Dr. Walter K. Dodds
Division of Biology
Kansas State University
Manhattan, KS 66506
785 532 6998
wkdodds@ksu.edu
Dr. Margaret A. Palmer
Chesapeake Biological Lab
University of Maryland Center for Environmental Science
Solomons, MD 20688
410 326 7241
mpalmer@umd.edu
Dr Bradley J. Cardinale
Dept of Ecology, Evolution & Marine Biology
University of California - Santa Barbara
Santa Barbara, CA 93106
805 893 4157
cardinale@lifesci.ucsb.edu
Humans have greatly altered freshwaters, including depletion and degradation of supplies, increased nutrient loads, and diminished biodiversity. How these changes will individually and collectively influence the resistance and resilience of ecosystems, and their interaction with climate change, is unknown. We propose a long-term, large scale field experiment coupled to a continental observational network that will quantify how nutrient enrichment, reduced consumer diversity, and increased hydrologic variability interactively influence the resistance and resilience of stream ecosystems to global change. The information produced is central to ecological forecasting for freshwater resources. We focus on streams, rivers, and their related wetlands subject to water flow (hereafter referred to as streams). Streams are disproportionately important habitats for biodiversity and ecosystem services, as well as economic, recreational and aesthetic value. Our overarching question is: how will chronic nutrient inputs (nitrogen or phosphorus), higher probabilities of extreme events (droughts and floods), and simplification of food webs (loss of consumers) impact the resistance and resilience of stream ecosystem function (stream-wide respiration, production, and nutrient retention)? We define resistance and resilience as the proportional change in ecosystem functions following a disturbance and the return interval, respectively. We will assess linkages and feedbacks among the ecosystem drivers that are the subject of this proposal.
Streams are ideal ecosystems for the proposed research because they (i) have well delimited inputs and outputs, thus allowing quantification of the focal ecosystem-level processes, (ii) are dynamic systems that respond to disturbances over periods of weeks to years - time scales amenable to observation and experimentation, (iii) can be studied with comparable methods spanning the entire continent, (iv) are important sites of biogeochemical processing, and (v) integrate watershed processes that occur at the same scales as NEON network sensor platforms. Many of these attributes are unique to streams and allow cross-continental research efforts that are more difficult in other ecosystems. In a companion document, we propose 30 aquatic sites for a continental scale observational network. Here we propose 19 experimental sites that are closely paired with a subset of observational sites (controls). Experimental sites will receive a continuous doubling of limiting nutrient(s), and in situ experiments will remove top consumers. Natural hydrologic variation will be used to explore long term climate-induced change of flood and drought, and a subset of the experimental sites will be subjected to direct hydrologic manipulations. Control and paired experimental sites will be arrayed across the continent to capture variation in hydrology, temperature, nutrient loading, elevation, and biogeographical context. To adequately characterize stream ecosystem processes, each observational site will require 2 aquatic sensor packages and each experimental site will require 2 more (a total of 60 + 38 sensor packages and 30 aquatic biodiversity units). We will require additional measurements not in the ISEP (total N and total P, decomposition, nutrient limitation, dissolved gas analyses, and stable isotope and gut-content food web analyses) to effectively characterize community structure and ecosystem function.
This is one of two linked responses to the NEON RFI’s for observational and experimental continental networks of stream research. Some sections of these responses are similar and others are different given the contrasting RFI requirements. The observational network is a very strong design to assess major drivers influencing streams and wetlands and how they link to other habitats. The experimental network builds on the observational one to strengthen the inference for a relatively low added cost.