I will be investigating three alternative mechanisms explaining the declines of understory insectivorous birds:

I. Food-limitation resulting from indirect effects of mesoherbivore release
The food limitation hypothesis states that understory insectivorous birds decline within forest patches due to limited availability of arthropod prey. This guild of birds has very highly specialized dietary and foraging habits, a characteristic which is strongly correlated with extinction risk generally, and at La Selva specifically, where the declines were linked to diet, with insectivores accounting for half of the declining or extirpated species (Sigel et al. 2006). Although arthropod abundance, avian abundance, and forest area were not correlated in either southern Costa Rican or Panamanian lowland rainforest (Karr and Brawn 1990, Sekercioglu et al. 2002), arthropod responses to changes in climate or vegetation differ widely by taxonomic order. Alterations to vegetation may impact insectivorous birds via altering abundances or behaviors of the relatively few prey types to which some bird species are specialized, in addition to the alteration of foraging substrate. The long-term, increase in rainfall and mean minimum temperatures observed at La Selva is also likely to impact arthropod populations both directly and indirectly, via declining tree growth rates and potentially lower leaf production (Clark et al. 2003). Strong but short-term climatic events such as El Ninos, which are associated extreme droughts in seasonal forests such as Barro Colorado Island (Leigh et al. 1990) and higher temperatures at La Selva (Clark et al. 2003) are also likely to have dramatic impacts on arthropod populations, perhaps leading to environmental bottlenecks which are not detectable over the long-term.

I will be investigating foraging rates and dietary preferences of five members of the understory insectivorous bird guild: ruddy-tailed flycatcher (Terenotriccus erythrurus), white-flanked antwren (Myrmotherula axillaris), checker-throated antwren (M. fulviventris), dot-winged antwren (Microrhopias quixensis), and golden-crowned spadebill (Platyrinchus coronatus). I will investigate abundances of arthropods, in total and of known prey species, at La Selva Biological Station (within and outside large mammal exclosures), Tirimbina Rainforest Center, and Refugio Bartola, a control site adjacent to the 500,000 hectare Indio Maiz Biological Reserve in southeastern Nicaragua. Additionally, I will be mentoring an REU student, Marvin Morales, who will be investigating dietary preferences and prey availability for an understory insectivore which has increased at La Selva over the previous 30 years, the white-breasted wood-wren (Henicorhina leucosticta).

II. Limited availability of preferred microhabitats and microclimatic regimes, potentially resulting from mesoherbivore release
The microhabitat and microclimatic availability hypothesis states that the limited availability of preferred foraging and/or nesting habitats results in reduced fitness and, eventually, declines or extirpations of understory insectivorous birds. Vegetation structure has long been known as an important factor influencing the distribution and abundance of birds, including insectivores (Karr and Freemark 1983, Terborgh 1985, Marra and Remsen 1997). Microhabitat preferences are demonstrably adaptive, resulting in higher fitness for those individuals in preferred habitats (Martin 1998). Further, small understory insectivores were amongst the birds most affected by selective logging in New Guinea, and habitat selection was the major determinant of vulnerability (Thiollay 1997).

While broad habitat use of this guild is known (i.e., they are typically limited to interior understory rainforest), microhabitat selection and use has been little-studied. Given their high dietary specialization and relatively "slow" life-histories (including low metabolic rates), it is likely that they have specialized microhabitat needs as well. In fact, those understory insectivore species whose microhabitat use has been studied (e.g., antwrens) show high habitat specialization (Greenberg and Gradwohl 1980, Gradwohl and Greenberg 1980, Karr and Freemark 1983, Terborgh 1985, Marra and Remsen 1997).

Disturbance of understory vegetation is also likely to alter the microclimatic regime, including light availability, temperature, and humidity (Pearson 1977, Saunders et al. 1991, Didham and Lawton 1999, Pohlman et al. 2007). Tropical understory insectivorous birds have low basal metabolic rates and a narrow thermoneutral range (Wiersma et al. 2007), as they are adapted to the relatively-constant abiotic environment of the understory. They are known to respond to changes in temperature and humidity (e.g., moving to humid sites on hot dry season days; Karr and Freemark 1983). Thus alterations to understory vegetation density could impact fitness and persistence of this guild via loss of both preferred microhabitats and required microclimatic regimes. As described on the Mesoherbivore Release page, vegetation density is significantly lower in sites where peccaries are abundant, thus hinting at a potential mechanism.

I will be investigating microhabitat and microclimate selection by the five target species listed above at Refugio Bartola and Tirimbina Rainforest Center. I will also estimate availability of preferred microhabitats and microclimatic regimes at Refugio Bartola, Tirimbina, and La Selva Biological Station (within and outside large mammal exclosures). Additionally, I will be mentoring an REU student, Marvin Morales, who will be investigating microhabitat selection by the still-abundant white-breasted wood-wren (Henicorhina leucosticta) at La Selva.

III. Elevated nest predation resulting from mesopredator and/or mesoherbivore release Nest predation rates are generally higher within fragments, due to edge effects and increased abundance of nest predators (mesopredator release) within disturbed habitats, following the loss of large predators (Crooks and Soulé 1999). However, nest predation rates of understory birds are higher than predicted by reserve size at La Selva (Young et al., in press). Snakes, particularly the bird-eating snake (Pseustes poecilonotus), are considered to be the primary nest predators in tropical rainforest (Gradwohl and Greenberg 1982a, Skutch 1985, Robinson and Robinson 2001, Weatherhead and Blouin-Demers 2004, Robinson et al. 2005, Stratford and Robinson 2005), however anecdotal evidence suggests this species may not be as common at La Selva today as previously (Dennis Wasko, pers. comm.). Conversely, mid-sized mammalian nest predators such as monkeys and white-nosed coatimundi (Nasua narica) may have increased in that same time period (Tom Sherry, pers. comm.). However, current ignorance of the identity of the primary nest predators at La Selva hampers our ability to understand the mechanisms driving the elevated nest predation rates there. It is also not clear how elevated nest predation rates at a landscape scale, with similar rates across nest types and distance from forest edge, would result in declines of only some species, notably understory insectivores, unless one species of nest predator is both disproportionately abundant and disproportionately affects understory insectivorous birds, for whatever reason. Conversely, lower vegetation density due to mesoherbivore (peccary) release may result in lower availability of sheltered nest sites, exposing nesting birds to higher predation rates without alterations to the nest predator community.

Myself and two field assistants, James Lewis and Julie Jackson, will begin monitoring nests using video cameras at La Selva Biological Station in 2008 with the goal of identifying the common nest predators there.