Reviewing hook degradation to promote ejection after ingestion by marine fish Shane P. McGrath A , Paul A. Butcher B , Matt K. Broadhurst B,D and Stuart C. Cairns C A School of Environmental Science and Management, Southern Cross University, National Marine Science Centre, PO Box 4321, Coffs Harbour, NSW 2450, Australia. B NSW Department of Primary Industries, Fisheries Conservation Technology Unit, National Marine Science Centre, PO Box 4321, Coffs Harbour, NSW 2450, Australia. C Department of Zoology, School of Environmental and Rural Sciences, University of New England, Armidale, NSW 2350, Australia. D Corresponding author. Email: [email protected]Abstract. A widely recommended strategy for releasing fish that have ingested hooks is to simply cut the line. The utility of this approach is based on the premise that the individual will eventually eject the hook following sufficient oxidation. However, few quantitative data are available describing the mechanisms affecting hook decay. We addressed this issue by testing the independence of various technical factors on the degradation of 828 hooks comprising 23 designs (absolute sizes 227–611 mm 2 ) after protracted submersion in seawater. Twelve replicates of each hook were destructively assessed for compression and tensile strengths (using a force gauge) and 24 replicates were weighed, photographed and submersed in seawater. After submersion for 8 and 28 days, 12 replicate hooks were removed, re-photographed, re-weighed and tested for compression and tensile strengths to provide indices of decay. Hook degradation was mainly affected by the wire material and diameter and could be significantly promoted by choosing carbon steel designs, either with a wire diameter of #0.9 mm for the examined sizes or, alternatively, bait-holder barbs (or similar modifications) along the shaft. By rapidly oxidising and weakening after ingestion, such designs could ultimately help to reduce negative impacts of hooks on released fish. Additional keywords: angling, carbon steel, catch-and-release, fishing hooks, oxidation. Manuscript received 7 April 2011, accepted 27 July 2011 Published online 4 October 2011 Introduction Globally, recreational fishing (and especially angling) is popu- lar, with participation rates estimated at .40% in some devel- oped countries (Cooke and Cowx 2004). Most anglers release some of their catch, either voluntarily or in response to legal sizes and personal quotas (Arlinghaus et al. 2007). Such ‘catch- and-release’ fishing is widely promoted as being socially and environmentally responsible, although, to ultimately satisfy this definition, there should be few associated mortalities or negative welfare impacts among survivors (Arlinghaus et al. 2007). Recognition of the need to validate these assumptions has resulted in numerous relevant studies; many of which have at- tributed fatalities to consistently recurring factors, particularly anatomical hook location (for reviews see Bartholomew and Bohnsack 2005; Arlinghaus et al. 2007). Specifically, those fish that ingest hooks usually have a much greater probability of dying than those that are hooked in the mouth (Bartholomew and Bohnsack 2005; Arlinghaus et al. 2007). Currently, the available management options for mitigating mortalities caused by hook ingestion are limited to: (i) spatial or temporal closures to fishing (Bohnsack et al. 2004) or main- taining fishing and promoting either (ii) shallow mouth hooking (Cooke and Suski 2004) or (iii) hook ejection after ingestion (Aalbers et al. 2004; Broadhurst et al. 2007). Of these strategies, closures to fishing are the most extreme, but may be warranted where there is considerable risk to a particular species (Powles et al. 2000). The second and third approaches are less contro- versial and consequently have received the most attention (Willis and Millar 2001; Broadhurst et al. 2007). Considerable work has been done to investigate ways that hook ingestion can be minimised (i.e. (ii) above). Strategies range from simply changing the type of bait (Pauley and Thomas 1993) and the method of fishing (Grixti et al. 2007), to more complex terminal rig modifications, including modified J-hooks (Willis and Millar 2001; Butcher et al. 2008) or specific designs such as circle hooks (Cooke and Suski 2004). In some cases, using such hooks can dramatically reduce ingestion, although many have lower hooking efficiencies than conventional con- figurations, which often restricts their acceptance by anglers (Butcher et al. 2008). CSIRO PUBLISHING Marine and Freshwater Research www.publish.csiro.au/journals/mfr Ó CSIRO 2011 http://dx.doi.org/10.1071/MF11082
Transcript
Reviewing hook degradation to promote ejection afteringestion by marine fish
Shane P. McGrathA, Paul A. ButcherB, Matt K. BroadhurstB,D
and Stuart C. CairnsC
ASchool of Environmental Science andManagement, Southern Cross University, National Marine
Science Centre, PO Box 4321, Coffs Harbour, NSW 2450, Australia.BNSW Department of Primary Industries, Fisheries Conservation Technology Unit, National
Marine Science Centre, PO Box 4321, Coffs Harbour, NSW 2450, Australia.CDepartment of Zoology, School of Environmental and Rural Sciences, University
of New England, Armidale, NSW 2350, Australia.DCorresponding author. Email: [email protected]
Abstract. Awidely recommended strategy for releasing fish that have ingested hooks is to simply cut the line. The utilityof this approach is based on the premise that the individual will eventually eject the hook following sufficient oxidation.However, few quantitative data are available describing the mechanisms affecting hook decay.We addressed this issue by
testing the independence of various technical factors on the degradation of 828 hooks comprising 23 designs (absolutesizes 227–611mm2) after protracted submersion in seawater. Twelve replicates of each hook were destructively assessedfor compression and tensile strengths (using a force gauge) and 24 replicates were weighed, photographed and submersed
in seawater. After submersion for 8 and 28 days, 12 replicate hookswere removed, re-photographed, re-weighed and testedfor compression and tensile strengths to provide indices of decay. Hook degradation was mainly affected by the wirematerial and diameter and could be significantly promoted by choosing carbon steel designs, either with a wire diameter of
#0.9mm for the examined sizes or, alternatively, bait-holder barbs (or similar modifications) along the shaft. By rapidlyoxidising and weakening after ingestion, such designs could ultimately help to reduce negative impacts of hooks onreleased fish.
Manuscript received 7 April 2011, accepted 27 July 2011
Published online 4 October 2011
Introduction
Globally, recreational fishing (and especially angling) is popu-
lar, with participation rates estimated at .40% in some devel-oped countries (Cooke and Cowx 2004). Most anglers releasesome of their catch, either voluntarily or in response to legalsizes and personal quotas (Arlinghaus et al. 2007). Such ‘catch-
and-release’ fishing is widely promoted as being socially andenvironmentally responsible, although, to ultimately satisfy thisdefinition, there should be few associatedmortalities or negative
welfare impacts among survivors (Arlinghaus et al. 2007).Recognition of the need to validate these assumptions hasresulted in numerous relevant studies; many of which have at-
tributed fatalities to consistently recurring factors, particularlyanatomical hook location (for reviews see Bartholomew andBohnsack 2005; Arlinghaus et al. 2007). Specifically, those fishthat ingest hooks usually have a much greater probability of
dying than those that are hooked in themouth (Bartholomew andBohnsack 2005; Arlinghaus et al. 2007).
Currently, the available management options for mitigating
mortalities caused by hook ingestion are limited to: (i) spatial or
temporal closures to fishing (Bohnsack et al. 2004) or main-taining fishing and promoting either (ii) shallowmouth hooking
(Cooke and Suski 2004) or (iii) hook ejection after ingestion(Aalbers et al. 2004; Broadhurst et al. 2007). Of these strategies,closures to fishing are the most extreme, but may be warrantedwhere there is considerable risk to a particular species (Powles
et al. 2000). The second and third approaches are less contro-versial and consequently have received the most attention(Willis and Millar 2001; Broadhurst et al. 2007).
Considerable work has been done to investigate ways thathook ingestion can be minimised (i.e. (ii) above). Strategiesrange from simply changing the type of bait (Pauley andThomas
1993) and the method of fishing (Grixti et al. 2007), to morecomplex terminal rigmodifications, includingmodified J-hooks(Willis andMillar 2001; Butcher et al. 2008) or specific designssuch as circle hooks (Cooke and Suski 2004). In some cases,
using such hooks can dramatically reduce ingestion, althoughmany have lower hooking efficiencies than conventional con-figurations, which often restricts their acceptance by anglers
(Butcher et al. 2008).
CSIRO PUBLISHING
Marine and Freshwater Research www.publish.csiro.au/journals/mfr
A more appropriate strategy to mitigate the mortality assoc-iated with anatomical hook location in some fish might be to
promote the use of hooks that are rapidly ejected (i.e. (iii)above). The concept for such an approach is supported byseveral studies which have demonstrated that, compared with
removing ingested hooks, releasing fish with their lines cutresulted in a lower mortality rate, with most individuals subse-quently ejecting their hooks (reviewed by Hall et al. 2009).
However, very little information is available on the associatedmechanisms and whether or not ejection can be promotedthrough the use of particular designs and materials. The consen-sus from the few available studies is that hook ejection is
influenced by hook decay, especially at the point and barb, bothof which can cause extensive internal injuries to fish (Aalberset al. 2004; Broadhurst et al. 2007). For example, Broadhurst
et al. (2007) observed that over 105 days, nickel-plated carbon-steel hooks that were ejected by yellowfin bream (Acanthopa-grus australis) had, on average, oxidised 4% more than those
that were not ejected. Further, many of the ejected hooks haddecayed sufficiently to break apart, often at or near bait-holderbarbs located on the shaft, which may have facilitated theirejection, and most had blunt points.
Intuitively, the rate of hook decay is likely to be affected byseveral technical factors, including the hook type, diameter andwire length. In any study that seeks to assess the full range of
such potential explanatory variables, it is important that there issufficient replication. One problem, however, is that there aresevere logistical and ethical issues associated with forcefully
causing the required number of fish to ingest hooks. Ideally,appropriate indices of decay might serve as proxies, beforesmaller scale trials with fish to test hypothesised explanatory
variables. Because Broadhurst et al. (2007) demonstrated nosignificant differences between the rates of oxidation (measuredas weight loss and breakage) for hooks that were ingested byyellowfin bream and those submersed in seawater, the latter
could be used to acquire relevant data on hook decay. Althoughsuch an approach does not replicate all of the forces applied to aningested hook, measuring the tensile and compression strengths
of the hook would at least facilitate a comparison amongdifferent designs (e.g. Edappazham et al. 2008) and providedirection for more refined studies with fish.
Given the above, the principal aim of this studywas to test thehypothesis that various technical characteristics (and theirinteractions), including weight, shape, design and wire con-struction would affect the temporal oxidation and integrity
of a range of hooks (absolute sizes – defined by Ralston 1982;– between 227 and 611mm2) commonly used to target small(,40 cm total length (TL)) coastal teleosts. A second aimwas to
use this information to provide directions for the refinement anddesign of hooks that might more readily decay and be ejectedby fish.
Materials and methods
Hook types
Thirty-six replicates of 20 different hooks commonly sold inAustralia and overseas were purchased for the experiment(Table 1). All hooks were measured for their wire diameter (ø)
and shaft, front, gape and bend lengths to the nearest 0.01mm,
weighed (to the nearest 0.0001 g) and had their steel type,coating, presence or absence of bait-holder barbs (for details see
Broadhurst et al. 2007) and design (i.e. J- or circle) recorded(Fig. 1; Table 1). Absolute hook size (mm2) was calculated asthe product of the shaft and bend lengths (Ralston 1982). The
hooks were classified as either J- or circle designs, based onwhether the point was approximately perpendicular or parallelto the shaft (Fig. 1a, b; Cooke and Suski 2004).
Three of the J-hook types had virtually identical sizes andshapes (Table 1; hooks 1–3), butwere constructed from differentmaterials: stainless steel, red-lacquer carbon steel and nickel-plated carbon steel. Based on their design uniformity, these three
hook types were chosen for a more detailed assessment of theutility of modifications to increase degradation among differentmaterials. Using a rotary tool (24-mm disc), 36 of the hooks
(termed ‘modified’ – hooks 1a, 2a and 3a; Table 1) made fromeach of the three materials had three small notches cut into theshaft, bend and point (to ,80% of the wire diameter, Fig. 1a).
The notches were designed to increase the surface area andsubsequent oxidation of the hooks and, therefore reduce theirstrength (Broadhurst et al. 2007). The remaining 36 hooks madefrom each material were left unmodified (Table 1). The modi-
fied hooks were also weighed after being altered and were onlyused if they comprised .98% of their original mass.
Hook assessment
On the first day (T0) of the experiment, 12 replicates of all
23 hook types (including the threemodified designs) were testedseparately for their compression (n¼ 6) and tensile strengths(n¼ 6) using a ChatillonDFX-100 (Brooklyn, NY,USA) digital
force gauge attached to an adjustable Chatillon LTCM-100-EUmotorised tester (see Fig. S1a available as an AccessoryPublication to this paper). The force gauge was rated to500N (accurate to� 2.5N) and recorded the strength at the
point of maximum hook elasticity or separation within a spec-ified sensitivity of� 0.1N. The speed of the tester was set at100mm min�1 (with a speed accuracy of� 15mmmin�1).
Compression was determined by positioning each hook betweentwo purpose-built stainless steel cylinders (40-mm ø� 50-mmlength) that were fitted with grooves in one of their ends to
secure the bend and eye and the other ends were screwed (5-mmthread) to the base plate and force gauge, respectively(see Fig. S1b). The tensile strength was tested by attaching theeye of the hook to a clip (491-N strength rated) secured in a small
vice screwed to the base plate and the bend of the hook to astainless steel (5-mm ø) hook screwed into the force gauge(see Fig. S1c).
The remaining 24 replicates of each hook design werephotographed using a stereoscopic microscope (Leica, S6D,Heerbrugg, Switzerland) fitted with a micron scale before being
individually placed into a 70mL perforated cylindrical plasticcontainer following the methodology of Broadhurst et al.(2007). The images were then used to measure the length of
each hook point to the nearest 0.01mm. The containers witheach hook were evenly distributed among three 3000L tanks(i.e. eight replicates of each hook type per tank). Each tank wassupplied with seawater (,188C) at a rate of 30Lmin�1. After
8 days (T8), half of the hooks were removed from the tanks,
B Marine and Freshwater Research S. P. McGrath et al.
Reviewing hook degradation Marine and Freshwater Research C
cleaned of any oxidised metal (using a paper towel) andre-weighed to determine their proportional oxidation (expressed
as the percentage of weight remaining). Photographs were takento determine the point length remaining after oxidation for eachhook. The difference between the oxidised and original pointlength was then converted to the proportion remaining. Half of
the hooks were tested for their maximum tensile strength,
whereas the remaining replicates were tested for maximumcompression strength (as above). At the end of 28 days (T28),
the remaining hooks in the tanks were sampled as above.
Statistical analyses
The null hypotheses of no effects of the various technical
parameters (and their appropriate interactions) on the observed
Gape
(a ) (b )
Gape
Shaft Shaft
1
3
2
Bend Bend
FrontFront
PointPoint
Fig. 1. The measurements recorded from all (a) J- and (b) circle hooks and the location of the
three small notches (1–3) cut into the shaft, bend and point of themodified J-hooks (numbers 1a,
2a and 3a in Table 1).
Table 2. Summary of main effects and interactions in parsimonious multiple regression models for which there were significant effects on
either percentage total hook weight and point remaining or tensile and compression strengths (N, newtons of force) following submersion in seawater
for 8 (T8) and then 28 (T28) days
***, P, 0.001; **, P, 0.01; Na, not included in model (not considered in the model-fitting process), Ns, not significant at P. 0.01; ø, diameter; T0, before
submersion
% weight remaining % point remaining Tensile strength Compression strength
Intercept *** *** ** Ns
Sample time 1 (T8 vs T0) – – *** Ns
Sample time 2 (T28 vs T0) – – ** **
Sample time 3 (T8 vs T28) *** *** – –
Steel type (stainless vs carbon) *** *** *** Na
Bait-holder barbs (presence vs absence) ** – ** ***
Modification (presence vs absence) ** *** ** **
Hook design (circle vs J) *** *** Na ***
Wire ø (mm) *** Na ** **
Shaft (mm) *** – Ns **
Front (mm) *** – Na ***
Bend (mm) *** – Na Na
Gape (mm) *** – Na Na
Wire ø� sample time 1 – – Na Na
Wire ø� sample time 2 – – *** Na
Wire ø� steel type *** Na Na Ns
Wire ø� bait-holder barbs Na – ** ***
Wire ø�modification ** *** *** Na
Wire ø� hook design *** *** Na ***
Wire ø� shaft *** – Na **
Shaft� sample time 2 – – Na Ns
Shaft� bait-holder barbs ** – Ns Na
Shaft�modification Na – Ns **
D Marine and Freshwater Research S. P. McGrath et al.
variability among each of the four independent response vari-ables (the percentages of weight and point remaining and tensile
and compression strengths) were assessed using multiple linearregression analysis. Categorical parameters (sample time, steeltype and the presence or absence of modifications and bait-
holder barbs) were incorporated into the models as ‘dummy’variables. Of the dummy variables, only sample time had morethan two levels (i.e. T8 and T28 referenced against T0) when used
to assess compression and tensile strengths. All percentage datawere arcsine–square root-transformed.
The most parsimonious models fitted to the results werederived using a step-down procedure of model reduction,
beginning with a full model comprising all main effects andthe first-order interactions between the categorical variables andwire diameter and the categorical variables and shaft length,
plus the interaction between wire diameter and shaft length.The coefficients of those parameters included in the final(reduced) models were significant at P, 0.01 (see Table 2).
The final models were compared with the full models using ananalysis of variance to ensure parsimony. All analyses werecompleted using the R statistical package (R Development CoreTeam 2009).
Results
Percentage of hook weight remaining
At the end of the experiment, none of stainless-steel hooksappeared to be oxidised, whereas the carbon steel designs werereduced to as little as 95% of their original weights. The most
parsimonious model accounted for 76% of the variation in thepercentage of hook weight remaining and was reduced to 10significant main effects and five significant first-order interac-
tions (with wire diameter or shaft length, Table 2). The fourmain effect variables that did not have a significant interactionwith either wire diameter or shaft length were sample time 3 andthe front, bend and gape lengths (Table 2). These presented as a
significantly lower percentage of weight remaining for all hookssampled at T28 (mean� s.e., 98.81� 0.07%) than sampled at T8(99.55� 0.03%) and a lower percentage of hook weight
remainingwith increasing front lengths (,12mm) and narrowerbends and gapes (P , 0.001; Fig. 2a–d).
Scatter plots of the interactions between wire diameter and
various categorical variables showed that, irrespective of wirediameter, the few stainless-steel hooks used retained theirinitial weights, whereas for the carbon-steel hooks, there wasa positive relationship between wire diameter and the percent-
age of weight remaining (Fig. 3a, b). Hook modification alsohad an effect on the percentage of weight remaining across therange of wire diameters, with unmodified hooks showing a
positive relationship similar to that for carbon steel (Fig. 3d).Although there were few data, this relationship was not main-tained for modified hooks (Fig. 3c). There were positive
relationships between wire diameter and the percentage ofhook weight remaining for both circle and J-hooks, althoughfewer data meant that the relationship was weaker for circle
hooks (Fig. 3e, f). The scatter plots of the interaction betweenshaft length and bait-holder barbs on the percentage of hookweight remaining showed that there was a stronger positiverelationship among hooks with bait-holder barbs than those
without (Fig. 3g, h).
Percentage of hook point remaining
After 28 days submersion in seawater, as little as 38% of thepoints remained on the carbon-steel hooks. The model that bestdescribed the percentage of hook point remaining was reduced
to four main effects and two first-order interactions (bothinvolving wire diameter) and accounted for 32% of the variation(P, 0.001, Table 2). The two main effect variables that did not
interact with wire diameter were sample time 3 and steel type(P. 0.01, Table 2). Specifically, there was lower percentage ofthe hook point remaining at T28 (90.49� 0.75%) than at T8(95.70� 0.38%). Although stainless-steel hooks retained theirinitial point length, carbon-steel hooks had only 92.06� 0.48%remaining.
An interaction existed between hook modification and
wire diameter, with unmodified hooks showing a slightly
94
96
98
100
9.0 10.0 11.0 12.0 13.0
Front (mm)
8.5 11.0 13.5 16.0 18.55
Gape (mm)
10.0 13.0 16.0 19.0 22.0 25.0Per
cent
age
of w
eigh
t rem
aini
ng
Bend (mm)
(a )
(b)
(c)
94
96
98
100
94
96
98
100
Fig. 2. The relationships between the percentage of hook weight remain-
ing and hook (a) front (b) bend and (c) gape lengths.
Reviewing hook degradation Marine and Freshwater Research E
positive relationship, especially for wire diameters ,1.24mm(P, 0.001, Fig. 4b). There were few data for modified hooks,which limited interpretation (Fig. 4a). Hook design also affected
the relationship between wire diameter and the percentage ofhook point remaining, with J-hooks displaying a trend similar tothat for unmodified designs (P, 0.001, Fig. 4d). Few data were
available for circle hooks (Fig. 4c).
Tensile strength
The range of tensile strengths of hooks before and after sub-mersion in seawater for 28 days was reduced from 80–370N to25–361N respectively. During testing, the majority of hooksstraightened at maximum elasticity, with breakages limited
to 16, 14 and 18% at T0, T8 and T28, respectively. The mostparsimonious model accounted for 72% of the variation intensile strength and comprised seven main effects and five first-
order interactions (three and two involving wire diameter orshaft length respectively, Table 2). Six of the seven main effectsincluded in themodel were significant atP, 0.01, as were three
first-order interactions with wire diameter. The two interactionswith shaft length were not significant at P, 0.01 and are notdiscussed further.
The main effects of sample time 1 and steel type werenot involved in significant interactions and presented asreductions in tensile strength among all hooks between T0(161.99� 4.99N) and T8 (144.78� 4.81N) and, irrespective
94
96
98
100
Stainless steel Carbon steel
94
96
98
100
Per
cent
age
of w
eigh
t rem
aini
ng
Modified Unmodified
Wire diameter (mm)
J
94
96
98
100
15 20 25 30 35 40 15 20 25 30 35 40
Shaft (mm)
Bait-holder barbs No bait-holder barbs
(a ) (b )
(c ) (d )
94
96
98
100
0.9 1 1.1 1.2 1.3 1.4 0.9 1 1.1 1.2 1.3 1.4
Circle (e ) (f )
(g ) (h )
Fig. 3. Scatter plots of the relationships between the percentage of hook weight remaining and wire diameter for
hooks that were (a) stainless steel, (b) carbon steel, (c) modified, (d) unmodified and (e) circle and (f) J-designs and
the relationship between the percentage of hook weight remaining and shaft length for hooks (g) with and
(h) without bait-holder barbs.
F Marine and Freshwater Research S. P. McGrath et al.
35
48
61
74
87
100
Per
cent
age
of p
oint
rem
aini
ngModified
Wire diameter (mm)
Unmodified(a) (b )
(c ) (d )
35
48
61
74
87
100
0.9 1.0 1.1 1.2 1.3 1.4 0.9 1.0 1.1 1.2 1.3 1.4
Circle J
Fig. 4. Scatter plots of the relationships between the percentage of the point remaining and wire diameter for hooks that were
(a) modified, (b) unmodified and (c) circle and (d) J-designs.
070
140210280350420
(a ) (b )
Ten
sile
str
engt
h (N
)
070
140210280350420
(c ) (d ) Unmodified
070
140210280350
420
0.9 1.0 1.1 1.2 1.3 1.4 0.9 1.0 1.1 1.2 1.3 1.4
Wire diameter (mm)
(e ) Bait-holder barbs (f ) No bait-holder barbs
T0 T28
Modified
Fig. 5. Scatter plots of the relationships between the tensile strength and wire diameter for hooks sampled
at (a) T0 and (b) T28 and those that were (c) modified, (d) unmodified and (e) with and (f) without bait holders.
Reviewing hook degradation Marine and Freshwater Research G
of all other variables, between stainless (148.65� 3.38N) andcarbon steel (145.27� 3.22N). Scatter plots of the wire diame-
ter interactions revealed that this variable had a stronger positiverelationship with tensile strength at T0 than at T28 (Fig. 5a, b) andamong unmodified hooks or those without bait-holder barbs
(compared with modified hooks and those with bait-holderbarbs respectively) (Fig. 5c–f).
Compression strength
The sampled hooks had compression strengths of 29–405Nbefore the experiment and 24–365N after 28 days submersion in
seawater. The percentages of hooks that broke during com-pression testing were 24, 27 and 30% for T0, T8 and T28respectively. The model that best explained the variation (71%)among compression strength comprised eight main effects andsix interactions (all involving wire diameter or shaft length); of
which seven and four, respectively, were significant at P, 0.01(Table 2). Two significant main effects that did not interact witheither wire diameter or shaft length were sample time 2 and front
length. These effects presented as significantly greater reduc-tions in compression strength at T28 (105.53� 5.87N) than at T0(133.58� 6.74N) and with shorter front lengths (P, 0.01,Fig. 6).
Scatter plots of the interactions indicated a slightly strongerpositive relationship between wire diameter and the compres-sion strength of circle hooks than for J-hooks (Fig. 7a, b),
although there were few data. Conversely, the presence ofbait-holder barbs negated any positive influence that wirediameter had on compression strength (Fig. 7c, d). A significant
interaction (P, 0.01) was found between shaft length and thepresence or absence of modifications, with a clear negativerelationship for unmodified hooks. The lack of a reasonablerange of shaft lengths among the modified hooks precluded
coherent interpretation of this result (Fig. 7e, f). Although notplotted, for any given wire diameter, there was a trend of greaterreductions in compression strength with increasing shaft length.
0
70
140
210
280
350
420
9 10 11 12 13
Com
pres
sion
str
engt
h (N
)
Front (mm)
Fig. 6. The relationship between hook front length and compression
strength.
(f ) Unmodified
(b) J
070
140210280350420
(a ) Circle
070
140210280350420
0.9 1.0 1.1 1.2 1.3 1.4 0.9 1.0 1.1 1.2 1.3 1.4
Com
pres
sion
str
engt
h (N
) (c ) Bait-holder barbs
Wire diameter (mm)
(d ) No bait-holder barbs
070
140210280350420
15 20 25 30 35 40 15 20 25 30 35 40
Shaft (mm)
(e ) Modified
Fig. 7. Scatter plots of the relationships between the compression strength andwire diameter for hooks that were
(a) circle, (b) J-designs and (c) with and (d) without bait-holder barbs and scatter plots of the relationship between
the compression strength and shaft length for hooks that were (e) modified and (f) unmodified.
H Marine and Freshwater Research S. P. McGrath et al.
Discussion
This study has isolated some of the key technical factorsaffecting the temporal oxidation and associated strengthreduction in commonly used hooks following submersion in
seawater and, in doing so, helped to identify simple modifica-tions by which decay could be promoted. Before discussingthese results and their implications in terms of promoting hookejection by fish, it is necessary to first consider the forces
imposed on hooks during the catching process. This informationcan then be used to give some indication of the necessary initialthreshold for hook strength (and any associated modifications).
Possible minimum hook strength required
There are few published studies quantifying the forces involvedin hooking fish (Sakazume and Kanamori 1971; Mitsugi and
Inoue 1985; Edappazham et al. 2008). However, intuitively,once the barb penetrates the flesh or bone and the fish pulls awayfrom the angler, most of the load should be tensile strain.
Fridman (1986) suggested that the approximate maximumtractive force (kg) of a fish could be estimated by dividing itsweight in air (kg) by the cube root of its length (m). Applyingthis formula to the largest sizes of fish (i.e.,40 cmTL) targeted
using the various hooks in the present study and assumingthat such fish typically weigh ,1 kg (e.g. Santos et al. 2002;Broadhurst et al. 2006), their continuous tractive force would be
,13N. Additional force from kinetic energy during the initialhooking might increase this estimate slightly (Fridman 1986),but it would still be less than almost 85% of the minimum initial
tensile strength detected here for the weakest hooks (i.e. thosemodified with notches – mean� s.e. of 92� 4N) and 92% lessthan the average tensile strength (169� 5N) of the remaining
conventional designs (Table 1).Based on the above estimate, all of the hooks tested in this
study were considerably stronger than required and, therefore,could be modified or redesigned to reduce their initial strength
and increase their temporal decay during ingestion. It is alsoclear that the starting point for any such changes should involvethose main factors identified to affect overall hook degradation,
in particular the wire material and diameter.
Reducing hook strength and maximising decay
Irrespective of all other parameters, the wire material had thestrongest impact on oxidation and strength. Unlike the carbon-steel hooks, after 28 days in seawater, the stainless steel designs
retained most of their initial strengths, weights and pointlengths; which means that the latter remained sharp. This resultis important since, once ingested, stainless-steel hooks would be
expected to have a greater probability of penetrating soft tissueand vital organs during progression through the digestive tract(Broadhurst et al. 2007). Further, the maintenance of initial
high tensile and compression strengths (between 127� 6 and197� 7 N) among the unmodified stainless-steel hooks wouldgreatly reduce their chances of breaking after ingestion. Both of
these factors could translate to an increased mortality amongfish that ingest hooks.
By comparison, after 28 days of submersion, the carbon-steelhooks were significantly oxidised to a condition whereby some
had less than 95 and 38% of their initial weights and hook points
respectively. The amount of degradation and concomitant lossof strength of carbon-steel hooks varied among designs and was
probably at least partly due to the alloy composition and the typeof protective coating (Edappazham et al. 2008); both of whichremain unknown. But, it is also clear that a significant propor-
tion of the degradation variability can be attributed to thediameter of the wire.
In support of the few relevant previous studies (e.g.
Edappazham et al. 2008), wire diameter had a consistentpositive relationship with tensile and compression strengthsand the percentages of weight and point remaining for mosthooks, especially conventional carbon steel J-designs. Owing
to the linear increase in the surface area-to-volume ratioand reduction in mass, at narrower diameters the latter hooksoxidised faster and were weaker. The importance of wire
diameter was reiterated by its consistent significant interactionswith the presence or absence of modifications and bait-holderbarbs. By reducing the diameter of the wire at multiple locations
on the shaft, bend or front (by,20%) and effectively providinga weak point, these changes largely negated any relationshipbetween increasing wire diameter and strength. Bait-holderbarbs and similar modifications to wire diameter could be an
appropriate strategy for increasing the probability of bothstainless- and carbon-steel hooks breaking and the latteroxidising, after ingestion. In support of such effects, Broadhurst
et al. (2007) attributed the bait-holder barbs on nickel-platedcarbon-steel hooks to rapid decay and weak points on theshaft which facilitated breakage (often within 3 weeks) during
ingestion by yellowfin bream.Although changes to wire material and diameter would have
the most benefit in promoting temporal hook decay, there were
also significant effects of the length of the bend, gape and front.The impacts of the bend and gape were restricted to thepercentage of weight remaining and presented as a trend ofgreater oxidation among narrower lengths. Such a result is
difficult to explain, although it might reflect a greater curvatureof the wire and the associated slightly increased outside surfacearea. A similar increase in surface area associated with longer
shaft lengths of up to 25mm and front lengths of up to 12mmmight also explain the slightly greater relative oxidation,although hooks with front and shaft lengths beyond these
estimates were oxidised at a lower rate. More data across widerranges of hook sizes and types would be required to elucidate therelationship between the bend, gape and front of hooks and theiroxidation.
In addition to oxidation, the shaft and front lengths alsoaffected the compression strength. As the shaft length increased,hooks generally became weaker owing to the load deflection
(causing the shaft to bend, Case et al. 1999). However, anypotential benefits associated with a longer, weak shaftduring ingestion might be negated since McGrath et al.
(2009) indicated that such hooks could not easily rotate in thedigestive tract, minimising their ejection from sand whiting(Sillago cillata). Hooks with long front lengths might be
similarly difficult to eject. Also, unlike for shaft length, longerfront lengths had significantly greater compression strengths.However, this result might reflect the experimental design sincesuch hooks were held more securely in the compression
cylinders.
Reviewing hook degradation Marine and Freshwater Research I
The latter result highlights one of the limitations of this studyand reiterates the conditional interpretation of the results.
Each of the four response variables were selected to assesstemporal hook degradation which, based on the observations byBroadhurst et al. (2007), was assumed to be the same in seawater
as that in the digestive tract of a fish. However, in some cases,hooks can become imbedded in soft tissue, resulting in loweroxidation rates (Aalbers et al. 2004) and, therefore, a greater
maintenance of strength than that observed here. Also, althoughthe tensile force measured at T0 might correlate with the forceimposed by a fish during capture (Edappazham et al. 2008), thedata collected at T8 and T28 are unlikely to represent
the same forces on hooks after ingestion (although these canbe quite extreme, Broadhurst et al. 2007). Such differencesmean that the results from this study should not be extrapolated
beyond simple indices of degradation.
Conclusions
There are some limitations of the experimental design, but thisstudy still supports recommendations for ongoing work into
hook construction and provides direction for anglers who wishto minimise impacts to fish using conventional hook designs.For example, future research might warrant examining designs
made from composite materials that comprise a sufficientlyresilient wire for the shaft, bend and front, but a weak pointand barb that oxidise very quickly. It might also be feasible
to incorporate weaker points on the wire, so that the hookbreaks down quickly after ingestion. The utility of such changesneed to be investigated in close consultation with hook manu-facturers since ultimately, their feasibility will be dictated by
economics.In the interim, anglers should be encouraged to avoid
stainless-steel hooks and choose carbon steel designs with the
narrowest wire diameter. Based on the results here, hooks withsimilar absolute sizes, but made from a wire diameter of lessthan ,0.9mm should still provide sufficient tensile strength,
while promoting rapid oxidation and subsequent breakage.Existing hooks (including stainless steel designs) with thickerwire could be easily modified by anglers to make them weaker,
by incorporating notches similar to bait-holder barbs. Suchmodifications could help to reduce unaccounted fishing mortal-ity and further validate catch-and-release as an appropriatemanagement tool for conserving stocks.
Acknowledgements
The experiment was funded by the New SouthWales (NSW) Department of
Primary Industries, the NSWRecreational Fishing Trusts and the University
of New England. We thank Richard Faulkner for extensive discussions,
Andy Revill and an anonymous reviewer for their comments and Chris
Dowling, Hester Bushell, LachlanRoberts andMilanDuwenhogger for their
technical assistance.
References
Aalbers, S. A., Stutzer, G. M., and Drawbridge, M. A. (2004). The effects of
catch-and-release angling on the growth and survival of juvenile white
seabass captured on offset circle and J-type hooks. North American
Journal of Fisheries Management 24, 793–800. doi:10.1577/M03-034.1
Arlinghaus, R., Cooke, S. J., Lyman, J., Policansky, D., Schwab, A., et al.
(2007). Understanding the complexity of catch-and-release in recrea-
tional fishing: an integrative synthesis of global knowledge from
historical, ethical, social and biological perspectives. Reviews in Fisher-
