Effects of fishes on algal response to storms in a tropical stream

Ecology, Dec, 1997 by Catherine M. Pringle, Toshihide Hamazaki

INTRODUCTION

A major challenge for ecologists is to determine how trophic factors interact with disturbance to affect community response. While much research has focused on the effects of natural disturbances on biotic communities (e.g., Pickett and White 1985), we know very little regarding how biotic factors can influence effects of disturbance. Traditional ecological theory treats abiotic disturbance as an external force (e.g., Smith 1990), typically overlooking how biological aspects of the community can influence disturbance effects (but see Naiman 1988, Jones et al. 1994). As natural communities become increasingly impoverished from human activities (e.g., Ehrlich and Ehrlich 1981, Soule 1991, Moyle and Leidy 1992), such knowledge becomes critical to the development of a predictive understanding of disturbance.

Increasing evidence shows that biotic factors can influence community response to disturbance effects. For example, by creating large pools or "wallows" that retain water during dry periods, American alligators in the Florida Everglades mitigate effects of disturbances caused by drought by providing refuges for fishes and fish-eating birds (Finlayson and Moser 1991). Herbivorous elephants in African savannas reduce the accumulation of woody debris and thus the frequency and severity of fires (McNaughton et al. 1988, Naiman 1988, Owen-Smith 1992). Similarly, grazing by domestic cattle has reduced the organic mat in rangelands of the southwestern United States, reducing the incidence of fires with dramatic effects on the structure of terrestrial vegetation (Phillips 1965, Box et al. 1967). In contrast, infestations of bark beetles in forests of the southeastern United States can enhance the intensity and incidence of fires (Schowalter 1985, Rykiel et al. 1988).

In aquatic ecosystems, little information exists regarding how natural assemblages of macroconsumers interact with natural disturbances to affect the stability and community structure of lower trophic levels. In streams, it is well known that high discharge during storms is a major disturbance that structures benthic communities (Fisher et al. 1982, Minshall et al. 1988, Resh et al. 1988, Steinman and McIntire 1990). Also, a growing number of studies have shown that stream macrobiota such as fishes and shrimps can play a major role in shaping the benthic community through feeding and/or removing sediments (e.g., Power et al. 1985, Power 1990, Flecker 1992, Gelwick and Matthews 1992, Pringle et al. 1993a, Pringle 1996). However, these trophic studies have been conducted during periods of relatively low discharge, intentionally eliminating interactions between biotic (e.g., trophic) and abiotic (e.g., disturbance) factors. For example, a review of 89 experimental studies of stream herbivory (Feminella and Hawkins 1995) concludes that most experiments were conducted during, or were designed to simulate, base flow conditions. The lack of data on effects of natural biotic and abiotic interactions is largely due to the unpredictability and severity of physical conditions during high discharge events which impose logistical constraints on in situ experiments (e.g., flooding/destruction).

In the present study, we use a storm-resistant electric exclusion technique (Pringle and Blake 1994) to manipulate fishes in situ and to quantify their relatively long-term (i.e., 7 wk) effects on algal response to high-discharge events caused by storms. Our primary objective was to examine how natural densities of fishes interact with natural disturbance events to affect primary producers. Specifically, we examined how omnivorous fish assemblages affected the response of algal assemblages (i.e., biomass, taxon composition, diversity, and richness) to storms in a lowland tropical stream. Our secondary objective was to examine effects of fishes on insect densities.

STUDY SITE

This study was carried out in a stream draining the lowland ([approximately]60 m above sea level) foothills of Costa Rica's Caribbean slope. Streams in this region are characterized by a diversity and abundance of omnivorous fishes (Burcham 1988, Wootton and Oemke 1992, Bussing 1994) and are often subject to dramatic fluctuations in discharge as a result of tropical storms (e.g., Stout 1982).

Our study site was located in the lower reaches of the Sabalo River, a third- to fourth-order stream that drains the eastern boundary of La Selva Biological Station (10 [degrees] 26 [minutes] N, 84 [degrees] 01 [minutes] W), which is owned and operated by the Organization for Tropical Studies. La Selva receives an average annual precipitation of 4000 mm, with peaks occurring both in June-July (mean = 460 mm/mo) and November-December (mean = 400 mm/mo) based on a 25-yr record. During the 'dry' season from February to April, precipitation averages 170 mm/mo (Sanford et al. 1994).

The Sabalo River lies within the Guacimo basin, which extends into the forested highlands south of La Selva. The river is subject to dramatic discharge fluctuations from storms in both the wet and dry seasons and water levels frequently rise 2-3 m within a few hours (C. M. Pringle, personal observation). Most of the Sabalo basin is in primary forest, but there is actively grazed pasture on its eastern side in the lower reaches near our study site. The stream bottom consists of medium-sized cobbles and gravel, with finer sediments and leaves in depositional areas.