David Brown's Research

My research addresses questions about the behavioral, population, and physiological ecology of Neotropical migrant and tropical forest birds at multiple spatial scales. To better manage and conserve populations of forest birds, we need to improve our understanding of how environmental factors affect basic population processes.  Along with my collaborators, I use basic field research and sophisticated laboratory techniques focused on individuals to derive empirical estimates of population processes.  These estimates are then fed into computer-based models along with geographic information about distributions and climate patterns to simulate and predict population patterns at continental scales.

 Population Ecology 

Most of my field research has been conducted during the winter non-breeding season in Louisiana and Jamaica. This was traditionally thought of as a ‘holding-over’ period, during which they simply needed to sustain themselves until the next breeding season.  However recent work suggests that events occurring on the winter grounds can carry over and affect individuals and populations in subsequent seasons (Marra et al. 1998, Norris et al. 2004).  My dissertation research  focused on untangling the mechanisms by which food availability influences over-winter success and thus contributes to individual fitness and population limitation.  I am approached this problem with observational data of environmental conditions, and Ovenbird (Seiurus auracapilla) behavior and body condition in replicated plots where food availability has been experimentally manipulated.  Specifically, I developed a statistical model to link the following:

                                                 

The annual dry season (Jan- Apr) throughout much of the Caribbean and Latin America, including Jamaica, coincides with the end of the over-wintering period and the onset of migration for many long-distance neotropical migrants.  In many habitats occupied by high abundance of migrants, the decreased precipitation and leaf abscission associated with the late winter dry season results in reduced arthropod activity; at a time when migrants need to accumulate fat in preparation for migration.  Thus, this period is a source of environmental and life-history related physiological stresses that have important demographic consequences.   

On top of this natural system, we maintain small (1 ha) replicate plots of control (N = 4), food reduction (N = 5) and food supplementation (N = 6) treatments. Food is reduced through one-time broadcast distribution of AMDRO ant bait, and supplemented with scattered piles of sliced citrus fruit.  Arthropod surveys revealed a strong, linear response by ants and other arthropods in both treatments (see fig. below),  indicating that we successfully manipulated food availability at the plot level.

 

 

 

 

 

 

 

 So far we have been able to support direct effects of climate on arthropod availability, and strong effects of food availability on foraging behavior and body condition.  We continue to investigate the other pathways:

- Ovenbirds are monomorphic and sexing is ongoing using a simple molecular technique, so we    currently don’t know how behavior and physiology vary between males and females.

- Departure timing (as determined by radio-telemetry) appears to be highly variable among individuals; resolution of the role of food is likely dependent on controlling for sex- so again we’re working on this one.

- We need additional years of data to determine annual survivorship: work will continue this winter. 

Continental Patterns

In our rapidly changing world, we need powerful and accurate predictive tools to inform conservation planners about continental scale patterns.  We plan to use the empirical annual survivorship estimates from different food treatments to simulate population changes in response to historic variation in environmental conditions.  Modeled population estimates will then be compared to observed population changes as derived from national Breeding Bird Survey data. After adjusting the model to better fit observed data we can simulate the effects of global climate change on Ovenbird population size.

We are also approaching these large scale questions by statistically modeling correlative patterns of breeding season population change in response to winter climate for multiple species of landbirds. Preliminary results suggest that some songbirds respond negatively to dry, warm winters that are characteristic of La Niña cycles, while two raptor species have an positive response, perhaps because of a time lag.  

 Life History of Tropical Species

The experimental system described above is the first successful manipulation of plot-level food availability in the tropics. Food availability has been manipulated many times for temperate zone birds, and much has been learned about the life history of these birds as a result, but the life history of tropical birds differs in fundamental ways from their temperate counterparts.  However, the behavioral and evolutionary foundations for these differences remain largely unexplored.  I have taken advantage of this system to address some basic question about how food availability affects abundance, body condition and the timing of breeding in resident Jamaican forest birds. Many more questions remain to be addressed such as the effect of food on reproductive success, survival, and recruitment.  This should be a fruitful line of research.

 

 Physiological Ecology 

            Theoretical models predict that winter body composition is determined by competing pressures of predation and the unpredictability of food availability or weather.  However,  migration-state, sex, age, dominance, and endogenous variation among individuals may also contribute to the energetic state and body composition of migratory songbirds. Once again, empirical insight into these patterns is largely based on temperate-zone, non-migratory wintering species.  Earlier work by Strong et al. (2000) with Ovenbirds on Jamaica demonstrated correlations between food availability and body condition  The results suggest that birds in high quality habitat maintain lower body fat, as predicted by temperate models; however, these patterns were largely driven by variation among habitats.  We are using our system of food manipulation to formally test the hypotheses of Adaptive Body Mass models (Rogers).  In collaboration with Jen Long  of the University of Maine we are using statistical models to understand how winter food availability affects body mass, fat, breast muscle size, plasma metabolites, and stress hormones. So far, we know that body mass (adjusted for skeletal size) and breast muscle size are positively related to food availability.  The other variables show predictable variation through the winter season and into the migratory preparation period, but do not show clear patterns in direct response to food. 

  Habitat Selection                              

For most Temperate-Tropical migratory bird species we have a good background knowledge of which habitats they use.  However, we know very little about the relative quality of these habitats and the consequences for birds of using sub-optimal habitats.  Conservation and management of migratory birds requires not only an understanding of which habitats are used, but more importantly, the quality and quantity of the different available habitats. For a select few model species, we know that occupancy of poor habitat leads to diminished physiological body condition, which can carry over to subsequent seasons, via variation in migration departure timing, to effect reproductive success.  Thus, we need to better understand how birds select habitat; and for successful management, we need to know the relative quality of available habitats.

 

Non-breeding Season:  Using Hermit Thrush (Catharus guttatus) as a model we (in collaboration with Phil Stouffer, LSU) determined that a managed forest can provide high quality habitat during the non-breeding, winter season in Louisiana. Specifically we compared arrival times, dominance patterns, movements, body condition, and survival among three stage-classes of Loblolly Pine plantation and a bottomland hardwood forest habitat. 

 

Breeding Season:  As a side project to my dissertation research, I am collaborating with my Ph.D. advisor, Tom Sherry, to investigate multi-scale habitat selection by American Redstarts (Setophaga ruticilla) in a Louisiana bottomland hardwood forest.  Louisiana is on the fringe of the breeding distribution of Redstarts, but they are locally extremely abundant, including in the Bogue Chitto National Wildlife Refuge, Louisiana.  What make this and other occupied patches so special?  They also appear to select Sweetgum as a nesting tree disproportionate to its availability. Redstarts in Louisiana nest higher and farther from the main tree trunk than in the northern part of their distribution, presenting an interesting case within-species variation in breeding behavior.  By addressing these questions of multiple scales we hope to better understand the structural cues within the landscape that migratory birds use to select habitat. 

 

 

Social Behavior and Spatial Use

 We are also making in-roads into understanding winter habitat selection at the scale of a home-range.  In the absence of breeding interests, home range size is thought to principally reflect food availability. However, it appears that although birds are able to monitor real-time changes in food availability, most individuals are not able to make adjustments to the size and position of their home range after the initial settlement period in the early fall. We think that social mechanisms are the main constraint on adjustments. We have experimental and correlative data from both Ovenbirds in Jamaica and Hermit Thrush in Louisiana to support this hypothesis.   This work has largely been conducted with the aid of radio telemetry equipment.  This technology has also allowed observation of otherwise inconspicuous behaviors such as exter-territorial junkets, floating, and roosting.