Pacific Southwest Research Station

Pacific Southwest Research Station
800 Buchanan Street
West Annex Building
Albany, CA 94710-0011

(510) 559-6300

Research Topics Ecosystem Processes

Teakettle Ecosystem Experiment

Teakettle Experimental Forest

The Teakettle Ecosystem Experiment, administered by the Sierra Nevada Research Center (SNRC), is a replicated experimental manipulation comparing the effects of thinning and prescribed fire on ecosystem structure and function in a mixed-conifer forest. The ongoing experiment is occurring in the Teakettle Experimental Forest in the southern Sierra Nevada east of Fresno, California and is directed by Malcolm North of the SNRC.

Background

"Although silvicultural treatments can mimic the effects of fire on structural patterns of woody vegetation, virtually no data exist on the ability to mimic ecological functions of natural fire. Silvicultural treatments can create patterns of woody vegetation that appear similar to those that fire would create, but the consequence for nutrient cycling, hydrology, seed scarification, nonwoody vegetation response, plant diversity, disease and insect infestation, and genetic diversity are mostly unknown."

Critical Findings Section of Sierra Nevada Ecosystem Project, 1996, p. 4-5

Purpose of Study

Plants are the essential stuff of terrestrial ecosystems, establishing a scaffold of structure and composition that influence most ecological processes. Disturbance-induced changes in the size, diversity and distribution of a plant community can have cascading effects on microclimate, soil, and biota, triggering functional responses that mitigate or exacerbate disturbance effects. For many forest communities, periodic disturbance is a defining influence on ecosystem function. Within the last century, the type, interval and intensity of forest disturbance has been dramatically altered by humans around the world. Although recognized as a global problem, we have only a rudimentary understanding of the consequences of this change, largely from independent studies of a forest’s biotic components. An integrated study of forest response is needed which focuses on basic ecological processes and their cumulative effect on ecosystem productivity and diversity.

One practical measure of forest response to disturbance is to assess changes in carbon pools and flows. Disturbance type determines which pools are affected, while intensity indicates how much carbon is removed. In this context, old-growth forests are opportune for studying disturbance because carbon pools are stable and turnover rates range from the big and slow (i.e. large wood pieces and soil organic matter) to the small and fast (i.e. litter, insect frass and herbs). Using disturbance to manipulate this gradient of pools provides a powerful tool for teasing apart the functional response of a forest ecosystem.

Forest carbon pools have been dramatically altered in much of the western U.S. where fire suppression has increased stem densities, modified species composition and age distributions, and increased fine and coarse woody debris. In many areas, forest are now thickets of shade-tolerant species which can "ladder" fire into the crowns of the overstory canopy. Combinations of mechanical thinning and controlled burning are often used in an attempt to mimic the structural effects of historic fire events. In the controlled setting of an old-growth Experimental Forest, these manipulations provide an opportunity to examine ecosystem response to stand-level changes in forest structure.

We propose manipulating stand structure with six treatments uniquely affecting different carbon pools (Table 1). Treatments are a factorial design combining two levels of burning (no burn and ground fire) with three levels of thinning (none, understory and overstory). The treatments will be applied on 18 replicated plots in old-growth Sierra mixed-conifer, providing a disturbance gradient to compare ecosystem response. Coordinated sampling of forest structure, soil nutrients, microclimate, plant and invertebrate responses will occur for two years before and four years after treatments. Data from these component studies will be pooled to quantify changes in net ecosystem productivity and functional group composition and diversity. Our central hypothesis is: Disturbance type and intensity determine the degree, direction (positive or negative) and temporal trend of the relationship between productivity and diversity in forest ecosystems.

The combination of burning and thinning treatments provides a fundamental contrast in how the type of disturbance affects forest structure, composition and function. Thinning treatments may mimic fire in reducing ladder fuels, tree density and the number of shade-tolerant species, but differ from fire by removing large woody pieces and leaving the litter layer intact. Fire reduces soil seed banks, consumes much of the litter layer providing an ephemeral nutrient pulse and more bare mineral soil. These direct effects on forest conditions will initiate responses that affect productivity and diversity. For example productivity may vary between treatment types due to higher nitrogen (N) concentrations in burn plots produced by abundant regeneration of Ceanothus spp., a serotinous shrub capable of fixing atmospheric N2. The strength of this design is that it can systematically address responses to individual treatments or treatments in combination. For example, is shrub-layer herbivory by invertebrates similar in the burn treatments regardless of thinning level, or does the response vary with the interaction of burning and thinning treatments?

Integrating Ecosystem Response

To assess forest response to disturbance, this experiment uses a conceptual model focused on the interaction of forest structure, soil, microclimate, plant responses and invertebrates. Following treatments, we expect tree and shrub mortality, changes in litter quality and quantity, and reduced microclimate buffering to alter the allocation of resources among surviving organisms. The most immediate response to this shift should be observed in differential regeneration, litterfall, and plant growth, and changes in invertebrate diversity and abundance . The five components interact by affecting core processes which have feedback effects on ecosystem structure, composition and function. We will measure changes in the constituent states of the five components and the process rates that drive these changes.

Two metrics will be used to assess overall ecosystem response: net ecosystem productivity (NEP) and functional group composition and diversity. NEP measures the accrual of matter and energy in biomass, effectively synthesizing respiration, primary and secondary productivity. Diversity will be assessed by changes in composition within functional groups, and relative abundance (evenness) between groups. Recent studies in grass ecosystems suggest that composition and diversity of functional groups may be more influential on ecosystem processes than total richness.