ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on April 3, 2008
ICES Journal of Marine Science: Journal du Conseil 2008 65(4):612-622; doi:10.1093/icesjms/fsn032
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
School-based indicators of tuna population status
1 School of Zoology, University of Tasmania, Hobart, Tasmania 7005, Australia
2 CSIRO Marine and Atmospheric Research, Castray Esplanade, Hobart, Tasmania 7000, Australia
Correspondence to J. T. Dell: tel: +61 03 62325182; fax: +61 62325187; e-mail: james.dell{at}csiro.au
Dell, J. T. and Hobday, A. J. 2008. School-based indicators of tuna population status. – ICES Journal of Marine Science, 65: 612–622.Theory and limited observation suggest that fish schools consist of more individuals of similar size when populations are large than when they are small. The hypothesis that population size might be indicated by school structure is tested for southern bluefin tuna (SBT), a commercially important large pelagic species, which has undergone an estimated 60% reduction in juvenile biomass since 1960. Fish size data are used to determine whether there have been changes in schooling behaviour that can be used as simple indicators of abundance. During tagging studies, juvenile SBT are removed sequentially from a school, measured, tagged, and released. These sequential size measurements are used here to describe school composition from different years in two locations using simple school metrics (including mean fish size, variance in size, and mean difference in size between sequential fish). Trends were significant in most metrics over the 40-year period analysed, and were inversely related to independent estimates of population size. Simple school metrics are cost-efficient and easily interpreted by stakeholders. Monitoring population trends in near real time through school composition metrics may indicate further decline or recovery of SBT and, therefore, assist future management of tuna and other schooling species.
Keywords: population monitoring, population trends, schooling behaviour, size-based index, southern bluefin tuna, Thunnus maccoyii
Received 27 February 2007; accepted 3 February 2008; advance access publication 3 April 2008.
 |
Introduction |
|---|
 Top Introduction MethodsResultsDiscussionAppendixReferences |
|---|
Recent studies suggest that the abundance of many of the world’s large marine predators have declined since the mid-20th century (Christensen et al., 2003; Ward and Myers, 2005; Polacheck, 2006; Sibert et al., 2006). Declines are in fact seen at many trophic levels of the marine foodweb, with reductions in catch per unit effort, major fishery collapses, and serial depletions reported globally, although mainly in the northern hemisphere (Orensanz et al., 1998; Pauly et al., 1998; Hutchings, 2000; Jackson et al., 2001; Shepherd and Rodda, 2001; Baum et al., 2003; Roman and Palumbi, 2003; Springer et al., 2003). Overexploitation by commercial fishing fleets (including illegal, unregulated, and unreported catches) has been suggested as the main cause of the declines, although increased pollution and climate change may also have played a role in some areas (Dayton et al., 1995; Goodyear, 1999; Kock, 2001; MacKenzie, 2002; Roberts, 2002; Pearce, 2003). A common element in the decline of fished species is that much of the understanding of the mechanisms and indicators of the declines occurs in hindsight when the damage has been done: it is clear that more timely indicators are required urgently.
Population status is often determined using stock assessment techniques. Most current stock assessment methods are costly, data-intensive, and generally provide information on the past status of the stock rather than the current state. This is generally because information from catch records is for older cohorts of fish, rather than for the juveniles on which future population status depends. There is a need for more real-time indicators of stock status; these are of great importance to sustainable fisheries management (Dahl, 2000; Seijo and Caddy, 2000).
The development of new management approaches melding the disciplines of behavioural ecology and population dynamics have been recognized as vital in the future of sustainable fisheries (Pitcher, 1995; Fréon and Misund, 1999). The importance of understanding fish behaviour in relation to catching and managing fish stocks is not a new concept; the origins can be found in aboriginal and traditional fishing practices in the Pacific and Indian Oceans (Fréon and Misund, 1999; Parrish, 1999; Johannes et al., 2000). Behavioural science, however, has been applied more to the development of more-efficient fishing techniques than to management (Wardle, 1993; Fréon and Misund, 1999; Parrish, 1999). Here, we use metrics of schooling behaviour to develop easily collected and calculated indices of school composition. We argue that this behaviour-based approach has particular application to the real-time management of fish stocks.
We demonstrate this approach using southern bluefin tuna (SBT), Thunnus maccoyii, a long-lived, highly migratory species with an extended juvenile period. Mature fish, older than 10–12 years, spawn exclusively in waters south of Java during the austral summer (Farley and Davis, 1998; Davis and Stanley, 2002; Farley et al., 2007). During the subsequent year, juveniles travel down the West Australian (WA) coast with the aid of the Leeuwin Current, and typically inhabit shelf and inshore waters between Perth and Esperance between December and April. The summer months December–April of the following 3–5 years are spent in the waters of the Great Australian Bight (GAB). While in the warm waters (19–22°C) of the inshore and shelf regions of the GAB, juvenile SBT form surface schools or "patches" (Cowling and Gunn, 1999). An Australian purse-seine fishery in the GAB depends on this schooling behaviour for efficiency of fishing operations. Currently, some 98% of the Australian quota for SBT (5265 t; more than one-third of the global total allowable catch set by the Commission for the Conservation of Southern Bluefin Tuna, CCSBT) is caught from schools in that area. The live fish are then towed back to farms in Port Lincoln where they are fattened before export to Japan. Research supporting the management of this fishery has involved a number of mark-recapture programmes, with generally consistent tagging protocols from the 1960s to the present.
Schooling is a common feature in the life histories of most animal resources harvested from aquatic environments (Fréon and Misund, 1999). Most of the large volume fisheries rely on their target species’ tendency to aggregate in schools. Some marine predators also rely on the schooling behaviour of their prey to facilitate feeding (Cherel et al., 1999; Noettestad and Similae, 2001). Therefore, changes in schooling behaviour driven by human activities are potentially of great economic and ecological concern, especially if those changes are linked to changes in abundance.
The breadth of the study regarding the associative behaviour of fish has generated many definitions relating to the groupings of fish (Pitcher, 1991; Pitcher and Parrish, 1993). Here, a school is defined as a group of fish moving together with an element of organization (Fréon and Misund, 1999). Individual fish join schools to (i) enhance predator vigilance and avoidance, (ii) increase feeding and breeding opportunities, and (iii) maximize energetic efficiency. The trade-off between the costs and benefits of these factors control size, composition, structure, and maintenance of schools (Pitcher and Parrish, 1993). The key point for this study, supported by extensive experimental work, is that fish preferentially school with morphologically similar conspecifics (Crook, 1999; Hoare et al., 2000; Krause et al., 2000). The decline in the abundance of a fish population may alter the proportion of individuals within size classes available to form schools. Individuals from depleted size classes tend to join schools of fish of dissimilar size, rather than remain single until they encounter a school of fish of similar size (Krause et al., 2000; Croft et al., 2003).
The collection of behavioural and biological data from the marine environment, in particular from large, fast-moving, highly migratory fish, is both complex and costly. Some aspects of the behaviour of these larger species have been inferred from data collected using archival data-logging and acoustics tags (Gunn et al., 1994; Klimley and Holloway, 1999; Gunn and Block, 2001; Kirby, 2001; Davis and Stanley, 2002). Such technology focuses on the movement and feeding behaviours of individual fish and cannot (yet) resolve how conspecifics associate with each other. Here, we investigate social behaviour inferred from a long-term mark-recapture dataset to track changes in size structure of fish schools over time. Specifically, we use size information from schools of juvenile SBT, collected as part of CSIRO tagging operations conducted since the 1960s, from the GAB and southern WA. These tagging programmes were primarily undertaken to aid the study of SBT migration routes and determining age, growth, and mortality for stock assessments (Murphy and Majkowski, 1981; Majkowski and Murphy, 1983; Majkowski et al., 1988; Hampton, 1989; Caton, 1991; Robins and Tsutomu, 1998; Clear et al., 2000; Leigh and Hearn, 2000; Hearn and Polacheck, 2003; Polacheck et al., 2004). Their data have not previously been used to examine school behaviour or school composition in relation to population size.
Our objectives were to quantify the schooling behaviour of juvenile SBT in Australian waters, and to develop a fisheries-independent index of juvenile abundance. This involved the retrospective analysis of sizes of SBT in schools to test if (i) there has been a change in school composition of juvenile SBT over the last 40 years, and if so, (ii) the changes in the size composition of schools are consistent with the decline in juvenile SBT abundance from 1960 to 2000. Testing these hypotheses also provided new perspectives on the response of schooling fish to prolonged exploitation. This approach may provide an additional alternative assessment index that is easy to collect and may be applicable to a range of exploited species.
|
Methods |
|---|
|
TopIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Fish size (fork length) data were collected from schools sampled during juvenile SBT tagging studies conducted in southern Australian waters between 1960 and 2004. Two regions were targeted: southern WA, and the eastern Great Australia Bight (GAB; Figure 1). Tagging operations involved "pole and line" fishing to catch individual fish from SBT schools (Hampton, 1989). Schools of fish were located visually or by trolling, then attracted to the poling vessel with live or dead bait. Individuals were sequentially removed by poling one fish at a time to the deck, where each was measured, tagged, and released back to the school. Tagging bouts lasted up to 5 h and involved many hundreds of fish. An assumption in the approach is that pole-and-line sampling is a non-selective method, so SBT of any size in a school behind the tagging vessel are equally susceptible to capture. Data from the tagging operations support this assumption (see Results below). Although reports on the schooling behaviour of SBT are few, tagging observations support the assumption that aggregations behind the tagging vessels represent a school, as defined in the literature (Fréon and Misund, 1999).
|
Data from the CSIRO and the CCSBT tagging programmes formed the basis of the study. The time of fish capture was essential for the analysis. Fish tagged in quick succession (e.g. within 1–2 min) were assumed to be from the same school. Sequential fish captures on datasheets separated by longer periods may not have come from the same school or tagging bout, so records that did not have associated time data were omitted from further analysis. Initial review of the full CSIRO/CCSBT dataset indicated that approximately half the tagging records were of sufficient quality to be included. Further detail regarding the quality assurance and inclusion of data is reported in Dell (2004).
Tagging data were collected over a long period of time, in several regions, and by many different taggers. Three types of potential bias may impact the derivation of school-based metrics to describe population status: (i) changes in the ocean environment, (ii) biological and/or behavioural variability of SBT, and (iii) variation in the tagging protocol. These three areas are briefly addressed below.
Oceanographic variability is recognized as a key driver in fluctuations of distribution, abundance, and behaviour of many marine species (Bakun, 1996; Gunn and Young, 1999; Hobday and Boehlert, 2001; Sharp, 2001). For example, water clarity, surface conditions, bathymetry, dissolved oxygen, primary productivity, ocean temperature, and circulation can affect the composition and behaviour of schools (Young and Lyne, 1993; Fréon and Misund, 1999; Robinson, 2004). In an attempt to limit the potential bias of intra-annual variation on the behaviour and composition of SBT schools, only schools sampled between October and April were considered in our analysis. Oceanographic conditions were generally similar from year to year in the whole study region, but schools from both regions tended to be tagged increasingly from warmer inshore waters from the 1960s through to the 2000s (Dell, 2004). This trend was attributable to a combination of factors related to the objectives of the original studies, as well as a decline in availability of fish in some regions.
We assumed that SBT schools behind tagging vessels behaved similarly throughout the tagging period, irrespective of year or vessel dimension, but the motivations for juvenile SBT to form schools may differ between the two study regions, so separate analyses were carried out. Juvenile SBT in WA (ages 1 and 2 years) are generally thought to be migrating east during the summer tagging period (Hampton, 1989; Caton, 1994; AJH, unpublished). Fish in the eastern GAB (ages 2–5 years) are thought to be quasi-resident. This may have some bearing on the size or stability of schools formed in the two regions. For example, clusters of schools often form in the eastern GAB. Each cluster may contain many schools separated by <500 m and each containing from 500 to >5000 fish (2.5 to >50 t; Kloser et al., 1998; Bravington, 2002). Fishers exploit these clusters, joining schools to maximize catches or to prolong tagging bouts. School clusters have not been recorded in recent years in the WA region, but may have occurred in the past. In both locations, it is likely that schools undergo fusion and fission events when other schools are encountered (Pitcher and Parrish, 1993). The duration and fidelity of schooling in juvenile SBT has not been quantified, but tagging events longer than 30 min may involve individuals or schools within a cluster joining and leaving the school associated with the tagging vessel. This might increase the variation in the size of fish tagged during a tagging event. The biases associated with the potential fusion of schools were reduced by limiting the fish selected to describe the school to only the first portion of a tagging event, when a single school was being sampled. Only schools where >10 fish were tagged in the first 20 min of each tagging bout, at rates of 0.5–3 fish min–1, were used for the calculation of school metrics. A range of time cut-offs was explored (Dell, 2004), and results were consistent with those presented here.
The tagging protocol was generally consistent over the period considered. Where deviations (such as "only tagging large fish") were noted in the cruise log or tagging sheets, data were excluded. Following careful checking of the original datasheets and data exclusion (
50% of the data were excluded), we contend that the subset we selected was consistent with regard to tagging protocol over the 40-year period, and suitable for calculation of school metrics.
Several simple metrics to characterize the selected schools from each region were evaluated. Metrics such as the mean, maximum, and minimum size summarize the distribution of fish sizes within schools. Variance and the mean first difference (MFD: the average difference in fish size between successive tagged fish) were used to provide an indication of the overall and sequential variability in the size of fish within each school. Linear regression was used to determine the percentage change in school metrics over the period of tagging studies for each region (GAB 1960–2004; WA 1980–2004). Higher-order fits were investigated but are not reported, because data were sparse owing to the sporadic pattern of tagging programmes over the 40 years of sampling. As the movement of juvenile SBT from WA is believed to be predominantly west to east (Hampton, 1989; Caton, 1994; but see Farley et al., 2007), we tested relationships between indices from WA and the GAB. In other words, 1–2-year-old fish schooling in WA may move to GAB waters in subsequent years, so we investigated correlations between the seasonal mean of the metrics from each region using lags of 0–4 years.
The school metrics generated for each region were compared with two other abundance indices used to monitor SBT population size. School metrics were compared with the estimated annual abundance of 2–5-year-old SBT determined by virtual population analysis (VPA; Polacheck et al., 2001; Basson et al., 2005). This population estimate is for the global population for each year (1960–2003), but it is assumed that the major part of juvenile SBT population is in southern Australia at the time of the tagging surveys (Cowling et al., 2003; but see Farley et al., 2007). A second independent estimate of relative abundance derived from an aerial survey for juvenile SBT is available for the GAB for the period 1995–2001 (Cowling et al., 2003). The seasonal mean of each school metric was compared with the aerial survey index (ASI) during overlapping years, using regression analysis. The seasonal means from WA were compared with the ASI using a lag of 1–4 years, based on the assumption that fish aged 1 year would be present in the GAB during following summers when they would be counted in the aerial survey. Although we assumed that the two alternative indicators (VPA and ASI) measure relative abundance, it should be stressed that they too are estimates.
|
Results |
|---|
|
TopIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
More than 139 000 SBT were tagged and released off southern Australia during the 40 years of tagging, but not all fish had suitable information to allow inclusion in the analysis. Less than half of the fish tagged (
45%) had time and school data of sufficient quality to be included. Moreover, the school metrics were only calculated from individual fish captured during the first 20 min of a tagging bout between the months of October and April: this represented <22% of the total SBT tagged. Therefore, the final dataset used to generate the SBT school metrics consisted of 30 000 individual fish drawn from 500 schools tagged in the GAB (155 schools) and southern WA (345 schools; Figure 1). However, we used data from very long tagging bouts (hundreds of fish over >4 h) to investigate the assumption that the sampling was non-selective. Analysis of those data showed that there was no change in the metrics calculated from the first 20 min of sampling, compared with metrics from the whole period. We also calculated a slope measure, to investigate trends in fish size over a sampling interval for a school. School slope was occasionally positive, with larger fish, or sometimes smaller fish, caught later in the tagging bout. However, most often there was no significant trend in the size of fish caught during the school interval. These investigations support our assumption that fish in the designated sampling interval were representative of the school as a whole. The number of available records differed between decades (Figure 2, Table 1). For example, in the GAB, almost 80% of the final data were from the 1990s.
|
|
The size distribution of the original GAB dataset (55 960 fish) was compared with the subset with sufficient information (26 445 fish) and the 9300 fish ultimately used to calculate school metrics (first 20 min of tagging; Table 1). The WA dataset included 24 years of records suitable for analysis. Despite there being fewer sampling years, 83 401 fish were tagged, of which 35 395 records had suitable information, and of these, 20 700 fish were used to determine the school metrics for the region (Table 1). The size distributions of the fish in both areas used in the analyses were representative of the original datasets (Dell, 2004). The size frequency distributions of all fish tagged shifted upwards slightly between decades in the GAB (Figure 2), but in WA, similar size distributions are evident across the decades.
Mean, maximum, and minimum size within a school, size variance and MFD metrics all increased over the time considered (Figure 3; Appendix). The net percentage changes in the seasonal mean of each metric, relative to the value calculated in the first year (1963 and 1980 for the GAB and WA, respectively), are listed in Table 2. For both regions, all metrics changed by between 26% and 259%: mean fish size (GAB 31% and WA 29%), maximum fish size (28% and 39%), minimum fish size (27% and 26%), size variance within schools (48% and 259%), and MFD (45% and 35%; Table 2). The change in metric value was significant for all five metrics in WA, and for all but two metrics (variance and MFD) for the GAB schools. Although the biggest percentage change in the GAB was the variance, this change was not significant; there was wide variation in the estimate over time (Figure 3). The maximum size of fish in a school increased more than the minimum size in both locations, indicating a greater range of sizes within a school. The explanatory power (r2) of each relationship was always <20%, in part because of the range of schools encountered in any year. Variation in the metric values in any 1 year was a common feature to the schools from both regions.
|
|
Lack of data prevented greater investigation of intra-seasonal variation in school composition. There were only three seasons during which tagging operations were conducted at both ends of the summer season (1963/1964, 1995/1996, and 2002/2003), and the number of schools was insufficient to ensure acceptable statistical power. The obvious solution to the issue of seasonal variability in the present dataset would involve selecting only schools tagged at similar parts of the season. Unfortunately, there was not a common month of data collection between all years tagged, so the data period was restricted to October–April.
Analyses on lagged correlations were based on 7 years where schools were tagged in both regions. As a result, the five (0–4 years) lagged correlations were based on 7, 7, 6, 4, and 4 sets of metric values. The only strong correlations between annual metric values in WA and the GAB (r2 > 0.5) were for maximum fish size lagged by 2 years and for minimum fish size lagged by 3 years. The trend in the correlation scores (r2) of the lagged mean fish size was strongly influenced by interannual variation in the metrics (data not shown).
The VPA estimate of juvenile SBT abundance (ages 2–5 years) over the same period shows a decline of 65% (Figure 3), compared with a lesser change of 28–48% in significant school metrics for the GAB region. The 24-year period investigated in WA produced a 20% decline in the VPA population estimate, and changes in the significant school metrics of 26–35% (Table 2).
The comparison between school metrics from each region and the GAB ASI was restricted by the limited overlap of shared years between datasets (only 5 years for GAB metrics and the ASI). The correlations between the GAB and WA metrics and the ASI were weak and not significant (r2 < 0.5, p > 0.05).
|
Discussion |
|---|
|
TopIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Change in composition of juvenile SBT schools
We compared school composition over a period of declining abundance using 40 years of SBT tagging data from southern Australia. Significant changes occurred in metrics of size composition of juvenile SBT schools in the two regions considered, indicating that schools have increasingly included larger fish and a greater range of sizes over time. Variance and MFD analyses indicate that schools are composed of more variable sizes and a less size-ordered structure. The reduced proportion of smaller fish in the GAB schools, as seen by the increase in minimum fish size in schools, is possibly due to a slowed migration from the west, suggesting that younger fish may be spending longer in southern WA before entering the GAB (Figure 4). This is supported by the increase in the mean and maximum size of fish in schools in WA (Figure 3).
|
Although we could not consider this aspect, school size composition and related metrics may vary seasonally owing to the growth of individual fish within a population. Juvenile SBT grow extremely rapidly in the first 2 years of life (>20 cm year–1). Growth rates slow in the third year (<20 cm year–1), and decline further in the following years (Hearn and Polacheck, 2003). Fastest growth is in summer (Caton, 1991), so we might expect an increase in the average mean, minimum, and maximum fish size of the juvenile population from October to April. The effect on variance and MFD is harder to predict. For example, if there is strong fidelity within SBT schools, younger (small) fish within each school would grow faster than older (larger) fish, reducing the size difference between the individuals in a school. It would therefore be likely that the MFD and variance metrics would decline over the summer. There is, however, currently no published data on the school fidelity of SBT, and this represents a notable gap in our understanding of the schooling behaviour of this species.
Potential biases associated with the spatial variability, depth, and sea surface temperature (SST) of sampled schools in the GAB could not be completely removed from the analyses (see Dell, 2004). The average SST and depth of the GAB locations where schools were sampled differed over the 40 years of study (Figure 1); tagging operations in later years focused inshore. Therefore, an alternative explanation for the change in school metrics is that schools found inshore maybe less size-structured and contain a broader range of fish sizes than those found around the shelf break. Studies have shown that fish move between inshore regions and the shelf break (Davis and Stanley, 2002; AJH, unpublished), so this may not be a major issue. For WA, there were not the same potential biases, although the location of tagging has varied longitudinally in response to perceived changes in fish distribution.
The most likely explanation for a change in the composition of juvenile SBT schools is a decline in local abundance in the WA and the GAB. Hypotheses for the decline in abundance include the impact of commercial fishing, long-term environmental change, modifications to historical migration patterns, or reduced juvenile recruitment as a consequence of population decline. No significant environmental trend in the study regions was identified using a variety of environmental variables (Dell, 2004).
Reduced recruitment of SBT aged 1 year to WA would result in a lower density of SBT there, which may have two consequences: (i) less competition for forage, so SBT spend more time there before moving into the GAB, and (ii) with fewer conspecifics available to form schools, the impetus to move into the GAB may be reduced such that fish remain longer in WA (Figure 4), or disperse away from Australian waters (Farley et al., 2007). Our analysis of the WA schools showed that there has been a significant increase in the mean size of fish within schools, with concurrent increases in the size variance and MFD, both indicating more size-mixed schools, suggesting that fish remain longer. Given the sparse temporal overlaps in data, conclusions regarding the similarity in the school metrics from the two regions and the ASI are premature, and the issue should be revisited as both time-series are extended.
A decline in abundance and therefore in the number of fish per cohort may also have altered the migratory behaviour of older SBT. For example, the altered size composition may be due to a larger proportion of older SBT (5+ years) returning to the GAB following winter foraging in the Southern Ocean. Similar changes in schooling and migration behaviour, as a result of changes in cohort strength, have been suggested for other commercial fish species (Hutchings, 1996; McQuinn, 1997; Huse et al., 2002). For example, McQuinn (1997) and Huse et al. (2002) investigated the impact of mixed aged schools on local populations of Atlantic herring (Clupea harengus), and proposed that migration patterns were reinforced and transferred through social interactions between schools (an adopted migrant hypothesis). Similar mechanisms may be occurring in SBT, with older juveniles returning to the GAB; this may have the undesirable consequence of increasing their availability to the Australian purse-seine fishery. A similar range collapse apparently contributed to the collapse of the Atlantic cod (Gadus morhua) fishery off Newfoundland (Hutchings and Myers, 1994).
Population decline and changes in school composition
Juvenile SBT recruitment is estimated to have fallen by between 60% and 80% between 1960 and 2000 (Polacheck and Preece, 2001). The results from the school metric analyses indicated significant changes in school composition over the same period in the GAB, and since the 1980s in WA. These schooling changes are inversely related to the decline in SBT abundance over the same 24–40-year period, supporting the premise that a decline in the abundance of a fish population may alter the behaviour and size distribution within schools.
Changes in juvenile SBT growth rates (Polacheck et al., 2004) are an additional factor potentially affecting school size composition. Changes in growth may be a density-dependent response to a decline in juvenile abundance, leading to decreased competition for resources (Polacheck et al., 2004). Although observed growth rate changes alone are not sufficient to account for the trends observed here, they may be a contributing factor. The estimated changes in growth can explain some of the changes in mean, maximum, and minimum fish size, but not those in the MFD and variance metrics.
We suggest that an increase in variance or MFD indicates a declining population, as mixed schools from different age classes are formed. Obviously, if a population decline continued to such a level that only a single cohort of fish (e.g. age 2) was present in a region, then an indicator such as school-size variance might in fact decrease. If considered in isolation, this might be taken as evidence of recovery, because individual fish could all find similar size conspecifics with which to school. Tracking the other school metrics such as mean size in schools or alternate indicators, such as the number of size classes present in a region, would preclude reaching this misleading conclusion and emphasize the need to consider a suite of indicators of population status. With regard to additional simple metrics, we did consider CV (standard deviation/mean) of fish sizes in a school, but that metric was confounded by changes in minimum and maximum fish size, so was less sensitive than those presented here.
School composition as a population indicator
Indicators have been recognized globally as a central issue in the future sustainability and precautionary management of fisheries and resources (Gunn et al., 1998; Dahl, 2000; Seijo and Caddy, 2000). The successful application of sustainability indicators to a system relies on information being monitored at many levels (biological, economic, and social), and at many temporal and spatial scales within the system (Dahl, 2000). For example, fishery indicators can be variables that monitor aspects associated with a resource (the fish), resource users (the fishers), and/or resource managers (policy-makers and fisheries scientists). Indicators are usually monitored relative to predetermined reference points, and may assume discrete values that represent critical states from a biological perspective (Seijo and Caddy, 2000). The approach presented here, used in conjunction with additional indicators, may provide a useful tool for the early identification of behavioural changes related to population status.
In 1998, the CCSBT reviewed a list of 12 fishery-based indicators, removing some redundant fisheries-based indicators, and adding biologically based indicators (Gunn et al., 1998). Some additional indicators were based on the rates of ageing and growth, determined from direct age determination of otoliths, tagging studies, and modal analysis of catches. The GAB ASI was also incorporated into the list of indicators. SBT school composition is a potential new population indicator, because it is nested at the first level of organization of a fisheries resource, the level of individual schools. The existing indicators operate at large temporal and spatial scales, making it difficult to incorporate the behaviour of the fish, previously identified as an important aspect in the management of fish populations. Recognizing, quantifying and recording relevant changes at a school level may well have ramifications throughout a fishery.
The metrics derived here reveal a change in school composition over the 40 years and 24 years of data collection in the two study regions. The lack of correlation between the regional metrics and other indicators at an annual scale requires further investigation, to determine the applicability of these metrics in quantifying changes in composition at shorter time-scales. For example, a practical fishery indicator would require the collection of data, their analysis, and the presentation of management advice within one fishing season (Fréon and Misund, 1999; Flynn and Hilborn, 2004). A number of positive factors may facilitate the creation of a SBT population indicator based on these metrics of school composition. First, the required data are already collected through existing scientific programmes, ensuring the economic use of funds allocated to tagging and survey programmes. Long-term tagging programmes using conventional spaghetti tags are expensive operations when the logistical costs of tagging, retrieval, rewards for return of the tag, and the administration and storage of data are included. Re-analysis of archived tagging records, chiefly through the application of modern data management and statistical techniques, can help justify the costs associated with the collection and storage of such valuable datasets. For example, preliminary investigations into fisheries management innovations can be modelled using archived data. Questions raised can also help to guide the collection of raw data in future field-based programmes. Second, the data collected during these programmes is fisheries-independent, which reduces issues of size-selective biases common to other fishery-based indicators (Pennington and Stromme, 1998). Third, the ease of data collection and the swiftness of analysis may allow a measure of abundance to be available within the year of sampling, providing a short-term alternative to the current 3-year lag in the comprehensive SBT stock assessments presented to the CCSBT (Polacheck et al., 2001). The procedures and metrics used in our study would be valuable components in the development of an abundance indicator based on changes in school behaviour. The school composition data compiled here provide a reference point to future investigations. Moreover, the simple structure of these indicators makes them accessible to fishers, industry, management, and public sectors, bringing transparency to the assessment process and potentially improving the implementation and enforcement of sustainable management strategies (Froese, 2004). Finally, metrics such as those presented here can be quickly obtained, calculated, and understood by a wide range of stakeholders, so constituting an important tool in the sustainable management of fisheries worldwide (Froese, 2004).
|
Appendix |
|---|
|
TopIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
|
|
|
Acknowledgements |
|---|
We sincerely thank the generations of CSIRO scientists who supported the research programmes and collected the data used in this project, in particular Tom Polacheck, Kevin Williams, John Gunn, Clive Stanley, and Thor Carter. The research was supported by the SBT Recruitment Monitoring Programme, CSIRO Marine and Atmospheric Research, the Australian SBT industry, and the CCSBT. We thank Bob Kennedy at CCSBT for providing data access post-2001, and Louise Bell for creating Figure 4. The staff of the CSIRO Pelagic Ecosystems group provided technical support and encouragement, particularly Scott Cooper, Klaas Hartmann, Sophie Bestley, Anne Preece, Jessica Farley, and Toby Patterson. Reviews by Chris Wilcox and two anonymous referees improved the manuscript.
|
References |
|---|
|
TopIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
-
Bakun A. Patterns in the ocean: ocean processes and marine population dynamics. (1996) California USA: University of California Sea Grant La Jolla. in cooperation with Centro de Investigaciones Biológicas de Noroeste, La Paz, Baja California Sur, Mexico.
Basson M., Bravington M., Eveson P., Farley J. Southern bluefin tuna recruitment monitoring program 2004–05. In: Milestone Report CSIRO Marine Research, Hobart (2005).
Baum J. K., Myers R. A., Kehler D. G., Worm B., Harley S. J., Doherty P. A. Collapse and conservation of shark populations in the Northwest Atlantic. Science (2003) 299:389–392.
Bravington M. V. Further considerations on the analysis and design of aerial surveys for juvenile SBT in the Great Australian Bight. In: Development of a Fishery Independent Index of Juvenile Southern Bluefin Tuna and Improvement of the Index through Integration of Environmental, Archival and Aerial Survey Data—Cowling A., Hobday A., Gunn J., eds. (2002) Hobart: CSIRO Marine Research and Fisheries Research and Development Corporation (FRDC). 145–185.
Caton A. E. Review of aspects of southern bluefin tuna biology, population and fisheries. In: World Meeting on Stock Assessments of Bluefin Tunas: Strengths and Weaknesses—Deriso R. B., Bayliff W. H., eds. (1991) La Jolla, CA, USA: I-ATTC. 181–357.
Caton A. E. Review of aspects of southern bluefin tuna biology, population, and fisheries. FAO Fisheries Technical Paper, 336/2 (1994).
Cherel Y., Waugh S., Hanchet S. Albatross predation of juvenile southern blue whiting (Micromesistius australis) on the Campbell Plateau. New Zealand Journal of Marine and Freshwater Research (1999) 33:437–441.[Web of Science]
Christensen V., Guenette S., Heymans J. J., Walters C. J., Watson R., Zeller D., Pauly D. Hundred-year decline of North Atlantic predatory fishes. Fish and Fisheries (2003) 4:1–24.[CrossRef][Web of Science]
Clear N. P., Gunn J. S., Rees A. J. Direct validation of annual increments in the otoliths of juvenile southern bluefin tuna, Thunnus maccoyii, by means of a large-scale mark-recapture experiment with strontium chloride. Fishery Bulletin US (2000) 98:25–40.
Cowling A., Gunn J. 1999–2000 Integrated Analysis Project Proposal (1999) Hobart, Tasmania, Australia: CSIRO.
Cowling A., Hobday A., Gunn J. Development of a fishery-independent index of abundance for juvenile southern bluefin tuna and improvement of the index through integration of environmental, archival tag and aerial survey data. (2003) Hobart, Tasmania: CSIRO Marine Research Document.
Croft D. P., Krause J., Couzin I. D., Pitcher T. J. When fish shoals meet: outcomes for evolution and fisheries. Fish and Fisheries (2003) 4:138–146.[CrossRef][Web of Science]
Crook A. C. Quantitative evidence for assortative schooling in a coral reef fish. Marine Ecology Progress Series (1999) 176:17–23.[CrossRef][Web of Science]
Dahl A. L. Using indicators to measure sustainability: recent methodological and conceptual developments. Marine and Freshwater Research (2000) 51:427–433.[CrossRef][Web of Science]
Davis T. L. O., Stanley C. A. Vertical and horizontal movements of southern bluefin tuna (Thunnus maccoyii) in the Great Australian Bight observed with ultrasonic telemetry. Fishery Bulletin US (2002) 100:448–465.
Dayton P. K., Thrush S. F., Agardy M. T., Hofman R. J. Environmental effects of marine fishing. Aquatic Conservation: Marine and Freshwater Ecosystems (1995) 5:205–232.[CrossRef][Web of Science]
Dell J. T. Long-term changes in the school composition of a declining tuna species. (2004) Hobart: Honours thesis, University of Tasmania.
Farley J. H., Davis T., Gunn J., Clear N. P., Preece A. Demographic patterns of southern bluefin tuna, Thunnus maccoyii, as inferred from direct age data. Fisheries Research (2007) 83:151–161.[CrossRef][Web of Science]
Farley J. H., Davis T. L. O. Reproductive dynamics of southern bluefin tuna, Thunnus maccoyii. Fishery Bulletin US (1998) 96:223–236.
Flynn L., Hilborn R. Test fishery indices for sockeye salmon (Oncorhynchus nerka) as affected by age composition and environmental variables. Canadian Journal of Fisheries and Aquatic Sciences (2004) 61:80–92.
Fréon P., Misund O. A. Dynamics of Pelagic Fish Distribution and Behaviour: Effects on Fisheries and Stock Assessment. (1999) Oxford: Fishing News Books.
Froese R. Keep it simple: three indicators to deal with overfishing. Fish and Fisheries (2004) 5:86–91.[CrossRef][Web of Science]
Goodyear C. P. An analysis of the possible utility of time-area closures to minimize billfish bycatch by U.S. pelagic longlines. Fishery Bulletin US (1999) 97:243–255.
Gunn J., Block B. A. Advances in acoustic, archival and satellite tagging of tunas. In: Tuna: Physiology, Ecology and Evolution—Block B. A., Stevens E. D., eds. (2001) San Diego: Academic Press. 167–224.
Gunn J., Polacheck T., Davis T., Klaer N. L., Cowling A., Farley J. H., Caton A. E., et al. Fishery indicators for the SBT stock: an update of 12 indicators first used in 1988 plus additional indicators from the 1990s. In: Report from the 1998 Trilateral Scientific Committee for the Commission for the Conservation of Southern Bluefin Tuna, Hobart, (1998) CCSBT-SC/98/40.
Gunn J., Polacheck T., Davis T., Sherlock M., Betlehem A. The development and use of archival tags for studying the migration, behaviour and physiology of southern bluefin tuna, with an assessment of the potential for transfer of the technology to groundfish research. Proceedings of the ICES Mini Symposium on Fish Migration (1994) 21:23.
Gunn J., Young J. Environmental determinants of the movement of juvenile southern bluefin tuna. In: Fish Movement and Migration. Australian Society for Fish Biology—Hancock D. A., Smith D. C., Koehn J. D., eds. (1999) Victoria: Bendigo. 123–128.
Hampton W. J. Population dynamics, stock assessment and fishery management of the southern bluefin tuna (Thunnus maccoyii). (1989) Kensington: University of New South Wales. PhD thesis.
Hearn W. S., Polacheck T. Estimating long-term growth-rate changes of southern bluefin tuna (Thunnus maccoyii) from two periods of tag-return data. Fishery Bulletin US (2003) 101:58–74.
Hoare D. J., Ruxton G. D., Godin J-G. J., Krause J. The social organization of free-ranging fish shoals. Oikos (2000) 89:546–554.[CrossRef][Web of Science]
Hobday A. J., Boehlert G. W. The role of coastal ocean variation in spatial and temporal patterns in survival and size of coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheries and Aquatic Sciences (2001) 58:2021–2036.
Huse G., Railsback S., Fernø A. Modeling changes in migration pattern of herring: collective behaviour and numerical domination. Journal of Fish Biology (2002) 60:571–582.[CrossRef][Web of Science]
Hutchings J. A. Spatial and temporal variation in the density of northern cod and a review of hypotheses for the stock’s collapse. Canadian Journal of Fisheries and Aquatic Sciences (1996) 53:943–962.
Hutchings J. A. Collapse and recovery of marine fishes. Nature (2000) 406:882–885.[CrossRef][Medline]
Hutchings J. A., Myers R. A. What can be learned from the collapse of a renewable resource? Atlantic cod, Gadus morhua, of Newfoundland and Labrador. Canadian Journal of Fisheries and Aquatic Sciences (1994) 51:2126–2146.
Jackson J. B. C., Kirby M. X., Berger W. H., Bjorndal K. A., Botsford L. W., Bourque B. J., Bradbury R. H., et al. Historical overfishing and the recent collapse of coastal ecosystems. Science (2001) 293:629–638.
Johannes R. E., Freeman M., Hamilton R. Ignore fishers’ knowledge and miss the boat. Fish and Fisheries (2000) 1:257–271.[CrossRef]
Kirby D. S. On the integrated study of tuna behaviour and spatial dynamics: tagging and modelling as complementary tools. In: Electronic Tagging and Tracking in Marine Fisheries—Sibert J., Nielson J., eds. (2001) Dordrecht: Kluwer Academic Publishers. 407–420.
Klimley A., Holloway C. School fidelity and homing synchronicity of yellowfin tuna, Thunnus albacares. Marine Biology (1999) 133:307–317.[CrossRef]
Kloser R., Davis T., Young J., Stanley C., Ryan T., Sakov P. Assessment of acoustics as a research tool to determine the distribution, biomass and behaviour of southern bluefin tuna schools and their prey in the Great Australian Bight (1998) Hobart: CSIRO. RMWS/98/8.
Kock H. The direct influence of fishing and fishery-related activities on non-target species in the Southern Ocean with particular emphasis on longline fishing and its impact on albatrosses and petrels—a review. Reviews in Fish Biology and Fisheries (2001) 11:31–56.[CrossRef][Web of Science]
Krause J., Hoare D. J., Croft D. P., Lawrence J., Ward A., Ruxton G. D., Godin J-G. J., et al. Fish shoal composition: mechanisms and constraints. Proceedings of the Royal Society of London (2000) 267:2011–2017.[Medline]
Leigh G. M., Hearn W. Changes in growth of juvenile southern bluefin tuna (Thunnus maccoyii): an analysis of length-frequency data from the Australian fishery. Marine and Freshwater Research (2000) 51:143–154.[CrossRef][Web of Science]
MacKenzie D. Fish out the politicians—it’s time for the experts to take control of Europe’s failing fish stocks. New Scientist (2002) 174:4.[Medline]
Majkowski J., Hearn W. S., Sandland R. L. A tag-release/recovery method for predicting the effect of changing the catch of one component of a fishery upon the remaining components. Canadian Journal of Fisheries and Aquatic Sciences (1988) 45:675–684.
Majkowski J., Murphy G. CSIRO southern bluefin tuna tagging: scientific achievements and future objectives. Australian Fisheries (1983) 42:26–27.
McQuinn I. H. Metapopulations and the Atlantic herring. Reviews in Fish Biology and Fisheries (1997) 7:297–329.[CrossRef][Web of Science]
Murphy G., Majkowski J. State of the southern bluefin tuna population: fully exploited. Australian Fisheries (1981) 40:20–29.
Noettestad L., Similae T. Killer whales attacking schooling fish: why force herring from deep water to the surface? Marine Mammal Science (2001) 17:343–352.[CrossRef][Web of Science]
Orensanz L. M. L., Armstrong J., Armstrong D., Hilborn R. Crustacean resources are vulnerable to serial depletion—the multifaceted decline of crab and shrimp fisheries in the Greater Gulf of Alaska. Reviews in Fish Biology and Fisheries (1998) 8:117–176.[CrossRef][Web of Science]
Parrish J. K. Using behaviour and ecology to exploit schooling fishes. Environmental Biology of Fishes (1999) 55:157–181.[CrossRef][Web of Science]
Pauly D., Christensen V., Dalsgaard J., Froese R., Torres F. J. Fishing down marine food webs. Science (1998) 279:860–863.
Pearce F. A high price to pay for caviar. New Scientist (2003) 2413:6–7.
Pennington M., Stromme T. Surveys as a research tool for managing dynamic stocks. Fisheries Research (1998) 37:97–106.[CrossRef][Web of Science]
Pitcher T. J. Behaviour of Teleost Fishes (1991) 2nd edn. London: Chapman and Hall.
Pitcher T. J. The impact of pelagic fish behavior on fisheries. Scientia Marina (1995) 59:295–306.[Web of Science]
Pitcher T. J., Parrish J. K. Functions of shoaling behaviour in teleosts. In: Behaviour of Teleost Fishes—Pitcher T. J., ed. (1993) London: Chapman and Hall. 263–439. 703 pp.
Polacheck T. Tuna longline catch rates in the Indian Ocean: did industrial fishing result in a 90% decline in the abundance of large predatory species? Marine Policy (2006) 30:470–482.[CrossRef][Web of Science]
Polacheck T., Eveson J. P., Laslett G. M. Increase in growth rates of southern bluefin tuna (Thunnus maccoyii) over four decades: 1960 to 2000. Canadian Journal of Fisheries and Aquatic Sciences (2004) 61:307–322.
Polacheck T., Preece A. An integrated statistical time series assessment of the southern bluefin tuna stock based on the catch at age data. Commission for the Conservation of Southern Bluefin Tuna (2001) CCSBT-SC/0108/13.
Polacheck T., Preece A., Ricard D. Assessment of the status of the southern bluefin tuna stock using virtual population analyses. CCSBT Scientific Meeting and Stock Assessment Group Meeting (2001) CCSBT/0108/0120.
Roberts C. M. Deep impact: the rising toll of fishing in the deep sea. Trends in Ecology and Evolution (2002) 17:242–245.[CrossRef]
Robins J., Tsutomu N. Considerations on the migration route of juvenile SBT based on the results of the tagging experiments conducted in waters off WA during 1960s–70s: a brief note to consider juvenile SBT movements around the acoustic survey area. In: Southern Bluefin Tuna Recruitment Monitoring and Tagging Program: Report of the 10th Workshop, Recruitment Monitoring Workshop 98/20 (1998) Hobart, Australia: CSIRO Marine Research.
Robinson C. J. Responses of northern anchovy to the dynamics of the pelagic environment; identification of fish behaviours that may leave the population under risk of overexploitation. Journal of Fish Biology (2004) 64:1072–1087.[CrossRef][Web of Science]
Roman J., Palumbi S. R. Whales before whaling in the North Atlantic. Science (2003) 301:508–510.
Seijo J., Caddy J. F. Uncertainty in bio-economic reference points and indicators of marine fisheries. Marine and Freshwater Research (2000) 51:477–483.[CrossRef][Web of Science]
Sharp G. D. Tuna oceanography—an applied science. In: Tuna: Physiology, Ecology and Evolution—Block B. A., Stevens E. D., eds. (2001) San Diego: Academic Press. 345–389.
Shepherd S. A., Rodda K. R. Sustainability demands vigilance: evidence for serial decline of the greenlip abalone fishery and a review of management. Journal of Shellfish Research (2001) 20:829–841.[Web of Science]
Sibert J., Hampton J., Kleiber P., Maunder M. N. Biomass, size and trophic status of top predators in the Pacific Ocean. Science (2006) 314:1773–1776.
Springer A., Estes J., Vilet G. V., Williams T., Doak D., Danner E., Forney K., et al. Sequential megafaunal collapse in the North Pacific Ocean: an ongoing legacy of industrial whaling? Proceedings of the National Academy of Sciences of the USA (2003) 100:12223–12228.
Ward P., Myers R. A. Shifts in open-ocean fish communities coinciding with the commencement of commercial fishing. Ecology (2005) 86:835–847.[CrossRef][Web of Science]
Wardle C. S. Fish behaviour and fishing gear. In: Behaviour of Teleost Fishes—Pitcher T. J., ed. (1993) London: Chapman and Hall. 609–643. 703 pp.
Young J., Lyne V. Ocean conditions affect Tasmanian tuna aggregations. Australian Fisheries (1993) 52:24–26.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||