Here I provide the first community-level assessment of avian elevational migration in South America. I used point counts and mist netting and focal observations of a mid-elevation mixed species flock to document patterns of avian elevational migration along a 2.7 km elevational gradient in Manu National Park, southeastern Peru.
To effectively study elevational migration, some knowledge of the annual cycles of birds is required. Despite the tropics often being perceived as aseasonal, climate varies significantly throughout the year at our study sites. Measurements of daily temperature and rainfall along the gradient reveal a cool dry season from May to August, and a warmer rainy season from August to April, with rainfall peaking in February. Patterns of monthly fruit production along the gradient are complicated, but on average fruit volume peaks in February and fruit number peaks in July. Breeding of birds begins to increase at the start of the wet season and peaks in October and November. Molt, the annual replacement of feathers, was rare during the dry season and early wet season, and increased thereafter to peak during the height of the wet season, 2-3 months after the peak in breeding, resulting in a strong correlation between molt and rainfall. Molt timing varied little along the elevational gradient, except that perhaps middle elevation birds complete their molt earlier than high or low elevation birds. Elevational migrants appear to molt more quickly than nonmigrants, suggesting a possible mechanism for the apparent differences among elevations. Feeding guild also appears to affect molt timing, with insectivores and nectarivores molting earlier than frugivores and omnivores.
Mixed-species bird flocks are a common phenomenon throughout the world, reaching their highest diversity and size in the tropics, where flocks often defend permanent territories from neighboring flocks. Through a combination of mist netting, color banding, resighting, and behavioral observations I documented the composition of one particularly large and diverse mixed-species flock from a middle elevation site at 1400 m a.s.l. Individuals of many species observed in the flock appeared to be elevational migrants, including the two most abundant species in the flock, the Deep-blue Flowerpiercer Diglossa glauca and the Streak-necked Flycatcher Mionectes striaticollis . A polyspecific subgroup originally captured in the mixed species flock was recaptured at a nearby ridge, suggesting the possibility that individuals in this subgroup maintained the interspecific bonds created on their nonbreeding grounds during part of their upslope migration, a first for a polyspecific group of terrestrial migrants.
To identify other elevationally migratory species I first needed to outline a method of analyzing abundance estimates from point count data. First I divided the year into three seasons (dry, early wet, late wet) and divided the elevational range of each species into 2 to 5 elevational zones, depending on the number of point counts I conducted within the species’ range. I estimated the relative abundance of each species in each season at two spatial levels: the elevational zone and the survey point. For each season I calculated the mean elevation of occurrence as the mean elevation of zone midpoints weighted by relative abundance. In paired seasonal comparisons I measured the shift in mean elevation of occurrence and tested for the key characteristics of a down- or upslope movement (an increase at one elevation and a corresponding decrease at another elevation) by measuring the significance of a season*elevation interaction effect in a two-way ANOVA. A combination of a significant interaction effect and a significant shift down- or upslope was taken as indication of a seasonal movement. This is the first time abundance has been used in this way to statistically test for movement along an environmental gradient. I scored each species as a migrant or resident based on the results of the point count data analysis. In addition, I scored each species using relative abundance estimates from mist netting, flock observations, and using a combination of all datasets, and averaged all available scores for each species. The cumulative score represents the combined weight of evidence in favor of designating each species as an elevational migrant or a resident. I discuss alternative hypotheses to explain the seasonal variation in density distributions, and offer comments on study design and the interpretation of my results. (Abstract shortened by UMI.)