northern Bering Sea—I. Seabird food habits

Two distinct environmental settings in the Bering Strait region of the northern Bering Sea lead to characteristic pathways of energy flow through primarily pelagic food webs to avian consumers. In Norton Sound, a large, shallow embayment on the northeastern coast, the physical environment is dominated by the discharge of the Yukon River and by a large seasonal temperature signal. Seabirds breeding at Bluff, the largest colony in Norton Sound, number in the order of 5 × 104 and require 1.2 × 106 g C d−1. Two piscivorous species constitute the bulk of all seabirds there and are supported by a pelagic food web typical of the coastal zone of the Bering and Chukchi seas. This food web also is present around St. Lawrence Island, on the northwestern shelf, and is important to at least one species of seabird there. In addition, and generally more important, St. Lawrence Island is in a biologically rich environment resulting from the northward flow of water that originates along the continental shelf break of the Bering Sea. This flow apparently accounts for the unexpected presence of oceanic zooplankton and a diversity of forage fishes on the shallow northern shelf that support an abundant and taxonomically rich avifauna. In comparison to Norton Sound, breeding seabirds on St. Lawrence Island number in the order of 2 × 106, with planktivores consuming about 8 × 106 g C d−1 and piscivores consuming about 16 × 106 g C d−1.

Seasonal Variation in Reproductive Output of House Sparrows: The Determination of Clutch Size

Seasonal changes in reproductive output of house sparrows, Passer domesticus, were studied at 2 localities: Lawrence, Kansas, USA and Calgary, Alberta, Canada. In both localities the most productive clutch (6 eggs) was larger than the modal clutch (5 eggs). Most variation in clutch size could be attributed to within-individual seasonal changes. Laying of 6-egg clutches was generally restricted to the middle of the breeding season. Clutch size changes were not consistently synchro- nized with trends in predictable environmental variables such as photoperiod and temperature. Be- cause seasonal trends in clutch size and fledging success were poorly correlated, there is no evidence that seasonal changes in clutch size represent adaptations to maximize the number of adequately nourished offspring. Although clutch size is ultimately determined by contribution to future genera-

Synchrony in egg-laying and reproductive success of neighboring common murres, Uria aalge

Abstract Common murres (Uria aalge) are highly colonial; pairs often breed at the highest possible densities, in bodily contact with neighbors. At Bluff, a colony in western Alaska, we tested for synchrony in egg laying at various spatial scales and found little evidence for higher synchrony, either within study plots of 15–195 pairs, or within subplots containing several pairs, than among plots in a 5-year study. Egg laying of neighbors generally was more synchronous than expected based on overall frequency distributions in laying dates, however. Breeding success was positively correlated with the number of breeding neighbors and the number of neighbors tending eggs or nestlings at the time of egg laying. Breeding success of pairs with neighbors was positively related to the breeding success of neighbors. Pairs that produced eggs synchronously with at least one neighboring pair had higher success than those that began breeding either before or after their neighbors. Most reproductive failures at Bluff are due to accidental egg loss and predation on eggs by common ravens, Corvus corax, soon after laying. By occupying space where a raven might otherwise land and defending their own eggs, active breeding neighbors locally reduce the probability of egg predation. Active breeding neighbors also are less likely to flush and accidentally dislodge nearby eggs when disturbed than are nonbreeders. Murres breeding synchronously with neighbors have the highest assurance of the presence of active breeding neighbors both at the time of egg laying and throughout their reproductive attempts. Groups of neighboring murres can be considered small “selfish herds,” demonstrating by-product mutualism through their continued presence and defense of their own eggs and nestlings. Despite the advantages of breeding synchronously with neighbors, early breeding may often be favored, however.

Seasonal declines in duration of incubation and chick periods of Common Murres at Bluff, Alaska in 1987-1991

--The Common Murre (Uria aalge) produces a single-egg clutch, and the female may replace the egg if it is lost. Even at breeding colonies in highly seasonal environments, egg laying may span several weeks. For example, the mean range of egg-laying dates was 41 days in a five-year period at Bluff, a coastal colony in western Alaska. In each of those five years, incubation periods of eggs were negatively related to laying date (i.e. incubation periods of eggs laid later in season were shorter). This is the first documentation of a seasonal decline in the incubation period of this species, and one of the few for birds in general Analyses of seasonal changes in weather patterns indicated that the seasonal shortening of the incubation period cannot be explained by changes in environmental conditions. The duration of the chick period also declined seasonally in all five years, similar to results reported elsewhere. Overall, duration of the period between egg laying and fledging averaged about six days less for the latest breeding pairs than for the earliest breeding pairs. There is no evidence of a seasonal decline in hatching or fledging success at this colony. Consequently, these results indicate that murres breeding later in the season at Bluff reduce the duration of both the incubation and chick periods well in advance of seasonal deterioration of environmental conditions. Chicks develop more rapidly than embryos in eggs, and chicks at sea likely grow and develop faster than chicks at the colony. As the time remaining for completion of development decreases through the summer, quicker hatching of eggs and sea-going of chicks will be advantageous. Formation of sea ice throughout the region in late fall likely is the primary time constraint selecting for accelerated developmental patterns of eggs and chicks of late-breeding murres at Bluff. In highly seasonal environments, seasonal time constraints may influence most or all phases of avian breeding cycles. Received 18 April 1994, accepted 12

Breeding Ecology of House Sparrows: Spatial Variation

tury ago. The geographic patterns in body size and dimensions, and those in plumage characters (Johnston and Selander 1964), agree with Bergmanns, Allens and Glogers ecogeographic rules, respectively. Geographic trends in thermal tolerances (Hudson and Kimzey 1966), lipid deposition (Blem 1973), insulation and metabolism (Blem 1974, Kendeigh and Blem 1974, Kendeigh 1976) have been documented also.

Dall Sheep Demography in McKinley Park and a Reevaluation of Murie's Data

Population numbers, age-sex composition, and the male age structure of Dall sheep (Ovis dalli) were studied in Mount McKinley National Park, Alaska, in the summer of 1972. Age of living males was estimated either by counting horn annuli from photographs taken in the field or, if annuli could not be discerned in the photographs, on the basis of horn size. The correlation of age with horn size first was determined for the former group by multiple regression analysis; the resultant equation then was used to estimate age in the latter group. Composition ratios and the sample of males whose ages were estimated were used to establish the age structure of the living male segment of the population. The results demonstrated that the age structure had not been stationary over the past several years. Differences in numbers in each age-class were due to marked variability in initial size and in mortality rates of the cohorts comprising the population. Quantitative analysis of skull data collected by A. Murie from 1937 to 1941, which have been used widely in the construction of life tables, demonstrated that the age structure was not stationary during the period the skulls accumulated. Analyses of changes in population size and of variation in natality rates suggested that the pattern of the life tables was due not only to variation in age-specific survivorship but also to marked variation in initial size of the actual cohorts comprising the samples. Natality was inversely correlated with snowfall and density. First-year survivorship was not correlated with natality rate, indicating that the factors affecting these demographic parameters varied substantially from year to year. J. WILDL. MANAGE. 40(4):597-609 The age structure of Dall sheep has played an important role in discussions of mortality patterns in mammals (e.g., Deevey 1947, Kurten 1953, Taber and Dasmann 1957, Caughley 1966) and is frequently cited in ecology texts (e.g., Emlen 1973:243-244, Pianka 1974:70-72, Ricklefs 1973:393-395). These discussions are based on Muries (1944) Mount McKinley National Park skull data. Murie determined age at death by counting horn annuli; Geist (1966) demonstrated that annuli are accurate indices of age of males but provide only a minimum estimate of female age. Assumptions in the construction of time-specific (Southwood 1966:316) or vertical life tables are basically two: (1) there is an equal (or known) probability of finding remains of an individual of any age-sex class, and (2) the age structure is stationary (Caughley 1966). It is known, however, that there is differential perishability of skulls depending both on sex and age (e.g., Buechner 1960:84, Geist 1971:293), and that the areas searched may not represent equally the home ranges of all age-sex classes (Geist 1971:294). For example, skulls of young individuals of both sexes and of adult females are extremely perishable and less easily found than those of adult males. Unless two or more sources of data are available, the second assumption, that the age structure is stationary, cannot be evaluated critically (Caughley 1966). Murie (1944) divided the skulls he collected into two groups according to the time of death (pre-1937 and 1937-41); therefore, his data could be used to determine if the second assumption was met during the two intervals in which the skulls accumulated. There remain two problems with the skull remains data: the cohort to which a dead individual 1 Present address: Museum of Natural History and Department of Systematics and Ecology, The University of Kansas, Lawrence 66045. 2 Present address: Alaska Department of Fish and Game, 1301 College Road, Fairbanks 99701. J. Wildl. Manage. 40 (4):1976 597 This content downloaded from 207.46.13.124 on Wed, 22 Jun 2016 05:12:10 UTC All use subject to http://about.jstor.org/terms 598 DALL SHEEP DEMOGRAPHY -Murphy and Whitten belonged cannot be determined because the time (year) of death usually is unknown, and therefore the data show little indication of numerical change in the population. Life tables constructed from such data are composed of synthetic, rather than real, cohorts (Spinage 1972) and indicate generalized survivorship patterns of a population over an indefinite time span. If one is to study either the mortality patterns of an actual cohort or the age structure of a population at any one time, it is necessary that a different technique of gathering mortality data (d,) or survivorship data (l1) be developed. The objective of this study was to develop such a technique to estimate age of living males from horn characteristics and to use these data to evaluate the male age structure of Dall sheep in McKinley Park in the summer of 1972. Over the past few decades much information on demography of Dall sheep has been gathered in the study area (Murie 1944, 1946, 1961, Sumner 1948, Nichols and Erickson 1969, P-R Proj. W-15-R-3 and W17-1, Work Plan N, Jobs 3, 4, 5, 6, and 7, Alaska Dept. Fish Game, Juneau). We also culled data from several unpublished censuses (Branges 1955, Erickson 1968, Gross 1957, Haber 1972, Murie 1951, Nancarrow 1952, Nichols 1968, Peters 1953, Prasil 1966, Reid 1957, Smith 1972, Sumner 1952) that are filed as reports in National Park Service files in Anchorage. The trends shown by the data provide perspective on the life tables constructed from Muries (1944) data, on the present age structure, and on the correlation of fluctuations in natality and mortality with variations in disease, predation, and weather. This report is a portion of an M.S. study (Murphy 1974) that was funded by the National Science Foundation and the Society of the Sigma Xi. G. Murphy, T. Gregson, and G. Gregson aided in fieldwork. R. D. Guthrie, S. F. MacLean, Jr., D. R. Klein, V. Geist, R. S. Hoffmann, and W. Heimer gave valuable criticism of the manuscript; S. J. Harbo and J. Dunlap provided much help in the data analyses; and A. DiCiccio and J. Robertson prepared several of the figures. G. Haber generously shared his unpublished data and R. Prasil searched through National Park Service files to find many reports concerning Dall sheep in McKinley Park. We thank the National Park Service for its cooperation.

Activity in a hunted and an unhunted herd of Dall sheep

Abstract In three sampling areas (two unhunted, one hunted), we compared activity patterns of Dall sheep rams (Ovis dalli) of different horn size during the rut. In all three areas rams with horns 3 4 - curled and smaller rams were less frequently seen in ewe bands as the rut progressed. In one area where we documented activity budgets, we observed equivalent budgets for ewes and rams with horns 1 4 - curled . Older rams spent less time foraging than young rams, except that rams with horns 5 8 – 7 8 curled spent less time foraging than full-curl rams late in the rut. Display frequencies of 7 8 - curl rams were similar to those of full-curl rams and considerably higher than those of smaller-horned rams early in the rut. These results suggest that the rut is more energy costly for old rams than for young rams, but that it may be most costly for the next-to-largest horn-curl class rams (e.g. 7 8 - curl rams in unhunted populations) rather than the dominant rams (full-curl and larger). Rams in the next-to-largest horn-curl class in the hunted herd, however, did not acquire the higher display rates (e.g., 3 4 - curl rams had equally low display rates in both the hunted and unhunted herds).