NOT TO BE CITED WITHOUT ICES CM 1998/J:11 PRIOR REFERENCE TO THE Session (J): Variation in the Pattern ofFish aggregation: Measurement AUTHORS and analysis at different spatial and temporal scales and implications

Using local and scientific knowledge to predict distribution and structure of herring shoals

Steven Macldnson! and Nathaniel Newlands

(Fisheries Centre, 22 04 Main Mall, University of British Columbia, Vancouver, Canada, V6T 1Z4,: corresponding author:email: [email protected],phone: (604) 822-2731,fax: (604) 822-8934)

, Abstract

Attempting to predict the distribution and structure of herring shoals is an ominous task given our incomplete understanding of ecological mechanisms. Despite recent attempts to link cross scale behaviour dynamics and distribution studies, there are still large gaps in our basic knowledge of shoal movements. Nonetheless, the knowledge of fishers and fishery managers is not incorporated in to our scientific analysis but such information is rich in observation since knowledge of fish behaviour and distribution is a pre-requisite for their profession. Combining such observations with more conventional scientific studies and theoretical interpretations provides a means by which, we may bridge some gaps in our knowledge. We present a framework in the form of a fuzzy logic expert systemwhose knO\vledge base incorporates both local and scientific knowledge in the form of heuristic rules, The knowledge base is flexible in the ,sense that it can easily be modified to add additional information or change current information. Additional 'operational rules' govern how the system utilises, its knowledge by setting priorities to certain factors under specific situations. The fuzzy system is driven by responses from an external user who provides information on biological and environmental aspects pertaining to the locality and time of year in which they wish to learn about In response, the system predicts structure and distribution patterns of herring shoals. It is context sensitive and 'intelligent', asking the user the minimum questions required to make its predictions. Hypertext screens ". and-graphical- mtert"ace providesuser-rnindl:Y·inriufaJi.oorifpiit 'toileiliei\Vifh-the-capacit)'-to-fiilly--" explain facility of how the system arrived atits conclusion.

Keywords: Heuristics, fuzzy logic, model, prediction, herring, school, distribution

1 Introduction

On broad geographic scales, herring populations world-wide display remarkable pattern in their spatial and temporal distribution. The annual migration paths of migratory stocks may cover 1000's of kilometeres between spawning, feeding and overwintering sites, many of which have been known for quite some time (Harden-Jones 1968, McKeown 1984). As ecologists, our challenge is to fathom how such repeatable pattern can be derived from the complex, day to day behaviour of individual herring. It is an onerous task; one that is matched in complexity by the tremendous inherent variations within and between our scales of observation. To develop spatially explicit predictive models needed for management and to allow us to respond to change, we must learn how to interface disparate scales of interest and understand how information is transferred from fine to broad scale and vice versa (Levins 1992). A key element guiding our research on the distribution of herring, focuses on understanding how selection has favoured schooling as an adaptive behaviour. Experimental behavioural studies (review by Pitcher 1993) together with detailed in situ acoustic observations (e.g. Gerlotto et al. 1994 Masse et al. 1996, N0ttestad et al. 1996, Pitcher et al. 1996, Petitgas and Levenez 1996, Maraveliasand Reid 1997, Misund et al. 1997) offer insight on which to base our ecological interpretations. The structure and distribution of herring shoals is currently believed to be governed by the strive to achieve 3 main goals; survival; growth arid reproduction (Ferno et al. 1998). Previously, most attentibnhas been directed to small scale (0.1 to 10 m, seconds to minutes), school organisation and dynamics; or, large scale (100' s km, weeks to months), stock structure and migration studies. Information on the mesoscale (0.1 to 100km, day to weeks) distribution pattern of schools and school clusters is particularly lacking. . Since many herring fisheries are typically conducted at spatial scales of one to tens of kilometres and occur for periods of days to weeks, both fishers and fishery managers alike operate within themesoscale realm. By virtue of their profession it is prerequisite that they have knowledge regarding the distribution and behaviour of herring. Such a rich information source should not be unutilised. Typically however, local ecological knowledge has been considered by moretniditional analytical fisheries science as 'anecdotal', and for the most part been absent from fish stock assessment or during development of management' plans. In contrast, social science emphasises the importance of local knowledge in a contextual and historical perspective, but again for the most part, has been unsuccessful in directly incorporating this informatiori into fisheries management (see Neis et at 1996 for exception). The relt!ctance and inability to,lltilise J1911-scientificknoWledgeasdata is asignifiqmt affliction hindering the progress of fisheries science and management (Mackinson and N0ttestad, in press). There is an urgent need to maximise the usefulness ofinformation by incorporating alternative sources of data and finding new ways of using traditional ones. We present a formal framework for combining local ecological knowledge and scientific knowledge in the form of heuristic rules and demonstrate how it can be used to predict the structure and distribution of adult herring shoals for different phases during their annual life cycle.

Heuristic Data Sources (developing the knowledge base)

This heuristic model is developed within the framework of an expert system (Figure 1) and incorporates two fundamental data sources of information on fish distribution and behaviour; (i) 'practical' data: local knowledge of interviewed fishers, fishery managers and First Nations people; (ii) 'hard' data: scientific information from interviewed fisheries scientists, field work studies and published literature sources.

2 Briet1y, expert systems are a branch of artificial intelligence; theories .and methods for automating intelligent behaviour. They are computer programs that provide assistance in solving complex problems normally handled by experts. They use rules to· store. knowledge. When the system is asked to solve a problem, it uses this knowledge to infer solutions. The software used to develop the system presented here is Exsys®' Professional by Multilogic.

Figure 1. near here

Interviews A total of 30 formal interviews were conducted, half with fishery scientists (8) and fishery managers (7) who generally provided technical or hard scientific knowledge, and half with fishers (9) and First Nations people (6, all previous or currently herring fishers) who provided professional practical local knowledge. With the exception of one gillnetter who specifically undertook herring surveys, most fishers interviewed were seine fishers, with a collective experience (CE) of approximately 270 fishing years. Seiners were specifically chosen as candidates in contrast to gillnetters or roe-on-kelpfishers, since seine-fishing typically requires specific knowledge of fish distribution and movements when searching broad areas. The First nations CE amounted to approximately 290 fishing years. Selection of interviewees was deliberately non-random. An attempt was made' to interview those fishers who had the most experience fishing herring during different seasons, at different locations and who were held with respect by other fishers in the community. For this reason, progressive selection of interviewees was conducted by word of mouth, one candidate suggesting another to talk to. This method proved to be very successful. Fishery scientists and managers were selected based ontheir experience with herring. The current regional herring co-ordinator and 3 long time fishery managers (CE approx. 160 years) offered a more technical 'field based' perspective that complimented observations by fishers. Three herring scientists from the Pacific biological sta.tion, Department of Fisheries and Oceans (CE approx .. 75 years), and a further 5 from Norway (Institute of Marine Research and University of Bergen; CE approx. 80 years) provided hard scientific observations from field and experimental studies. Typical interView duration was 2 hours but some ranged from 1- 4. With two exceptions, all interviews were conducted on a separate individual basis at the preferred location of the candidate. For help with interpretation, it was necessary on one occasion to interview two fishers together. In another . instance, a meeting was held withtl1e. first Nations Sliamm()nband.elders, that included men and. women who had traditionally been involved in herring fishing prior to the demise of their localised hei:ring stock. . . Two basic questions were asked; what? and why? Firstly, interviews were asked to recount what theY'had observed regarding distribution and behaviour of herring and secondly to offer possible explanations to account for their obserVations. All candidates were asked the same type of questions although specific interViews were 'free range' or 'adaptive' in the sense that the format and directness in which the questions were presented depended upon the context of discussion. Allowing discussion to cohtinue openly in this manner prOVided insight in to many aspects which would have been overlooked by a simple questionnaire offering only a fixed set of responses. On almost all occasions, completely new knowledge was acquired. Using the same technique of Johannes (1978), reliability of the subjects' answers was tested by asldn,g two tyPes ofquestions at a~_onYenil:mJ tilllecluDngcliscussiQn.. Ihej'irst""ere question.s tow.hjch._. the answer was already well-known. Responses to these questions were almost always the correct answer or that they didn't know. The second type of question sounded plausible, but which were ones that the fishers were unlikely to be able to answer. In virtually every instance, the response to this type of question was "I don't know", although on two occasions plausible explanations were offered.

3 Interviews were recorded by hand written notes and subsequently typed and resent to the candidate within 48 hrs for verification of accuracy, ;corrections and additions. Prior to. interviewing, candidates signed an ethical review form affirming that the information received would remain their property and that reference would be given directly to them when cited. At a later stage, details ofal! candidates together with a full account of their interviews was recorded in a database. In addition to interviews, 18 questionnaires were posted bye-mail to researchers involved with herring populations world-wide. Despite a second reminder, there was only one response that contributed to the knowledge base. There was clear demarcation in the type of responses from different interviewees. Typically fishers were particularly strong .on observation providing detailed accounts of school structure, distribution and behaviour including; school sizes, shape, density, depth distribution, association with specific features, ease of capture and specific behaviour patterns relating to season, tide, weather, fishiI).g vessels, time of day, feeding, occurrence of predators. However, when asked why;. they were generally reluctant or found it difficult to offer an interpretation. It was difficult to elicit a rank order of factors they· considered important in determining the observed shoal structure, distribution and behaviours. In contrast, fishery scientists were more familiar and at e.ase with offering. interpretations for their .observations or experimental findings and for the most part, were able to assign an order of relative importance to the factors contributing to shoal structure, distribution and behaviour patterns. Responses of fishery managers were more akin to those of fishers, being grounded firmly in field observations. Due to the nature of their job, most were uncomfortable with ascribing behaviours to any particular factor. They tended to err on the side of caution and uncertainty, usually offering provisos and comments of exception to any of their observations. They were however, more willing than fishers to offer potential interpretations, and it was apparent that these were frequently guided by scientific understanding from colleague fishery scientists. Remarkably there were no instances. in which knowledge accumulated from any single source diverged. Information either complimented previous knowledge or added additional understanding. Moreover, the majority of comments contained information pertaining to mesoscale spatialanc\ temporal distribution pattern of schools.

Field work and literature sources Information from four research cruises is compiled within the knowledge base .. Two cruises were cond)lcted in Pacific coastal waters of British Columbia during pre-spawning and spawning period; one in the Strait of Georgia (1997) and another in the Central Coast (1998). A specific attempt was made in .the 1998 survey to. valicjate observations from fishers. They were used to form hypotheses on which the survey was analysed (Mackinsor\!n prepfTwo further cruises werecondilci:edm-the­ Norwegian sea during May 1997 and 1998, during a period of migration and ocean feeding for Norwegian spring-spawning herring (see Mackinson et al. this conference for further details). Literature sources concerning· field and experimental studies on behaviour and .distibution of herring and other schooling pelagic fish was reviewed. Relevant information was recorded directly in a database cross-referenced to its original source. The 'hard' data obtained from field work and 1iteratur~ sources was used for defming ranges for descriptors of structure and distribution .. Through the process of defuzzification these .values are combined with information from heuristic rules to provide numeric output from the system. This is described in detail in a later section.

Structuring tbe heuristic system (k.nowledge Imgineering) ..

Structuring of the heuristic system involves the process of building a multi-layer decision tree, that relates information from different spatial and temporal scales. After considering alternative strategies for dealing with information, it was deemed most appropriate that all knowledge, would be

4 contribute equally in building the knowledge base. Therefore, an assumption is equality in the degree of belief in a piece of information from either fishers, fishery managers, fisheryscientists, First Nations people or from literature. This assumption helped to maximise all of the data sources (Mackinson and N0ttestad, in press).

Combining knowledge using heuristics Knowledge from interviews and literature was sorted in detail to identify information relating descriptions of behaviour and distribution pattern (descriptors) to those factors having influence upon them (attributes) A complete classified list of the extracted descriptors and attributes is provided in Figure 2.

Figure 2. near here

Using a series of heuristic rules capturing knowledge, the relationships between attributes and descriptors were defmed in .the expert system. Many of these rules came directly from source whilst others were defined on best inference. A rule has the form IF this THEN that and may contain .several conditions in the IF part linked by AND, OR, NOT, and one or more .elements in the THEN part linked by" an AND. Each item contained in the THEN part is then assigned an associated confidence.value according to the relative contribution of information relating to that particular rule. For the most part, attributes are used in the IF part and descriptors are used in the THEN part of rules since the goal is to conclude upon how behaviour and distribution is determined by the influence of various combined attributes. A example rule is:

IF type of predator is an aquatic predator (fish or mammal) AND abundance of aquatic predators is high THEN' packing density of school is high (item conf = x) AND internal dynamics schooling (item conf= y)

Within the expert system, attributes and descriptors are classified according to their properties (strings or variables), in the way they are used in rules, and whether they are required as input or output to the expert system. There are two basic types with subdivisions: (i) Qualifiers (lists of options/ choices to chose from) may be designated as single choice, multi-choice or as fuzzy sets associated with a variable, (ii) Variables may be defined as numeric, text or strings but are most commonly applied as numeric variablesusedin association for outputting descriptors. In the example-rille above,­ type of predator operates as. a multi-choice input with two options,. predator type is either 'aquatic predator' or 'avian predator'. Abundance of aquatic predators is designated as fuzzy sets associated with a fuzzy numeric variable, 'predator abundance scale', Figure 3 . . The use of fuzzy sets as qualifiers allows us to build a system that. directly represents linguistic expressions containing useful and relevant knowledge. It is unreasonable, impractical andalrnost impossible to relate information in a purely quantitative way without making gross assumptions. More importantly, implementation of fuzzy qualifiers allows us to define relationships between attributes and descriptors on a continuous scale; This aspect provides particular benefit in managing the whole system since it slashes the number of rules required to describe relations. Furthermore, fuzzy rules provide a .. direct connection for combining quantitative. and qualitative knowledge and expressing uncertainty. They are the key to achieving quantitative output from qualitative understanding as will be shown later. Interconnected associations between fuzzy qualifiers can be conveniently expressed using a fuzzy associative map (F AM) (Figure 4)

Figure 3. near here Figure 4. near here 5 Imposing hierarchy and uncertainty Several techniques are implemented to represent hierarchy in the behavioural decision chain and temporal changes in the importance of factors influencing distribution and structure. We assert that specific behaviours are dependent on seasonal and diurnal changes in motivational state. There are four ways in which the importance of various attributes are altered. 1. Adjusting weights through a series or rules that define seasonal priorities to motivational states of hunger, risk avoidance, reprodliction and energy saving. These are manifest by assigning a weight factor (low, medium, high) that adjusts the influence of particular attributes on the overall outcome. For example, when risk avoidance is a high priority, a high weight is assigned to increas.e the abundance of predators. This is a pseudo-weighting, that is based on the assumption that increasing predation risk can be represented by increasing the abundance of predators. Ideally, weight factors should be directly applied to confidence factors associated with rules, but it is not possible to implement such a method using the current software. . 2. By assigning confidence factors to each rule we influence their importance when fired and propagated through the system. This method imposes hierarchy in the importance of different attributes on descriptors whilst simultaneously representing. the uncertainty. or degree of belief in - the rule. 3. Using operational logic including rules and commands that defme how the system fires under specific circumstances .. In this instance, particular rules may be ignored whilst others are followed. Variables may be assigned values directly as a consequence of operational rules (in particular, when they are deemed oflow importance) 4. Providing the user the opportunity to assign low importance or ignore a particular element. By offering placebos (choices that do not lead to ,any conclusions), such as 'not sure' or 'low importance' the user can exclude or reduce the influence of attributes. However, if the user answers 'not sure' where possible they are gnided by a series of hypertext screens and nues to make a conclusion or decide between ignoring the attribute or letting the system assign a default value.

Predicting distribution and structure

When queried, the expert system guides the user through a series of questions, each step comparing the input information with theru\t;s in ihekn.owledge base. OPerationallogicensures.tI1aL each run is unique. Based on previous answers the system prompts the user, asking only those questions pertinent to draw conclusions relevant to· the initial conditions specified by the user (Figure 1). Ru1es using qualifiers defined' as fuzzy sets· are followed in parallel and to' a partial extent since all rules may be true to a certain degree. The operation can be viewed as a parallel mu1ti-layer decision" tree and this aspect separates a fuziysystem from a more simplistic non-fuzzy decision-tree system in which branches of the tree are triggered independently.

Propagation ofconfidence The confidence 'associatedwilh specific qualifiers is set in one of three ways; taken as being 1 when setby the user; determined'by a fuzzy membership function for fuzzy qualifiers; if it is assigned a value in the THEN part, the confidence is propagated from the IF part and combined: with any confidence in the THEN part. Confidence in the IF part The IF part may be made up of one or more conditions linked by AND or OR. Depending on whether the IF part is made up of only one condition or a single qualifier value the confidence is simply the confidence associated with that condition or qualifier. If there are more than one qualifier values .set

6 using an OR such as; IF food abundance almost non OR sparse, we combine the confidence of 'food abundance almost non' and 'food abundance sparse', using the default formula:

Combined conf. = combined conf. +(Value conf. *(l-combined conf.)) Eq.l

The formula is known as the MYCIN method (after the first system that used this method) and has the advantages that the confidence never exceeds I, each positive confidence increases the final confidence, and it not sequence specific. The same method is used for combining the confidence of blocks of OR statements. in the IF part of a rule. The overall confidence in the IF statement is calculated as the product of the parts. This simply means multiplying the combined confidence of conditions linked with an OR with those conditions linked with AND. By setting a threshold, rules are only considered to be true when the overall IF confidence is above 0.01. Confidence in the THEN part .. In the fuzzy system, when a rule fires, the IF part will be TRUE, meaning all the parts in the IF statement have a confidence >0. The final THEN confidence is achieved by multiplying the overall IF confidence with the confidence assigned in the THEN part and combining it with the current confidence by a series of formulas (again, the MYCIN method);

Current conf. = (IF conf.* assigned THEN conf.) +(Curr. Conf.*(l-( IF conf.* assigned THEN conf.)) Eq.2

The formula has the characteristics that any positive value increases the finalconfi.;lence, and when a THEN part has a low confidence, a high value significantly increases it, but when a THEN part has a fairly high confidence, additional high values only increase it slightly. There is one important shortcoming in the use of the above formulas for propagating confidence. When the IF part has an overall confidence of 1, which occurs when a user directly chooses a qualifier value rather than it being calculated during inferencing,and this is combined with a THEN item whose assigned confidence is also 1, the final IF confidence is pushed immediately to 1. Thus, the confidence combined by firing of other rules has no apparent effect. The simple pragmatic solution used here is to ensure that all· items in the THEN part are assigned confidence values <1.

Output format Combined in a graphical user interface, two types of output format are used in the predicting the strUctllre andmeso~scaIedisiributlono(slioi!1s.-- ...... _- ...._- .-.-.-- (1) Descriptive diagnostics: Text-only qualifiers or variables describing features such as shoal shape, internal dynamics and likelihood of shifting location, together with other general comments. (2) Numeric descriptors: by applying defuzzification (see below) on output fuzzy sets, numeric variables describing the. mean and range values are output for the descriptors; shoal size, packing density, catchability, relative shoal depth, nearest neighbour distance and inter-school distance.

Defuzzification Using the formulas in equations 1 and 2, detailed confidence in each output fuzzy set is propagated through the system. For each fuzzy set of a qualifier weare given an associated confidence. e.g. the system gives the fuzzy conclusion that shoal size is small (conf= 0.2), medium (conf= 0.6) and large (conf =0.3). The process of obtaining a single non-fuzzy value as output is defuzzification. The method used here is weighted-average defuzzification based on a multiplication between the degree of membership to the output fuzzy sets and the supremum value of each set (Figure 5). By applying the same procedure to maximum and minimum ranges associated with each of the output fuzzy sets we also. obtain a range around the discrete output value (Figure 5). Using the example in Figure 5, the discrete defuzzified weighted output would be: .. . 7 Mean = (0.2*Smallsup)+(0.6*Medsup)+(0.3*Largesup) Range minimum = (0.2*Smallmin)+(0.6*Medmin)+(0.3*Largemin) Range maximum = (0.2 *S mallm";)+( 0.6 *Medmax )+( 0.3 *Largemax)

FigureS. nearhere

GraphicalUser Inteiface There are two separate components to the user interface. The input interface is designed to be user-friendly and robust, using custom screens that guide the user through a series of questions. Some are in the style of a form whilst others pop-up and ask specific questions (Figure 6). Bylhe use of special help and hypertext screens, there is facility for the user to' ask why' a question is being asked, receive guidance on answering questions, or view the current known data or specific rules at any time. Limits on variables and qualifiers prevent the user from inputting nonsensical values, and should they wish 10 correct an undesired input it is simple to clear the previous input and re-enter it. The output interface contains two elements: (i) a text results screen with full explanation capability so the user may query how the system derived at its answers, (ii) a Visual Basic graphic display of the structure and distribution of shoals!. Furthermore, data may be output to a spreadsheet so that specific output variables can be plotted (Figure 7).

Figure 6 .. near here Figure 7. near here

Discussion

. Although the magnitude and relevance of local knowledge in resource management has been recognised f()r some time (Johannes 1978, Dahl 1989, Maquire et al. 1994, RIFM symposium 1996) a mismatch between that which is known and that which is used for any practical sense remains. Resource management decisions are typically based on detailed, yet limited studies ofa more traditional scientific nature, 'hard science'. Despite recognising the obvious need to incorporate local knowledge into science and management, two important barriers still exist; the reluctance to give it . respect equal to that given to 'hard science' .and the. inabilitytojnc()rporateit in a.holisticIheaningful way (Mackinson and N0ttestad, in press). The former is deep rooted and requires a fundamental change in our scientific approach; a pointrecognisedby Chambers (1980), who comments "the most difficult thing for an educated expert to accept is that poor farmers may often understand their situations better than he does. Modern scientific knowledge and the indigenous technical knowledge of rural people are grotesquely unequal in leverage. It is difficult for some professions to accept that they have anything to leam from rural people, or to recognise that there is a parallel system of knowledge to their own which is complementary, that is usually valid and in some aspects superior". Recently, there have been several attempts at achieving the latter. For example, Nei~ et al. (1996) conducted a thorough series of interviews with cod fishers in Newfoundland, the results of which convincingly demonstrated that local knowledge, formerly treated as anecdotal and then ()verlooked, was capable of contributirigdetailed scientific information on stock structure, changes in catchability, abundance during a closed fishery and potential impacts of a reopened capelin fishery on northeru cod recruitment. Case studies by Pinkerton and Weinstein (1995) highlight how local knowledge can be applied to great benefit under a system of community based management. In a study of the Pacific herring bait fishery, Schweigert and Linekin

1 Refer to accompanying poster "Generating meso-scale pattern of fish shoals", Newlands and Mackinson, for further 'details of method for Visual Basic display of results. 8 (1990) also recognise the value of local knowledge. Questiounaireswere used to obtain information on spatial distribution of non-migratory herring that are not sampled or .assessed aspartof the routine monitoring of the major adult migratory populations. The approach taken in .this study compliments these approaches, but goes one step further by combining both local and scientific knowledge using a formal framework. Despite potential biased perceptions of resource abundance and their impacts, knowledge of fishers can bea fountain of information (see for example Johannes 1978). Frequently their knowledge is compiled over time based on shared knowledge with others they have fished with. The interviewsjn this study reveal that fishers closely observed physical environmental conditions and temporal. changes resulting in variation in distribution, size and ease of capture of schools. However, in contrast to interviewed scientists and fishery mangers, fewer were prepared to suggest behavioural inte)1lretations for their observations. With the exception of several enthusiastic individuals, it did npt appear 'necessary' that they should ask why? N eiset al. (1996) found a similar response from interviews with cod fishers;" ... fishers ' knowledge of fish stocks is primarily acquired to optimise catches while minimising effort. Therefore, they tend to closely observe those environmental features which are linked to fishirJ.g Snccess: seasonal movements, habitat preferences, feeding behavionr and.abundance dynamics; as well as those physical attributes that affect fish distribution, the performance of gear and fishing. time:. wind direction, currents, water temperature and clarity, bottom characteristics. and local assemblage structures, as well as gear fouling". Remarkably, there was no conflict in the information obtained from fishers, scientists and literature sources that could not be explained by observ.ations at different scales. More 'unique' instances of information were obtained occasionally from fishers. Information from scientists, fishery mangers, field observations and literature accounts tended to support andcompliment knowledge given by fishers rather extend it. On assessment of the responses to specific questions. used as test controls for .assessing the reliability of answers, it was deemed that all information relating. to distibution and structure of shoals was accurate according to repeatability. Where pecnliar or unique observations were made, these were deliberately verified with other subjects in subsequent interviews. Further validation was conducted during the 1998 Pacific herring survey, in which comments were used to form hypotheses on which the survey was then analysed (Mackinson in prep). In support of this approach, Hutchings (1996) noted how improved communication with fishers can lead to testable hypotheses regarding the biology of northern. cod. Use of heuristics makes it possible to combine qualitative knowledge sources in a.series of rules that constitute a knowledge base. Application of fuzzy logic (Zadeh 1965, 1978), takes us one step further, interlocking both qualitative and quantitativeinforrnation. Since fuzzy sets are able to model words mathematically, they link the numerical withthe verbalandtllus clmbringthese-tWo-----­ worlds in to sync (McNeill and Freiberger 1993). Several important benefits are realised by using a fuzzy model approach. Firstly, by use of asymmetric membership functions we are able to easily approximate non-linear relationships. A fuzzy system allows to guess at the non-linear world and yet does not make us write down a mathematical model of it (Kosko 1993). Secondly, since they are able to show the continuum between 'yes' and 'no' fuzzy models are ironically, more precise. Finally, they do not require us to describe the universe of possible events, as is the case with models founded in probability. Although both. describe uncertainty numerically, probability treats only yes/no occurrences, and is inherently statistical. Fuzziness deals with degrees and is completely non-statistical (Zadeh 1965). Defuzzification is the process that allows us to obtain a discrete numeric output used to desc.ribe structure and distribution of herring shoals. There are many techniques of defuzzification (Meech 1992, Meech and Kumar 1995). Typically, only the supremum value and membership values of each output fuzzy set is required to produce a discrete output. We demonstrate how defuzzification can be applied to the support and membership functions of each fuzzy sets, to provide a 'weighted range' associated with the single weighted average output value. These results are used in generating the mesoscale shoaling pattern in the graphical output interface. 9 -

Typically expert systems are used to solve problems' that cannot be solved using a purely algorithmic approach; those that have an irregular structure, contain incomplete, qualitative or uncertain knowledge, are considerably complex and where solutions must be obtained by ,reasoning from available evidence and sometimes making "best guesses" (Dabrowski and Fong 1991). Recognising that much of our understanding of fish ecology, ecosystem effects of fishing, and social and economic effects of fishing exhibit these qualities, it is somewhat surprising that the use of expert systems and fuzzy logic has been given limited attention for application in fisheries. In a review of fishery related expert systems, Saila (1996) offers a list of only 18 noteworthy systems. Of these, only two (Aoki 1989, Fuchs 1991), both of which are non-fuzzy systems, consider interactions between fish and envirenmental factors. Furthermore, Saila comments, "through building and testing we move toward practicality, recognising that decisions based on qualitative and sometimes incomplete knowledge is still better than makingdecisiens without any understanding". Warwick et al. (1993) identify 98 references to expert systems in the field, of environmental management. A system that is similar in' some respects to the one developed here is that presented by Veiga and Meech (1994). Combining knowledge of prefessionals and field observations, the fuzzy system accommodates imprecise data input and uses lingnisticexpressions 'in place of a complex mathematical'model te diagnose mercury bioaccumulation risk from gold mining operations. . The fuzzy expert system is capable of predicting state dependent, mesoscalespatioctemporal changes in'the structure' and distribution of herring shoals. When queried by the user, the interactive input interface follows a series 'of questions constantly updating its current knowledge based on the users answers and information contained within the knowledge base. A temporally fixed graphic interface is provided to generate the predicted structure and distribution of shoals together with the descriptive diagnostic characteristics. This interface is currently being developed to display dynamic seasonal changes in shoal structure and distribution.' Four approaches will be applied during validation; (i) automatic validation routine -' an internal consistency checker that is part of the system (ii) sensitivity analysis -sequentially changing an input variable and observing the consequences; (iii) expert validation- going back to interviewed 'experts' and letting them test it; (iii) comparison with literature and field work observations. : Heuristic models in general offer great potential, a point emphasised by Hilborn and Mangel (1996), who comment, "although the output of most models is numerical, the most influential models are the ones in the numerical output is not needed te gnide the qualitative understanding": Potential uses for this system include; training herring fishery managers on distribution and behaviour of herring; .... abuilding ground for development of nonclinearbehavioural models in 'other schooling fish (may include additions such- aslloeuraYnetW6rk); anadaptiveformal frariiework for combining locM and scientific knowledge.

Acknowledgements

Sincere thallksto all those who participated in interviews; Bill Wilson and crew of the FN'Kynoc' for helpatid humour during Pacific field sunreys; Dr. Daniel Ware, Department of Fisheries and Oceans for financial support of the expert system and interviews; Norwegian colleagnes from the Institute of Marine Research and the University of Bergen for invitation to participate in field studies in the Norwegian Sea; Also funded by the International Council for Canadian Studies.

10 References

Aoki, I., Inagaki, T., Mitani, I. and Ishii,T. 1989. A prototype expert system for predicting fishing conditions of anchovy of the coast of Kanagawa Prefecture. Bulletin of the Japanese Society. for Fisheries and Oceanography, 55 (10), 1777-1783. Chambers, R. 1980. The small farmer is a professional. Ceres, March~April: 19-23. Dabrowski, C.E. and Fong, E.N. 1991. Guide to expert system building tools for computers. National Institute of Standards and Technology, Gaithersburg,MD 20899. Special Publication 500-188, 141pp. Dahl, A.L. 1989. Traditional environmental knowledge and resource management in New Caledonia. In Johannes, R.E. [ed.] 1989. Traditional ecological knowledge: a collection of essays. Gland, Switzerland, IUCN, World Conservation Union, pp. 57-66. Ferno, A.,T.J. Pitcher, L. N0ttestad, W. Melle, S. Mackinson, C.Hollingworth. and O.A. Misund. . 1998. The challenge .of the herring in the Norwegian Sea: making optimal collective spatial decisions. Sarsia 83:149-167. Fuchs, F. 1991. An expert system for analysing relationships between fish and the environment, InternationaI.Council fortheExploration of the Sea, C.M. 19911D:5, 6pp. Gerlotto, F and. Freon,P., Soria, M., Cottais, P-H., and Ronzier, L. 1994. Exhaustive observation of 3D school structure using multibeam side scan sonar: potential use for school classification, biomass estimation and behaviour studies. ICES CM 1994B: 26, 12 pp. Harden Jones, F.R. 1968. Fish Migration. Edvard Arnold (publishers) Ltd. London. 325 pp. Hilborn, R. and Mangel, M. 1997. The ecological detective: confronting models with data. Princeton UniversityPress,Princeton, New Jersey. 315pp. Hutchings, J.A. 1996. Spatial and temporal variation in the.(iensity of northern cod and a review of the hypotheses for the stock's collapse. Can.J. Fish Aquat Sqi. 5~:943-962. Johannes,RE. 1978. Words of the Lagoon.: Fishing and marinelaw in the Palau district of Micronesia. University of California press, Berkeley, 245 pp ..• Johannes, R.E. 1989. Traditional ecological knowledge: a collection of essays. Gland, Switzerland, ruCN, World Conservation Union, 77 pp. Kosko, B. 1993. Fuzzy thinking: the new science of fuzzy logic. Pub!. Hyperion, New York, USA. 318p. Levin, S.A. 1992. The problem of pattern and scale in ecology. Ecology 73 (6): 1943-1967. Mackinson, S. and N0ttestad, L. In press. Combining local and scientific knowledge. Reviews in Fish Biology and Fisheries. .. -- ... . - - .. _._... - -_._-- Maguire, J-J., Neis, B. and Sinclair, P.R. 1994. What are we managing anyway?: the need for an interdisciplinary approach to managing fisheries ecosystems. ICES-CM-1994/T:48 Maravelias CD & Reid DG. 1997. Identifying the effects of oceanographic features and zooplankton on prespawning herring abundance using generalized additive models. Marine Ecology Progress Series, 147: 1-9. Masse, J. Koutsikopoulos, C., and Patty, W. 1996. The structure and spatial distribution of pelagic fish schools in multispecies clusters: an acoustic study. ICES j. Mar. Sci. 53: 155-160. McKeown, B.A. 1984. Fish Migration. Croom helm, Timber press, Portland Oregon, 224pp. McNeill, D. and P. Freiberger. 1993. Fuzzy Logic. Publ. Touchstone, New York. ISBN:0-671-73843-7, . 320 pp. Meech, J.A. 1992. Managing uncertainty in expert systems: a fuzzy logic approach. AI in mineral processing, 31't CIM Metsoc conf., Edmonton, Alta, 77-85. Meech, J.A. and Kumar, S. 1995. A hypermanual on Expert Systems. V.5. Ed. Canadian Centre for Mineral and Energy Technology, Ottawa, Ont., Canada. 3200 ep (electronic pages) Misund, O.A., A. Ferno, T. Pitcher, and Bjorn Totland. 1997. Tracking herring schools with high resolution sonar. Variations in horizontal area and relative echo intensity. ICES J. Mar. Sci. 54: 11 Neis, B., Felt, L., Schneider, D.C., Haedrich, R., Hutchings, J. and Fischer, 1. 1996. Northern cod stock assessment: what can be learned from interviewing resource users? DFO Atlantic Fisheries. Research document 96/45, 28 pp. . Nliltlestad, L., Aksland, M., Beltestad, A., Ferno, A, Johannessen, A., & Misund, O. A. 1996. Schooling dynamics of Norwegian spring spawning herring (Clupea harengus 1.) in a .coastal spawning area. Sarsia 80, 277-284. Petitgas,P. and J.J. Levenez. 1990. Spatial organisation of pelagic fish: echogram struclIire, spatio­ temporal condition, and biomass in Senegalese waters. ICES J.Mar.Sci: 53147-153. Pinkerton, E.and Weinstein M. 1995. Fisheries that work. Sustainability through community based management. A report to the DavidSuzuki foundation, Vancouver, BC, Canada. 199pp. Pitcher, T.1. 1993. Functions of schooling behaviour in teleosts. In: The Behaviour of Teleost Fishes, 2nd ed., Ed. by T.1. Pitcher: Croom Helm, London & 'Sidney, 364-439. Pitcher, T. J., Misund, O. A, Ferno, A., Totland, B., & Melle, V. 1996. Adaptive behaviour of herring schools in the Norwegian Sea as revealed by high-resolution sonar. ICES Joum. Mar. Sci.· 53, 449-452. Reinventing Fisheries Management. 1996. Fisheries Centre Research Report Vol 4 (2). 84p. Saila;'S.B. 1996. Guide to some computerised artificial intelligence methods. In Megrey, B.Aand Moksness E. [eds.] Computers in Fisheries Research. Chapman and Hall, London, pp.8~37. Schweigert, IF. and M. Linekin. 1990. The Georgia and Johnstone straits herring bat fishery in 1986: Results of a questionnaire survey. Can. Tech. Rep. Fish. Aquat. Sci. 1721: 44pp. Veiga, M.M. and Meech, J.A. 1994. Application of fuzzy logic to environmental risk assessment. Paper presented at IV Encuentro Hemisferio Sur sobre Tecnologia Mineral, Concepcion, Chile, Nov. 20-23, 1994. Warwick, CJ., Mumford, J.M. and Norton, G.A 1993. Environmental management expert systems. Journal of environmental management, 39: 251-270. Zadeh, L.A 1965. Fuzzy sets. Information and Control. 8 (3): 338~353. Zadeh, L.A. 1973. Outline of a new approach to the analysis of complex systems and decision processes. IEEE transactions on systems, man and cybernetics, Vol SMC-3, No.1. January 1973.

12 END USER

Facts about problem Solutions E "WoW."."."."'''"''''.'''''.".".'''''''.'''.'''''."'''"''''',1 X P E R T INFERENCE ENGINE S Rules y S KNOWLEDGE BASE T

.. ' .. ' E M engineer Practical data Hard data (scientific kno ledge)

Published data Field studies Fishers Fishery managers Scientists

Figure 1. Schematic of an expert system. Knowledge Base - Typically knowledge is stored in the form of heuristic rules although other more complex data structures may be used. Rules have the format "IF a certain situation occurs THEN a known outcome is likely" and are gathered from expert sources, manipulated and input to the knowledge base by the "knowledge engineer". Inference Engine - The inference engine compares rules against known facts (input by user), to determineifnew facts can be inferred. Since the system developed here is predictive, a forward chaining (or 'data driven') inference strategy is used. Given the conditions, it searches forward for possible solutions.' User Interface - provides the link from system to user. It is designed using graphics and hypertext. An important element is the facility to provide explanation of the actions of the system, so that at any time the end user can ask for help or ask, why?

13 Attributes Descriptors

A. Internal factors A. Structure i) Motivational (driving) factors I. Shape: horizontal and vertical extent I. Maturation 2. Size (mass) 2. Hunger 3. Area/extent 3. Energy saving (e.g. hibernation) 4. Packing density 4. Risk avoidance 5. Average speed of fish in shoal 6. Internal structure: schooling / shoaling ii) Biological (Intrinsic) factors 7. Depth in water column (relative to 4. Fish length bottom) 5. Experience (past history/learning) 8. Dynamic tendency - frequency of split, 6. Life history stage (season) join or change shape 7. Pheromones - during spawning 9. Size segregation: separation of agel 8. Health (diseases etc) size within schools 9. Condition (fat content) 10.Cohesion -likelihood of splitting 10. Abundance: stock size completely 11. Avoidance response B. Distribution II. Mean inter-school distance B. External factors 12 . Nearest neighbour distance 13. Stock range i) Biotic 14. Migration direction 12. Predators abundance 15. Distance to shore 13. predator type: fish, mammals and birds 16. Physical feature associations 14. predator distribution 15. Food abundance 16. food Size 17. food distribution 18. Oxygen 19. Competition

ii) Abiotic 20. Light (time of day)- related to season and ------latitude - - _,, __ n. ..---.-.-.--•• -". ------"- 21. Currents 22. Temperature 23. Vessel activity 24. Water depth 25. Bottom topography 26. Bottom substrate 27, Fisheries 28. Weather 29. Available spawning habitat

Figure 2. List of attributes and descriptors

14 Member sets of the fuzzy variable 'predator abundance scale'

o~----~------~~----~----~----- 1 5 10 Predator abundance scale

Figure 3. Membership functions of fuzzy sets on the variable 'predator abundance scale'. The member sets (also sometimes caned subsets) are the linguistic concepts: rare, occasional, frequent, common and abundant. The slope and degree of overlapping of the memberships functions is the key element determining how unique or 'fuzzy' the sets are. The confidence on the Y-axis shows our degree of belief in the linguistic concepts. For example, when predator abundance is 5 on the scale weare 0.8 confident that predator abundance frequent and also 0.2 confident that predator abundance is common. In an expert system both pieces of information are used simultaneously to make conclusions, thus avoiding the simplistic notion that something is or is not true, when in fact it may be both to different degrees. Thus it implicitly captures uncertainty. The value of predator abundance scale whose membership (confidence) is 1, is callec!:theJ~upremulI!vaJl.Ie.The range()tPI~ciator abuncllJJ1ce_s.qal(j values contained by a fuzzy set is caned the support. values

15 SPATIAL

Immature Collapsed

Adolescent Collapsed

Mature Collapsed Very small STOCK SIZE

Figure 4. F AM map associating stock size and age structure to spatial range occupied· by the stock. Each box in the map represent one rule in the expert system. Note that each choice of the qualifiers age structure and stock size are previously represented as fuzzy sets.

t

0) Medium ~. 0.6 ...... Olo.. 0)'- -0.1::.en L- L- a 0) .0 8E ·53 .'.~ O' 3 ...... l:..c::.rg e -0.... .' <;:: ...... 8(;)0.2 ... ,... ,...... /i .. all

Smal1min Medmin Smallmax Largemin Med",.x Large Smallsup Medsup sup Shoal size (t)

Figure 5. Output fuzzy sets for shoal size used in defuzzification. Smallsup, represents the supremum values of shoal size for the fuzzy set small. Similarly, Smallmin to Smallmax represents the support of the fuzzy set small.

16 -

Select the liillI!LEill from the menu below: r Pre-spawning r. Ocean feeding r Spawning r On-shore migration r Off-shore migrating r Overwintering Do you wish to know about a specific time of day? r.r Yes Dawn Please c.heck the box IL______ill=El ... below atter making your r NO selection r.r Done Enter the number of in the stock, a rough estimate of the size of the stock you are and select a description that best describes the YEAR CLII.SSES (1-15) 7 F==L----, STOCK SIZE (tannes)

RELATIVE STOCK SIZE ~~~!p1]

Figure 6. Example of user-friendly input interface.

17 .-

20 o a. Relative depth· . b.· Packing density 10 14 20 12 %30 c

100~c-~~~~---,------,------~ 0+------,------r------T7~--_, Day Dusk Night Dawn Day Dusk Night Dawn

c. Ease of capture d. ShoalNND 1.4

1.2

E0.8 -'" 0.6

0.4 0.2

O+------,--~--._~~_.--~__. Day Dusk Night Dawn Day Dusk Night Dawn

Figure 7. Example results showing changes in shoal structure and distribution during different periods within one day. In addition to time of day, other important conditions that are considered in making the above conclusions are: life phase - overwintering, moon stage - not full, weather conditions - fair, .. ·""predators"abUlldarice·-low;"higJi Pri6trity on riskavoidance and energy savirtg:Theresrnts show a ... reasonable depiction of the mesoscale dirunal changes known to occur in herring. Furthermore, the run also predicts that during overwintering the average shoal size is large (1580 t), average swimming l speed is low (0.25 ms· ) (presumably due to energy saving) and shoals exhibit very restricted movement (i.e. holding).

18