In addition to the migration monitoring research at MBO, banding activities provide the opportunity for any number of other potential studies.  Listed below are brief abstracts of projects that have already been investigated to some extent, most of which have been undertaken as term projects by undergraduate wildlife biology students at McGill University.  Below them are a number of other research questions of interest that we have not yet had an opportunity to pursue.  We welcome anyone interested in these questions to contact us for more information on how to get involved, and also encourage anyone to suggest additions to the list. 

Past research topics:

Assessment of flank colour on American Redstarts
McGill undergraduate student project by Christine Barrie and Meaghan McDermott

The American Redstart (Setophaga ruticilla) is a migratory member of the wood-warbler family Parulidae. Young males display delayed plumage maturation, appearing relatively similar to females until after their prebasic molt (approximately July the year after hatching). There have been a number of different hypotheses proposed to explain the female-like plumage of HY and SY male American Redstarts and the possible evolutionary advantages of it, such as Female Mimicry Hypothesis, Delayed Plumage Maturation Hypothesis, and Juvenile Mimicry Hypothesis. McGill Bird Observatory recorded flank colour of American Redstarts during the 2009 fall migration by comparison with the Naturalist's Color Guide. We hypothesized that flank colour proportions of HY and SY American Redstart males would be similar to flank colour proportions of female American Redstarts of all ages. A chi-square analysis was used to compare the proportions of HY American Redstart male flank colours to the proportions of all American Redstart female flank colours. A total of 56 individuals were included in the analysis; of them, 36 were HY male and 20 were female. Only three different colours were recorded for the males: the majority of males (66.7%) were orange-yellow. Eight of the males (22.2%) were spectrum orange and four (11.1%) were spectrum yellow. Although fewer females were caught and recorded, there was more variability in the females' flank colours, with a total of six different colours observed. Nine of the thirty-six (25%) had spectrum yellow flanks, four (20%) had buff yellow and another four (20%) had straw yellow. One (5%) female had trogon yellow flanks, another one (5%) had sulfur yellow flanks and one (5%) female had orange-yellow flanks. Despite this variability, flank colour proportions between HY males and all females were significantly different c2(6, n = 56) = 39.71, p = 0.05. Only 4 out of the 36 males captured showed plumage similar to females. We documented HY males with orange flanks, which generally only occurs in ASY males. Our results do not discount hypotheses explaining the advantages of delayed plumage maturation, as HY and SY males may still resemble females on the basis of their similar back colours. Although our results show that flank colour of HY and SY males is not consistently similar to that of females, there is no dispute that the dullish brown colour of HY males' backs resembles that of females. Although much literature has been published speculating about the evolutionary advantages of HY and SY American Redstart males having a female-like plumage, there is clearly more research needed before evaluating these hypotheses.

Temporal correlations among fall migrants and influence of observer effort on abundance estimates
McGill graduate level statistics student project by Kristen Keyes and Demetrios Kolibiris

Avian migration is traditionally defined as a regular, seasonal, large-scale, movement of a population twice a year between fixed breeding and non-breeding areas. The majority of studies focusing on migratory cues have been conducted in the laboratory, but data from migration monitoring stations offer a valuable alternative. The Canadian Migration Monitoring Network (CMMN) found predominantly declining population trends for all guilds examined in both spring and fall. Unexplained patterns of change (positive and negative) in abundance over time were detected at some stations. The analysis noted that while such changes may reflect real overall population changes, the extent to which they reflect changes in habitat, personnel, effort or methodology needs to be better understood. These findings have significant implications for the conservation of avian diversity in Canada, but more information is needed regarding the time each species may spend on its breeding and wintering grounds in order to develop effective management plans. This may be most feasible by identifying which species respond to endogenous or weather-related cues for migration, and when they are expected to show annual northerly/southerly movements. As an initial step to that end, the first objective of this study was to investigate whether peaks and troughs of species’ daily capture rates (i.e. banding data) are correlated across the 13-week fall migration season. By extension, the second objective was to determine to what extent effort (i.e. Coverage Code or CC) is correlated with species richness (i.e. DET). Five species were selected for this study based on their classification as boreal breeders, nocturnal migrants, and their consistent occurrence within the top 10 most commonly banded species during 2006-2008 fall seasons at McGill Bird Observatory (MBO): American Redstart (Setophaga ruticilla), White-throated Sparrow (Zonotrichia albicollis), Magnolia Warbler (Dendroica magnolia), Nashville Warbler (Oreothlypis ruficapilla), and Ruby-crowned Kinglet (Regulus calendula). Fall data were collected from August 1 to October 30 in 2006-2008. We used a temporal cross-correlation analysis to determine the correlation between each pair of species in each year, and between the DET and CC in each year. American Redstart and Magnolia Warbler were strongly cross-correlated across years, whereas White-throated Sparrow was strongly cross-correlated with Nashville Warbler and Ruby-crowned Kinglet. These two groups likely respond to different migratory cues since there is little overlap between their wintering ranges. This is consistent with other studies which have found that the migration regulation systems employed by any species may be ultimately determined by wintering latitude and associated environmental cues. We also found strong evidence that the Daily Estimated Total is highly correlated with the Coverage Code, or with the level of effort by observers and banders on the same day. Since each migration monitoring station in the CMMN uses a similar code, it is likely the recent species declines across Canada are not a reflection of decreased monitoring effort, and are likely due to real overall population changes, changes in habitat, and/or changes in methodology. We concluded that at MBO, migrants that display overlapping wintering ranges also respond to similar migratory cues. In particular, American Redstart and Magnolia Warbler over-winter in the tropics, and appear to display characteristics of ‘calendar’ migrants, thus most likely relying on endogenous cues for migration. White-throated Sparrow, Nashville Warbler and Ruby-crowned Kinglet, on the other hand, over-winter primarily in temperate North America, and are best described as ‘weather’ migrants that rely on ambient weather conditions for migration. However, based on the findings of this as well as previous studies, it is likely that many species respond to a combination of endogenous and ambient cues, in particular the White-throated Sparrow and Nashville Warbler. A lack of significant species pair-wise cross-correlations in 2007 suggests yearly variability in cue-response for migrating birds.

Influence of weather on the efficiency and accuracy of passerine migration banding studies
McGill undergraduate student project by Katleen Robert (April 2008)

Significant declines in many passerine species have been reported in recent years. Banding studies are a frequently used and efficient method to monitor fluctuations in population size. However, external factors such as weather can influence numbers of individuals caught, and thus bias the results. In this study, we investigated two potential ways by which weather might influence banding success. First, certain weather factors during the night preceding banding operations may favour migration and increase the number of birds staging within the study area. Second, environmental conditions during the sampling period may impede or enhance the catching of birds with mist nets. We used banding records from fall 2005-2007 collected by staff at McGill Bird Observatory, and examined them for potential associations with weather measurements taken hourly by Environment Canada. Birds were grouped into guilds to facilitate the analyses. In total, 273 sampling days were considered, with 10,402 birds banded and 139,165 birds observed on site. We used multiple linear correlations to determine the relative importance of each weather variable and compared the models obtained using Akaike’s Information Criterion. When on-site abundance was assessed, the average temperature of the previous night consistently appeared as the variable with the most explanatory power. When temperature decreased, the number of birds on site increased. No environmental variables influenced the number of birds caught in 2005 and 2006, but temperature was found to be significant in 2007. These results suggest that biases associated with environmental conditions may not affect this sampling method as significantly as previously believed. This supports claims that banding studies are effective and reliable tools which have the ability to contribute meaningfully to conservation initiatives.

Differential timing of migration in passerines in spring and fall
McGill undergraduate student project by Jacinthe Daprato (December 2007)

Differential timing of migration takes two forms: protandry in spring and protogyny in autumn, In spring, protandry occurs when adults and males arrive earlier than females and immatures on the breeding grounds. In autumn, protogyny occurs when females and immatures depart before males and adults from the breeding grounds. There are several direct and indirect hypotheses which attempt to explain protandry and protogyny. The least-studied one, at least for avian species, is the susceptibility hypothesis, which stems from Bergmann’s Rule and states that because most males are larger than females, males are better adapted to survive the cold, harsh weather conditions on the breeding grounds in early spring and late autumn. Thus, adverse environmental conditions are associated with higher levels of protandry and protogyny. We predicted that migratory arrival and departure dates would differ between the sexes, with males arriving earlier in spring at McGill Bird Observatory and females leaving earlier than males in fall. We also predicted that the degree of protandry and protogyny among species would be positively related to sexual size dimorphism. Thus, as the level of dimorphism increases, larger differences in arrival/departure dates between the sexes would be observed. Data from MBO’s 2005-2006 spring and fall seasons were used to address these hypotheses. A total of seven species in spring and 14 species in fall were analyzed from four different families (Cardinalidae, Icteridae, Parulidae and Regulidae). Median arrival and departure dates were calculated for each of the four seasons; the difference between male and female median dates were calculated to quantify the degree of protandry or protogyny for each species. We chose wing chord as a measure of sexual size dimorphism and used simple linear regressions to assess the relationship between the degree of protandry and sexual size dimorphism. The Ruby-crowned Kinglet was the only species to show protandry in spring in both years, while the Common Yellowthroat showed protandry in the spring of 2006. In autumn of 2005 and 2006, Yellow-rumped Warblers and Ruby-crowned Kinglet showed protogyny. None of the analyses showed a statistically significant relationship between sexual size dimorphism and the degree of protandry in spring or the degree of protogyny in fall. Inconsistencies between our preliminary results and previous findings may be explained by sample size (n = two years), weather, analytical methods and sex ratios, though the latter two are less likely. Additional studied with larger sample sizes are required to adequately address these questions and to determine whether a changing climate might affect protandry and protogyny.

Net avoidance in passerine birds
McGill undergraduate student project by Michael Mayerhofer (May 2007)

Mist-nets are frequently used to assess or monitor avian populations because they greatly reduce observer bias and are non-selective when compared with visual or aural census methods. However, various factors may influence the capture rate and variety of species caught, including weather, time of day, net location, net tension, habitat structure, net avoidance and species-specific behaviour (e.g. territoriality and flight patterns). Very few studies have examined the impact of bird activity, or the total movement of birds within an area, and net avoidance, or specific modifications in behavioural patterns of birds due to the presence of mist-nets which cause a decrease in capture rates. Net avoidance can be broken down further into passive net avoidance (avoiding an area with a net) and net evasion (a sudden shift in flight patterns). The objective of this project was to determine evasion rates with respect to species and weather. Three nets were observed from behind a blind during 30 to 45-minute periods. Net evasion due to flight pattern changes were recorded, as were escaped birds, perching birds and successful captures. Wind speed and shading of the net were also recorded. Principle Components Analyses were used to assess the importance of wind and shade for several species groups. Though differences were found in evasion rates in sun vs. shade, variation among species precluded identification of any significant trends.  For example, sparrows had lower evasion rates in the sun than in the shade; the opposite was true for orioles. Overall, swallows showed the highest evasion rates, icterids had average rates and warblers showed the lowest evasion rates. Overall, no significant effects of wind and shade were found on capture rates, which may be due to habitat, net location and exposure, and overall species composition. Unlike other studies, we found substantial variation in evasion rates among species groups, which has implications for the assessment of relative abundance of birds within an area.  Click here for an expanded version of this report.

Refining techniques for ageing and sexing of Black-capped Chickadees
McGill undergraduate student project by Limoilou-Amélie Renaud (December 2006)

Discriminant analysis of morphometric measurements has been used for determining the sex of monomorphic bird species. Such an analysis of physical characteristics of Eastern Black-capped Chickadees (Poecile atricapillus atricapillus) provided a useful means to sex these birds, as compared to single plumage or morphological characters. Staff at McGill Bird Observatory collected information on mouth colour, mass and wing length of all individuals captured, and took photographs of bibs and caps of many individuals. Measurements from museum specimens (35 males and 37 females) were used to produce a linear function which was applied to a sample of 66 chickadees banded at MBO and successfully classified 79.2% of individuals to the correct sex using a combination of wing length, weight, and tail length. Wing length contributed most to the classification by sex, but validation with sexual characters is needed before it can be used on unknown individuals.  The colour of the roof of the mouth is variable in Black-capped Chickadees, and in some other species this can provide a helpful guide to age.  In this test, over 70% of HY/SY chickadees had a plain white roof lining, while over 70% of AHY/ASY individuals had a dark chevron on the roof.  While there was too much variability in this test to suggest that this characteristic is reliable, there nonetheless appears to be a tendency for difference by age which merits further investigation. Methods for quantifying bib and cap shapes based on photographs and museum specimens were not successful; these methods are therefore still needed for refining chickadee ageing and sexing in the field. Click here for an expanded version of this report.

Current research projects:
•  Refining techniques for ageing and sexing of American Goldfinches

  Variations in the timing and extent of moult in Song Sparrows
•  Assessing whether the extent of white on Magnolia Warbler tails can be used to determine sex

Potential research topics:
  Nest attendance using data loggers in nest boxes
  Revised ageing and sexing guides using banding data, photos
  Analyzing the variation in migrant American Robins in relation to geographic origin
  Comparing isotopic signatures and plumage/moult variation between local breeders and migrants
       for common species such as Song Sparrow, Yellow Warbler, and Red-winged Blackbird
  Importance of nest placement for different species
  Seasonal patterns of occurrence (by species / age / sex)
  Influence of local weather conditions on banding success (and monitoring effectiveness)
  Long-term indices of breeding success
  Analyzing the variation in Baltimore Oriole plumage and relating it to moult


2002- The Migration Research Foundation Inc.