Research Areas > Wetlands > Statewide Validation of CRAM for Riverine and Estuarine Wetlands

Project: Statewide Validation of CRAM for Riverine and Estuarine Wetlands

Background and Objectives

The California Rapid Assessment Method (CRAM) was developed to provide consistent, scientifically defensible, affordable information about the conditions of wetlands and riparian habitats throughout California. To achieve this goal, CRAM needed to be standardized, cost-effective, scientifically defensible, and broadly applicable to the variety of wetland types across California. Development of CRAM consisted of nine steps with distinct products organized into three phases: basic method design, calibration, and validation. Basic method design was completed in a previous study. Because of CRAM’s integrative nature and reliance on translating ecological theory into field indicators that reflect wetland condition, it was important that CRAM be validated against other independent measures of wetland condition in order to establish its scientific validity.

Objectives for this study were:

1) Evaluation - ensure CRAM produces meaningful results based on a comparison between CRAM scores and independent measures of condition
2) Calibration - make adjustments to the method, where needed, to improve the ability of CRAM to discern differences in wetland condition
3) Standardization - minimize observer bias by assessing repeatability among independent assessment teams and then modifying metrics that lead to inconsistencies

Status

This project was conducted from 2002 to 2007.

Methods

In the summer of 2005 the riverine and estuarine components of CRAM were tested against intensive measures of condition. The two CRAM "modules" were validated based on the results of this analysis. CRAM was evaluated in terms of its responsiveness to “good” vs. “poor” wetland condition, its ability to represent a range of conditions, internal redundancy between its component metrics, alternative models to integrate the metrics into overall scores, and in terms of reproducibility of results between independent assessment teams.

Given the complexity and diversity of wetland function and the inherent simplifications associated with CRAM, there was no single “gold standard” measure of wetland condition that directly reflected the same elements as the CRAM attributes and could be used for validation. Consequently, researchers took advantage of data from existing monitoring and assessment programs based on avian diversity, benthic macroinvertebrate indices, and plant community composition to validate CRAM. Decisions regarding modification of metrics or attributes were made based on a “weight-of-evidence” approach. Weight-of-evidence is the process of combining information from multiple lines of evidence to reach a conclusion about an environmental system or stressor.

Findings

Significant findings of the CRAM calibration and validation study were as follows:

• Results for riverine and estuarine wetlands indicated that CRAM is an effective tool for assessing general wetland condition based on its correspondence with multiple independent assessments of condition.

• Most CRAM attributes captured a range of wetland conditions. The one exception, Buffer and Landscape Context, was modified based on the calibration analysis to improve its representativeness.

• Several metric combination models were tested for each CRAM attribute, and in most cases the “neutral” model (i.e., a linear combination of metrics) was comparable to alternative models based on more complex computations.

• Reproducibility analysis revealed several problematic metrics where ambiguous language or metric construction led to high inter-team error rates. Clarification of metric construction and inclusion of additional guidance rectified these problems and improved the overall average error between independent assessment teams to ±5%.

• This study demonstrated that when calibrated and validated, rapid assessment methods provide a reliable tool for assessing wetland condition that has potential application for general condition assessments, screening-level evaluations, and assessment of program performance.

Next steps include validation of the depressional wetland portion of CRAM, implementation of a CRAM training program for agency staff and practitioners, and development of guidelines for use of CRAM in various wetland regulatory and management programs. This work was continued in a subsequent projects (see Development of a Statewide Framework for Wetland and Riparian Monitoring and Assessment).

Partners

This project was conducted in collaboration with California State University at Northridge, the San Francisco Estuary Institute, San Jose State University Foundation, US EPA Region IX and numerous other partner agencies and institutions.

Research Brief

• Diagnosing Wetland Health with Rapid Assessment Methods - E Stein, M Sutula, AE Fetscher, RP Clark, JN Collins, JL Grenier, C Grosso