Repeatability is a useful tool for the population geneticist or genetical ecologist, but several papers have carried errors in its calculation. We outline the correct calculation of repeatability, point out the common mistake, show how the incorrectly calculated value relates to repeatability, and provide a method for checking published values and calculating approximate repeatability values from the F ratio (mean squares among groups/ mean squares within groups).

A problem frequently faced by researchers involved in collecting tissues for DNA isolation is the preservation of samples in the field prior to and during their transportation to the laboratory. Prevention of DNA degradation is usually achieved through freezing. As this is not always practical, we have tested the efficiency of chemical solutions containing high concentrations of salts (e.g., NaCl, EDTA, and diaminocyclohexanetetraacetate) and detergent at preserving DNA in bird tissue and blood samples stored at ambient temperature for extended periods of time. For blood samples, we recommend the use of a buffer that lyses the cells and nuclei and contains 0.01 M Tris, 0.01 M NaCl, 0.01 M EDTA, and 1% n-lauroylsarcosine, Ph 7.5. Tissue samples are best preserved as small pieces in a saline solution made of 20% dimethyl sulfoxyde, 0.25 M EDTA, saturated with NaCl, pH 8.0. DNA extracted from samples preserved in these solutions for up to 24 weeks was compared with DNA recovered from tissue samples stored at −70 °C and blood samples stored at −70 and −20 °C. Yields were similar, averaging 300 μg/0.2 g of tissue and 500 μg/50 μL of blood. Quality of DNA in terms of fragment size, ability to be cut by restriction enzymes, and ability to hybridize to radioactive probes was also similar between cryopreserved and chemically preserved samples. Yields of DNA recovered from tissue samples preserved in 70% ethanol for 6 or 11 weeks was very low and significant degradation was observed. We have also examined how DNA contained in crude avian blood samples withstands freeze–thaw cycles. We found normal yields and no significant degradation of DNA in samples that experienced up to six cycles. We encourage field researchers who refrain from preserving tissue samples because of logistical problems, such as transporting liquid nitrogen containers in the field, to consider using these solutions. Both solutions can also facilitate exchanges of samples between laboratories, and they form an alternative to storage of samples at −70 °C for laboratories and museums with limited access to deep-freezing facilities.

Survival of Darwin's finches through a drought on Daphne Major Island was nonrandom. Large birds, especially males with large beaks, survived best because they were able to crack the large and hard seeds that predominated in the drought. Selection intensities, calculated by O'Donald's method, are the highest yet recorded for a vertebrate population.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecological Monographs.Abstract. The Darwin's finches on Isla Daphne Major, Galapagos, were studied between July 1975 and June 1978. Geospiza fortis and G. scandens are residents, while G. fuliginosa and G. magnirostris are regular immigrants. The Daphne climate is unpredictably dry. The island has a simple plant community displaying marked annual and spatial variation in the foods which form finch diets. Breeding is stimulated by rain falling irregularly between January and April; G. scandens laid eggs with as little as 16 mm of rain, but G. fortis required 35 mm or more before laying. G. scandens consistently bred prior to the rains, associated with specialized exploitation of dry season Opuntia cactus flowers. The breeding system of both species was similar to that of other Geospiza species: monogamous matings on small, permanent, all-purpose territories.Reproductive output of both species varied. In 1976 single broods were produced by both species at high densities, with modes of three young. In 1977, only 24 mm of rain fell during the breeding season and G. scandens alone bred, with poor success. By 1978, G. scandens populations had declined by 66% and G. fortis by 85%. In 1978, both species laid an average of three clutches per pair, with a mode of four young per brood. During the 1977 drought, the sex ratio became skewed in favor of males in both species, and as a consequence some females bred successively with up to three different males in 1978. The skewed sex ratio retarded population recovery following the drought.The population decline was associated with a decline in seed abundance in the drought. As food disappeared, G. fortis diets widened to include a broad selection of food items, while G. scandens diets contracted to Opuntia seeds. Even in normal years, both species showed pronounced seasonal variation in feeding habits. Both fed heavily on Opuntia flowers in the late dry season, followed by a mixed diet of insects, fresh seed, and other green matter during the breeding season. After breeding, the birds fed primarily on seeds, the two species selecting different proportions of the available range of seed sizes. Overall, G. fortis displayed the feeding, breeding, and population dynamics characteristic of an ecologically generalized species, whereas G. scandens is a highly specialized species.Large variation in food supply, caused by large variation in rainfall, is responsible for the presence of only two species with breeding populations, for large fluctuations in their population sizes, and for their large clutch sizes and opportunistic breeding. Long-term studies...

Male song reflects the quality of the singe many animals and plays a role in female ch, of social and copulation partners. Eavesdropi on male-male vocal interactions is a means which females can compare different ma singing behavior directly and make immed comparisons between potential partners the basis of their relative vocal performa (1, 2). Using an interactive playback exp ment followed by microsatellite pater analysis, we investigated whether female black-capped chickadees (Poecile atricapilla) base their reproductive decisions on information gained through eavesdropping.Black-capped chickadees are socially monogamous songbirds that follow a mixed reproductive strategy in which one-third of broods include young that are not related to their social father (3). From 1999 to 2001, we assessed dominance ranks in a free-living population of chickadees at Queen's University Biological Station, Canada, to predict which males were likely to be sought for extrapair copulations (high-ranking males) and which males were likely to lose paternity within their nests (low-Fi ranking males) (3, 4). sl At the start of the breeding season, (t when male-male song contests are corn-s mon and females actively solicit copu-tf lations, we used interactive song play-al back to engage territorial male chicka-ti dees in countersinging interactions with rz a simulated intruder (5). We performed 6.0-min playback trials to dyads of neighboring high-ranking and lowranking males from the same winter flock control treatments, we mimicked natural ten rial encounters; we simulated an intruder sang submissively (Fig. 1A) with the high-n ing playback subject and sang aggressively ( 1B) with the low-ranking neighbor. In exl mental treatments, we attempted to alter ea dropping females' perceptions of their sc mates; we simulated an intruder that sang gressively with the high-ranking playback ject and sang submissively with the low-ranl neighbor. To test whether interactive playl altered the normal pattern of paternity in nests of subject males, we conducted pater analysis on blood samples collected from spring (6).High-ranking males that lost song cont with a simulated intruder lost paternity in their nests (Fig. 1C); high-ranking males that received playback simulating an aggressive intruder showed a significantly greater level of paternity loss than high-ranking males that received playback simulating a submissive intruder (control I; Fisher's exact test, P = 0.05) and a significantly greater level than a control group of high-ranking males that received no playback (control II; P = 0.05). As predicted (3, 4), we observed little A Submissive playback N 4.5: Y 4.0-) 302 2.5 U-*1-B Aggressive playback -1