THE NEED FOR LONG-TERM STREAM MONITORING PROGRAMS IN FOREST ECOSYSTEMS OF THE PACIFIC NORTHWEST ROBERT C. W I S S M A R School of Fisheries (WH-IO) and Center for Streamside Studies (AR-IO), University of Washington, Seattle, WA 98195, U.S.A. (Received: 11 March 1992; accepted: 26 February 1993)

Abstract. Concepts, planning and design procedures are examined that are needed in the development of long-term stream monitoring programs in forested regions. A long-term stream monitoring program is viewed as the key component for bringing together management organizations, researchers and decision-makers to improve the management of natural resources. The keystones of such ecosystem monitoring are long-term data records that provide the basis for analysis of environmental assessment objectives, predictions and analysis of outcomes which in-turn can be used to modify and improve future projects. Management organizations that initiate long-term monitoring programs are urged to use monitoring actions and information to facilitate decision-making processes that pertain to conserving and allocating resources for future beneficial uses. Recommendations are provided for careful planning and definition of interactive activities of monitoring programs and that should provide information feedbacks that can be used to evaluate issues pertaining to beneficial uses of resources. Procedural requirements and literature sources are suggested for developing long-term stream monitoring programs. They include reviews of background and historical information to provide precise definitions of long-term objectives, planning considerations and monitoring methods. Examples are given of specific procedures that need to be identified during the planning process. They include the application of management standards to variable conditions encountered within natural ecosystems and the detection of the timing of recovery phases of stream ecosystem development following a disturbance. These procedures are viewed as being essential for improving applications of management standards and perceived thresholds to stream and watershed ecosystems monitoring programs.

1. Introduction

This paper examines some of the issues and challenges involved in monitoring changing stream and watershed conditions in forest ecosystems of the Pacific Northwest. A broad review of the literature on stream and watershed ecosystem monitoring and assessment is not the objective of this paper. Such reviews can be found in numerous articles (Ausmus, 1984; Karr, 1987; Schindler, 1987; Daggan etal., 1987; Schaeffer etal., 1988) and in a recent report by the National Research Council (1990). The working assumption of this paper is that, in order for management organizations to fulfil responsibilities of improving and protecting stream ecosystems, better procedures are needed for developing plans, implementing and executing long-term monitoring programs. Furthermore, a functioning long-term monitoring program can become the key component for bringing together the efforts of management organizations, decision-makers and researchers that intend to improve and protect natural ecosystems. The reality of the development of a truly functional and sustainable monitoring program is a difficult and unusual endeavor. Environmental Monitoring and Assessment 26: 219-234, 1993. @ 1993 Kluwer Academic Publishers. Printed in the Netherlands.

220

ROBERT C. WISSMAR

This paper begins by describing historical attributes of the USDA Forest Service that appear favorable for the initiation of sustained monitoring programs for streams of forested ecosystems. The Forest Service is used as a case history because it is the principal agency managing federal forest lands in the Pacific Northwest. The Forest Service is responsible for monitoring and improving forest and stream conditions. A summary is given of the common issues and challenges that can be encountered when an organization like the Forest Service becomes involved in monitoring stream ecosystems. Recommendations are then provided of the important procedural steps in developing a long-term monitoring plan. An example is given of plan components needed to apply a restoration program to degraded streams and watershed ecosystems. The paper concludes by discussing some of the challenges of: (a) applying standards to variable and changing spatial and temporal conditions encountered in stream ecosystems; and (b) detecting recovery phases of stream ecosystem development following a disturbance.

2. Management History and Long-TermMonitoring Programs A management organization that has a history that favors the development of long-term monitoring programs for forest and stream ecosystems is the National Forest Service. Historically, the Forest Service has the longest history of land-use planning of any government agency. The initial legal basis was laid by the Organic Act of 1897. The Organic Act of 1897 gave the Forest Service a wide management window to plan for improvement and protection of national forests for the purpose of "securing favorable conditions of waterflow, and to furnish a continuous supply of timber". This act's requirement for continuous supplies of resources, implies the need for sustained monitoring activities with objectives being better management and conservation. The Forest Service was managed within the original mandate of the Organic Act until after World War II when increased timber and recreational demands necessitated a more comprehensive planning process. In 1960, the Multiple-Use, Sustained Yield Act established for the Forest Service a multiple-use management policy by requiring the agency to develop formal district and regional Planning Guides. The intent of the act was to give equal status to timber, watershed, range, recreation, wildlife and fish resources. The importance of planning also increased with the need for compliance with the National Environmental Policy Act of 1969 (NEPA) which required the preparation of environmental impact statements (EIS). To improve Forest Service planning, the Congress passed the Forest and Rangelands Renewable Resources Planning Act in 1974 (RPA). Many of the above concepts have been incorporated into the National Forest Management Act of 1976 (NFMA) which presents a strong precedent for developing long-term programs for both stream and terrestrial ecosystems. NFMA focused the Forest Service management by requiring a comprehensive planning process for each of the 154 national forests. NFMA requires several planning documents:

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

(a)

221

a Forest Plan or programs that cover 5 years and contain objectives that apply to the next 45 years for all Forest Service activities;

(b) annual reports that evaluate Forest activities in relation to the Plan's objectives; (c)

assessments at 10 year intervals that describe the renewable resources of a national forest; and

(d) a Presidential Statement of Policy every 5 to 10 years that describes budget requests for its objectives. Most attributes of Forest Plan as required by NFMA can be implemented and improved by developing long-term monitoring programs. Most importantly, the objectives for attaining desired future conditions should be rigorously attached to long-term time-frames and activities. This can be accomplished by providing for continuous evaluations of objectives and their refinements through iterative amendment processes. In the Pacific Northwest, examples of Forest Service implementation plans that attempt to address the needs of stream monitoring programs can be found in Steps of the Journey: Forest Plan Implementation Strategy: USDA Region 6 (1990) and the Columbia River Basin Anadromous Fish Habitat Management Policy Implementation Guide: USDA Regions 1, 4, 6 (1991). A summary of the types of information needed to implement such plans can be found in Meehan (1991). This publication presents a 10 step Forest Service planning procedure for streams of forest and rangeland ecosystems as well as reviews other U.S. federal agency planning processes (Brouha, 1991).

3. Issues and Challenges The issues and challenges involved in monitoring the changing conditions of stream ecosystems are common to those facing other environmental situations and assessment activities. The most common problems are exhibited in the shortcomings of environmental impact assessments (Culhane, 1987). The major problems include poorly stated objectives and plans for: (a)

coordination between disciplines;

(b) definition and understanding of the time constraints for accomplishing objectives; (c) accommodation of probability and uncertainty; (d) utilization of predictive methods; and (e) evaluation of indirect and direct cumulative effects.

222

ROBERT C. WISSMAR

Many of these problems relate to a management organization's need for immediate answers which precludes lengthy study. In other words, crisis situations that are commonly driven by institutional and fiscal constraints that force quick fixes. As a consequence, a variety of operational modes and problems appear during short-term activities for managing stream and watershed ecosystems. First, many short-term projects tend to use scientific advice and information for making rapid decisions that are then perceived as final solutions. Secondly, the need for immediate solutions leads to the unwillingness to alter management practices and regulations in response to changing environmental conditions, management needs, scientific advice and knowledge. Finally, short-term management activities may set precedents where the perceived goal becomes the desire to control the evolution of an ecosystem. All these operational patterns and their inherent problems can be better addressed by frequent monitoring over sufficiently long time periods to attain the required information needed for evaluating stated objectives and the numerous natural and disturbance induced changes that occur in dynamic ecosystems like streams (e.g. flooding and livestock disturbances in streams and riparian areas). To further counter these operational problems, management organizations need to recognize the values of creating policies and projects that deal with changing stream ecosystem processes (e.g. sedimentation and erosion) that operate both naturally and unnaturally over long time spans covering several years to decades. Key elements for the development, implementation and evaluation of effective long-term management projects include monitoring activities and the keeping of records. The sustainment of monitoring records and evaluation processes facilitates problem solving by providing an increased degree of certainty of information through long-term validation. To implement stream ecosystem monitoring programs several obstacles need to be overcome. Some major obstacles include: (a)

convincing policy makers, managers and scientific advisors of the importance of sustained data acquisition and records in managing and solving many major ecosystems problems;

(b) improving communications between policy makers, managers and scientific advisors to better understand institutional constraints, to identify mutual concerns and questions; (c) providing new commitment of funds and longer funding cycles by both management and research organizations; and (d) recognizing for stream and watershed ecosystems the time required to acquire long-term data that can be used to evaluate the multivariate nature of ecosystems and their instabilities induced by both natural and man-induced disturbances. The above issues and challenges also point to the inappropriate amount of scientific efforts in ecological research being directed towards working with the

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

223

decision-making process (Likens, 1988). Research conducted in conjunction with long-term monitoring programs could have increased value if the program and results are coordinated and presented in a manner that is useful for the decisionmaking processes and policy-makers. Tomlinson and Atkinson (1987) suggest several research activities that could be better coordinated with policy development and environmental assessment projects. These activities include: (a)

assessing the accuracy of predicted results;

(b) identifying approaches for inclusion of probability and uncertainty; (c) defining the desirable error limits in prediction; and (d) developing mechanisms for using long-term records to improve the planning process.

4. Development of Long-Term Monitoring Plans Once the issues and challenges of an environmental problem have been identified, the next step is the development of a long-term monitoring plan. The development of a long-term monitoring plan can be described by presenting plan components required for applying a restoration program to degraded stream and watershed ecosystems (Maurizi and Poillon, 1992). The first step is stating the general objective of the monitoring plan, for example, to assess the effectiveness of ecosystem restoration program. More specific objectives include assessing if the restored ecosystem is providing through time the planned functions and beneficial uses (Hildebrand et al., 1987). These objectives can be addressed by conducting a longterm monitoring program that includes separate implementation, evaluation and validation monitoring procedures (MacDonald et al., 1991). The next step is to define what the criteria is for success. Criteria for success of the restoration program can be defined by scrutinizing of plan objectives and components. Important components needed in the development of the restoration plan include: (a) background information; (b) long-term objectives; (c) planning consideration; and (d) long-term monitoring procedures. 4.1. BACKGROUND INFORMATION Background information required for developing the ecosystem restoration plan includes the present status and historical review of disturbances induced by natural events and human activities. This information can be obtained by defining the landscape in terms of bedrock geology, geomorphic landforms, hydrologic regimes,

224

ROBERT C. WISSMAR

and distribution of stream and riparian habitats (Platts et al., 1983; MacDonald et al., 1991). The use of background information to develop plans with stream ecosystem perspectives should include available stream and riparian habitat inventories. For riparian areas, consideration needs to be given to the stability of streambanks without vegetation. Attention should also be paid to the feasibility of conserving riparian and transitional upland vegetation buffers as filters for surface flows of stormwater and habitats or corridors for wildlife (Gregory and Ashkenas, 1990). Desirable hydrologic information includes: (a)

stream flow data and weir locations-controls (e.g. USGS records);

(b) flooding regimes; (c) groundwater elevations and seasonality; and (d) water storage capacity in a watershed's groundwaters, riparian areas, wetlands and lakes (Platts et al., 1983; MacDonald et al., 1991; Peterson et al., 1992). The synthesis of background information needs to include topographic scales and the location of benchmarks that are useful for planning long-terms monitoring procedures. This background information is very useful in planning the timing of construction phases in relation to critical biotic processes. In stream ecosystems an important consideration is the timing of earthwork and road building to prevent impacts to fish spawning and rearing habitats. The background information is also useful in defining the criteria for success of the restoration plan. Attention should be given to defining the principal functions and beneficial uses that the stream and riparian areas and watershed condition are intended to provide. Important functions and beneficial uses include fish and wildlife habitats, recreational use, water supply, educational and aesthetics needs. The monitoring plans should strive to assess the success of select functions and beneficial uses of a restored ecosystem by following the capacities of the restored watershed's interacting stream, riparian, and floodplain components to provide for: (a)

acceptable water quality and flood control through water storage;

(b) stormwater management through retention and/or attenuation of water flow; (c) stabilization of streambank erosion processes; and (d) opportunities for enhancement of wildlife and fisheries habitat. 4.2. LONG-TERMOBJECTIVES The formation of long-term objectives can be the most important component of the planning process. Long-term objectives can include criteria for success which provides a basis for assessing the future ecosystem conditions after restoration

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

225

(Gore, 1985; Turner, 1987; Maurizi and Poillon, 1992). The long-term objectives can be defined in three ways: (a) endpoints or points in time; (b) plan predictions; and (c) monitoring parameters expressed as standards required for the assessment of the success of the predictions. The key question to be answered when forming these objectives is: what is the desired recovery of developmental stage in the ecosystem that meets your objectives? The formation of the objectives sets the stage for other planning considerations. 4.3. PLANNINGCONSIDERATIONS Planning considerations include legal requirements, land acquisition, conflict resolution, availability of physical and ecological information, suitability of construction designs, timing and degree of maintenance required for restoration, and problems unique to a particular ecosystem. Potential problems encountered in developing and implementing stream and watershed ecosystem restoration plans include: (a) erosion damage (bank stability and peak flows); (b) water logged and flooded soils (flood frequency and duration): (c) ice damage (anchor ice): (d) animal depredation of streambanks by livestock; (e) logging on unstable hillslopes; (f)

excessive secondary road miles in the watershed (age, maintenance conditions);

(g) excessive recreational use of riparian areas; and (h) poor definition of monitoring procedures for assessing cumulative effects created by many of the above problems. 4.4.

LONG-TERM MONITORING PROCEDURES

Long-term monitoring procedures should give consideration to the scale of the program. Stream and watershed restoration programs have different monitoring requirements than smaller site-specific projects. Ecosystem level programs require monitoring procedures that can assess influences of natural and man-induced disturbances and land-water interactions that play important roles in causing cumulative impacts.

226

ROBERT C. WISSMAR

The procedures should be able to address the objectives using key parameters relevant to long-term evaluation and validation monitoring. These parameters need to have reliable and robust qualities applicable to recording changes in natural conditions of stream ecosystems and providing variability statements. Changes in stream ecosystems can be caused by large infrequent events like floods as well as other complex phenomena involving rapid and slow processes of erosion and sedimentation. Extremely important considerations in the planning of monitoring procedures are the standards used in managing natural resources. The selection of standards needs to include specific definitions of: (a) parameters and measured units; and (b) methods for providing variability statements for a standard's parameters and data records. These definitions are required to improve the application of standards to dynamic ecosystems by management organizations. Management organizations commonly apply standards in terms of desired threshold concepts with undefined limits of acceptability. The use of thresholds is questionable when they do not take into account the natural variability of an ecosystem. A more meaningful procedure is the application of standards that define the limits of acceptable and unacceptable conditions within the spatial and temporal variabilities inherent in dynamic ecosystems. A useful approach for defining variability statements involves the analysis of monitoring records.

5. Variability of Standards for Managing Stream Ecosystems A useful and simple definition of the variability of a standard involved is the confidence interval (C.I.) of a sample mean. The sample mean and confidence intervals can be obtained from the analysis of monitoring records of stream ecosystems. The use of these statistics brings greater certainty to decision-making processes. The definition of the confidence intervals (C.I.) of a sample mean can be demonstrated through the analysis of monitoring data for large woody debris (LWD) in stream channels. In the Pacific Northwest, LWD is extremely important to the structure and function of stream and riparian habitats (Bilby and Ward, 1989). Both the retention and mobility of LWD within stream ecosystems can alter flow patterns that create pool and other habitat types by influencing sediment storage and transport, streambank stabilities and associated fluvial-geomorphic conditions. The recruitment of large wood to stream channels depends on riparian forest stand characteristics, hillslope gradients and other landform characteristics and the frequency of natural disturbance events (Wissmar and Swanson, 1990). For LWD, an important objective for the analysis of monitoring data can be the development of a LWD standard applicable to stream channel reaches throughout a watershed (Bilby and Wasserman, 1989). This LWD standard can be expressed as the amounts and distributions of different sizes of LWD stratified by channel width categories for upstream and downstream reaches of a stream. The first step involves calculating the LWD sample mean for each of the width categories of a

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

227

reach. The next step provides the confidence intervals of sample means. At this stage a decision needs to be made about the desired level of significance, for example, 10% or p = 0.1. This probability level can be used to evaluate if the mean of a LWD population lies between the confidence intervals, if so, there may be only 1 chance in 10 that the LWD value and the related decision are wrong. This procedure uses a t value and calculated confidence intervals. The C.I. is the t value times the standard deviation of sample mean. The application of C.I.s as variability statements to the management of stream ecosystems is most useful when the confidence intervals for the mean of a standard represents natural baseline conditions of unmanaged ecosystems, such as LWD in streams within wilderness areas. Confidence intervals for natural variations can be compared to LWD values from altered stream ecosystems. LWD values falling outside the C.I. can be denoted as exceeding the means expected for natural baseline conditions. Attention can also be given to extreme deviations that can be attributed to conditions caused by both natural and unnatural disturbances. An example of potentially variable conditions for altered stream habitats includes very low amounts of LWD per 100 meters of channel. The evaluation and validation of the variability of standards also applies to the use of multiple standards for identifying and forecasting stream ecosystem conditions. Multiple standards like LWD densities, riparian forest areas, stream habitat types and areas and sediment characteristics can be expected to vary according to their spatial positions in landscapes. The application of these standards should consider the entire landscape of a stream's watershed. These considerations should include changes in scale of stream channel widths, gradients, entrenchment and sinuosity relative to larger spatial features like geomorphic deposits, valley landforms, watershed reliefs and drainage density. Multiple standards permit the assessment of potentials for degradation in environmental conditions due to both natural and management induced disturbances. Potentials for degradation are indicated when several standards simultaneously approach or fall outside the confidence intervals for their respective means. If the confidence intervals are exceeded for different standards of a stream, this stream ecosystem may be very sensitive to disturbances. The use of different standards to identify sensitive stream ecosystems can be very useful to future management activities in ecosystems with different management histories and from different geologic, hydrologic, and climatic regions. In such circumstances, monitoring records of management practices and changing natural conditions provide information feedbacks useful in future plans that avoids sensitive areas in ecosystems. Sensitive areas of stream ecosystems can span several spatial scales in a watershed. The largest sensitive areas might include stream reaches whose geomorphic conditions are susceptible to potential cumulative impacts of land-use practices. Smaller landscape units can include stream habitats that are sensitive in terms of an animal's life history requirements, for example, fish spawning and rearing habitats.

228

ROBERTC. WISSMAR

Logging Period:

Pre ::~=~::::~>

Stream Phases:

Disturbances Stages:

Stable

Persistence ==::~====:>

Post =======================

Degraded

Reaction ::::~:====>

Recovery ====::::~:>

Persistence =====:~>

Time for Disturbance Recurrence

Fig. 1. Stable and degrading phases in stream ecosystems. In stream ecosystems influenced by management activities, stable and degrading phases can occur in response to land-use disturbances such as logging. Stable and degrading phases may also be synonymous with stages of natural disturbance regimes for ecosystems.Natural disturbanceregimes can displaythree stages: persistence, reaction and recover (Wissmarand Swanson, 1990). The persistencestagecorresponds with the stable phase while the reaction and recovery stages include the degraded phase.

6. Monitoring Records and Detection of Long-Term Trends To address many of the issues and challenges discussed in this paper, a simple procedure is needed that detects long-term trends of phases of changing ecosystem conditions. Most conventional sampling design and statistical methods require varying resolution levels of data and complex analytical approaches that can be difficult to apply to large stream ecosystem monitoring programs (Green, 1979; Gilbert, 1987). Here, a simple procedure is suggested for detecting stable and degraded phases in stream ecosystems. In stream ecosystems influenced by management activities, stable and degrading phases can occur in response to land-use disturbances such as logging (Figure 1). Stable and degrading phases may also by synonymous with stages of natural disturbance regimes for ecosystems. Natural disturbance regimes may display three stages: persistence, reaction and recovery (Wissmar and Swanson, 1990). The persistence stage may correspond with the stable phase while the reaction and recovery stages include the degraded phase (Figure 1). A variety of management concerns point to the need for a simple procedure to detect phases of changing ecosystem conditions. Managers need to know when to decide to alter monitoring frequencies to adjust to short-term and long-term changes in stable and degraded phases. As noted above, these phases can reflect both the natural variability of streams and watersheds as well as influences of

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

229

management activities. Other management concems include problems introduced by the slow acquisition and low precision of data and high costs of long-term monitoring. Managers need to know if a system is in an unstable phase in order to decide if more monitoring activity may be needed. Where stable phases occur, or where longer periods of little change can be identified, less monitoring and lowering of costs may be possible. A procedure for detecting ecosystem phases that might facilitate decisions that need to be made sequentially over time in long-term monitoring situations is presented in Equation 1, the sequential probability ratio test of the mean. It should be noted that this test is an unconventional ecological method that will need further evaluation for proper application to the different needs of various ecosystem monitoring programs. Some important features of Equation 1 include the comparison of current measurements to historical records and the assumption of the removal of the effects of a monitoring station's location from the detection process. The sequential probability ratio test of the mean is based on quality control schemes (Lucas, 1985). This one-sited test is an estimate of the cumulative sum S, defined as Si for the ith sample: S~ = max{0, zi - k + Si - 1 }

(1)

where the zi value is the quantile of the standard normal distribution (zi = {'~? # } / s n ) for the ith sample and s,~ is the standard error (s/~//~) of the mean; # is defined as the mean of parent population and ,~ is the sample mean. The procedure can be demonstrated using the results of a 17-year study of the effects of logging activities on a small ecosystem in British Columbia (Hartman and Scrivner, 1990). The goal is to detect stable and degrading phases in the stream ecosystem. The data includes annual densities of fry emergence for coho salmon (Oncorhynchus kisutch) in Camation Creek. The densities of coho fry that emerged declined during the 17-year study period. The decrease in fry emergence suggested influences of erosion related to logging practices (Figure 2). Mean annual coho fry emergence from 1972 to 1986 approached 30,000 fry. The highest fry emergences occurred during the pre-logging period (1972-1976) mad the lowest following logging (1977-1986) (Figure 2). Survival of eggs to emergence during pre-logging years (1972-1976) was 28.8%. After logging, the 1977-1986 time interval, survival was reduced to 10.9%. Lower percentages of coho eggs surviving to fry was correlated with the reduction in mean particle size of spawning gravel. To apply the procedure (Equation 1), annual emergence is assumed to represent :~?,the sample mean from the annual population of emerging fry. The mean of the parent population (#) is assumed to be 30,000 fry. Estimates of the cumulative sum S, suggests that the distribution of stable and degraded phases in the stream ecosystem corresponds with sequential linear time trends in Si (e.g. 4 or more values) (Figure 3). The stable phase is the increasing trend during the pre-logging period. After logging the 5:i trend indicates two degraded phases, R- 1 and R-2. Si

230

ROBERT C. WISSMAR

Pre-Logging

Post-Logging

19 e19 t~ ~L19

E UJ

60000

50000

40000

LJ

ii 0 tO

30000

0

20000

0

it._

19 ..CI

E

10000

z 1970

!

!

1980

1990

Year Fig. 2. Annual densities of fry emergence for coho salmon (Oncorhynchus kisutch) in Carnation Creek, British Columbia (Hartman and Scrivner, 1990). The densities of coho fry that emerged declined during the 17-year study period suggesting impacts of logging practices. Mean annual coho fry emergence from 1972 to 1986 approached 30,000 fry.

remains in a relatively constant R-1 phase for four years followed by an R-2 phase of six years of sequential decline. R-1 and R-2 could correspond with reaction and recovery stages following disturbance (Figure 1), suggesting that the health of the stream ecosystem continues to decline ten years after logging. The R-I plateau most likely signifies a lag or the beginning of the reaction phase while the subsequent six year decline represents the negative character of the response of the recovery phase (R-2). Both phases corresponded with high inter-annual variability or percentage change in fry emergence (%N) (Figure 3). The percent change in fry emergence is defined as: %N = {No + A N } / N o

(2)

where A N = Nt - No and Nt is the fry emergence one year after No. The negative response of the R-2 phase may signify that the management decisions need to be made with regard to altering long-term monitoring program.

231

LONG-TERM STREAM MONITORING IN THE PACIFIC NORTHWEST

20

a-1

- 300

Z

R-2

0 0 0 O'J

t.n

100

iI '

• 200

,.0

I=

1,1,,I C m

,100

O e-"

-10

O

-20 1970

1980

0

1990

Year Fig. 3. Cumulativesum (Si) and percent change (%N) in coho salmon (Oncorhynchus kisutch) fry emergence in Carnation Creek. The stable phase is the increasing Si trend during the pre-logging period. After logging the S~ trend indicates two degraded phases, R-1 and R-2. Si remains in a relatively constant R-1 phase for four years followedby an R-2 phase of six years of sequential decline. Both the R-1 and R-2 phases correspond with high inter-annual variabilityof percentage change in fry emergence (%N). Si values (Equation 1) are denoted as solid squares and %N (Equation 2) as open circles. R-2 predominates where successive sample means appear below the population mean (#). # is represented by a horizontal line at 30,000 fry in Figure 4. The level for making the decision to take additional monitoring actions can be set at u - z s n , where annual fry emergence falls below 23,676 fry. In this situation z is assumed to be 1.61, where u is 30,000, s~ is 4,361 and Y: is 4,150 or the lowest fry emergence in 1986. It should be noted that the above example for applying the s e q u e n t i a l p r o b a b i l i t y ratio test o f the m e a n to a stream ecosystem is a suggested concept that needs further evaluation for ecosystem level monitoring programs. All the values presented are examples only and will vary with different ecosystems and management objectives. 7. Conclusion Long-term monitoring programs are viewed as key components for bringing together with management organizations, researchers and decision-makers to improve the management of natural resources. Long-term records provide the basis for analysis of environmental assessment objectives, predictions and outcomes which in-turn can be used to modify and improve future projects. Management

232

ROBERT C. WlSSMAR

0 0

0 O3 !._

60000

- 20

0

E UJ

50000 ,10

E

40000'

:3

>, ill

0 tO

0

U

30000

u

+ 1.6(se)

t~ "-z

20000 -10

0 0 JO

0 :> °n

E -! 0

10000

E z

0

,

1970

1980

I -20

1990

Year Fig. 4.

Annual densities of fry emergence and cumulative sum (Si) values for coho salmon in Carnation Creek. The S¢ values are the lowest where successive sample mean values are below the population mean (#). # is represented by a horizontal line at 30,000 fry. The level for making the decision to take additional monitoring actions can be set at u - z s ~ , where annual fry emergence falls below 23,676 fry. In this situation z is assumed to be 1.61. Densities of fry emergence are denoted as closed circles and the cumulative sum Si values as solid squares. (Oncorhynchus kisutch)

organizations that initiate long-term monitoring programs should strive to use monitoring actions and information to facilitate decision-making processes that pertain to conserving and allocating resources for future beneficial use. Careful planning and definition of interactive activities of monitoring programs and information feedbacks allows for evaluation of perceptions of issues pertaining to beneficial uses of resources. The development of long-term monitoring programs requires reviews of background and historical information to provide precise definitions of long-term objectives, planning considerations and monitoring procedures. Procedures need to be identified during the planning process that apply management standards to variable conditions encountered within natural ecosystems and that detect the timing of stable and degraded phases of stream ecosystem development following a disturbance. These procedures, and additional provisions, are essential for improved applications of management standards and perceived thresholds to dynamic stream and watershed ecosystems. Systems where wide spatial and temporal variabilities in environmental conditions prevail.

LONG-TERM STREAMMONITORINGIN THE PACIFICNORTHWEST

233

Acknowledgement This w o r k w a s s u p p o r t e d b y Pacific N o r t h w e s t R e s e a r c h Station G r a n t ( P N W 9 2 0224), U S D A - F o r e s t Service. Contribution no 877, Univ. W a s h i n g t o n S c h o o l o f Fisheries, Seatle, WA.

References Ausmus, B. S.: 1984, 'An argument for ecosystem level monitoring' Environ. Monit. Assess. 4, 275-293. Culhane, E J.: 1987, 'The precision and accuracy of U.S. Environmental Impact Statements', Environ. Monit. Assess. 8, 217-238. Bilby, R. E. and Wasserman, L. J.: 1989, 'Forest practices and riparian management in Washington State: Data-based regulation development', in Gresswell, R. E., Barton, B. A., and Kershner, J. L. (eds.), Practical Approaches to Riparian Resource Management, U.S. Dept. of the Interior, BLM, Billings, MT, pp. 87-94. Bilby, R. E. and Ward, J. W.: 1989, 'Changes in characteristic and function of woody debris with increasing size in streams in western Washington', Trans. Am. Fish. Soc. 118, 368-378. Brouha, R: 1991, 'Fish habitat planning', in W. R. Meehan (ed.), Influences of Forest and Rangeland Management on Salmonid Fishes and Their Habitats, Amer. Fish. Soc. Spec. Pub., 19, pp. 587597. Daggan, S., Cohrssen, J. J., and Morrison, R. E. (eds.): 1987, Preserving Ecological Systems: The Agenda for Long-Term Research and Development, Praeger, NY. Gilbert, R. O.: 1987, Statistical Methods for Environmental Pollution Monitoring, Van Nostrand Reinhold. Gore, J. A. (ed.): 1985, The Restoration of Rivers and Streams: Theories and Experience, Butterworth Publishers, Boston, 280pp. Green, R. H.: 1979, Sampling Design and Statistical Methods for Environmental Biologists, Wiley, New York. Gregory, S. and Ashkenas, L.: 1990, Riparian Management Guide; Willamette National Forest, USDA Forest Service, Pacific Northwest Region. Hartman, G. H. and Scrivner, J. C.: 1990, 'Impacts of forestry practices on a coastal stream ecosystem, Carnation Creek, British Columbia', Can. Bull. Fish. Aquat. Sci. 223, 148. Hildebrand, S. G., Barnthouse, L. W., and Suter, G. W.: 1987, 'The role of basic ecological knowledge in environmental assessment', in Daggan, S., Cohrssen, J. J., and Morrison, R. E. (eds.), Preserving Ecological Systems: The Agenda for Long-Term Research and Development, Praeger, NY, pp. 51-69. Karr, J. R.: 1987, 'Biological monitoring and environmental assessment: Conceptual framework', Environ. Management 11(2), 249-256. Likens, G. E. (ed.): 1988, Long-Term Studies in Ecology: Approaches and Alternatives, Springer Verlag, NY, 214pp. Lucas, J. M.: 1985, 'Cumulative Sum (CUSUM) Control Schemes', Comm. Statistics - Theory and Methods 14, 2689-2704. MacDonald, L., Smart, A., and Wissmar, R. C.: 1991, Monitoring Guidelines to Evaluate Effects of Forestry Activities on Streams in the Pacific Northwest and Alaska, Edward Brothers Press, Ann Arbor, MI. Maurizi, S. and Poillon, E (eds.): 1992, Restoration of Aquatic Ecosystems: Science, Technology and Public Policy, National Academy Press, Washington, DC, 552pp. Meehan, W. R. (ed.): 1991, Influences of Forest and Rangeland Management on Salmonid Fishes and Their Habitats, Amer. Fish. Soc. Spec. Pub., 19, 751pp. National Research Council: 1990, Forestry Research: A Mandate for Change, National Academy Press, Washington, DC, 84pp. Peterson, J., Schmoldt, D., Peterson, D, Eilers, J., Fisher, R., and Bachman, R.: 1992, Guidelines for Evaluating Air Pollution Impacts on Class I Wilderness Areas in the Pacific Northwest, USDA

234

ROBERT C. WlSSMAR

Forest Service, Pacific Northwest Station, Portland, OR, Gen. Tech. Report PNW-GTR-299. Platts, W. S. et al. (12 co-authors): 1983, Methods for Evaluating Stream, Riparian and Biotic Conditions, USDA Forest Service, Intermountain Forest and Range Experiment Station, Ogden, UT Gen. Tech. Report, INT-138. Schaeffer, D. J., Herricks, E. E., and Kerster, H. W.: 1988, 'Ecosystem health: I. Measuring ecosystem health', Environ. Management 12, 445-455. Schindler, E. W.: 1987, 'Detecting ecosystem responses to anthropogenic stress'. Can. J. Aquat. Sci. 44, 6-25. Tomlinson, P. and Atkinson, S. E: 1987, 'Environmental audits: Proposed terminology', Environ. Monit. Assess. 8, 187-198. Turner, M. G. (ed.): 1987, Landscape Heterogeneity and Disturbance, Springer Verlag, NY, 214pp. Wissmar, R. C. and Swanson, E J.: 1990, 'Landscape disturbance and lotic ecotones', in Naiman, R. J. and Decamps, H. (eds.), Ecology and Management of Aquatic-Terrestrial Ecotones, Parthenon Press, London, pp. 65-89.

The need for long-term stream monitoring programs in forest ecosystems of the Pacific Northwest.

Concepts, planning and design procedures are examined that are needed in the development of long-term stream monitoring programs in forested regions. ...
969KB Sizes 0 Downloads 0 Views