Elsevier

Fisheries Research

Volume 232, December 2020, 105735
Fisheries Research

Benefits of a restricted spatial and temporal survey design for determining average weight of recreational catches

https://doi.org/10.1016/j.fishres.2020.105735Get rights and content

Abstract

Surveys of recreational fishing are conducted in many jurisdictions to provide catch estimates, usually measured as number of individuals. Converting these catch estimates to weight is necessary for comparability with catch data from other sectors to support resource allocation and other management strategies. Weight data are rarely obtained during these surveys to minimise respondent burden and accommodate the logistical and resource constraints on the survey design. On-site Boat Ramp Surveys of boat-based recreational fishers for this purpose were implemented across four marine bioregions in Western Australia in 2011/12 using a randomised design (12-months) and in 2013/14 using a randomised design restricted to peak periods of fishing activity (4-months). Both surveys achieved high response rates (>97%) and, even though the 2013/14 survey was implemented for a shorter time period, it obtained a greater number of measurements (>10,000) than in 2011/12 (>5,000). Data were collected from >280 species (or family groupings), with more individuals measured in every bioregion in 2013/14 when compared to 2011/12. Average weights were calculated for 30 key species aligned with fisheries management, of which 75% of nearshore and 30% of demersal species had significant differences between fishing seasons and survey years. Reasons for such differences may be due to biological factors (i.e., migration), fisher behaviour (i.e., avidity) or management actions. Implementing recreational fishing surveys with the specific aim of obtaining data on average weight for this sector is rare. Average weights calculated from both designs can be used to provide current estimates of catch (by weight) for comparison with other sectors, thereby reducing reliance on secondary data sources drawn from different spatial areas, temporal periods or related species. However, the restricted survey design was more cost-effective (AUD16 vs AUD46 per individual measured) and obtained higher samples from a broader suite of species. This design is therefore recommended for the future collection of weight data, especially in locations that are remote from major population centres and have highly seasonal peaks in fishing activity.

Introduction

The collection of biological data for fish and other marine organisms are essential for many aspects of fisheries management including resource allocation, stock assessment, monitoring and evaluation of fish condition (including the development of indices), geographic comparisons of life histories and morphometrics (Ghosn et al., 2012; Goncalves et al., 1997; Lai and Helser, 2004; Mendes et al., 2004; Petrakis and Stergiou, 1995). Measurements of length (which may include total length, fork length, carapace length and shell width) and weight (which may include whole weight, gutted weight and wet weight) are biological metrics that make a valuable contribution to these aforementioned aspects of fisheries management, especially resource allocation, by enabling calculation of average weights for species of interest, either via conversion of length measurements from known length-weight equations or from direct measurement of weight. These metrics are also less demanding to collect accurately in terms of cost and time, and in conjunction with on-site boat ramp interviews, when compared to more intensive techniques required to determine age (otolith samples), or fecundity and maturity (gonad samples) of marine organisms.

Globally, in the commercial sector, it is often a licence requirement that fish are weighed on landing and fisheries observer programs may also provide an additional avenue for the collection of biological data (Hartill and Davey, 2015; Marriott et al., 2012; Stanley et al., 2009). Similarly, charter operators within the recreational sector in some countries also have a mandatory licence requirement to submit logbooks containing data on fish length and/or weight as well as information on catch and effort (Chan et al., 2018; Gray and Kennelly, 2017; Hartill et al., 2012). Conversely, recreational fisheries are predominately open access with no mandatory reporting requirements and are undertaken in diverse environments with multiple, diffuse access points. Establishing a survey design which allows representative samples of length and weight to be collected can therefore be challenging, and requires significant resources and funding (Morato et al., 2001).

Depending on the objective of survey programs, biological data (including weight) are also collected during other fishery dependent and independent surveys, which may vary from one-off studies to long-term monitoring using a variety of techniques. Underwater visual census is also commonly used to estimate fish abundance and length, although it is difficult to obtain weight using this technique (Morey et al., 2003). However, depending on the spatial and temporal scales of such research, the average weight of species generated may not always be reliably applied to broader spatial areas or species distributions. For example, research may be focused on a small spatial area (i.e., river or estuary) or a small number of target species for small-scale fisheries (Brouwer and Buxton, 2002; Gray and Kennelly, 2003; Marriott et al., 2012; West and Gordon, 1994).

Length and weight data for recreational fisheries can be collected and verified by trained interviewers during on-site (i.e., boat ramp) surveys, which are generally utilised when a fishery is open access and are conducted over smaller spatial scales (Pollock et al., 1994). This differs from off-site surveys (i.e., phone/diary, mail) which are generally utilised for fisheries operating over broad-spatial scales or have a licence sampling frame and are designed to minimise potential errors, such as recall bias via regular telephone contact or use of a diary card (Henry and Lyle, 2003; Hyder et al., 2018; Ryan et al., 2016). Off-site surveys rely on self-reporting by recreational fishers and potential errors with reporting of length and weights (i.e., interpretation of different measurement instruments and rounding bias) (Bellanger and Levrel, 2017; Fairclough et al., 2014; Sparrevohn, 2013) can be mitigated by providing respondents with training material and measuring tools (i.e., calipers) (Bradford, 2000; Lyle et al., 2019). Volunteer programs, during which recreational fishers report their fishing activity in a logbook, or donate fish skeletons which are then utilised by researchers, can also be useful sources of biological data (Fairclough et al., 2014; Rocklin et al., 2014; Smith et al., 2007). Ensuring self-reported length and weight data is representative of the broader population is difficult due to the specific characteristics and behaviours of each fisher, although this effect can be minimised via randomisation and stratification.

The practicalities of collecting biological data from recreational fishers can also be challenging, especially when considering the additional burden on respondents (Georgeson et al., 2015; Hartill and Davey, 2015; Henry and Lyle, 2003). These include: time limitations (i.e., fishers are eager to depart a boat ramp); personal preference (i.e., fishers reluctant to have fish removed from storage) (Morales-Nin et al., 2005); difficulty in setting up scales in exposed environments (i.e., need to be on a level surface for accurate reading); and transporting equipment to remote locations to access fishers (i.e., scales may be bulky and sensitive to knocks). Recreational fishers may also be suspicious that length information will be used to report non-compliance with slot limits. For these reasons, as well as others relating to the practicality of weighing large fish which are often released, collection of biological data has been excluded from many surveys, or only estimated based on a visual inspection of the catch (Ghosn et al., 2012; Herfaut et al., 2013; Rangel and Erzini, 2007).

If average weight data are not available for a particular species or time period, this information may instead be acquired from secondary sources or imputed from random length samples. These data may be for the same species but from different stock distributions, sectors (i.e., commercial or charter), jurisdictions or an earlier sampling period (Hamer et al., 2019; Lloret et al., 2008; Ma and Ogawa, 2016; Radford et al., 2018; Strehlow et al., 2012; Tracey et al., 2020, 2013; Veiga et al., 2010). Although length data are generally easier to collect (i.e. a measuring tape or mat is more transportable), obtaining such measurements may only be secondary after asking questions relating to primary objectives of fishing effort and total catch. Furthermore, length-weight relationships are still then required to impute weight from measured length, and are often only known for the most common and frequently studied species (Torres et al., 2012). Such approaches may produce estimates of average weight that are not representative (Henry and Lyle, 2003; Lyle et al., 2009; Ryan et al., 2016).

Exploring differences in on-site survey designs may provide opportunities to improve the collection of weight data from recreational fishers. Randomised survey designs use probability-based sampling regimes to collect data across broad spatial and temporal strata, which are often defined by periods of similar activity levels to improve the variance of estimates when expanding to a population (Pollock et al., 1994). It may be difficult to obtain measurements when sampling during periods of low activity (i.e., winter months), due to low survey effort and few interviews being completed. Conversely, a probability-based randomised survey which is restricted to spatial and temporal periods of peak fishing activity (herein referred to as a restricted survey) are more likely to obtain greater numbers of interviews (MRIP Design and Analysis Workgroup, 2012); although this may influence the representativeness of data, as the biological and behavioural characteristics of fish and fishers may change temporally (Al Nahdi et al., 2016; Hartill and Davey, 2015). Considering the differences in survey designs, and the implications this may have on data, is important to ensure comparability between surveys, and over the long-term.

This paper investigates randomised and restricted survey designs and their ability to collect length and weight data from finfish and invertebrates caught by boat-based recreational fishers. These data are essential for converting estimates of recreational catch (by numbers) to catch (by weight) and can also be developed into an index of recreational catch to monitor change over time (similar to catch rate indices). Implementing recreational fishing surveys with the specific aim of obtaining data on average weight for this sector is rare, as is the use of a restricted spatial and temporal survey design. The two designs are evaluated in terms of survey effort, species composition and cost-effectiveness of the survey designs as well as average weight and relative precision of estimates for individual species. The findings from this study will assist with the development of an index of average weight, provide more timely and accurate catch estimates (by weight) for recreational fisheries and improve the efficiency of survey designs for the collection of weight data from the recreational sector. This, in turn, will benefit fisheries management policies, particularly those that allocate resources across sectors.

Section snippets

Study Area

In Western Australia, the coastline extends 12,889 km and recreational fishing is widespread across four broad marine bioregions from the tropical north to the temperate south (Fig. 1). These waters contain approximately 3,000 fish species (Hutchins, 2001), of which ∼10% are harvested by recreational fishers. However, a small number of these species contribute a large proportion of the recreational catch. For example, Glaucosoma hebraicum, Chrysophrys auratus, Choerodon rubescens and

Survey Effort

The restricted survey design in 2013/14 had the desired effect of providing greater survey hours in the North Coast, Gascoyne Coast and South Coast when compared to 2011/12 (Table 2). Statewide, 41% of survey hours in 2011/12 were undertaken in the 4-month peak fishing season corresponding to the 2013/14 survey, with the North Coast having the highest proportion (50%).

As expected from the increased number of survey hours (which is the amount of time which was spent by the interviewer at each

Discussion

Worldwide, robust estimates of recreational catch (by weight) are essential for facilitating the equitable allocation of catch between sectors and are an important element of sustainable fisheries management (Crowe et al., 2013; Hyder et al., 2018; Ryan et al., 2016). The dispersed nature of recreational fisheries, which are often characterised by numerous access points, means that collecting weight measurements from this sector is challenging. However, without such information it is not

Conclusion

The randomised and restricted Boat Ramp Surveys were both successful in achieving their objective of obtaining biological data on length and weight for key species caught by boat-based recreational fishers. Although the restricted survey sampled only peak activity periods, this design performed better when evaluated with respect to survey effort, species composition, cost-effectiveness and relative precision of average weights, especially for demersal species. These benefits were readily

CRediT authorship contribution statement

Claire B. Smallwood: Methodology, Project administration, Formal analysis, Writing - original draft. Karina L. Ryan: Conceptualization, Methodology, Project administration, Writing - review & editing.

Declaration of Competing Interest

The authors report no declarations of interest.

Acknowledgements

This study would not be possible without contributions from the many recreational fishers who took the time to participate in the boat ramps interviews throughout Western Australia. Thanks also to the Research Survey Officers and Technical Officers who conducted all the fieldwork. Stuart Blight, Vangie Gerginis, Tim Leary, Elena Sulin, Adrian Thomson, Veronique Vanderklift, Karen Williams and Brent Wise from the Department of Primary Industries and Regional Development (formerly Department of

References (89)

  • T. Morato et al.

    Length–weight relationships for 21 coastal fish species of the Azores, north-eastern Atlantic

    Fish. Res.

    (2001)
  • G. Morey et al.

    Weight–length relationships of littoral to lower slope fishes from the western Mediterranean

    Fish. Res.

    (2003)
  • S.J. Newman et al.

    A risk assessment and prioritisation approach to the selection of indicator species for the assessment of multi-species, multi-gear, multi-sector fishery resources

    Mar. Policy

    (2018)
  • G. Petrakis et al.

    Weight-length relationships for 33 fish species in Greek waters

    Fish. Res.

    (1995)
  • N. Pleizier et al.

    Variation in whole-, landed- and trimmed-carcass and fin-weight ratios for various sharks captured on demersal set-lines off eastern Australia

    Fish. Res.

    (2015)
  • M.A. Torres et al.

    Length–weight relationships of 76 fish species from the Gulf of Cadiz (SW Spain)

    Fish. Res.

    (2012)
  • M.T. Zischke et al.

    Time-location sampling with capture-recapture to assess specialised recreational fisheries

    Fish. Res.

    (2014)
  • ABS

    3218.0 - Regional Population Grown, Australia 2016/17 [WWW Document]. Aust. Bur. Stat.

    (2018)
  • A. Al Nahdi et al.

    Spatio-Temporal Variation in Length-Weight Relationships and Condition of the Ribbonfish Trichiurus lepturus (Linnaeus, 1758): Implications for Fisheries Management

    PLoS One

    (2016)
  • D. Bearham et al.

    Molecular evidence of three species in the Pseudocaranx dentex complex (Carangidae) in Australian waters

    Mar. Freshw. Res. Freshw. Res.

    (2019)
  • E. Bradford

    Sample sizes needed for reliable estimates in marine recreational fishing surveys, New Zealand Fisheries Assessment Report 2000/36

    (2000)
  • S.L. Brouwer et al.

    Catch and effort of the shore and skiboat linefisheries along the South African Eastern Cape Coast

    South African J. Mar. Sci.

    (2002)
  • J.W. Brownscombe et al.

    Canadian recreational fisheries: 35 years of social, biological and economic dynamics from a national survey

    Fisheries

    (2014)
  • N. Caputi et al.

    The marine heat wave off Western Australia during the summer of 2010/11–12 years on Management implications of climate change effects on fisheries in WA: an example of an extreme event. Fisheries Research Report No. 250

    (2014)
  • W.G. Cochran

    Sampling Techniques

    (1977)
  • F.M. Crowe et al.

    Development and implementation of allocation arrangements for recreational and commercial fishing sectors in Western Australia

    Fish. Manag. Ecol.

    (2013)
  • Department of Fish and Wildlife

    California Recreational Fisheries Survey Methods

    (2017)
  • DPIRD

    Department of Primary Industries and Regional Development Annual Report 2018

    (2018)
  • C. Edwards et al.

    Updated estimates of Southern Bluefin Tuna Catch by CCSBT Non-Members states

    (2016)
  • D.V. Fairclough et al.

    Breathing life into fisheries stock assessments with citizen science

    Sci. Rep.

    (2014)
  • K. Ferter et al.

    Angler behaviour and implications for management - catch-and-release among marine angling tourists in Norway

    Fish. Manag. Ecol.

    (2013)
  • W.J. Fletcher et al.

    Status reports of the fisheries and aquatic resources of Western Australia 2015/16: State of the Fisheries

    (2017)
  • R. Froese et al.

    Editorial note on weight-length relations of fishes

    Acta Ichthyol. Piscat.

    (2011)
  • L. Georgeson et al.

    A framework for regular national recreational fishing surveys

    (2015)
  • D. Ghosn et al.

    Catch per unit effort and size composition of striped marlin caught by recreational fisheries in southeast Australian waters

    (2012)
  • C.A. Gray et al.

    Recreational charter fishery attributes and variation in key species catches and discards: Resource management considerations

    Fish. Manag. Ecol.

    (2017)
  • C.P. Green et al.

    Quantifying recreational catch and effort from the Corio Bay region

    (2015)
  • P. Hamer et al.

    Developing tools to inform management risk and improve recreational fishery monitoring for a complex multi-sector, multi-jurisdiction fishery: the ‘Western Victorian Snapper Stock,’ Final Report, FRDC Project 2013/201

    (2019)
  • B.W. Hartill et al.

    Estimation of mean fish weights from the recreational catch landed at boat ramps in 1996, NIWA Technical Report 31

    (1998)
  • B.W. Hartill et al.

    Scale-and context-dependent selection of recreational harvest estimation methods: the Australasian experience

    North Am. J. Fish. Manag.

    (2012)
  • B.W. Hartill et al.

    Mean weight estimates for recreational fisheries in 2011-12, New Zealand Fisheries Assessment Report 2015/25

    (2015)
  • G.W. Henry et al.

    The National Recreational and Indigenous Fishing Survey, Final Report for FRDC Project No. 99/158

    (2003)
  • S.H. Hurlbert

    Pseudoreplication and the design of ecological field experiments

    Ecol. Monogr.

    (1984)
  • J.B. Hutchins

    Checklist of the fishes of Western Australia. Rec. West. Aust. Museum Suppl. No. 63

    (2001)
  • Cited by (4)

    • Improving design-based estimates of biological data collected from a restricted spatio-temporal access point survey of recreational fishers

      2022, Fisheries Research
      Citation Excerpt :

      Even so, a probability-based survey design that utilizes all strata within the available sampling frame may not provide sufficient data within the constraints imposed by budgets, the availability of resources, or characteristics (e.g., remote/small-scale/unlicenced) of the fishery (Griffiths et al., 2013; Smallwood and Ryan, 2020; Taylor et al., 2021). To address these constraints, survey designs used to collect data from recreational fishers may use a restricted spatio-temporal sampling frame (Smallwood and Ryan, 2020), with an unequal probability of selection and non-independent samples. Within this context of varied survey design and data collection methods, appropriate stratification, cluster sampling (Nelson, 2014), pseudoreplication (Hurlbert, 1984; Millar and Anderson, 2004) and sampling biases (Pollock et al., 1994) need to be considered, so that statistical analyses and subsequent estimates are not confounded (Thomas and Heck, 2001).

    • Increasing confidence in estimates of average weight and recreational harvest ranges

      2022, Fisheries Research
      Citation Excerpt :

      The lengths and weights of fish in the retained catch were measured (Smallwood et al., 2017). Around 55% of fishers in the WCB interviewed agreed to have their completed catch weighed and measured (Smallwood and Ryan, 2020). The first survey in 2011/12 was a randomised design of key boat ramps throughout Western Australia over 12 months across a large number of boat ramps.

    View full text