Utility of gillnets for selectively targeting penaeids off Iran

Aref Hout, Seye Yousef Pighmbri,＊, Mortez Eighni, Mtt K. Brohurst,Shnnon M. Byse

a Fisheries Department, Gorgan University of Agricultural Science and Natural Resources, Gorgan, 517212, Iran

b New South Wales Department of Primary Industries, Fisheries Conservation Technology Unit, National Marine Science Centre, Southern Cross University, Coffs Harbour, NSW, 2450, Australia

c Marine and Estuarine Ecology Unit, School of Biological Sciences, University of Queensland, Brisbane, QLD, 4072, Australia

d Fisheries and Marine Institute, Memorial University of Newfoundland, St. John’s, NL, A1C 5R3, Canada

Keywords:Shrimp Gillnet Catch rate Bycatch Small-scale fisheries

ABSTRACT In response to concerns over the mortality of discards and other negative environmental impacts of otter trawling for penaeids, the utility of bottom-set gillnets deployed by artisanal vessels to target Penaeus monodon, Metapenaeus affinis and other retained species by-product was assessed in the Sea of Oman. Two gillnets (each comprising several 80-m panels for total lengths of up to 1.3 km) were diurnally fished at two locations (off Beris and Govater, eastern Iran) within established penaeid fishing grounds. During 30 days, the total catch was 2721 kg, comprising penaeids (981 kg), eight retained by-product species (598 kg and mostly Otolithes ruber, Cynoglossus bilinearus and Carangoides talamparoides) and 19 discarded species (1142 kg). The catch-per-unit-of-effort(CPUE, kg/gillnet-panel ⋅ hour-fished) of targeted penaeids and by-product varied significantly between areas(0.5–1.0 kg), but nevertheless reached economically viable levels. The CPUEs of total bycatch were similar at both sites (<0.5 kg) and no marine turtles were caught. The concomitant ratios of retained-catches-to-bycatch were considerably less than those for regional penaeid-trawl fisheries. Nevertheless, while gillnet selectivity was satisfactory for M. affinis, most of the P. monodon were smaller than their carapace length at maturity at both fished sites. The results support using gillnets to target penaeids and by-product species as an alternative to otter trawling in the Sea of Oman; albeit with ongoing assessments of modifications to maximise selectivity.

1. Introduction

Small-scale (artisanal) fisheries are globally important, supporting the employment of >50 million fishers and accounting for more than half of the world’s fisheries production (Chuenpagdee & Jentoft, 2018;Peckham et al., 2007; Teh & Sumalia, 2013). The largest subset of small-scale fisheries is in developing nations throughout tropic and temperate regions which typically are characterised by considerable biodiversity. Cumulatively, these fisheries face a plethora of sustainability issues that require attention, and especially among penaeid fisheries which contribute the most towards global discards and wastage(Jones et al., 2018; Zeller at al., 2018).

Like for other coastal countries throughout the middle east, smallscale penaeid fisheries are economically important for Iran, and especially in the Gulf/Sea of Oman where 30,000 people derive their livelihood directly from this sector, which comprises ～1500 small wooden vessels (<10 m) and 950 larger dhows (up to ～23 m). The key targets arePenaeus monodon(sought at carapace lengths; CLs >47 mm) andMetapenaeus affinis(>25 mm CL). These species are mostly caught,along with several economically important teleosts (e.g.Otolithes ruber,Cynoglossus bilinearusandCarangoides talamparoides) (by-product), by small single otter trawls (～26-m head ropes and made from polyethylene twine, with a 25-mm stretched mesh opening—SMO in the codend) deployed inshore. Currently, trawling is permitted during daylight only from mid-October to late December.

In addition to retained catches, Iranian trawlers catch and discard large numbers of unwanted species (bycatch), comprising various species, and mostly teleosts such as scombrids, carangids, soleids, and clupeids, but also marine turtles (especiallyChelonia mydas). Typically,the total retained penaeid-to-discard ratios in the otter-trawl fisheries exceed 1:4 (Eighani & Paighambari, 2013). The mortality of bycatches,and especially threatened megafauna such as marine turtles and juveniles of elasmobranch species caught elsewhere has raised concerns over sustainability (Eighani et al., 2016).

Numerous studies have identified various applied (and successful)technical solutions to reduce bycatch and/or its mortality from otter trawlers, ranging from modifying the operation of existing gears to more commonly fitting physical devices to the posterior trawl designed to improve selectivity (termed bycatch reduction devices or ‘BRDs’)(Broadhurst, 2000). A plethora of BRD designs and configurations are available—typically reflecting the unique characteristics of assessed fisheries, including their target, by-product and bycatch species(Broadhurst, 2000; Broadhurst et al., 2007). In addition to BRDs, more recently there has been an impetus to assess other modifications to otter trawls that holistically address the additional perceived environmental consequences due to poor fuel efficiencies and habitat impacts (McHugh et al., 2016).

While modifications to problematic trawls can be quite effective for reducing unwanted impacts, another option is to simply examine alternative fishing gears that because of their particular fishing mechanisms, are inherently more environmentally benign (Broadhurst et al.,2007). Depending on the species and their ecology, there exist various alternatives to otter trawling for harvesting penaeids, including cast nets, seines and stow and trap nets in estuaries (e.g. Gray et al., 2006),and traps and bottom-set gillnets (or trammel nets, and often drifting) in coastal or offshore areas (e.g. Thomas et al., 2003). In some cases, these non-trawl gears can be associated with improved selectivity, and for traps and gillnets, minimal benthic impacts and considerably lower operational and fuel costs (Broadhurst et al., 2007; Gray et al., 2006).Nevertheless, like for modifications to otter trawls, alternative methods for targeting penaeids need to be assessed on a fishery-specific basis across a range of conditions to determine not only their utility for minimising environmental impacts, but also their technical practicality and economic justification.

Considering the above theoretical framework supporting simply substituting problematic fishing gears for others that might be more environmentally benign, the objective of this study was to preliminarily assess the utility and viability of drifting bottom-set gillnets for targeting key penaeids (and their sizes) and by-product, while minimising unwanted bycatches by artisanal vessels fishing in the Sea of Oman. More specifically, our research aim was to quantify and compare catches of coherent gillnet configurations at two key locations within the fishery as a measure of consistency, and discuss these with reference to published catches from regional trawlers as a precursor to possible future fishery development and refinement.

2. Material and methods

The study was completed in October and November 2018 using two artisanal vessels (both 7 m long with 33 kW motors) each deploying a single bottom-set drift gillnet in the Sea of Oman, immediately adjacent to two ports in the Sistan and Baluchestan province: Beris (25◦13′N, 61◦18′E) and Govater (25◦02′N, 62◦19′E) (Fig. 1). The fishing grounds were ～5 km from the coast, in water <10-m deep and characterised by a combination of sand and mud substrate.

Identical gillnets were constructed following the general specifications (e.g. hanging ratio and material) of similar configurations used elsewhere to target penaeids (e.g. Thomas et al., 2003). Both gillnets comprised several panels (each 80 m long × 2.6 m stretched depth) of meshes (50 mm stretched mesh opening) that were made from multifilament polyamide twine (210 denier and 18 ply with a diameter–Ø of 0.7 mm) and hung on polypropylene floatlines and polyamide foot ropes at a ratio of 50% (Fig. 2). Each panel had 20 (120-mm Ø) floats and 20 small weights (total of 480g) clipped at the floatlines and foot ropes,respectively (Fig. 2). The configuration meant the entire gillnet made contact with the seabed, but could drift freely and at a height of <～1 m(Fig. 2). Depending on operational logistics (e.g. weather conditions or other regional fishing activity), between 10 and 17 panels were sequentially attached (i.e. sewn immediately adjacent to each other) to form each gillnet, which was deployed (between 7:00 and 13:00 each day) with one end fastened to a marker buoy and the other end attached to the vessel (Fig. 2).

Immediately following gillnet retrieval, all catches were sorted,identified to the species level and separately weighed (nearest 10 g).Subsamples of the targeted penaeids were measured for carapace length(CL to the nearest 1 mm). All catches were then defined into categories of targeted penaeids (retained), by-product (retained non-penaeid species) and bycatch (discarded species). These categories and the individual weights of species were standardised to catch-per-unit-of-effort(CPUE):

Fig. 1. Map of Iran, showing the Sistan and Baluchestan province with the ports of Govater and Beris, in the Sea of Oman.

Fig. 2. Diagram of the generic drifting bottom-set gillnet configuration and mesh size used to target penaeids in the Sea of Oman.T: meshes in the transverse direction; N: meshes in the normal direction; Ø: diameter.

SIMPER analysis (cut off 90%) was used to identify those taxa that contributed to overall dissimilarity between study areas (Clarke, 1993).Linear mixed models (LMM) were used to test the hypothesis of no differences in the CPUEs of the broad categories of catches (which were first assessed for normality using QQ plots) between locations, withsignificance determined by WaldFs. In these analyses, the ‘paired days’were considered random effects.

Table 1 List of species caught in drifting bottom-set gillnets at two locations in the Sea of Oman (off Beris and Govater) with the total weights (kg), percentage of the catches and mean (SE) CPUE.

Carapace-length data forP. monodonandM. affiniswere also analysed by comparing the proportion of the catch between locations, Beris and Govater. Generalised linear mixed models (GLMMs) were used to fit curves of the expected proportions (logit; Beris/(Beris +Govater)) of CL with a binomial error. Using the glmer function of the lme4 package(Bates et al., 2015), low-order polynomial (cubic, quadratic, linear or constant) models were fit by maximum likelihood of proportions at length retained at each location. The dependent variable was the proportion per CL class, the independent variable was ‘length’, and the random effect was ‘paired day’ on the intercept. The best model was selected based on the minimum AICc value using the function AICctab from the bblme package (Bolker, 2017). After selecting the best model,bootstrapping was applied via the bootMer function in the lme4 package with 1000 simulations deriving 95% confidence intervals (CIs). The results are interpreted by inspecting the model proportion and CIs. A proportion of 0.5 indicates no location-specific difference in catch for a length, while 0.75 indicates 75% of penaeids at a specific length were caught at Beris and 25% at Govater. Where CIs contain 0.0, there is no significant difference in size selection between locations. All statistical analyses were performed using R software (version 3.3.2, R Development Core Team, 2016).

3. Results

In total, the gillnets were each deployed for 15 days off Beris and Govater for a cumulative 170 fishing hours and caught 2721 kg comprising 29 species, including two species that were targeted penaieds (981 kg), eight others retained as by-product (598 kg) and 19 discarded as bycatch (1142 kg; Table 1). The key species caught off Beris wereP. monodon,O. ruberandM. affinis(52.9% of the total catch), while the same two penaeids along with Scyphozoa (76.4% of the total catch)dominated catches off Govater (Table 1). Pairwise comparisons(SIMPER) showed a significant difference in catch compositions between locations, with a dissimilarity average of 15.3% that was mostly attributed to Scyphozoa,Pterois russellii,Sepia pharaonis, Portunus pelagicus,andTrichiurus lepturus(P<0.05; Table 1).

Linear mixed models detected significant differences in the CPUEs of the total retained catch, targeted penaeids and by-product between locations, with 50%, 33% and 69% lower relative catches, respectively off Govater (P< 0.001, Table 2). However, there was no significant difference in the CPUE of total bycatch between locations (LMM,P>0.05,Table 1). Considering the mean CPUEs, the ratios of retained-todiscarded catches were 1:0.43 and 1:0.94 off Beris and Govater(Table 2).

There were inter-specific differences in the sizes of targeted penaeids retained at each location (Table 3). Generalised linear mixed models revealed that proportionally moreM. affinis<43 mm CL were captured at Govater, but larger sizes were similarly represented between locations(P< 0.01; Fig. 3a). ForP. monodon, proportionally more individuals were captured at sizes between 27 and 65 mm CL at Govater (GLMM,P< 0.01; Fig. 3b). In terms of sizes at maturity, onlyM. affiniswas satisfactorily retained in the gillnets at mostly mature sizes (Table 3). By comparison, >65% ofP. monodonretained in the gillnets at both locations were juveniles (Table 3; Fig. 3b).

Table 2 Summary of Wald Fs and predicted means (±SE) from linear mixed models assessing the independence of location (Beris vs Govater) on the CPUE of drifting bottom-set gillnets in the Sea of Oman.

Table 3 Carapace lengths (CL in mm) at maturity for M. affinis and P. monodon, their ranges of CLs, and the percentages of individuals caught below CLs at maturity at two locations (off Beris and Govater) in the Sea of Oman.

Fig. 3. Proportions of (a) Metapenaeus affinis and (b) Penaeus monodon retained at each mm CL (black, solid circles; Beris/(Beris +Govater)) whereby a value of 0.5 indicates an even split between locations for the specific length.

4. Discussion

The data here contribute to the few previous studies assessing the utility of drifting bottom-set gillnets for commercially targeting penaeids, and support a general trend of low retained catch-to-bycatch ratios (i.e. <～1:1) for these types of gears tested elsewhere (e.g.Akyol, 2008; Akyol & Ceyhan, 2009; Daliri et al., 2015; Thomas et al.,2003). Further, the ratios observed here were considerably lower than those recorded for regional penaeid trawlers (1:4 to 1:6; Paighambari &Daliri, 2012; Eighani & Paighambari, 2013), and the historical ranges noted by Andrew and Pepperell (1992) of between ～1:2 and >1:20 for other tropical penaeid-trawl fisheries. And, unlike for regional otter-trawl fisheries, no marine turtles were caught in the present study(Abdulqader et al., 2017).

In addition to the inherent improved species selectivity of penaeid gillnets compared to otter trawls, are their lower associated fuel costs and habitat impacts. However, for the Sea of Oman (like all fisheries),such benefits nevertheless need to be rationalised in terms of the total gillnet harvests and efficiencies, which can be discussed here by first considering their observed utility of bottom-set gillnets for the target and by-product species, and then possible modifications or developments to maximise selectivity.

The vessels used in this study deployed up to 17 panels for total gillnet lengths of 1.3 km which achieved variable CPUEs and total retained catches and sizes between locations. Bycatches were spatially consistent, but the most productive area for the targeted penaeids and by-product was off Beris, with a mean of ～0.5 kg retained for each panel fished for an hour. Extrapolating these CPUEs would provide an average of ～100 kg of retained catch per day, which is well within profitability expectations of fishers. However, the same calculations off Govater yielded only 50% of the retained catch (and included proportionally more of the smaller penaeids)—although this appeared to be at least partially attributable to the excessive amounts of Scyphozoa encountered.

The medusa of Scyphozoa were easily entangled in gillnet meshes,and rapidly reduced meshing efficiency. Scyphozoa have been noted to similarly affect the efficiencies of various gears in other fisheries (e.g.Conley & Sutherland, 2015; Kim et al., 2012; Nagata et al., 2009), and in fact precipitated installing BRDs in some penaeid trawls (e.g. Broadhurst& Kennelly, 1996). While there are no existing retroactive modifications that might similarly reduce Scyphoza among gillnets, it is well established that the selectivity of entangling gears can be controlled by altering several key technical factors, including the twine material and diameter, mesh size, hanging ratio and vertical orientation in the water column (Hamley, 1975). Of these factors, simply increasing the hanging ratio and/or reducing the gillnet height might have benefit in minimising Scyphoza entanglement and/or allow some to drift over the headline, respectively. Separating the gillnet panels might also allow some Scyphozoa to pass through and escape.

While changes to the other technical factors affecting gillnet selectivity (listed above) might not be expected to influence Scyphoza catches, some could also warrant assessment for improving overall gillnet catchability. Specifically, several previous studies have demonstrated that compared to multifilament gillnets (such as those tested here), those made from monofilament often are characterised by considerably greater CPUEs (e.g. up to 2 folds) for various species(Simasiku et al., 2017; Washington, 1973), including penaeids (e.g.Thomas et al., 2003). Such benefits often are attributed to lower visibility (Hamley, 1975). Any inherent improvements in catchability associated with monofilament might be justified to offset some of the efficiency reductions due to Scyphoza off Govater. But, the influence of these changes on retained and other discarded catches would require concomitant assessment.

Although possibly unlikely to affect Scyphoza catches, a change in mesh size might also warrant assessment for improving gillnet selectivity at certain times. Specifically, it was clear that although the chosen mesh size is appropriate for harvestingM. affinisat CLs mostly greater than that of maturity across both locations, many of theP. monodonwere juveniles and especially off Govater (where proportionally more of the smallerM. affiniswere also caught). Increasing the mesh size might reduce juvenileP. monodonfishing mortality, although anything beyond a subtle increase would likely have concomitant impacts on at least some of the by-product and/or bycatch. Of concern, larger mesh might increase the catchability of marine turtles, which are a problematic bycatch in many costal gillnet fisheries targeting teleosts throughout tropic/temperate regions (Gray & Kennelly, 2018). Considering the strong spatial influence on the catches ofP. monodon, area or time closures (pending sufficient data), might be more appropriate than technical modifications to gear.

5. Conclusions

This work here represents a definitive study assessing the commercial utility of bottom-set gillnets as a possible substitute for trawling in coastal waters off the Sea of Oman. Notwithstanding the clear need for additional research to refine gillnets for targeting penaeids and byproduct to maximise size and species selectivity and efficiency, the data support a conclusion that this fishing method is an alternative,more environmentally benign option to inshore otter trawling at the studied areas. Future research should establish not only optimal gillnet configurations, but also appropriate effort in terms of the number of operators and gillnet lengths to ensure profitability, and with absolute bycatches that remain lower than those from trawling. Management mechanisms already exist for spatio-temporal trawl closures in the Sea of Oman, and these could be easily applied to gillnetters.

CRediT authorship contribution statement

Aref Hout: Investigation. Seyed Yousef Paighambari: Project administration, Conceptualization. Morteza Eighani: Methodology,Writing - original draft. Matt K. Broadhurst: Writing - review & editing,Formal analysis, Validation. Shannon M. Bayse: Formal analysis.

Declaration of competing interest

The authors declare no conflict of interest.

Acknowledgments

Thanks are extended to The Gorgan University of Agricultural Science and Natural Resources for supporting the study, and to the participating fishers, without whom the data collection would not have been possible.