SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 19

2. clothing. Name probably connotes the long, bending appear- Buburupoto: Oplismenus compositus (A grass which ance of this soft-fleshed sea cucumber. is common in disturbed areas, such as the fringes of gardens). A calendar plant. The presence of the Plants sticky seeds of this species (March, April, May) indicate a bad time for fishing. Etymology: Buburu = Ahoaho: Premna corymbosa (Family Verbanaceae). grass; poto = generic for some types of Damselfishes. The name A beach-side tree, the leaves and small branches of connotes the sticky, clinging nature of the seeds of this grass, which are often taken by women on long canoe which is likened to the pugnacious behaviour of Poto. trips to ward off sea-devils (Asi). Koga: a species of mangrove (No identification) Alite: Terminalia catapa. Also called Tahile and Kulikuli: (Seagrass {generic}). Lengga. A calendar plant. The presence of red Busu: a type of green alga preferred by hawksbill leaves on this semi-deciduous species (usually turtles (possibly Chlorodesmus chloroticus). twice a yearÑaround June and December) indi- Tongo: (Mangrove {generic}). cate a time when Ôred fishÕ aggregate and/or have Tongo bua: a species of mangrove (No identifica- oily flesh (mona). This might include holocentrids tion). (Sori, Talaa, etc.), serranids (Sivari, Taburara, etc.) Tingale: a species of mangrove with small leaves. and lutjanids (Koukoru, Uvoro, etc.).

The use of optimal foraging theory to assess the fishing strategies of Pacific Island artisanal fishers: A methodological review by Shankar Aswani 1

In this paper, foraging theory and its methodology are presented as a complementary framework to the study of Pacific Island artisanal fisheries. It is expected that such inclusion will allow for the development of a clearer anthro- pological model describing the relationship between human foraging and fishery management.

Introduction From the standpoint of maritime anthropology, any comprehensive study of the integration of Artisanal fisheries play a major role in the marine and terrestrial biotic components requires social, cultural, and economic life of most Pacific the parallel consideration of human activities, Islanders, particularly in rural communities where including existing property regimes, resource people are highly dependent on marine resources access and distribution rules, and resource for subsistence and commercial purposes. Yet, exploitation strategies. Although numerous stud- marine resources are being threatened by pressure ies have concentrated on the social aspects of from exploding human populations and the Pacific Island artisanal fisheries (e.g. Johannes, increasing commercialisation of the subsistence 1981; Hviding, 1996; Lieber, 1994), few have dealt fisheryÑcircumstances which are now forcing explicitly with the micro-ecology of daily human- researchers to find novel ways to examine issues marine interactions (see Aswani, 1997; Bird & of coastal management and marine resource con- Bird, 1997). Such neglect has hampered attempts servation. Among the most recent approaches to to fully integrate studies of environmental coastal coastal management has been to study marine eco- processes with those of human activities. logical processes in conjunction with those of the In this paper, I examine the utility of optimal contiguous shoreline and upland habitats, or what foraging theory and its methodology, as applied to has been termed Integrated Coastal Zone the study of Pacific Island artisanal fishers. The Management (ICZM). inclusion of foraging theory can contribute to

1. Mailing Address: 3093 Pualei Cr. #309, Honolulu, Hawaii 96815, USA. The author is a Research Associate (consultant) for the Pelagic Fisheries Research Program (Western Pacific Regional Fishery Management Council) Project: ÔThe Hawaii Troll and Handline Fishery: FishermenÕs Motivations and Fishing ActionÕ 20 SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 building a clearer anthropological model to The former choice is examined by diet breadth describe the relationship between human foraging and patch choice models (Charnov & Orians, 1973; and fishery management. MacArthur & Pianka, 1966; Stephen & Krebs, 1986) which solve for the decision component of Theoretical review food harvest, or the probability that a forager will select a given prey or patch upon encounter. The Since the mid-1970s, a growing number of two main model components are search time, or anthropologists have employed optimal foraging time spent looking for prey or patches, and handle theory as developed in evolutionary ecology to time or time employed following, capturing, and study the subsistence practices of indigenous peo- processing prey. ples. Evolutionary ecology explains human The second choice, time spent in a patch, is behavioural adaptations in ecological context addressed by the patch time allocation model through the use of natural selection theory. Briefly (Charnov, 1976) which examines the decision summarised: individuals exhibit genotypic varia- variable for how long to forage. The two main tion that affects their capacity to survive and components of this models are travel time, or time reproduce. Certain adaptive traits will dominate spent looking for adequate patches, and residence over time and become prevalent in a population. time, or time spent in a given patch (Stephens & The objective of evolutionary ecology is to exam- Krebs, 1986). ine the phenotype of an organism (i.e. biological Another significant axiom of optimal foraging and behavioural traits) and to explain why certain theory, and perhaps the most polemical, is the phenotypic traits (e.g. foraging strategies) evolve modelÕs currency assumption. In determining the in a given ecological context (Smith & optimal choice facing a forager, a currency, or the Winterhalder, 1992). cost and benefit decision variable, must be select- Evolutionary theory is too abstract to explain ed for the model. the presence of specific human traits, so a Ômiddle- Typically a currency can be expressed as units range theoryÕ is required to link observed of maximisation (e.g. kcal per hour of foraging), behaviour and general theory (Smith, 1991). In this minimisation (e.g. time, risk), or stability (e.g. respect, optimal foraging theory provides a con- energy versus risk). Most researchers using forag- ceptual link between empirical reality and theory. ing theory have employed a maximisation criteri- The purpose of foraging theory is to formulate on to evaluate foraging decisions (Stephens & testable predictions that can account for foragersÕ Krebs, 1986). decisions (choices) with regards to the types and If maximisation is the criterion, however, what abundance of food they consume (diet breadth), are foragers maximising? (e.g. survivorship, fertili- the areas utilised (patch choice), and the time ty, energy or protein intake, or even money). spent foraging in these areas (patch use). Optimal Anthropologists have commonly used energy foraging models assume that a foragerÕs decisions optimisation as a proxy for reproductive fitness made during foraging are formulated to maximise (e.g. Alvard 1995; Hames & Vickers 1982). Energy short-term energy gains (Stephens & Krebs, 1986). optimisation can be expressed as Ônet acquisition This is an evolutionary approach, because if for- rate,Õ Ônet rate of energy capture,Õ Ôreturn rate,Õ or agers successfully adapt to a long-term foraging Ôforaging efficiencyÕ (Smith, 1991: 46). Following strategy that maximises food returns and minimis- Smith (1991), this concept is best expressed as the es resource harvest time, their Darwinian fitness Ônet return rateÕ per capita, or equivalent to the may be enhanced. (For readers feeling uncomfort- energy gained during foraging (the kcal value of able with the Darwinian fitness postulate, strip- the catch) minus the labour input (labour cost ping foraging theory from its evolutionary impli- incurred during foraging including travel, search cations still leaves a operationally-defined set of and handling times) divided by the total residence cost-benefit models capable of empirically mea- time at a patch. The utilisation of calories as units suring particular foraging choices.) of energy maximisation permits the operationali- Foraging models have four identifiable ele- sation of the foraging models without relying on ments (Stephen & Krebs, 1986): nebulous concepts like ÔutilityÕ and ÔfitnessÕ (Smith ¥ the participating actors, & Winterhalder, 1992). ¥ a set of choices made by the foragers, Notwithstanding the conceptual value of ener- ¥ a currency, and gy as a unit of maximisation, numerous social ¥ a set of intrinsic and extrinsic constraints anthropologists have vehemently opposed the faced by the forager. idea of reducing human food disposition to mere All participating actors display a set of decisions caloric values. The general complaint is that forag- and/or choices while foraging. Ordinarily, foraging ing models do not account for cultural and ideo- theory models have examined two decisions: logical preferences of food (e.g. taste, or prestige ¥ what prey (or patch) to consume, and foods) (Smith, 1991). The objective of foraging ¥ when to abandon a patch. models is not to determine human proximate deci- SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 21 sions (i.e. food choice based on ideology) but to similar in several respects, differ because they elucidate the underlying causal structure of those analyse different decisions: where to forage versus decisions. In fact, the idea of calories as the unit of for how long. maximisation may not be so removed from the The patch choice model predicts that patches indigenous view of prey value. For instance, (e.g. habitats) will be selected by a fisher according Pacific Island fishers generally rank prey desirabil- to the mean productivity of that patch. Patches are ity according to its fat content. added to the foraging range until an increase in A more problematic and challenging issue is travel time (i.e. a cost) lowers the mean return rate selecting a unit of maximisation in a monetised for foraging in that patch (Winterhalder, 1981). economy. The classical foraging models focus on The marginal value theorem predicts that if a energy as a proxy for Ôreproductive fitness,Õ and do fisher is foraging optimally, a patch (e.g. fishing not include ÔutilityÕ measured in a monetary curren- ground) should be abandoned when the cy in their predictions. Yet it seems unrealistic to marginal rate of return for fishing in that ground deal exclusively with energy when evaluating for- is equal to the mean return for the entire habitat aging practices in an increasingly monetised global or set of visited patches. The model also fore- economy. Time spent foraging for food is time that casts that if habitat productivity decreases with- could be employed to produce income. out affecting the yield of a specific patch (i.e. This raises an important question: which cur- within the habitat), then foragers should spend rency are foragers actually trying to maximise more time on that patch, and that if productivity when foragingÑcash (per unit of effort) or calo- increases, less time should be allocated to each ries? If the population under study, as is the case ground (Smith, 1991). with some Pacific artisanal fishers, primarily The combined predictions of the models sug- engages in subsistence fishing, then calories are an gest that as seasonal productivity of a habitat type appropriate currency. Alternatively, if fishers (e.g. outer-reef drops) increases, more overall time equally engage in subsistence and commercial is assigned to the habitat, but less time is spent at fishing (i.e. small-scale) a common currency can be each particular fishing ground within it. Frequent developed by converting all foraging inputs and mobility between accessible grounds allows fish- outputs (including cash) into a single currency ers to sustain considerable catches before any of such as net energy capture per hour of labour (for the visited grounds undergoes resource depletion. further discussion see Smith, 1991, 357Ð397). Conversely, as seasonal habitat productivity The final tenet of optimisation models is of con- decreases, less overall time is assigned to the habi- straint assumptions. Briefly defined, constraints Ôare tat and, when visited, more time per bout is spent all those factors that limit and define the relation- at a fishing ground. It does not pay for fishers to ship between the currency and the decision vari- move elsewhere within the habitat if they cannot able(s)Õ (Stephens & Krebs, 1986: 9). Constraints can do better. Alternatively, fishers can search for be extrinsic and/or intrinsic to an organism. more productive habitat types (e.g. inner-lagoon Extrinsic factors which limit a fisherÕs foraging abil- reefs) as long as they are accessible and travelling ity include constraints such as changing patch pro- costs are not to high. ductivity, changing weather patterns, and even social constraints such as religious bans on working Analysing Pacific Island artisanal fishers: on Sundays. Intrinsic constraints are those which A case study physiologically limit the capacity of an organism to interact or tolerate environmental variables. This section describes the methodology employed to test the foraging models outlined in The foraging models: two examples this paper. The case study presented here is based on my own research conducted at the Roviana and Foraging theoryÕs analytical value is evaluated Vonavona Lagoons in South West New Georgia, in this paper by presenting two complementary Solomon Islands from April of 1994 through models: the patch choice (MacArthur-Pianka, December of 1995 (see Aswani, 1997). 1966; Charnov & Orians, 1973), and the marginal A major objective during this research was to value theorem (Charnov, 1976) patch time alloca- describe the behaviour of fishers and to account tion models. for the temporal variability of their activities. This The general theoretical objectives of these mod- required my direct participation in fishing forays els, as applied in a marine context, are to under- as well as that of my assistants. Participation in the stand the daily and seasonal movement of marine fishing activities of Islanders allowed me to under- foragers. The first model is designed to forecast a stand the complexities involved in their daily fisherÕs habitat selection, whereas the second com- choices, which could never have been attained by plements the former by predicting the time that a interviewing alone. fisher should spend harvesting prey in a fishing To elicit detailed comparative behavioural data ground or set of grounds. These models, although for other fishers, self-reporting diaries were hand- 22 SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 ed out to men and women 2 in villages across the in a canoe). In figuring expenditure rates for lagoons. These were important to understand labour input, the recorded times for behavioural regional variation in foraging strategies. categories are multiplied by standardised indirect Direct participation by my assistants and me in calorimetric measures. This procedure will be fishing forays produced records on 372 fishing explained below. During analysis of fishing for- trips to 978 fishing grounds, encompassing a total ays, work categories for each participating indi- of 751.4 hours of fishing. Including the self-report- vidual were broken down into two groups: travel ing diaries, a total of 2,203 fishing trips encom- and within-ground activities. These two were fur- passing 5,920.7 hours of fishing in 4,445 visits to ther subdivided into their respective behavioural fishing grounds were collected. Data compiled categories. Several stop watches were used con- during fishing forays included data categories currently to time observed behaviours. such as: In addition to recording the behaviour of ¥ name, sex, and age of participants, observed fishers, these detailed measurements ¥ date and village, served to calculate the behavioural ratios for each ¥ total time allocation and time-motion fishing method. These, in turn, were used as a records for all behavioural categories con- proxy to figure out the behaviour of fishers in ducted at each ground, trips that neither me nor my assistants had ¥ name and environmental characteristics of observed (i.e. foraging diaries). For instance, exploited grounds, anglers generally spend 27 per cent of their with- ¥ name and number of captured species, in-patch time in some handling activity (casting, ¥ total weight of catch by species and areas bating, unhooking fish etc.) whereas 73 per cent is visited, spent waiting. This ratio was applied to the forag- ¥ fishing methods employed, ing diaries, which did not have as much detailed ¥ mode of transportation, information as the focal diaries. If a fisher stayed ¥ expenses incurred (e.g. petrol cost when 50 minutes in a patch, it was assumed that 13.5 using outboard motors, hooks lost, etc.), minutes had been employed in handling, while ¥ income, if any, and 36.5 minutes were used in waiting (search time). ¥ weather patterns, including tidal cycle, lunar stage, wind direction, and other environ- The diary method mental variables. Moreover, during fishing trips I was able to To complement my own observations and to elicit other types of information such as the ethno- attain detailed comparative behavioural data for historical characteristics of the seascape, localised other fishers in different areas of the Roviana and temporal events (e.g. fish aggregations), and spe- Vonavona Lagoons, the diary method was cific data on prey species. employed. This method was indispensable in The data collected during the focal analysis accessing data on regional variation in habitat selec- and foraging diaries form the basis to test the for- tion, differences in methods used, seasonal influ- aging models presented in this paper. Foraging ence on fishing strategies, and the Ôforaging histo- effort (labour input) and foraging outputs (the riesÕ of particular individuals. Most importantly, catch) data are essential to estimate the foraging the use of this method allowed for the examination efficiency of Roviana fishers. Although measuring of seasonal cross-regional time allocation to various the output of fishing activities was not too diffi- habitat types and the correlation between time allo- cult, figuring out the labour input of fishers was cation and relative resource abundance (i.e. as mea- more complex. The primary analytical tool sured from recorded yields). employed to calculate labour inputs was time- The diary method consists of randomly select- motion analysis (see Nydon & Thomas, 1989). ing subjects to keep diaries of their foraging activi- ties. In this study, random selection of informants Time motion analysis was not always achievable. Selecting the appropri- ate subjects was hard because many fishers were Time motion analysis is a research strategy either unwilling to keep a log of their activities or used by ecologists and some anthropologists to simply could not handle the provided materials. determine the time and energy that an organism Also problematic was the fact that many fishers spends in an activity. The first analytical step is to were only interested in the provided materials and break down observed patterns of behaviour into did not care about the project. Those fishers will- work categories (e.g. paddling, walking, etc.) and ing to cooperate were given a watch, a scale, a to measure the specified behaviours by timing pen, and a set of standardised forms. individuals while they conduct them (e.g. sitting Approximately one hundred wrist watches and

2. Only nine women participated in writing the foraging diaries. Nevertheless, my assistants and I recorded the activities of over one hundred women across the region. SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 23 scales were handed out in hamlets throughout the participating individual. To do that, the standard Roviana and Vonavona Lagoons. Each subject measures for BMR, or Ôbasal metabolic rate,Õ for kept the material if they satisfactorily completed at different age, sex, and weight groups in the FAO least 25 fishing events. Fishers were not told that report (1985) were utilised. For instance, if the they could keep the materials to prevent people Papuan study tables indicated that the energy from falsifying their diaries. Diaries where I sus- expenditure for Ôpaddling canoeÕ for a male pected cheating (e.g. a record of ten full moons in between 25- and 65-years-old weighing 65 kg is a single month!) were discarded. Notwithstanding 3.2 kcal per minute of labour, and I had to correct the possible sampling bias, my own observations for a woman weighing 50 kg and 32-years-old, the of fishing patterns indicates that the chosen fishers following was carried out. If the BMR for that (both male and female) were a representative sam- individual was 1,290 kcal per day (FAO, 1985: 72), ple of the fishing population. Despite the prob- this number was divided by the number of min- lems, many fishers were interested and keen to utes in a 24-hour period, or by 1,440. The result cooperate in this project. In the 20-month duration was then multiplied by the standardised expendi- of this project, 1915 foraging diaries were collected ture rate for males to adjust for the age, weight, from more than one hundred participating fishers. and sex of the subject. Therefore, the energy To make sure that fishers in my village were being expenditure for Ôpaddling a canoeÕ for a 32-year- accurate in their self-reporting, I frequently old female weighting 50 kg was equal to recorded their movements while out fishing to 1,290 Ö 1,440 x 3.2 = 2.87 kcal per minute. Using cross-check their reported times. the Papua New Guinea study and other sources, coupled with the BMR calibration for specific age, Estimating the ‘foraging efficiency’ of fishers sex, and weight provided by the FAO (1985) tables, a range of energy expenditures for Roviana Energy maximisation as a unit of foraging effi- and Vonavona fishers was determined. ciency is best expressed as the Ônet return rateÕ per capita. This rate (R) is equivalent to the energy Estimating output—catch values acquired (Ea) during fishing (the kcal value of the catch) minus the labour input (Ee) (labour cost Energy outputs harvested during fishing are incurred during foraging including travel, search, equivalent to the edible weight of the catch multi- and handling times) divided by the total residence plied by standardised caloric values. In this study, time (t) at a patch multiplied by the number of the energy returns of each catch varied according participating foragers. This is mathematically to the caloric value of the constituent species. expressed in the following equation (Reproduced When possible, the catch harvested at each visited from Smith, 1991: 186): fishing ground was separated by species. For small

n catches dominated by multiple species of small reef fish, an averaged measure was used to deter- R = · (Ea Ð Ee)/(t) (n) i =1 mine the energy value of the catch. The literature on fish nutrition and seafood (e.g. Nettleton, 1985) Estimating labour costs indicates that the edible portion of a whole fish is about 60 per cent (for shellfish and crustaceans this The labour energy expenditures were calculat- measure varies between 10 and 40%). However, ed by taking the time-motion records for each vis- these measures are for edible portions considered ited fishing ground and multiplying them by by Western consumers, and do not include parts of established calorimetric values.3 Estimating ener- fish and crustaceans eaten by other populations gy expenditure rates from time-motion data is a (e.g. head, liver, eyes etc.). To adjust for difference proxy method to calculate human energy expendi- in feeding habits between Western and Melanesian ture. Reliable energy expenditure rates for a populations, a 10 per cent edibility portion was Melanesian population have been provided by added to fish, crabs, and crayfish. Norgan, Ferro-Luzi, and DurninÕs (1974) study of energy expenditure amongst the Kaul, a Papua Estimating the net return rate New Guinea coastal population. Measures attained from this study were complemented with Once the energy input (labour costs for an FAOÕs (1985) energy expenditure tables for subsis- activity) and the energy outputs (value of the tence societies. Because these tables only offer catch) were solved, the unit of foraging efficiency, mean energy expenditure measures, it was neces- or Ônet return rate,Õ was determined algebraically. sary to calibrate for age, weight, and sex of each As an example, if a male in his 40s weighing 65 kg

3. Labour expenditure for fishing at a fishing ground also includes energy expenditures incurred during searching for bait. If the fisher visited more than one ground, bait-search labour expenditure was factored among all visited grounds. 24 SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 paddled for 12 minutes, stayed in a fishing type were sorted by indigenous name and their ground angling for 47 minutes, and caught a bar- mean return rates determined. Each ground was racuda weighing 2.3 kg, and then paddled back further sorted by tidal season to see if patterns of to the village in 10 minutes, the net return rate time use simultaneously changed with seasonal was calculated as follows: The labour cost is shifts in localised mean productivity. A Pearson equal to a total of 22 minutes x 3.3 kcal per correlation coefficient test was utilised to analyse minute of paddle (at regular speed) + 47 minutes return rates and concurrent time use across sea- x 2.1 kcal per minute of angling time. This is sonal variation for habitats and specific grounds to equal to a labour cost of 171 kcal. check for negative and positive correlations in the The next step was to calculate the energy output data. A positive correlation between overall time of the catch. If the barracuda weighed 2.3 kg but allocation and habitat seasonal productivity indi- only 70 per cent of it was edible, the actual usable cated that the most productive habitats received portion was equal to 2.3 x 0.7 or 1.61 kg. The caloric the most attention in a given season. Concurrently, value was then calculated by multiplying 1,610 g a negative correlation between per-bout foraging (1.61 kg) by the energy value for Pacific barracuda, time in fishing grounds within the habitat type or 118 kcal per 100 g edible portion, so that and their seasonal mean productivity indicated an 1,610 x 118 Ö 100 = 1,899 kcal. Subsequently, the inverse relationship between time spent in a labour output was subtracted from the input to ground and its yields. A t-test was conducted on figure the net energy return, or 1,899 Ð 171 = all data sets to check for statistical significance. 1,728 kcal. To convert this measure into a rate, It should be noted that to uncover the the net return was divided by the time spent for- behavioural patterns of Roviana and Vonavona aging, so that 1,728 Ö 47 minutes = 37 kcal per fishers, data sets for each village were sorted in minute of foraging, is the net return rate. This, in many different ways. For instance, data were sort- turn, was multiplied by 60 minutes to find the ed by Ôspecial eventsÕ (e.g. fish aggregations) to hourly rate. Whereby 37 x 60 = 2,206 kcal would explore the effects of sudden changes in patch pro- be the hourly rate gained for fishing in this fish- ductivity on indigenous selection of fishing ing ground of a habitat type at that specific sea- grounds and subsequent uses of time. In assessing son and time of the day. individual responses to changing productivities, several fishers were analysed to trace their month- Calculating mean return rates for fishing ly selection of fishing methods, habitats, and fish- methods, habitats and fishing grounds ing grounds. Additionally, events that included income returns were sorted independently to see The previous section has shown the general if a changing currency (i.e. kcal to cash unit) method employed in this study to factor the net resulted in differences in time allocation. return rate. In this section the methodology employed to calculate mean net return rates for all Implications for the analysis of fishing methods, habitat types, and grounds are Pacific Island artisanal fishers outlined. The initial step to was to code all forag- ing events and to enter each respective visit to A question that remains to be answered is what fishing grounds as separate cases (4,445 cases). does confirmation or refutation of optimal forag- Once the data was coded, the next step was to find ing theory hypotheses tell us about the foraging out the seasonal mean net return rates for each strategies of Pacific Island artisanal fishers? The fishing method, the major habitats, and for specific first implication is a theoretical one. Confirmation fishing grounds within each habitat. Finding the of the foraging hypotheses suggests that fishers seasonal return rates for each method revealed the optimise their short-term self interests by harvest- effectiveness of each technique, and the geograph- ing resources as efficiently as possible. The models ical disparities in yield and effort for each method. presented in this paper hypothesise that individu- The environmental productivity (i.e. measure of als chose habitat types and the foraging times allo- relative abundance only) of each habitat was cated to them according to changes in habitat sea- assessed by sorting all bouts by habitat type and sonal productivity. Such a strategy can result in attaining their mean return rates. Subsequently, the conservation or depletion of resources, each habitat type was sorted by the three main depending on changing environmental conditions. tidal seasons in Roviana (see Aswani, 1997) to Resource depletion may occur during periods of attain seasonal yields and overall foraging effort resource scarcity when fishers increase pressure allocated to each. The overall time allocation on specific grounds (i.e. if there are no alterna- results for each habitat type illustrated whether tives), whereas conservation may occur during fishers were allocating more fishing effort to habi- periods of resource abundance when fishersÕ tats experiencing an increase in seasonal produc- movement between fishing grounds, to increase tivity. In fine tuning the analysis of seasonal pat- short-term foraging efficiency, results in the aban- tern, individual fishing grounds within a habitat donment of remaining prey. Foraging theory SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998 25 shows that the consequences of human foraging (Aswani, in press). Finally, data on the relative behaviour are conditional and dynamic. productivity of habitat types and specific fishing On the other hand, rejection of the foraging grounds, can assist fishery researchers in regional hypotheses can indicate inconsistencies with the stock assessment. modelsÕ assumptions (e.g. need for a new curren- cy), or can show that fishers are indeed practising Conclusion a resource management strategy. Fishers can miti- gate resource scarcity by controlling their short- For all its merits, optimal foraging theory is not term intake rates (i.e. stop resource exploitation) to a theoretical and methodological panacea, and increase long-term sustainable harvests. In this much can be said about its shortcomings. scenario, fishers will actively restrain their efforts, However, a growing number of anthropological whether aware or not, to reduce pressure from studies employing this approach are showing that habitats and fishing grounds experiencing a per- it is robust enough to understand the foraging ceived or absolute decline in productivity practice of subsistence and mixed economy soci- (Aswani, in press).4 Regardless of the results, the eties. It is hoped that the integration of optimal utilisation of foraging theory reveals foraging pat- foraging models to the study of Pacific Island arti- terns that cannot be revealed by conventional sanal fisheries will result in a clearer understand- qualitative ethnographic field methods alone. ing of human foraging activities and their impact A second implication is a methodological one. on the coastal ecosystem. A major problem faced by some anthropologists in the field is the lack of an organised methodology References and theoretical framework. In this respect, optimal foraging theory offers researchers a sound body of ALVARD, M. (1995). Intraspecific prey choice by theory and a systematic set of field research meth- Amazonian hunters. Current Anthropology 36: ods. As foraging models focus on human daily 789Ð818. actions, rather than exclusively on human beliefs ASWANI, S. (in press). Patterns of marine harvest and ideologies, they permit a detailed analysis of effort in SW New Georgia: Resource manage- human foraging practices. Besides acquiring quan- ment or optimal foraging? Special edition on titative measures of time use and yields, the appli- Pacific marine tenure. In: Ocean and Coastal cation of foraging models necessitate the investi- Management Journal, Ed. Kenneth Ruddle. gation of indigenous foraging choices and the ASWANI, S. (1997). Customary sea tenure and arti- extrinsic environmental and social forces shaping sanal fishing in the Roviana and Vonavona them. This requires, among other data sets, the lagoons: Solomon Islands. The evolutionary collection of indigenous ecological knowledge, the ecology of marine resource utilization. mapping of regional ecological characteristics, and Unpublished University of Hawaii Ph.D. dissert. the study of the local social-economyÑdata that BIRD, R.B. & D.W. BIRD. (1997). Delayed reciprocity can be useful for management purposes. and tolerated theft: The behavioral ecology of The final implication is a managerial one. food sharing strategies. Current Anthropology Because foraging models are able to predict the 38: 49Ð78. types and abundance of fish that fishers prey on, CHARNOV, E.L. (1976). Optimal foraging, the the frequency of visits to marine habitat, and the marginal value theorem. Theoretical Biology 9: changing intensification of fishing activities, they 129Ð136. are useful in linking anthropological studies with CHARNOV, E.L. & G.H. ORIANS. (1973). Optimal coastal management plans. Foraging data together Foraging: Some Theoretical Explorations. with local and western biological knowledge can Mimeograph, Department of Biology, Salt Lake be incorporated into management blueprints City: University of Utah. which mimic local seasonal resource exploitation FAO/WHO/UNU. (1985). Energy and protein patterns. For instance, during periods of declining requirements. Technical Report. Series. 724, exploitation, certain habitats could be temporarily Geneva: World Health Organization. closed. Access restrictions to habitats or grounds HAMES, R. & W.T. VICKERS. (1982). Optimal diet that are temporarily considered less desirable than breath theory as a model to explain variability other fishing grounds would likely be more in Amazonian hunting. American Ethnologist acceptable to local fishers than closing prime areas 9: 258Ð278.

4. An important distinction needs clarification. When fishers allocate less overall fishing time to habitats undergoing a seasonal decrease in yields, the behaviour suggests that they are either practicing a resource management strategy or an optimisation one. To distinguish the actual strategy, it is crucial to analyse time-use of fishers while foraging in specific grounds within the habitat experiencing a decrease in yields. An increase in per-bout time suggests a foraging strategy designed to maximise forag- ing efficiency (i.e. only if there are no alternative areas, or travel times elsewhere are too high), while a decrease in time suggests a strategy designed to manage resources. This is counterintuitive to the common notion that fishers decrease time during sea- sonal lows and increase per-bout foraging time during seasonal highs (Aswani, in press). 26 SPC Traditional Marine Resource Management and Knowledge Information Bulletin #9 – February 1998

HVIDING, E. (1996). Guardians of Marovo lagoon: SMITH, E.A. (1991). Inujjuamiut foraging strategies: practice, place, and politics in maritime evolutionary ecology of an Arctic economy. Melanesia. Honolulu: University of Hawaii New York: Aldine de Gruyter. Press. STEPHENS, D.W. & J.R. KREBS. (1986). Foraging the- JOHANNES, R.E. (1981). Words of the lagoon. ory. Princeton: Princeton University Press. Fishing and marine lore in the Palau district of SMITH, E.A & B. WINTERHALDER, eds. (1992). Micronesia. Berkeley: University of California Evolutionary ecology and human behavior. Press. New York: Aldine de Gruyter. LIEBER, D.M. (1994). More than a living: fishing WINTERHALDER, B. (1981). Foraging strategies in the and social order on a Polynesian atoll. Boulder: boreal forest: An analysis of Cree and gather- Westview Press. ing. In: Hunter-gatherer Foraging Strategies. MACARTHUR, R.H. & E.R. PIANKA. (1966). On opti- Eds. B. Winterhalder & E.A. Smith, Chicago: mal use of a patchy environment. American University of Chicago Press. 6Ð98. Nature 100: 603Ð609. NETTLETON, J. (1985). Seafood nutrition: Facts, Acknowledgements issues and marketing of nutrition in fish and shellfish. New York: Osprey Books. I am grateful to the people of Roviana and NORGAN, N.G., A. FERRO-LUZZI, & J.V.G.A. DURNIN. Vonavona lagoons for allowing me to live (1974). The energy and nutrient intake and the among them and study their fishing practises for energy expenditure of 204 New Guinean almost two years. This research was funded by adults. Philosophical Transactions of the Royal the National Science Foundation (SBR-9320498) Society of London B 268: 309Ð348. and Sea Grant, University of Hawaii (R/MA1 NYDON, J. & R.B. THOMAS. (1989). Methodological and NA36RG0507). Additional financial and procedures for analyzing energy expenditures. logistical support was provided by the WWF- In: Research methods in nutritional anthropol- Pacific, ICLARM, and the Solomon Islands ogy. Eds. G.H. Pelto, P.J. Pelto, & E. Messer. Development Trust. The United Nations University: Tokyo. 57Ð81.

Native title recognition of CMT and the implications for the GBRMPA and future management of marine areas by Julie Lahn 1

Introduction and use of the marine environment could be a solid basis from which to build management The Great Barrier Reef Marine Park (GBRMP) strategies. Traditional knowledge is acknowl- stretches along the Queensland coast of Australia. It edged as useful and the study suggested that fur- has often been showcased both locally and interna- ther research and consultation should be carried tionally as the worldÕs most successfully managed out around Australia to Ôtake stockÕ of the infor- marine park. However, in its management of this mation held by indigenous people and to listen to park, the Great Barrier Reef Marine Park Authority current concerns. (GBRMPA) has come under scrutiny by researchers Other research funded by GBRMPA also stress- and indigenous people alike. This paper presents es the importance of Aboriginal interests in marine an update on issues concerning indigenous rights, areas (Smith, 1987). Ethnobiological research car- management strategies and GBRMPA. ried out by Andrew Smith (ibid.) in two Cape York communities, Lockhart River and Hopevale, Indigenous interests in the Great Barrier documented Aboriginal interests in the Cairns and Reef Marine Park Far Northern Sections of the Marine Park. Smith carried out a comparative study of marine hunting The Great Barrier Reef Marine Park Authority and fishing practices of the Hopevale and has initiated research and workshops to examine Lockhart River communities and at the comple- Aboriginal and Torres Strait Islander interests in tion of his research, made suggestions for future the marine park area. One workshop (Gray & directions GBRMPA should take with regard to Zann, 1985) concluded that traditional knowledge Aboriginal and Torres Strait Islander peoples.

1. Doctoral candidate, Dept, of Anthropology & Archaeology, James Cook University, Townsville, QLD 4811, Australia. (E-mail: [email protected])