Regional Scale Ecological Risk Assessment - Chapter 8 docx

THE RELATIVE RISK MODEL Regional ecological risk assessment is concerned with describing and estimatingrisks to environmental resources at the regional scale or risks resulting from regi

Trang 1

Developing a Regional Ecological Risk Assessment: A Case Study of a Tasmanian Agricultural Catchment* Rachel Walker, Wayne G Landis, and Philip Brown

CONTENTS

Introduction 160

The Relative Risk Model 160

Problem Formulation 161

The Risk Region 161

Defining Assessment Endpoints within the Mountain River Catchment 162

Identifying Stressors in the Region 164

Identifying Habitats in the Region 164

Interaction of Stressors and Habitats — Risk Hypotheses in the Conceptual Model 166

Risk Analysis Using the Relative Risk Model 167

Identifying Risk Areas 168

Ranking Stressors 168

Ranking Habitats 170

Relative Risk Calculations Using the Conceptual Model 170

Risk Characterization 171

Sensitivity Analysis 172

A Basis for Action 175

Discussion: Regional Risk Assessment 175

Acknowledgments 176

References 176

* Previously published in the Journal of Human and Ecological Risk Assessment, 7(2), 2001 Reprinted with permission.

L1655_book.fm Page 159 Wednesday, September 22, 2004 10:18 AM

Trang 2

160 REGIONAL SCALE ECOLOGICAL RISK ASSESSMENT

INTRODUCTION

A regional ecological risk assessment was conducted for the Mountain Rivercatchment in Tasmania, Australia The relative risk model (RRM) was used inconjunction with geographic information systems (GIS) interpretations Stakeholdervalues were used to develop assessment endpoints, and regional stressors and habitatswere identified The risk hypotheses expressed in the conceptual model were thatagriculture and land clearing for rural residential development are producing multiplestressors that have potential for contamination of local water bodies, eutrophication,changes in hydrology, reduction in the habitat of native flora and fauna, reductions

in populations of beneficial insects in agricultural production systems, increasedweed competition in pastures, and loss of aesthetic value in residential areas In therisk analysis the catchment was divided into risk regions based on topography andlanduse Stressors were ranked on likelihood of occurrence, while habitats wereranked on percentage of land area Risk characterization showed risks to the main-tenance of productive primary industries were highest across all risk regions, fol-lowed by maintenance of a good residential environment and maintenance of fishpopulations Sensitivity analysis was conducted to show the variability in risk outcomesstemming from uncertainty about stressors and habitats Outcomes from this assessmentprovide a basis for planning regional environmental monitoring programs

THE RELATIVE RISK MODEL

Regional ecological risk assessment is concerned with describing and estimatingrisks to environmental resources at the regional scale or risks resulting from regional-scale pollution and physical disturbance(Hunsaker et al 1989) Within any catchmentregion there are various stressors impinging on the quality of the environment Without

a framework it is difficult to objectively assess the risks associated with multiplestressors The RRM as developed by Landis and Wiegers (1997) is a framework forranking and comparing the risks associated with multiple stressors It is a useful toolfor describing and comparing risks to valued resources within a catchment

The RRM was developed for a regional risk assessment for the fjord of Port Valdez,Alaska (Wiegers et al 1998) This chapter reports on the application of the RRM in amore localized region in southern Tasmania, Australia The aim of this work was to usethe RRM as a tool to put catchment issues in context and highlight issues that needed

to be further addressed

RRM methodology essentially mirrors the traditional three-phase risk assessmentapproach: problem formulation, analysis, and risk characterization, but requires amodification of the traditional approach Expanding an assessment to cover a regionrequires consideration of larger scale, regional components: sources that releasestressors, habitats where the receptors live, and impacts to the assessment endpoints

In the problem formulation phase of the relative risk assessment, the scope ofthe assessment is defined; at this stage the values of regional stakeholders areinfluential in determining assessment endpoints Generic goals for regional riskassessment include explanation of observed regional effects, evaluation of an actionL1655_book.fm Page 160 Wednesday, September 22, 2004 10:18 AM

Trang 3

DEVELOPING A REGIONAL ECOLOGICAL RISK ASSESSMENT 161

with regional implications, and evaluation of the state of a region (Suter 1990).Regional stressors and habitats are identified in the problem formulation phase

In the risk analysis phase the stressors and habitats are ranked based on theirlikelihood of occurrence within the risk region The interaction between stressorsand habitats is considered when total relative risk calculations are made for eachstressor and habitat In the risk characterization phase the risks for stressors andhabitats are compared Stressors with the greatest potential for ecological impactand habitats most at risk are both identified This provides a basis for discussionsabout management of the region

It is particularly apparent at the regional scale that not all components of theenvironment can be measured, tested, modeled, or otherwise assessed (Suter 1993a)

In addition, there is a large degree of spatial and temporal variability On a regionalscale there is a large degree of uncertainty preliminary risk assessment such as this.However, this should not stop the assessment from proceeding Uncertainty should

be recognized as an inherent component of each stage of the risk assessment andaddressed at each stage rather than at the conclusion of the risk analysis A sensitivityanalysis can be performed at the conclusion of the risk analysis to determine howuncertainty is influencing the overall risk rankings

PROBLEM FORMULATION The Risk Region

As noted by Suter (1993a), a catchment lends itself to being an easily defined riskregion for aquatic-borne contaminants The catchment considered in this assessment isthe Mountain River catchment in southern Tasmania, Australia It covers approxi-

2The Huon Valley is a major horticultural region The main horticultural cropsare apples, cherries, stone fruit, and berries Apples are by far the biggest crop, and65% of the Tasmanian apple crop is grown in the Valley, with an estimated marketvalue of $28 million Other primary industry enterprises include beef cattle produc-tion, mushroom farming, herbs, honey, and cut flowers The Huon Valley is a popularresidential locality for urban commuters who have no financial dependence on theland, but value the aesthetic and lifestyle benefits of living in a rural environment.There is a significant level of public interest and concern in the Huon Valleyabout environmental issues generally, and waterways in particular Catchment man-agement in the Huon Valley was formally instigated with the establishment of theHuon Healthy Rivers Project initiated in 1995 with funding provided through federaland local governments The Huon Healthy Rivers Project is an ongoing project thataims to promote environmental awareness and provide a resource base for commu-nity projects

Information in this assessment was obtained from a number of sources, ularly publications produced by the Huon Healthy Rivers Project and personalcommunication with Huon Healthy Rivers Project officers who facilitated variousL1655_book.fm Page 161 Wednesday, September 22, 2004 10:18 AM

partic-mately 190 km and is located in the Huon Valley region (Figure 8.1)

Trang 4

community forums There has been no extensive or consistent environmental itoring of freshwater bodies within the valley other than basic water quality dataavailable through state agencies

mon-Defining Assessment Endpoints within the Mountain River Catchment

Assessment endpoints represent the social values to be protected and serve as apoint of reference for the risk assessment The values to be protected in a regionmay be described in terms of characteristics of its component populations andecosystems or in terms of characteristics of the region as a whole (Suter 1993a).The goals of the local community were used as a starting point for developingassessment endpoints A community forum, held in 1998 to identify water valuesfor Mountain River as a starting point for setting environmental flows for the river,identified the following issues to be important: improve water quality (particularlydecreased E coli counts), maintain/establish water of drinkable and irrigable quality,maintain habitats for aquatic animals, maintain water in suitable volumes to sustainagriculture, maintain catchment quality for town water supply, maintain water forswimming, maintain water for trout fishing, maintain or improve beauty of the river,and maintain seasonal fluctuations between summer and winter flows of the river

In a 1999 catchment community forum, local residents created an image of theirprefered catchment having the following characteristics: clean water that is safe fordrinking and swimming, sustainable landuse practices, optimum stream flow, naturalvegetation along the riverbanks, an active and responsible community, and an attrac-tive setting for picnics As noted by Steel et al (1994), analysis of survey data shouldconsider relationships between survey responses and stakeholder backgrounds.Length and location of residence, occupation, education, and other factors caninfluence stakeholder values This particular “community” forum was not well

Figure 8.1 Location of Mountain River catchment in southern Tasmania, Australia.

Trang 5

attended by local farmers, and the values stated may not necessarily representpriorities for primary producers It is vital that assessment endpoints be determinedwith a conscientious and intelligent effort to represent the values of the entirecommunity

Beginning with the water body and expanding across the catchment, assessmentendpoints were identified based on the views expressed by stakeholders, discussionwith resource managers, and expert judgment The assessment endpoints were iden-tified as:

Guide-lines for Fresh Water Quality

the regulatory body, Tasmanian Inland Fisheries)

established by the regulatory body, Department of Primary Industries, Water and Environment)

to less than 10% ground cover

Suter (1990) states that good assessment endpoints should have the followingcharacteristics: social relevance, biological relevance (function of its implicationsfor the next higher level of biological organization), unambiguous operational def-inition, accessibility to prediction and measurement, and susceptibility to the hazard

We compared the above assessment endpoints to Suter’s criteria Water qualityparameters to meet or exceed Australian and New Zealand Guidelines for FreshWater Quality (2000) are currently the only assessment endpoint that meets all ofSuter’s criteria At the time of writing the state fisheries agency was in the process

of establishing quantitative goals for Tasmanian brown trout fisheries, which is thestate’s most popular inland fishery, and for establishing an environmental flow forMountain River Quantitative goals have currently only been set for the recoveryplans of the rare and endangered native galaxias (Crook and Sanger 1997), none ofwhich occur in Mountain River

The assessment endpoints of maintenance of productive primary industries andlandscape aesthetics are intuitively understood, but not well defined These endpoints

do not meet Suter’s criteria, but clearly an imperfect definition must not excludethem; maintenance of productive primary industries is of utmost importance in aprimarily agricultural catchment For the purposes of this preliminary relative riskranking, this assessment endpoint is not operationally defined; instead, generalknowledge of good soil and water management practices is applied to it Work byLandis and McLaughlin (2000) is providing a conceptual framework for quantifyingsustainability, although it is unlikely that an unambiguous operational definition forquantifying sustainable agriculture will be achieved because of the huge diversity

of inputs to agriculture It is possible, however, to quantify the sustainability eters of individual inputs to agriculture, for example, using water quality criteriaand regional soil databases

param-L1655_book.fm Page 163 Wednesday, September 22, 2004 10:18 AM

Trang 6

Similarly, landscape aesthetics is not operationally defined, but can be understood

as meaning that Mountain River is a nice place to live Other assessment endpointsdirectly impinge on this, particularly the quality of the natural environment asmeasured through water quality, water flow, weed infestation, aquatic life, and factorsaffecting agriculture such as soil stability and climate

Identifying Stressors in the Region

The issues of environmental concern identified in the Huon Healthy RiversProject were categorized in terms of the stressor and corresponding ecosystemstressors identified were anthropogenic The effects of seasonal floods can beenhanced or mitigated by regional land management practices

Out of all the stressors identified for the Huon catchment in Table 8.1, the onlystressors considered relevant in the risk assessment for Mountain River catchmentwere agriculture and land clearance for rural residential development No large-scaleforestry activities occur within the catchment, although it is possible there may besome paddock-scale tree plantations on individual farms No aquaculture occurswithin the catchment Mountain River is too small for boating, and recreationalpursuits in the catchment are mainly hiking, horse racing, fishing, and swimming,which were considered to have negligible impact

Agricultural stressors in the Mountain River catchment were identified as ticides used in orchards, fertilizers (pasture, orchards, and other cropping activities),pumping irrigation water from the river, weed infestation, and clearing of nativebush for farmland (Table 8.1) Another stressor that could be included under theumbrella heading of agriculture is contaminated sites because of possible copper, lead,and arsenic residues in the soil from previous use of orchard pesticides containing theseelements It was decided to omit contaminated sites from this risk assessment becausethe focus is on risks associated with current agricultural practices In addition,introducing contaminated sites into the risk assessment involves considerable uncer-tainty Currently the actual extent of contamination, if any, is unknown An intensiveregional soil testing program is required before contaminated sites should be con-sidered as a stressor

pes-Stressors resulting from land clearing for rural residential development wereidentified as bacteria from septic tank effluent, clearing of native bush for residentialpurposes, nutrients from households, pumping water from the river for garden andhousehold use, and weed infestation

Identifying Habitats in the Region

Human exclusion from ecosystems has been symbolic of a long-held belief thatsomewhere there exists a reference, pristine ecosystem It is more realistic to rec-ognize that humans are participants in most ecosystems; indeed agricultural ecosys-tems are created and maintained by humans It was decided in this risk assessment

to recognize anthropogenic habitats in the same way as natural habitats This hasrecently been considered as a valid risk assessment approach because changes inL1655_book.fm Page 164 Wednesday, September 22, 2004 10:18 AM

response variable (Table 8.1) With the exception of seasonal flooding, all the

Trang 7

Table 8.1 Anthropogenic Stressors and Ecosystem Response Variables Identified

in the Huon Valley Anthropogenic Stressor Ecosystem Response Variable

LAND CLEARANCE AND

RURAL SUBDIVISION

Species and habitat destruction Soil erosion and landslips Increase in frequency of erosive flood events Increase in environmental weeds — willows, blackberries, ragwort, gorse, pampas grass

INTENSIVE AGRICULTURE

Fertilizers and animal waste

Agricultural runoff causing eutrophication of freshwater bodies Toxic algal blooms in the estuary affecting estuarine species Pesticide contamination of

soils and water through

spray drift, spillage and

RURAL AND COASTAL

AREA DEVELOPMENT

River and coast modification altering the habitat of local species Wetland degradation; reduction of the “biological filtering” capacity of the estuary

Septic tank effluent — effluents from improperly maintained septic tanks have contaminated waterways and groundwater

in various locations Refuse disposal site leachate — current public sites are located

at Huonville, Geeveston, Cygnet; former sites were located at Glen Huon and Judbury; older and former public and private sites are spread throughout the municipal area; contaminants

of unknown types and quantities discharged to waterways Pumping drinking and household water from local waterways, reducing stream flow and changing hydrology affecting the microhabitat of aquatic species

Nutrient input from sewage; sewage treatment plants are located at Ranelagh, Cygnet, and Geeveston; sewage lagoons

at Huonville, Dover, Southport; Franklin sewage currently discharged into the Huon River

Solid waste management — public landfill facilities at Geeveston and Cygnet; waste transfer stations at Cygnet, Southport, Dover, and Huonville; private contractor also provides recycling facilities at each site

Untreated stormwater containing unknown types and quantities

of contaminants FORESTRY Soil erosion and landslips

Nutrient runoff Road building causing siltation of waterways Environmental weeds

AQUACULTURE Nutrients from fish waste, uneaten food, and disposal of net

wash effluent causing nutrient enrichment of the estuary and increasing probability of toxic algal blooms

Escaping fish possibly competing with native species RECREATIONAL

PURSUITS

IN THE VALLEY

Ballast water introducing pest species Boat pollution (fuel, sewage waste, rubbish) Information from the Huon Healthy Rivers Project (1997) was used as a basis for this table.

Trang 8

ecological systems result in risks to cultural resources, economic activity, and quality

of life because of the numerous and important services of nature (Suter 1999a).Moreover, ecological risks can often be considered as risks to the sustainability ofthe activities being assessed (Suter 1999b)

Based on landuse in the catchment, five different habitat categories were tified (Table 8.2) Given the diversity of stressors, there are a variety of impacts thatcould occur within each habitat

iden-Interaction of Stressors and Habitats — Risk Hypotheses in the Conceptual Model

At this point in our preliminary risk assessment, stressors and habitats in theregion have been identified The values of various stakeholder groups have beenconsidered in the formulation of assessment endpoints A conceptual model of theregion showing the interaction of stressors, habitats, and the potential for impactsdescribes the approach that will be used for the risk analysis phase It is a graphicalsummary of the risk hypotheses being assessed within the catchment (USEPA 1992).Conceptual models are representations of the assumed relationships between sourcesand effects (Suter 1999a) The conceptual model shown in Figure 8.2 representsassumed interactions of stressors and habitats within the catchment It containsuncertainty; however, it is adopted as an operating tool in the absence of morecomplete knowledge

The risk hypotheses shown in Figure 8.2 assume that agriculture and landclearing for rural residential areas produce multiple stressors that have potential for

Table 8.2 Habitats Identified within the Mountain River Catchment

Aquatic All water bodies are included in this category,

although the emphasis is on larger waterways in the catchment, in particular Mountain River and Crabtree Rivulet

Contamination of the water body

Eutrophication Changes in hydrology Native

vegetation

Includes all native vegetation types mapped

in the TasVeg™ 2000 series; priority vegetation associations in the Mountain River catchment are Eucalyptus ovata, E

amygdalina, E tenuiramis, E globulus

Reduction in the habitat of native flora and fauna

Orchard Includes all land mapped as orchard; major

orchard crops are apples, followed by cherries

Reductions in populations of beneficial insects

Weeds competing with orchard trees, especially during establishment Pasture Includes all pastures used for grazing sheep,

horses, goats, and for cutting hay; limited crop production in Mountain River catchment, but any occasional cropping that does occur is also included in this category

Weeds competing with pasture and crop species; weeds can also decrease quality of pasture and decrease price of cut hay Residential Includes the area around each residence that

is actively used or maintained by the resident; also includes the residence

Loss of aesthetic value

on chosen assessment endpoints is given in Figure 8.2 The conceptual model

Trang 9

contamination of local water bodies, eutrophication, changes in hydrology, reduction

in the habitat of native flora and fauna, reductions in populations of beneficial insects,increased weed competition in pastures, and loss of aesthetic value in residential areas.Particular emphasis has been placed in this regional risk assessment on theconceptual model as a tool for visually interpreting the relative risk calculations.This is described in the Risk Analysis section

RISK ANALYSIS USING THE RELATIVE RISK MODEL

Much of the input data for the risk analysis in this assessment came from landusepatterns shown in the Tasmania 1:25,000 Series The map sheets used were Longley

5024 (Edition 2, 1988) and Huonville 5023 (Edition 2, 1987) Digitized map dataare supplied to the Australian public on a cost recovery basis and there is only alimited amount of digitized data available Landuse themes, including vegetation,were not available in digital format so it was necessary to digitize vegetation patternsfrom paper maps The maps were scanned and on-screen digitized Vegetation themeswere transformed from scan unit coordinates to the Universal Transverse Mercatorprojection using Shape Warp 2.2 ArcView version 3.1 (Environmental SystemsResearch Institute, Redlands, CA) was the GIS software used in this assessment

Figure 8.2 Conceptual model — hypothesized interactions between stressors and habitats

in Mountain River The rankings and calculations shown here are for risk region

4 (Figure drawn by Angela Schuler.)

Pumping tion Water 2 Weed Infestation 0 Clearing of Native Bush for Farmland 4 Clearing of Native Bush for Resi- dential Purposes 6 Bacteria from Septic Tank Effluent 4 Nutrients from Households 4 Pumping Water for Garden and Household Use

Irriga-Weed Infestation 2

Sum of stressors

in risk area = 22 sure to habitats in risk area = 92Sum of potential stressor

expo-Total risk to assessment endpoints in risk area = 766

Loss of Aesthetic Value Residential 4

Competition from Weeds

2 Pasture

Weeds Competing with Orchard Trees

in Establishment

0 Orchard Reductions in Popu- lations of Beneficial Insects

4 Native Vegetation Reduction in Habitat

of Flora and Fauna

Changes in Hydrology Eutrophication

Contamination of Water Body

Effects

Event in habitat has potential to impect assessment endpoint

Assessment Endpoint

Water Quality Total risk to assessment endpoint

60

Maintenance of Local Flsh Populations Total risk to assessment endpoint

120

Maintenance of Adequate Stream Flow Total risk to assessment endpoint

60

Maintenance of Adequate Native Streambank Vegetation Total risk to assessment endpoint

Total risk to assessment endpoint

24

374

Maintenance of Productive Primary Industries

Maintenance of a Good Residential Environment 128

Trang 10

Identifying Risk Areas

The ranking criteria described below for stressors and habitats are primarilybased on landuse Landuse patterns generally change dramatically between the upperand lower reaches of a catchment It would be unrealistic and unachievable to attempt

a risk ranking for the entire catchment It is more practical and relevant to dividethe region into subareas or risk regions so that stressors and habitats within a specificsubarea can be better considered This also allows comparison of risks from differentstressors to specific habitats within different catchment areas

An incremental gradient of human activity occurs as Mountain River flows downthrough the catchment The intensity of agriculture, orcharding, and residentialdevelopment increases The risk regions in Figure 8.3 were chosen to match thisgradient of human activity and of the natural boundaries determined by contoursand tributaries flowing into Mountain River Aligning risk regions with the flow oftributaries to Mountain River was very important Even though two tributaries may

at some point only be separated by a few kilometers, they may flow through verydifferent landuse activities before they join the main channel, ultimately contributingvery different inputs to the main channel

Figure 8.3 Mountain River catchment risk regions Risk regions have been mapped according

to flow of tributaries into Mountain River and the incremental gradient of human activity in the lower reaches of the catchment.

Trang 11

although they could make approximations A total of 39 household water pumps areinstalled along the river, although locations of the pumps could only be estimatedbased on residential density along the river

To collect quantitative data about the extent and severity of each stressor in eachrisk region is a task not warranted for a preliminary risk assessment of the catchment.All data currently available about the catchment were collated, but no additionalfield data were collected for this preliminary risk assessment

We decided in this preliminary risk assessment to use expert knowledge of theregion and landuse maps to qualitatively rank stressors The ranking criteria andpoints assigned were:

The distance between risk categories is assumed equal, i.e., stressors ranked as 6are not by definition three times larger than those ranked as 2 This also applies tothe habitat ranking criteria given in Table 8.3

Uncertainty is obviously a significant consideration at this point of the riskassessment However, it was planned to undertake a sensitivity analysis to determine

Table 8.3 Ranking Criteria for Mountain River Habitats

Habitat Rank Criteria and Assigned Points Uncertainty

Aquatic 6; The aquatic habitat was given a single, high

ranking because all activities that occur within a catchment ultimately impinge on the waterway In addition, the risk regions all include a section of the waterway, so the aquatic habitat was considered to be a highly ranked habitat in all risk areas

Assumption that a high ranking is justified across all regions

Native

Vegetation

6; 23–37% (of total catchment native vegetation found within the risk region) 4; 11–22%

2; 10%

0; < 1%

Accuracy of 1:25,000 maps

Orchard 6; 41–46% (of total catchment orchards found

within the risk region) 4; 15–40%

2; 1–14%

0; <1%

Accuracy of 1:25,000 maps

Pasture 6; 29–35% (of total catchment pastures found

within the risk region) 4; 16–28%

2; 1–15%

0; < 1%

Accuracy of 1:25,000 maps Classification of pasture vs vacant land and home gardens

Residential 6; Many ratepayers (approx > 5000)

4; Not so many ratepayers (approx > 3000 2; Few ratepayers (approx > 1000) 0; No ratepayers

No localized population data available for different areas

of the catchment;

assumption is that all residences are contributing equally to the source

Trang 12

if the stressor rankings had a significant effect on the relative risk ranks In this way,the effect of an incorrect stressor ranking on the overall relative risk outcomes could

be compared The sensitivity analysis is described in the Risk Characterizationsection

The ranking criteria and points assigned for severity of weed infestation werebased on standard categories for mapping weed density as used in the surveyconducted by the Huon Healthy Rivers Project

Ranking Habitats

Habitats were ranked according to the proportion of a particular habitat within

a region To determine the proportion of a particular habitat within a risk region,map themes were manipulated and planimetric areas measured using ArcView soft-ware Habitat ranks and uncertainties associated with the ranking are described inThe major source of uncertainty in establishing the ranking criteria for habitatsstems from the 1:25,000 maps The content of these maps was determined fromaerial photography undertaken in 1986 Obviously, there would have been changes

in landuse since that time, so the exact proportion of different vegetation and landusetypes would have changed However, in the absence of other map data, the 1:25,000series maps must be used Since 1986, the major changes in landuse in the catchmenthave been the subdivision of pasture into rural residential blocks The number ofresidences in the catchment is now greater than indicated on the maps and, conse-quently, the true extent of pasture may have been overestimated However, therehave not been other significant landuse changes in the catchment, and for thepurposes of this preliminary risk assessment the 1:25,000 maps were consideredadequate

Relative Risk Calculations Using the Conceptual Model

all the interactions between stressors and habitats being considered in this riskassessment It has been produced as a spreadsheet so that it can simultaneouslymathematically describe the risks associated with the stressors and habitats found

in each risk region, based on the assumed interactions between stressors and habitats.These assumed interactions are indicated by the exposure and effects arrows.There is a number above each stressor and habitat category in Figure 8.2 Thesenumbers are the risk rankings for the stressor and habitat in a given risk region (theexample reproduced here is for risk region 4) This number describing risk ranking

is in a cell that is part of a spreadsheet formula Spreadsheet formulas are used tocalculate the risks indicated in Figure 8.2, that is, the sum of stressors within the

Trang 13

risk region, sum of potential stressor exposure within the risk region, total risk toassessment endpoints within the risk region, and total risk to each assessmentendpoint Incorporating spreadsheet calculations into the conceptual model means

it is easy to compare total risks between different risk regions and for differentrankings of stressors and habitats The assumed interactions between stressor andhabitat remain constant; only the risk rankings change

The spreadsheet formulas used for calculating risk are:

inter-actions where an exposure arrow indicates the stressor has potential to impact habitat

exposure arrow indicates the stressor has potential to impact habitat AND an effects arrow indicates that an event in the habitat has potential to impact assessment endpoint

endpoint)

The use of exposure and effects arrows serves to ensure that only realisticinteractions are included in the conceptual model and the risk ranking Not everystressor has the potential to impact every habitat, nor has every stressor the potential

to impact every assessment endpoint A relative risk ranking cannot simply be a sum

of stressor×habitat; the interactions assumed in the conceptual model must beWhen Wiegers et al (1998) did their relative risk calculations for the Port Valdezregional risk assessment, they used an exposure filter and effects filter to ensure thatonly realistic interactions were included in the risk calculations Their filteringmethod included 1 in the risk calculations that represented realistic interactions and

0 in the interactions that represented unrealistic interactions Their method did notuse the conceptual model as visual reference, so the filtering method involvedindividually assessing each stressor/habitat/impact interaction and questioningwhether it was a realistic scenario

RISK CHARACTERIZATION

has occurred Risks to productive primary industries are greatest, which is notsurprising considering the diversity of inputs to agriculture After primary industries,risks to the residential environment and maintenance of fish populations are greatest.Degradation of water quality in the region had the greatest impact on assessmentendpoints Initially it was surprising that risks to native vegetation were compara-tively low, given the development that has occurred in the region However, this riskoutcome is accurate because there is actually very little native vegetation remaining

in the region (habitat rank is 2), so risks to this habitat type are relatively low

accounted for These interactions are indicated by the linking arrows in Figure 8.2

A comparison of risks to assessment endpoints in risk region 4 is shown in Figure8.4 This is where the most agricultural and residential development in the catchment

Định dạng
Số trang	27
Dung lượng	1,6 MB