PARENT SESSION
Wednesday, August 9, 1:30-5:00 pm
Symposium 14 - The urban food web: how humans alter the state and interactions of trophic dynamics
Cotton Row, Mezzanine Level, Cook Convention Center
Organized by: PS Warren (pswarren@forwild.umass.edu), C Tripler, C Lepczyk, and J Walker

This symposium addresses the unification of food web theory, with its classic debates over the regulation of populations and communities, and the inventive new approaches by which urban ecologists examine the dynamics of coupled social-biophysical systems.

Trophic dynamics in urban habitat fragments: does fragmentation induce a top-down cascade?

Bolger, Douglas ^*,1, Diffendorfer, James², Walters, Eric¹, Anguiano, Michael³, Morin, Dana³, Patten, Michael^{1, 4}, ¹ Dartmouth College, Hanover, NH² Illinois Natural History Survey, Champaign, IL³ San Diego State University, San Diego, CA⁴ University of Oklahoma, Bartlesville, OK

ABSTRACT- Habitat fragmentation is thought to often cause changes at higher trophic levels that can cascade down to mid-trophic species like birds and small mammals. To test the top-down cascade hypothesis we conducted studies of avian reproduction and survival and small mammal recruitment and survival in coastal sage scrub habitat fragments in urban San Diego, CA, USA. Previously in this system, Crooks and Soulé (1999) presented indirect evidence of a top-down cascade on birds due to mesocarnivore release that occurs in fragments too small to support coyotes. Our studies were conducted along a fragmentation gradient that included the interior of large nature reserves (>2300 ha), the edge of those reserves, large fragments (50-100 ha) embedded within the urban matrix and small fragments (5-20 ha). Like Crooks and Soulé, we found mesocarnivores to be more common in habitat fragments than in unfragmented habitat. Other predator groups also differed in predictable ways across the landscape gradient, however, the response to fragmentation differed among the groups. Snakes were much less diverse and abundant in fragments, but raptors and corvids were more common. We focused on four passerine bird species: two shrub nesters and two ground-nesters. By videotaping active nests we found that ground-nesting birds were primarily vulnerable to snakes, while, shrub nesters were vulnerable to snakes as well as avian nest predators. Mesocarnivores do not appear to be important nest predators in this system. Despite changes in nest predator abundance, reproductive output did not differ consistently across this gradient and birds in fragments did not experience higher nest predation probability or lower reproductive output than birds in reserves. In fact ground-nesting species tended to have higher reproductive output in fragments, consistent with reduced snake abundance there. The most common small mammal species, Peromyscus eremicus , also showed no difference in survival and recruitment across the four landscape treatments. Thus, in terms of avian reproductive success and small mammal survival and recruitment we find no support for the top-down cascade hypothesis. It appears that the presence of a diverse predator community with divergent responses to fragmentation prevents a linear top-down cascade.

Key words: habitat fragmentation, coastal sage scrub, birds and small mammals