1. Read: The Case Study of Monitoring In Newfoundland - Beanlands and Duinker pages 110 - 117. When we were discussing WHO SHOULD PAY, we referred to bias and the general perception that if some sector pays for a study, the results are inherently biased in their favor

What is your opinion? Is there any solution to this real or perceived problem?

2. Will you be recommending a monitoring program as part of the mitigation plan for the new development you are working on from Assignment #1? If so, briefly explain what will be included in your monitoring plan.

.

bility of an oil well blowout (i.e., impacts were predicted be measured during and after an oil spill. While the propoassuming the occurrence of a worst-case blowout). As the nents supported this concept, there was little indication probability of such a blowout was claimed by the propo- that such rigorous 'baselines' had been established or were nents to be very low, the overriding conclusion of the being undertaken concurrent with drilling activities. (An assessment was that no significant adverse impacts were exception was the intensive monitoring programme on sealikely to accrue from the proposed project. bird distribution and abundance. This study began in 1978 The predictions can be considered to be cast in a semi- but was terminated after 1979, reportedly because of a sigquantitative form. For example, as indicated through the nificant change of personnel within a government departframework established for impact significance, a major ment). The ElS suggested that such studies would entail a impact on polar bears occupying ice habitat in late winter cooperative effort between industry and government. In the implies that the entire regional population may decline or opinion of one person interviewed, the bulk of the responsimay change in distribution to an extent beyond which the bility for these studies should lie with government, espeformer population status would not be achieved for several cially for species and resources for which various agencies generations. Such semi-quantitative forms are much more have the mandate of management. Regardless of who amenable to post-impact testing than are vague, qualitative should carry out such studies, their absence from this statements. Nevertheless, this does not imply that the pre- assessment seriously jeopardizes the interpretation of the dictions are fully testable, since testability depends on other factors such as adequate, pre-impact control data, the technical ability to measure changes of the magnitude pre- Mitigation and Contingency dicted and actual occurrence of the impact. Indeed, the testability of the predictions made in this assessment is The impact assessment studies provided the basis for the questionable on the grounds of an absence of adequate oil spill contingency plan for this project. The detailed conbaseline descriptions of natural variation. tingency manuals were based primarily on resource maps The proponents and consultants adopted a 'worst-case'_._ which identified high priority areas and species for protection. The manuals also outlined the measures most approinformation or insight or both. The documentation empha- priate in undertaking such protection. sized that, in such cases, impacts were placed in a cate- As well as helping to specify appropriate mitigation and gory of higher severity than what initially may have been countermeasures equipment, the assessment studies also thought appropriate. All cases where this occurred were were influential in choosing a site for base camp operations noted in an oil blowout impact matrix (Imperial Oil Ltd. et and in improving other aspects of the rig servicing pro- blowout on polar bears occupying ice in late winter was ini- ment and locations, these were determined largely by the tially labelled a moderate impact but was recast as a major engineering possibilities for the project and by the impact in light of insufficient data. There was undoubtedly probabilities of success in finding a hydrocarbon deposit some comfort underlying this strategy; the EIS dismissed within the exploration acreages. the risk posed by the project to the species of concern on the basis of the very low probability of impact occurrence. HYDROELECTRIC DEVELOPMENT Monitoring Background The EIS called for environmental monitoring to be undertaken, in the event of a major release of oil, to determine Early in 1975, Newfoundland and Labrador Hydro the fate of the oil and its environmental effects. It was (Hydro) informed the Newfoundland Department of Constated that the monitoring strategies and techniques as sumer Affairs and Environment (DCAE) of its plans to detailed by Cox and others (1980) would be followed. This examine four hydroelectric developments (including the manual summarized the proceedings of a workshop on oil Upper Salmon Hydroelectric Development) as generation spill studies and the main conclusion was as follows: options. A comparative preliminary environmental impact "The workshop participants strongly endorse the con-_ assessment of these proposed projects was undertaken cept of a few comprehensive, well-planned, statistically and completed in 1976 (Airphoto Analysis Associates Convalid studies of oil spillages rather than many inconclu- sultants Limited/Beak Consultants Limited, 1976). This inisive studies which are the current norm. Oil spill impact tial work concluded with recommendations for an array of analyses require highly sophisticated, expensive tech- more detailed studies in support of a full-scale impact niques which must be performed with sufficient replica- assessment for the Upper Salmon project. Hydro initiated a tion to provide data amenable to rigorous statistical number of these studies in 1978, and tabled an EIS for the testing." project early in 1980 (Newfoundland and Labrador Hydro, They also emphasized the need for time controls in oil given for the project, but the approval was contingent upon spill impact studies, that is, an indication of the range of the findings of certain studies and upon the continuation of natural variability, in time and space, of variables that will certain other investigations. Subsequent to the submission of the EIS, Hydro was that the high degree of control the committee wielded on required to prepare an environmental information report which studies must be undertaken and how they must be (Newfoundland and Labrador Hydro, 1981a) to address done, was rather stifling to innovative and creative thinking some of the deficiencies of the assessment; namely, to pro- on the part of the proponent and the consultants. Others vide greater detail on the monitoring, research and mitiga- maintained that since environmental assessment is a govtion of environmental impacts related to the project. Con- ernment process established to provide answers to governstruction of the project has proceeded generally as ment, then government bodies have the responsibility to expected, and operation is to commence late in 1982. Envi- ensure that what is done in an assessment meets their ronmental studies related to the project are still being expectations. undertaken and will be described later. This impact assessment was important for the Newfound- Scoping and Study Planning land environmental assessment community for a number of reasons. It was one of the first assessments to be adminis- The impact assessment for the Upper Salmon project tered under the new legislated procedures for impact gave the investigators the rare luxury of undertaking the assessment adopted by the Newfoundland Department of assessment in stages over a relatively long period prior to Environment. As well, a new concept, that of the 'environ- the beginning of construction. The scope of the investigamental monitor' or surveillance person, began to mature tion and the planning for studies in the full-scale impact with the assessment of the Upper Salmon development. assessment were based largely on recommendations arisWhile Hydro employed its first monitor for the earlier Hinds ing from the preliminary comparative assessment (Airphoto Lake Hydroelectric Project, it became customary with this Analysis Associates Consultants Limited/Beak Consultants assessment for DCAE to employ a monitor, in parallel with Limited, 1976). Even following the completion of the first a Hydro monitor, for each hydroelectric development. The round of studies, three years were still available before duties of the monitors are to ensure environmentally sound project operation (but during project construction) to conconstruction practices and to observe project-related envi- duct subsequent studies - studies designed to answer ronmental events. much more specific ecological questions. The approach exemplified by this assessment borders on Objectives the ideal. Seven or eight years will have been available for The EIS (Newfoundland and Labrador Hydro, 1980a) scientific investigations prior to the beginning of project clearly states that the objectives of the assessment were (i) operation. The tiered nature of the studies provided an excellent vehicle for effective study planning. Early prelim- or eliminate undesirable effects. On the other hand, the objectives for the individual studies supporting this assess- From among a host of possible concerns, emphasis was ment were somewhat less precise. In some cases, it was eventually focussed on caribou (the disturbance of migraunclear how the specific studies were to provide a meaning- tion patterns and the long-term viability of local herds) and ful contribution to the overall assessment. In this regard, a salmonids (reduced recruitment to lakes and reservoirs nested set of objectives would have been helpful. through loss of spawning habitat by inundation or by migration blockage). One might reflect on whether it was neces- Guidelines sary to undertake all of the earlier studies before these concerns could be identified. In retrospect, interviewees The guidance provided to Hydro for the Upper Salmon indicated that some of the individual studies did not conimpact assessment consisted of two elements. One con- tribute in any substantial way to identifying or addressing sisted of a four-page outline entitled "General Guidelines major concerns in the assessment. Lessons have been for the Content of an Environmental Impact Statement." As learned and a reduced effort will be needed for identifying the title suggests, these guidelines provided a generalized the major concerns with future hydroelectric developments table of contents for an EIS, and as such, provided little in Newfoundland. substantive guidance for the design and implementation of the specitic studies for the assessment. Hydro also received more specific direction from an assessment committee, Impact Significance comprised of individuals from various provincial and federal government agencies as well as from Memorial University of committee, in essence, controlled the assessment; that is, it ment. There was no indication in any of the documentation required certain studies to be undertaken and reviewed the of the meanings of the words used to describe the importerms of reference for studies undertaken by consultants on tance of impacts. Examples of such adjectives include the proponent's behalf. major and minor, significant and insignificant, and high and There was disagreement amongst interviewees on the low. While most of those interviewed recognized the need value of the committee as a means of guidance. Some felt respect to the importance of impacts, no one indicated that any organized effort to do so was undertaken. Indeed, it In general, spatial bounding for the studies was based was speculated that since decisions on the need and first on the physical changes to result from the project and design of specific studies were so difficult to obtain through secondly on the distribution of biota to be affected. Notable the collaborative committee process, then arriving at a examples of the latter include studies on caribou and salworking framework for impact significance through this monids, both of which have short migration routes that will same approach would have been next to impossible. be interrupted by the project. In the fish investigations It appears that the absence of a context for impact sig- (Beak Consultants Limited, 1980), the physical limits of nificance resulted from a combination of the following fac-_ upstream and downstream movement for the two major tors: species being studied (land-locked Atlantic saimon and brook trout) were used as boundaries for the study. The (a) the lack of guidance from any source about how to boundaries for the study on the Grey River caribou herd construct and use such a framework; (Mahoney, 1980) were set to include most of the limits of (b) the reluctance of proponents and consultants to be the annual range for the herd. definitive with respect to impact signficance; (c) the lack of consensus on the significance of impacts There was little evidence of any ecological rationale in the on environmental attributes not strictly regulated by temporal aspects of impact predictions. In fact, few predicgovernment; and tions of biotic impacts were described with any more than the general qualifiers short-term and long-term. These (d) the lack of recognition of the value of having such a terms may have been useful had they been defined. The framework for impact significance. interviews uncovered two reasons for the absence of more The only attempt in this assessment to place impact pre- specific estimations of the duration of impacts. First, it was dictions into context was described by Newfoundland and considered impossible to be more precise given our very Labrador Hydro (1981a). A large table was presented in limited understanding of natural phenomena, especially which each predicted impact was stated concisely and was biotic phenomena. Second, consultants and proponents in accompanied with the following terms: general are often unwilling to be specific, and thus commit- (a) type of impact - positive or negative: ted, when qualifying predictions. Such equivocacy has seldom been seriously questioned in the past. (b) severity - (provincially and locally) major, moderate and minor; and (c) duration - short-term or long-term. However, neither the table nor the text gave an indication Modelling of what was meant by any of these general terms. An explicit conceptual modelling exercise was not undertaken as part of this assessment. Interviewees did not indicate that any attempt at such modelling would have assisted them in understanding project-environment interactions or in planning studies. In general, it appears that Boundaries conceptual modelling is considered a tool most appropriate for addressing concerns at the ecosystem level, such as The spatial limits of the development area were precisely nutrient budgets or energy transfers. described in the EIS. Study boundaries however, were not consistently well described in the reports of supporting The only apparent application of quantitative modelling studies. In one extreme case, no study boundaries whatso- was in the study of the hydrologic regime and the changes ever were indicated. In another case, extensive discussion it would undergo with the proposed development. Some pertained to the exact definition of the area under investiga- simulation modelling is being planned in studying the water tion. flow characteristics through various alternative designs for a channel improvement downstream from the Upper Salmon Although the EIS established the study boundaries, no powerhouse. rationale for their location was given. While we consider such a rationale to be an essential part of the study report Quantitative modelling (simulation or otherwise) of the (thus allowing for critical review of this important exercise), environmental impacts of hydroelectric projects appears to one interviewee suggested that most study reports did not concentrate on impacts within new reservoirs (e.g., Thrien, include the rationale for setting space boundaries because 1981) or impacts from changed downstream physical and of the negotiated nature of the boundaries. As a result of chemical conditions or both. In the case of the Upper the influence of the assessment committee over the impact Salmon Hydroelectric Development, the main impact conassessment, the boundaries often were established through cerns involved altered migration patterns of caribou and compromise between proposals from the proponent, the saimonids; the chemical and physical characteristics of the consultants and the regulatory agencies. It was suggested reservoirs were expected to undergo little change. As a that such compromises were not amenable to description in result, there was little need for ecological modelling of the study reports. reservoirs. Population vs Community vs Ecosystem Salmon area prior to completion of the access road) may While the primary concerns in this assessment involved exceed the financial resources available for the impact undertaken were specifically focussed on this level. For example, the levels of community and ecosystem were addressed in the biophysical study (Northland Associates Limited, 1979a). Vegetation communities were mapped in Hypotheses and Experiments this exercise, as were ecological land units based on the principles of ecological land classification (Environmental Conservation Service Task Force, 1981). The results of this study were instrumental in (i) choosing a route for the main access road and (ii) quantifying certain types of wildlife habitat that would be lost by inundation. Rigorous hypothesis testing was not undertaken in this assessment. Those interviewed suggested that this deficiency was not the result of hypothesis testing being technically impossible (although the need for more study time was mentioned), but rather the result of attitudes of various study planners and assessment committee members and their perceptions of what constitutes an adequate environOther studies relied on a combination of investigations mental impact assessment. Since many practitioners and directed at the organism and population levels in order to ment studies should be descriptive and survey oriented, address population level questions. The best example con- there is little incentive to abandon those studies in favour of cerns the examination of effects on caribou (Mahoney, more directed studies designed to provide answers to spe1980; and E. L. Hill, pers. comm.). While part of the study cific questions. Indeed, one interviewee was of the opinion involved radiotelemetric tracking of tagged individuals, that a study approach dominated by hypothesis testing another aspect was designed to observe the behavioural may result in the inadvertent omission of studying, even and migratory patterns adopted by individuals or small superficially, an element of the environment that later turns groups of individuals as they respond to the construction out to be an important concern. activity and presence of the project. It was also evident from the interviews that many hypo- This impact assessment provides a good example of theses were being tested in a non-rigorous sense in the going beyond species distribution and abundance in study minds of the scientific investigators. However, such hypoplanning and design. While most of the important ecologi- theses and tests are seldom committed to the written cal questions pertained to the level of species populations, record. It was suggested that this easily could have been advantage was taken of approaching the problems at other done if required. levels that were more amenable to investigation. Most of the interviewees agreed that impact studies Baseline Studies should shift from emphasis on the descriptive survey to emphasis on hypothesis-testing studies. One interviewee cautioned that special care must be exercised in formulat- Few adequate baselines, as we have described them ear- ing hypotheses to be tested during an impact assessment; lier in this report, were established in this environmental specifically, the scale of the questions asked must reflect assessment. Two of the better pre-project baselines estab- the ability to provide answers within the limits posed by the lished include the quantification of salmonid spawning and timing, financial and procedural constraints on the assessrearing habitat expected to be lost, and the study of migration patterns and behaviour of caribou. Most of the remain- ing studies were, to varying degrees, snapshot descriptions major experiments conducted for this assessment of the environment. have used and are using the project itself as the source of Theinterviewsrevealedanumberofimpedimentstoperturbation.Forexample,thecariboumigrationandbehaviourstudiesarebasedonactualconstructionactivi- establishing adequate baselines, and also why these snap- ties and project structures as the "experimental manipulashot descriptions of the environment persist. First, many tion." The stream crossing-fish migration study (Shawmont practitioners and reviewers believe that qualitative environ- Newfoundland Ltd., 1981) has investigated fish passage at mental descriptions have an important role to play in actual culvert installations in the access road. impact assessments. Secondly, the universal limitations of inadequate time and money were offered as reasons why blitz-style surveys predominate over directed, quantitative baselines. There is a general feeling that at least three field seasons (years) are required to allow an adequate As in the case of the environmental assessment for offappreciation of natural variation. Consultants often are not shore petroleum exploration discussed earlier, the impact given this temporal luxury, being asked to complete studies concerns in the Upper Salmon impact assessment were at in as little time as a few months. As well, studies are usually the population level, specifically with respect to caribou and planned with a view to minimizing logistical complexity. salmonids. Consequently, the time-related ecological conresulting in intensive but short-term study operations. cepts used most frequently in predicting impacts on such Finally, the cost of establishing firm baselines in areas populations were not community and ecosystem level conaccessible only by helicopter transport (e.g., the Upper cepts. The frameworks of importance to this assessment involved population-habitat interactions and the importance (Newfoundland and Labrador Hydro, 1980a). The impact of unimpeded migration to the viability of populations. predictions were summarized in a table in Newfoundland and Labrador Hydro (1981a). Approximately one third of Habitat the predictions were quantified, and these included impacts on loss of fish and wildlife habitat, reservoir flushing rates, Most of the studles undertaken during this assessment permanent alteration of land, destruction of forest had some degree of focus on either fish or wildlife habitat. resources and some socio-economic benefits. The unquanThe biophysical study (Northlands Associates Limited, tified predictions dealt mainly with effects of the project on 1979a) dealt primarily with habitat loss and creation from biotic productivity and the presence-absence and abunthe proposed project. The reservoir preparation study dance of species. In this respect, the Upper Salmon envi(Northlands Associates Limited, 1979b) elaborated on the ronmental assessment is similar to most other assessments changes in shoreline and littoral habitat that would occur of hydroelectric projects in that quantified predictions were following alternative clearing strategies for the reservoirs. A provided only for direct physical changes. great deal of emphasis was placed on an interpretation of habitat, and its potential, in the wildlife survey (McLaren. This situation may reflect a number of underlying causes. 1979), and on the quantification of losses of salmonid First, it seems that physical environmental changes are relaspawning and rearing habitat in the aquatic studies (Beak tively easy to calculate whereas most biotic impacts remain Consultants Limited, 1980; 1981). As well, the stream speculative and in the realm of professional judgement. The crossing investigations (Shawmont Newfoundland Ltd., interviews suggested that quantified predictions for biotic 1981) were partly directed at assessing salmonid habitat in impacts are seldom possible because of (i) a lack of underthe streams being examined. Finally, part of the long-term standing of causal relationships for ecological phenomena, caribou study (Mahoney, 1980) has been directed at eluci- and (ii) the high degree of chance events in natural dating caribou preferences for various types of habitat. phenomena. As well, it was apparent that some consultants The degree to which a habitat approach can be taken and proponents are uncomfortable in committing themselves to quantified predictions. Thus, the expected direcviewee suggested that the survey of salmonid spawning and rearing habitat was not difficult to undertake in the Upper Salmon watershed where the streams are relatively homogeneous in a spatial sense. In other situations, such Upper Salmon watershed where the streams are relatively homogeneous in a spatial sense. In other situations, such Monitoring as the Cat Arm watershed in northern Newfoundland in The term monitoring has taken on a special meaning for which stream morphology varies considerably, the habitat some members of the impact assessment community in approach to fish impact studies would be much more dif- Newfoundland; it refers to the supervision and surveillance ficult and expensive to undertake. of construction activities by an environmental officer. In this report, monitoring refers simply to measurement of environ- Migration mental variables over time, often associated with studies undertaken during and after project initiation. Perhaps even more important than habitat impacts in this This environmental impact assessment is acknowledged assessment were the interruptions of the seasonal migra- for its commitment to extensive monitoring and research tion patterns of caribou and salmonids. In recognition of activities. As described by Newfoundland and Labrador this, several studies were undertaken, or are currently underway, to examine and quantify these migration disruptions. Hydro (1981a), these studies include: (a) a three-year stream crossing study to assess fish passage through various culvert installations; As is the case in using the habitat approach to predicting impacts on species populations, the gap between changes (b) extensive water quality and quantity monitoring in migration patterns and population variables must be (some of which is routinely called for by regulatory bridged by professional judgement. In this sense, few studies were undertaken to improve on such professional judgeagencies); ment, and those that were are considered somewhat incon- (c) a long-term study on the successional change of clusive. The assessment recognized the unpredictable vegetation in an important local delta; nature of the effect of migration changes on the viability of (d) a study to examine the effectiveness of water release the caribou population and in doing so initiated a study that from the West Salmon dam to preserve spawning will attempt to document these effects for the benefit of habitat in the lower West Salmon River; and future assessments of hydroelectric and other projects. (e) the long-term caribou migration and behaviour studies already described. Prediction This commitment to follow-up studies is accompanied by One of the prime objectives for this impact assessment written recognition (Newfoundland and Labrador Hydro, was to predict impacts from the proposed development ness of mitigation measures and (ii) to improve predictive capability. The interviews substantiated these views but Of particular significance are current studies examining they also pointed out that considerable pressure was the feasibility of substituting a fish stocking programme for needed from certain regulatory agencies to obtain a com- the stream habitat maintenance programme. Since the ultimitment for some of the studies. This leads to two general mate objective of the latter is to maintain viable stocks of conclusions on why impact assessments so often lack fol- sport fish (specifically, land-locked Atlantic salmon and low-up monitoring and research programmes: brook trout), in the reservoirs, then artificial stocking from a (a) proponents are generally unwilling to spend time and hydro-sponsored hatchery may be more cost-effective than money on a particular environmental assessment the water release programme. after the EIS is submitted; and (b) regulatory agencies often fail to take action in requiring such studies. SUMMARY OF FINDINGS Some Notable Achievements Mitigation and Compensation These case studies have revealed some notable applica- Both the assessment reports and the interviewees tions of a number of the concepts described earlier in this stressed the importance of the mitigation of impacts. For report. Along with several other examples cited earlier, they ease of description, Newfoundland and Labrador Hydro have indicated that many members of the impact assess(1981a) divided the discussion of impact mitigation into two ment community in Canada have been cognizant of the groups - (i) measures that normally are components of ecological shortcomings of environmental impact assesssound environmental planning and construction practice ment as it has developed over the past decade. and (ii) special actions and structures that a project nor- In many cases, the efforts to improve the ecological basis mally would not incorporate. The former category included for environmental assessment have exceeded the explicit the site selection process for the access road and the trans- requirements established by administrative agencies. For mission line, and special consideration for the Godaleich example, in the Davis Strait assessment, the development Pond delta in siting the powerhouse. and use of an ecological framework for impact significance Mitigative measures as special actions and structures resulted from the motivation of the proponents and consultwere more numerous and included: ants to upgrade the scientific integrity of the predictive (a) downstream channel improvements to prevent per- analysis. In other cases, it is clear that the inputs and requirements of government agencies have contributed to Most of the interviewees agreed that impact studies Baseline Studies should shift from emphasis on the descriptive survey to emphasis on hypothesis-testing studies. One interviewee cautioned that special care must be exercised in formulat- Few adequate baselines, as we have described them ear- ing hypotheses to be tested during an impact assessment; lier in this report, were established in this environmental specifically, the scale of the questions asked must reflect assessment. Two of the better pre-project baselines estab- the ability to provide answers within the limits posed by the lished include the quantification of salmonid spawning and timing, financial and procedural constraints on the assessrearing habitat expected to be lost, and the study of migration patterns and behaviour of caribou. Most of the remain- ing studies were, to varying degrees, snapshot descriptions major experiments conducted for this assessment of the environment. have used and are using the project itself as the source of Theinterviewsrevealedanumberofimpedimentstoperturbation.Forexample,thecariboumigrationandbehaviourstudiesarebasedonactualconstructionactivi- establishing adequate baselines, and also why these snap- ties and project structures as the "experimental manipulashot descriptions of the environment persist. First, many tion." The stream crossing-fish migration study (Shawmont practitioners and reviewers believe that qualitative environ- Newfoundland Ltd., 1981) has investigated fish passage at mental descriptions have an important role to play in actual culvert installations in the access road. impact assessments. Secondly, the universal limitations of inadequate time and money were offered as reasons why blitz-style surveys predominate over directed, quantitative baselines. There is a general feeling that at least three field seasons (years) are required to allow an adequate As in the case of the environmental assessment for offappreciation of natural variation. Consultants often are not shore petroleum exploration discussed earlier, the impact given this temporal luxury, being asked to complete studies concerns in the Upper Salmon impact assessment were at in as little time as a few months. As well, studies are usually the population level, specifically with respect to caribou and planned with a view to minimizing logistical complexity. salmonids. Consequently, the time-related ecological conresulting in intensive but short-term study operations. cepts used most frequently in predicting impacts on such Finally, the cost of establishing firm baselines in areas populations were not community and ecosystem level conaccessible only by helicopter transport (e.g., the Upper cepts. The frameworks of importance to this assessment factors underlying these inadequacies, it often was difficult the proponents or both) have either not been motivated or to distinguish between perceived and real constraints. As not set the requirements to establish rigorous baselines. well, it has been necessary to speculate on the reasons behind many of the responses to the questions posed. At Logistical constraints were also identified as having limtimes, interviewees doubtlessly were inclined to defend spe- ited the knowledge base for predicting impacts or specifycific interests or parties connected with the assessments, ing mitigation or both. Of greatest importance in this resulting in the inevitable shading of responses. Also, some respect is the difficulty and uncertainty in gaining access, at of those interviewed had only been exposed to the findings specified times, to the remote study areas by way of air or of this research project through an early progress report sea. As well, the technical requirements of undertaking cer(Beanlands and Duinker, 1981). Therefore, the implications tain studies would have been nearly impossible to meet. For of implementing some of the concepts discussed may not example, had pelagic fish been implicated in case of an oil have been fully realized. well blowout in the Davis Strait, it would have been a logisti- Nevertheless, cal nightmare to undertake sufficient surveys to document Nevertheless, from a combination of the explicit and fish distribution and abundance. While these limitations implicit messages arising from the interviews, it has been were identified in the interview discussions, none of the possible to identify a number of factors which largely interviewees indicated that the knowledge base for the preaccount for the character of the assessments reviewed. dictive analyses was seriously deficient purely on account First, on the technical side, both assessments were faced of such technical limitations. with the task of understanding environments for which no substantial knowledge base had previously been developed. Neither the Davis Strait nor the Upper Salmon water- applying many of the concepts discussed during the intershed had been relatively well studied prior to the assess- views were not largely technical in nature. On the contrary, ments and consequently the study programmes were the more important constraints appeared related to the attiforced to begin with a substantial effort at the reconnais- tudes and perceptions of the persons involved and to the sance level. This was especially so for the Davis Strait administrative and institutional forces at work. In particular, assessment which, by definition of the project, encom- the three most pervasive factors were: (i) lack of recognipassed a very large area of a poorly understood ocean. tion of, and agreement on, what elements comprise an eco- It is difficult to say whether this feature of the assess- tion, and (iii) a lack of requirement. ments should have limited the extent to which more detailed ecological studies, possibly incorporating simula- These constraints are particularly relevant for aspects of tion modelling, experimentation, or specific baseline stud- impact assessment to which field constraints do not directly ise wora undartakan Tho inwoctinatare undouhtadily ...suld abolv: for example. in establishing a framework for imnact assessment documents for review based on whatever infor- undertaken was determined by the perception that such mation could be collected in the time available. In the Davis studies provided the appropriate information base for the Strait assessment, the proponents claimed that further assessment. In the Davis Strait assessment, a lack of constudy (beyond that upon which the initial predictive analysis tinued government requirement for certain studies was based) would have served to amplify, but not signifi- appeared responsible either for their exclusion or their early cantly alter, the impact predictions. termination. For certain types of study, especially the establishment of baselines against which to measure project-induced perturbations, it appears that a shortage of time may not be an overriding constraint. Indeed, the fact that such a baseline This report has emphasized that substantial improvehas been incorporated into the caribou study of the Upper ments can be made in the contribution of ecology to enviSalmon assessment is evidence that temporal limitations ronmental assessment. It is recognized that analysis within can, in some cases, be overcome. The key seems to be to impact assessments will always be limited by the knowlmake appropriate decisions early and then undertake the edge base either already established or obtainable in the studies without delay. From a temporal perspective, oppor- appropriate period. Nonetheless, we have noted that much tunities for undertaking pre-impact monitoring for the Davis of the upgrading can be realized through greater effort at Strait assessment were even greater than in the case of the conceptualization, more effective study planning, and a Upper Salmon - that is, ongoing baseline studies would common, realistic expectation of what can be accomcontinually improve the data base for variables of interest plished through a focussed, applied research programme. until an oil well blowout finally (if ever) occurred or until an adequate baseline fix had been achieved. But such base- The case studies have substantiated these views. Techniline studies have not been undertaken. Perhaps because of cal limitations, whatever their form or magnitude, are unithe very low probability of such a blowout ever occurring, versal and will continue to apply. Practitioners must be the possible sponsors of such studies (i.e., government or aware of these when planning study programmes, as they bear directly on what can be achieved in the laboratory or broader scientific concepts to environmental assessment field. Despite these limitations, the foregoing analysis has are within the means of those who plan, undertake and identified some key opportunities where overcoming the review the assessments. As well, they have shown that connon-technical barriers can lead to an upgrading of the eco- straints posed by the attitudes and perceptions of the perlogical integrity of impact assessment. sons and organizations involved may be equally, if not In conclusion, the case studies have demonstrated that more, important than the technical and logistical limitations some improvements in the application of ecological and that mayply. (a) downstream channel improvements to prevent per- requirements of government agencies have contributed to manent flooding of the Godaleich Pond delta; sound impact assessment studies and analyses. Thus, it (b) preparation of the reservoir to remove barriers to was recognized by Newfoundland government agencies caribou migration and boat passage; and Hydro that impacts of the Upper Salmon Hydroelectric (c) water release facilities to protect salmonid stream Project on local caribou herds could not be predicted with habitatintheNorthSalmonandWestSalmonRiv-ers;sufficientaccuracy,andamajorco-operativeresearchpro-grammewaslaunchedtodocumenttheeffects. (d) special design of the access road to facilitate cross- Some combination of motivation and requirement also ing by caribou; appears to account for other strengths in these assess- (e) design improvements in the power canal, penstock ments. Specifically, one of the most important components and diversion channels to facilitate crossing by cari- of the Davis Strait assessment was the preparation of bou; and detailed oil spill contingency plans. A major effort in the (f) construction restrictions (i.e., work stoppage) on Upper Salmon studies was focussed on using the results to account of (i) nearby caribou or (ii) archaeological prescribe appropriate mitigation. In summary, three factors appear to be associated with Given that one of the primary objectives for the assess- improvements in the ecological basis for environmental ment was to identify mitigation measures, it is appropriate assessment: (i) recognition of the main problems and an to examine whether the studies undertaken were effective appreciation of the solutions, (ii) motivation on the part of in doing so. Two examples in particular demonstrate suc- practitioners to pursue improvements, and (iii) the position cess in this regard. The purpose of the reservoir preparation of government review agencies to accept nothing less than study (Northland Associates Limited, 1979b) was to recommend a clearing strategy that would minimize or eliminate impacts. The aquatic investigations (Beak Consultants Lim- Constraints Against Improvement ited, 1980) were instrumental, when combined with details of project design, in pointing out alternative viable mitiga- The case studies have also shown many areas where the tion techniques (especially water release for stream habitat two assessments have fallen short of standards that might maintenance and provision of fish passage around be set on the basis of a perceived ideal assessment. obstacles) for protection of the fish resources of the area. Through the use of personal interviews in investigating the