Editorial Manager(tm) for Journal of Fish Biology Manuscript Draft
Manuscript Number: MS 05-338R2
Title: Gape morphology of cod Gadus morhua L., haddock Melanogrammus aeglefinus (L.) and whiting
Merlangius merlangus (L.) through metamorphosis fom larvae to juveniles in the western Irish Sea.
Article Type: Regular paper Section/Category:
Keywords: gadoid; larvae; gape; jaw; development
Corresponding Author: Dr. William Llewelyn Rowlands, BSc, MSc, PhD Corresponding Author's Institution: University of Liverpool
First Author: William Llewelyn Rowlands, BSc, MSc, PhD
Order of Authors: William Llewelyn Rowlands, BSc, MSc, PhD; Mark Dickey-Collas, PhD; Audrey Geffen, PhD; Richard D.M. Nash , PhD
Manuscript Region of Origin:
Abstract: Variations in standard length, gape size and jaw length were studied in larval and juvenile gadoids from 4-70mm. The increase in gape size and jaw length was not linear with respect to standard length. The relationship was best described by segmented regression lines in all three species, with an inflection point at ca. 10.5mm. Gape size and jaw length increased more rapidly in relation to larval length for individuals smaller than this inflection point size. The rate of increase slowed significantly post-inflection, an effect more noticeable in gape size data compared to jaw length data. In each case the inflection point fell in the
intermediate period of development between the larval and juvenile stages, which could be considered as metamorphosis. Published equations that have been used to predict gape size from jaw length lead to the overestimation of gape. New relationships are presented which may be used to predict gape size from measurements of either standard length or upper jaw length in cod, haddock and whiting.
This piece of the submission is being sent via mail.
Copyright form
Dear Dr Craig,
I have re-submitted my manuscript (MS 05-338R1) with the following alterations:
1. Line numbers have been added
2. Authorities have been corrected to Gadus morhua L., Melanogrammus aeglefinus (L.) and Merlangius merlangus (L.)
3. Presentation of statistics in text changed from (GLM, F(2,536) = 8.6, p<<0.001) to (GLM, d.f. = 2, n = 536, P<0.001) in all cases
4. Variables changed as shown below - standard length: LS to LS
upper jaw length: LUJ to LUJ
lower jaw length: LLJ to LLJ
gape size: SG to SG
5. Indents corrected on paragraphs 6. References corrected in the text
7. Tables all converted from excel files to “word tables” so are now all doubled spaced (however they do not look as good)
8. Figures all converted to EPS files
9. Reference list corrected – there was something wrong with the formatting I hope this covers all aspects that needed revising
Yours,
William Rowlands
* Response to referees
Gape morphology of cod Gadus morhua L., haddock Melanogrammus
1
aeglefinus (L.) and whiting Merlangius merlangus (L.) through
2
metamorphosis from larvae to juveniles in the western Irish Sea.
3 4
W. LL. ROWLANDS*∞, M. DICKEY-COLLAS†, A. J. GEFFEN*‡AND R. D. M. NASH*• 5
6
* University of Liverpool, Port Erin Marine Laboratory, Port Erin, Isle of Man, IM9 6JA, British Isles.
7
† Netherlands Institute of Fisheries Research, P.O. Box 68, 1970 AB IJmuiden, The Netherlands.
8 9
Variations in standard length, gape size and jaw length were studied in larval and juvenile gadoids from 10
4-70mm. The increase in gape size and jaw length was not linear with respect to standard length. The 11
relationship was best described by segmented regression lines in all three species, with an inflection 12
point at ca. 10.5mm. Gape size and jaw length increased more rapidly in relation to larval length for 13
individuals smaller than this inflection point size. The rate of increase slowed significantly post- 14
inflection, an effect more noticeable in gape size data compared to jaw length data. In each case the 15
inflection point fell in the intermediate period of development between the larval and juvenile stages, 16
which could be considered as metamorphosis. Published equations that have been used to predict gape 17
size from jaw length lead to the overestimation of gape. New relationships are presented which may be 18
used to predict gape size from measurements of either standard length or upper jaw length in cod, 19
haddock and whiting.
20 21 22
Key words: gadoid; larvae; gape; jaw; development.
23
‡ Present address: Department of Biology, University of Bergen, P.O. Box 7800, 5020 Bergen, Norway.
• Present address: Institute of Marine Research, P.O. Box 1870, N-5817 Bergen, Norway.
∞ Author to whom correspondence should be addressed. Tel: +44 (0) 1624 831037; Fax: +44 (0) 1624 831001; email: [email protected]
Manuscript
24 25
INTRODUCTION 26
27
Cod Gadus morhua L., haddock Melanogrammus aeglefinus (L.) and whiting 28
Merlangius merlangus (L.) are commercially important white-fish which spawn 29
between March and May in the Irish Sea (Bal, 1941; Wheeler, 1969; Nichols et al., 30
1993; Brander, 1994). Due to similarities in spawning times and locations there is a 31
considerable overlap in the distribution of the pelagic larval stages of these three 32
species (Fox et al., 1997). This distributional overlap along with the utilization of 33
similar planktonic prey items through development (Last, 1978; Robb & Hislop, 34
1980; Economou, 1991) implies a high degree of inter- and intra-specific competition 35
for food; as previously suggested by Kane (1984) with >75% overlap between larval 36
gadoid diets.
37 38
The extent of competition for food is directly related to food availability (Last, 39
1978), and the size distribution of available prey (Bremigan & Stein, 1994). The array 40
of these available prey items is defined by prey detection, swimming ability, strike 41
ability and mouth size of the larvae (Sabatés & Saiz, 2000), all of which develop with 42
age (Hunter, 1980).
43 44
This study compared variations in gape size between species, as this is a major 45
factor limiting the maximum size of prey consumed by larvae (Arts & Evans, 1987;
46
Bremigan & Stein, 1994). Larval development proceeds through different growth 47
phases, leading to changes in allometric growth patterns (Osse & van den Boogaart, 48
1995; van Snik et al., 1997). If gape size increases in this manner with body length 49
then the rate of gape increase could be partially responsible for specific differences in 50
food size selection and thus niche prey size (Hunter, 1980) between the larvae of 51
different species. This is so as at the start of exogenous feeding larvae with larger 52
gape size have been shown to have faster growth rates (Shirota, 1970, 1978). The 53
faster growth rate is due to the larvae’s ability to take a larger size of prey with leads 54
to increased growth efficiency (Kerr, 1971a, b). A larger gape size also leads to a 55
larger size selection of prey available to the larvae and this prey niche size has been 56
shown to increase with larval development, though is not constant between 57
developmental periods or between different species (Pepin & Penny, 1997). The 58
importance the increase in prey niche size is that intraspecific competition between 59
early and late spawned members of the same year class will be reduced due to the 60
increase prey available (Robb & Hislop, 1980; Pepin & Penny, 2000). Therefore, with 61
the ability to both ingest a larger size of prey and have a larger prey niche size, the 62
effects of intra- and interspecific competition for food between larvae would be 63
reduced, increasing chances of survival. However, Pepin & Penny (1997) observed 64
that the rate of increase in prey size was slower than the increase in the physical 65
capacity to ingest larger prey. This suggests that perception and vulnerability of prey 66
along with environmental factors are further elements required for comparative 67
studies of larval feeding ecology.
68 69
To address the physical and morphological limitations, this study examines 70
and compares the development of gape size between cod, haddock and whiting during 71
larval development. Although work has previously been carried out on the gape sizes 72
of marine larvae (Shirota, 1970; Schael et al., 1991; Munk, 1997; Sabatés & Saiz, 73
2000), little has been done on comparisons in relation to metamorphosis in gadoid 74
larvae to juveniles. This is an important life history phase, as during this period larvae 75
undergo rapid internal and external changes (Kjørsvik et al., 1991; Osse & van den 76
Boogaart, 1995; Fuiman & Higgs, 1997). Morphological comparisons between the 77
species were undertaken through the gape measurement of these gadoids and thus 78
probable overlaps in available prey items evaluated.
79 80
MATERIALS AND METHODS 81
82
SAMPLE COLLECTION AND PRESERVATION 83
84
Samples of fish larvae were collected on two occasions (April and May 2001) 85
in the northwestern Irish Sea from the Department of Agriculture and Rural 86
Development – Northern Ireland (DARD) research vessel, the M.R.V. Lough Foyle.
87
The full range of developmental stages, yolk-sac larvae to juveniles, were sampled 88
with a Hydrobios Multiplankton (MP) Sampler with 280µm mesh (Weikert & John, 89
1981) to capture yolk-sac larvae in both April and May, a MOCNESS sampler 90
(Multiple Opening and Closing Net and Environmental Sensing System (Wiebe et al., 91
1976; Roe & Shale, 1979)) with 1000µm mesh to sample larger larvae in May only 92
and a German pelagic juvenile trawl in May only to sample juveniles. The first two 93
samplers were towed for 20 minutes using a stepwise oblique profile at 3 knots (1.5 94
m.s-1). The later gear was towed for 20 minutes at between four and five knots (2-2.6 95
m.s-1). Over a twenty-four hour period the MP was deployed with two replicates every 96
four hours, the MOCNESS six times, and the juvenile pelagic trawl only once. This 97
sampling strategy was part of a larger project and these deployments were carried out 98
at four different stations in the northwestern Irish Sea (Fig. 1). The high numbers of 99
deployments gave a relatively large catch of larval and juvenile gadoids.
100 101
All samples were preserved immediately in 4% buffered (sodium acetate 102
trihydrate) formaldehyde to reduce larval shrinkage (Blaxter, 1971; Hay, 1981).
103 104
MORPHOLOGICAL AND DEVELOPMENTAL MEASUREMENTS 105
106
Standard lengths of cod, haddock and whiting larvae were measured in the 107
laboratory, either under a dissecting microscope with graticule (with a precision of 108
0.02mm), or in the case of larger juveniles with vernier calipers (with a precision of 109
0.05mm).
110 111
Gape sizes were calculated using a method similar to that of Wong & Ward 112
(1972), which allowed the limiting dimension to be measured, in this case with 113
differently sized disposable plastic pipette tips ranging from 0.3mm to 10mm in 114
diameter. The pipette tips were placed under a microscope and scored along their 115
length using a scalpel at 0.5mm intervals. The scored marks were then rubbed with 116
permanent ink to leave a clearly identifiable line. The diameter of the pipette tip at 117
each scored interval was measured using a dissecting microscope with graticule to 118
give a range of sizes (diameters) along the pipette tip with an accuracy of 0.05mm.
119 120
Gape size was measured by holding a larva under the microscope, while 121
inserting a calibrated pipette tip into the mouth until a jaw angle of 90º was achieved.
122
The diameter of the cone at the scored mark where the front tips of the upper and 123
lower jaws touched the cone was taken to be the gape. The measurement was made at 124
a mouth angle of 90º as this has been assumed to give the maximum gape for larval 125
fish (Shirota, 1970). Care had to be taken not to insert the cones too far, as this caused 126
the pipette tip to reach the back of the mouth before the jaws were fully opened thus 127
giving false readings (Arts & Evans, 1987). In order to avoid this problem, a slightly 128
larger sized tip was used if a tip went too far into the mouth.
129 130
The upper and lower jaw lengths of the larvae were measured under a 131
dissection microscope with graticule and defined as the distance from the tip of the 132
jaw to where the upper and lower jaws joined. Jaw lengths were measured with the 133
mouth in the closed position.
134 135
The measurement of the jaws also allowed gape sizes to be calculated using 136
equations given by Shirota (1970) and later modified by Guma’a (1978);
137 138
( )
UJG
L
S = 2 ∗
(Equation 1) (Shirota, 1970)139
(
UJ2 LJ2)
G
L L
S = +
(Equation 2) (Guma’a, 1978)140 141
Where the calculated gape size (SG), was obtained from the length of the upper jaw 142
(LUJ) and the length of the lower jaw (LLJ).
143 144
Basic developmental features of each larva were also recorded using external 145
and internal morphology (Balon, 1975; Timmermans, 1987; Kjørsvik et al., 1991;
146
Osse & van den Boogaart, 1995). A score was allocated to each developmental 147
observation (Table I) which, when totaled, gave a generalized index of larval 148
development for each individual.
149 150
STATISTICAL ANALYSIS 151
152
Due to the allometric relationships observed between larval length and 153
measured structures, data were logarithmically transformed to give linear 154
relationships allowing ease of comparison between species (Peters, 1983). However, 155
the data best fitted a segmented regression, based on examination of the residuals 156
(Peck et al., 2005). The linear regression showed a clear trend in the plotted residuals 157
(Fig. 2 (Insert A)) suggesting a systematic error in the relationship, while the plotted 158
residuals of the segmented relationship (Fig. 2 (Insert B)) showed no pattern, 159
suggesting a good fit to the data. This was shown to be the case for all morphological 160
measurements on all three species, and so a segmented regression was used for all 161
subsequent analyses.
162 163
The points of inflection for the segmented relationships were found using the 164
computer program PRISM® (produced by GRAPHPAD® (www.graphpad.com)), 165
which calculated the residual sum of squares (RSS), testing each inflection point 166
iteratively to obtain the lowest RSS value (Somerton, 1980).
167 168
The significance of each segmented regression line was tested and 169
comparisons between regression lines were made using the General Linear Model 170
(GLM) assuming normality and homogeneity of variance. All calculations were made 171
using MINITAB® release 13.1 (www.minitab.com).
172
173
RESULTS 174
175
GAPE SIZE/STANDARD LENGTH RELATIONSHIP 176
177
Gape size increased allometrically with standard length and when transformed 178
using the natural log, a segmented relationship with a single inflection point was 179
observed (Fig. 3). Although a single line linear regression could be fitted to these data, 180
it obscured the true relationship between gape size and fish length (Table IIi – shown 181
by the lower F statistic values) . The inflection points were located at a standard 182
length of 9.84mm for whiting, 10.21mm for haddock and 11.72mm for cod. The 183
differences between species were not significant, suggesting a global gadoid 184
breakpoint of 10.56mm.
185 186
In all three species the slope of the regression of gape size on length for fish 187
pre-inflection point was >1 (positive allometry) and post-inflection point the slope 188
was <1 (negative allometry), (Table IIi). In all cases the relationship pre- and post- 189
inflection point was significantly different (GLM, d.f. = 2, n = 141, P<0.001; d.f. = 2, 190
n = 258, P<0.001; d.f. = 2, n = 553, P<0.001, for cod, haddock and whiting 191
respectively).
192 193
Among smaller larvae (pre-inflection point) the gape size of whiting increased 194
at a significantly slower rate than that of haddock (GLM, d.f. = 2, n = 520, P<0.001), 195
while the rate of gape size increase of cod fell between that of whiting and haddock 196
and was not significantly different to either. In larger larvae (post-inflection point) the 197
gape size of whiting increased at a significantly faster rate with length than for both 198
haddock and cod (GLM, d.f. = 2, n = 433, P<0.001); again no difference was seen 199
between cod and haddock.
200 201
JAW LENGTH/STANDARD LENGTH RELATIONSHIP 202
203
Upper jaw length and standard length data (Fig. 4) also showed an allometric 204
relationship and on transformation (natural log), a segmented regression explained 205
more variance compared to a standard single linear regression (Table IIii shown by 206
the higher F statistic values). A significant reduction in the rate of jaw length increase 207
with standard length post-inflection point was observed for all species (GLM, d.f. = 2, 208
n = 141, P<0.001; d.f. = 2, n = 259, P<0.001; d.f. = 2, n = 552, P<0.001, for cod, 209
haddock and whiting respectively). Inflection points for cod, haddock and whiting 210
occurred at standard lengths of 10.51mm, 8.95mm and 10.36mm respectively, with no 211
significant difference between species suggesting a global gadoid inflection point at 212
9.94mm.
213 214
Pre-inflection point regression comparisons between species, showed the rate 215
of jaw length increase with standard length of haddock significantly greater (GLM, 216
d.f. = 2, n = 602, P<0.001) than for either cod or whiting. Post-inflection point 217
regression comparisons showed haddock to have the slowest rate of jaw length 218
increase (GLM, d.f. = 2, n = 536, P<0.001).
219 220
GAPE SIZE/JAW LENGTH RELATIONSHIP 221
222
The upper jaw length was used for analysis, as it was the easier dimension to 223
measure, and therefore most reliable. When compared, upper jaw length and gape size 224
again produced a curved relationship, which was transformed using natural logs. A 225
segmented relationship with single inflection point (Fig. 5) was a better fit to the data 226
than a single linear regression (F-statistic greater in segmented relationships - Table 227
IIiii). The positions of these inflection points were compared and no differences were 228
observed between species, occurring at a jaw length of 1.27mm.
229 230
For each of the three species the rate of gape size increase with jaw length 231
decreased after the inflection point (GLM, d.f. = 2, n = 188, P<0.001; d.f. = 2, n = 232
317, P<0.001; d.f. = 2, n = 632, P<0.001, for cod, haddock and whiting respectively), 233
changing from positive to negative allometry. Between species, the slope of the 234
regression lines below the inflection point showed no differences. However, 235
regressions above the inflection point suggested that gape size increased at a 236
significantly faster rate with jaw length for cod compared to either haddock or whiting 237
(GLM, d.f. = 2, n = 411, P<0.05), with no difference between the haddock and 238
whiting.
239 240
MODELS FOR ESTIMATING GAPE 241
242
Gape size was calculated from jaw length, following equations in Shirota 243
(1970) and Guma’a (1978) and compared to the actual measurements of gape size.
244
The calculated gape sizes were significantly larger than actual gape measurements 245
(GLM, d.f. = 2, n = 1801, P<0.001), with post-inflection point data (those >9.84mm – 246
taken from Fig. 4) showing increased over-estimation.
247
248
From data in this study, individual models converting standard length (LS), 249
upper jaw length (LUJ) and lower jaw length (LLJ) to gape size (SG) could be calculated 250
and are shown below. The associated variables (a and b) for each gadoid species pre- 251
and post-inflection point are given in Table III:
252 253
b L
S
G= (
Sa) *
(Equation 3) 254b L
S
G= (
UJa) *
(Equation 4) 255b L
S
G= (
LJa) *
(Equation 5) 256257
Each model is subject to size constraints due to the segmented nature of the 258
data and the size range of larvae used. Pre-inflection equations for all models are 259
limited to larvae of standard lengths and upper and lower jaw lengths below the size at 260
inflection. The standard length model, post-inflection equations have an upper size 261
limit of 50mm for cod, and 65mm for haddock and whiting. The upper and lower jaw 262
length models, post-inflection equations have an upper limit of 4.5mm for cod, 6.5mm 263
for haddock and 7.5mm for whiting.
264 265
LARVAL DEVELOPMENT 266
267
The relationship between developmental index and larval length (natural log 268
standard length – Fig. 6) was well fitted by a segmented regression model with two 269
inflection points (ANOVA, d.f. = 2, n = 185, P<0.001; d.f. = 2, n = 255, P<0.001; d.f.
270
= 2, n = 482, P<0.001, for cod, haddock and whiting respectively). For cod, haddock 271
and whiting, development fell into three distinct categories named as larval, 272
intermediate and juvenile stages. The developmental index increased rapidly with 273
length following the first inflection point, suggesting that the transition from the larval 274
to intermediate stages was linked to the start of metamorphosis (Fig. 6). This occurred 275
at a standard length of 8.21mm for cod, 8.13mm for haddock and 7.71mm for whiting.
276
These were not found to be significantly different suggesting a general global gadoid 277
size of ca. 8mm could be considered as the start of metamorphosis.
278 279
The developmental index was calculated primarily from changes in fin and 280
stomach characteristics, and their period of most rapid development (intermediate 281
stage) ended at a standard length of 15.61mm for cod, 12.85mm for haddock and 282
13.67mm for whiting. Once again no significant difference was found between these 283
values (partly due to the large size variation in the data) implying a general global 284
gadoid size of ca. 14mm.
285 286
DISCUSSION 287
288
The relationships between gape size and standard length, and jaw length and 289
standard length for cod, haddock and whiting changed from positive to negative 290
allometry during development. A change in allometry has previously been shown to 291
occur in the head and tail of larvae, due to initial high-energy investment for primitive 292
functions and essential organs (Osse & van den Boogaart, 1995). In this study only 293
mouth morphology was studied, therefore what is the advantage of reducing the rate 294
of gape and jaw length increase? An explanation could be that newly hatched larvae 295
require rapid growth. The need for rapid growth is to avoid predation, as the smaller 296
the larvae the more likely they are to be predated (Scharf et al., 2000; Aljetlawi et al., 297
2004) and to allow them to obtain a greater size spectrum of prey rapidly, to avoid 298
competition from more recently hatched larvae (Sabatés & Saiz, 2000). The 299
implication to gape size is that it has been shown that a larger gape size leads to larger 300
prey items (Robb & Hislop, 1980), which in turn leads to more rapid growth (Hunter, 301
1980). Therefore the greater the ratio of gape size to body length the greater the 302
ability of a larva to take large prey, and grow more rapidly. A high gape size to body 303
length ratio would therefore be especially beneficial for first feeding larvae and 304
explain the observed positive allometry pre-inflection point in gape size and standard 305
length data. However, once at the juvenile stage the need for rapid growth is lessened 306
due to a reduction in competition for food, due to a shift to larger prey items (Pearre, 307
1986; Sabatés & Saiz, 2000), and reduced predation pressures (Scharf et al., 2000).
308
Thus the energetic expenditure for rapid gape increase could be reduced and 309
apportioned to other aspects of development.
310 311
A segmented relationship leading to a reduction in the rate of mouth growth 312
(as jaw length) with increasing body length was reported previously in a number of 313
marine larvae by Shirota (1978), where the point of inflection was seen to differ 314
between species. However, in this study the inflection point occurred at the same 315
standard length for all species for gape size, and upper jaw length data, implying a 316
general gadoid model, certainly for these gadoids in the Irish Sea. In these fish the 317
inflection points occurred at a standard length corresponding to the metamorphic 318
period of development (between ca. 8mm and ca. 14mm). This may also be seen in 319
other gadoids as Osse & van den Boogaart (1995) suggested that there was similarity 320
in particular allometries at an equal size range of fish larvae of distantly related taxa 321
and Mandali de Figueiredo (2003) showed that the larvae of two gadoid species 322
(whiting and rockling), had identical gape size to body length ratios. These 323
observations suggest that a global gadoid model could be assumed, where gape and 324
jaw allometry could be used as aids in the developmental staging of larvae into this 325
metamorphic stage, a poorly defined area in the life history of fish (Copp & Kováč, 326
1996).
327 328
Although similarities in standard lengths at allometric inflection points, and 329
developmental stages suggest a global gadoid model, some differences between 330
species did occur. The rate of gape increase with standard length both pre- and post- 331
inflection point were shown to differ between species as suggested previously by 332
Robb & Hislop (1980). Whiting were shown to have a slower rate of gape size 333
increase pre-inflection compared to haddock but a greater rate post-inflection 334
compared to both cod and haddock. Due to the slowest rate of gape increase, whiting 335
had the smallest gape size at the inflection point, while those of cod and haddock were 336
larger and did not differ (cod: 1.57mm; haddock: 1.43mm; whiting: 1.25mm – 337
obtained from Fig. 3 and equations in Table III). However, post inflection the gape 338
size of whiting rapidly became larger than that of both cod and haddock, suggesting 339
that in terms of gape size whiting differed consistently from both cod and haddock.
340
The similarities in gape size of cod and haddock throughout development could imply 341
that inter-specific competition would be high due to a similar prey niche size.
342
However, cod larvae take larger sized prey than both haddock and whiting, (Robb &
343
Hislop, 1980), and haddock in turn have been shown to be the least selective of these 344
larvae, having a tendency to go for more slow moving prey items (Economou, 1991).
345
This suggests that although gape size can give an indication of maximum prey size it 346
cannot be used by itself to determine prey niche size as other factors such as predator 347
and prey behavior will also influence prey selection (Pepin & Penny, 1997; Scharf et 348
al., 2000). A study of prey niche size requires gut analysis along with gape size 349
analysis, and knowledge of the prey available in the environment, especially for low 350
concentrations of larger prey which may lead to underestimating niche sizes (Munk, 351
1997; Gonzalez-Quiros & Anadon, 2001).
352 353
Gape size is an important parameter to measure for studies of feeding ecology 354
and potentially for studies of larval development. Gape size is often difficult to 355
measure directly and is often calculated from jaw measurements. This study compared 356
measured gape size to gape size calculated from the equations of Shirota (1970) and 357
Guma’a (1978), and showed clearly that gape sizes derived from the equations 358
consistently overestimated the actual maximum gape (as measured directly), being 359
especially pronounced after the point of inflection. This observation is directly 360
comparable with the work of Shirota (1970) who estimated gape size up to a larval 361
length of 30mm, well beyond the inflection points found in this study. As both mean 362
and maximum prey sizes are a proportion of gape size (Robb & Hislop, 1980), this 363
discrepancy would suggest that larvae could take a greater size distribution of prey 364
items than was actually the case. If used in comparisons such as with prey size (as in 365
Munk (1997)) an overestimation of prey available will occur and affect the prediction 366
of larval survival in relation to prey present.
367 368
Shirotas’ (1970) and Guma’as’ (1978) calculations rely on a single 369
relationship to derive gape size from jaw length. When post-inflection point data was 370
compared to pre-inflection point data from this study there was a 25% ± 5% reduction 371
in the rate of jaw length increase compared to a 50% ± 5% reduction in the rate of 372
gape size increase for the same increase in standard body length. Thus the real 373
segmented allometric effect was not fully translated into the gape size calculations of 374
Shirota (1970) and Guma’a (1978), leading to overestimations of gape size, especially 375
during both the later intermediate and juvenile stages. Two non-linear models (one 376
pre- and one post- inflection) for each gadoid species that incorporate the relative 377
changes of gape size in relation to either jaw length or fish length produces better 378
estimates where gape size cannot be measured directly. This method of gape size 379
calculation on unpreserved larvae would have to be carried out with caution as 10%
380
shrinkage has been observed in the standard length of other gadoid species preserved 381
in formalin (Porter et al., 2001).
382 383
In conclusion, gape size can be, a useful tool in establishing prey niche size of 384
fish larvae, and in helping to developmentally stage larval fish. However, if gape size 385
is calculated from jaw length, the allometric relationships as larvae grow and develop 386
must be incorporated. Although gape sizes and rates of gape size increase have been 387
shown to vary between species, similarities suggesting the existence of a global 388
gadoid model were also observed. These similarities were related to larval 389
development, where in all species, larval length at changes in jaw and gape allometry 390
did not differ significantly. In every case these changes in allometry occurred in the 391
same period of development, the intermediate stage, equating to the period of most 392
rapid development implying metamorphosis. This was found not to differ between 393
species implying a general gadoid length for the start of metamorphosis at 8mm, with 394
rapid development ceasing at a length of 14mm. Further experimentation of interest 395
would therefore be the larval study of other gadoid species to see if they too adhered 396
to these developmental patterns. Another interesting comparison would be the 397
application of the newly derived gape calculation equations to adult cod, haddock and 398
whiting.
399
REFERENCES 400
401
Aljetlawi, A. A., Sparrevik, E. & Leonardsson, K. (2004). Prey-predator size- 402
dependent functional response: Derivation and rescaling to the real world. Journal of 403
Animal Ecology 73, 239-252.
404 405
Arts, M. T. & Evans, D. O. (1987). Precision micrometer measurement of mouth gape 406
of larval fish. Canadian Journal of Fisheries and Aquatic Sciences 44, 1786-1791.
407 408
Bal, D. V. (1941). Observations on spawning periods and key to pelagic eggs of the 409
fishes in Manx waters. Proceedings and Transactions of the Liverpool Biological 410
Society 54, 8.
411 412
Balon, E. K. (1975). Terminology of intervals in fish development. Journal of the 413
Fisheries Research Board of Canada 32, 1663-1670.
414 415
Blaxter, J. H. S. (1971). Feeding and condition of Clyde herring larvae. Rapports et 416
Proces-Verbaux des Réunions Conseil International pour L'exploration de la Mer 417
160, 128-136.
418 419
Brander, K. (1994). The location and timing of cod spawning around the British Isles.
420
ICES Journal of Marine Science 51, 71-89.
421
422
Bremigan, M. T. & Stein, R. R. (1994). Gape-dependant larval foraging and 423
zooplankton size: Implications for fish recruitment across systems. Canadian Journal 424
of Fisheries and Aquatic Sciences 51, 913-922.
425 426
Copp, G. H. & Kovác, V. (1996). When do fish with indirect development become 427
juveniles? Canadian Journal of Fisheries and Aquatic Sciences 53, 746-752.
428 429
Economou, A. N. (1991). Food and feeding ecology of five gadoid larvae in the 430
northern North Sea. Journal du Conseil International pour l'Exploration de la Mer 431
47, 339-351.
432 433
Fox, C. J., Dickey-Collas, M. & Winpenny, A. J. (1997). Spring planktonic surveys of 434
the Irish Sea in 1995: The distribution of fish eggs and larvae. Science Series 435
Technical Reports, CEFAS Lowestoft 104, 148.
436 437
Fuiman, L. A. & Higgs, D. M. (1997). Early life history and recruitment in fish 438
populations. In Fish and fisheries series: Ontogeny, growth and the recruitment 439
process. (Chambers, R. C. & Trippel, E. A., eds). 21, pp. 596. London: Chapman and 440
Hall.
441 442
González-Quirós, R. & Anadón, R. (2001). Diet variability in the larval blue whiting 443
as a response to plankton size structure. Journal of Fish Biology. 59, 1111-1125. doi:
444
10.1006/jfbi.2001.1724.
445 446
Guma'a, S. A. (1978). The food and feeding habits of young perch, Perca fluviatilis, 447
in Windermere. Freshwater Biology 8, 177-187.
448 449
Hunter, J. R. (1980). The feeding behavior and ecology of marine fish larvae: Fish 450
behavior and its use in the capture and culture of fishes. ICLARM Conference 451
Proceedings, International Center for Living Aquatic Resources Management, 452
Manila, Philippines. pp: 287-326.
453 454
Kane, J. (1984). The feeding habits of co-occurring cod and haddock larvae from 455
Georges Bank. Marine Ecology Progress Series 16, 9-20.
456 457
Kerr, S. R. (1971a). Analysis of laboratory experiments on growth efficiency of 458
fishes. Journal of the Fisheries Research Board of Canada 28, 801-808.
459 460
Kerr, S. R. (1971b). Prediction of fish growth efficiency in nature. Journal of the 461
Fisheries Research Board of Canada 28, 809-814.
462 463
Kjørsvik, E., van de Meeren, T., Kryvi, H. J. A. & Kvenseth, P. G. (1991). Early 464
development of the digestive tract of cod larvae Gadus morhua L., during start- 465
feeding and starvation. Journal of Fish Biology 38, 1-15.
466 467
Last, J. M. (1978). The food of three species of gadoid larvae in the eastern English 468
Channel and southern North Sea. Marine Biology 48, 377-386.
469 470
Mandali de Figueiredo, G. (2003). Trophodynamics of the planktonic community in 471
the coastal areas of the central Irish Sea, with emphasis on fish larvae and their prey.
472
PhD Thesis, Liverpool. pp: 61-101.
473 474
Munk, P. (1997). Prey size spectra and prey availability of larval and small juvenile 475
cod. Journal of Fish Biology 51 (Supplement A), 340-351.
476 477
Nichols, J. H., Haynes, G. M., Fox, C. J., Milligan, S. P., Brander, K. M. & Chapman, 478
R.J. (1993). Spring plankton surveys of the Irish Sea in 1982, 1985, 1987, 1988 and 479
1989: hydrography and the distribution of fish eggs and larvae. pp: 111. Lowestoft:
480
Fisheries Research Technical Report. MAFF, Directorate of Fisheries Research.
481 482
Osse, J. W. M. & van den Boogart, J. G. M. (1995). Fish larvae, development, 483
allometric growth, and the aquatic environment. ICES Marine Science Symposia 201, 484
21-34.
485 486
Pearre, S., Jr. (1986). Ratio-based trophic niche breadths of fish, the Sheldon 487
spectrum, and the size-efficiency hypothesis. Marine Ecology Progress Series 27, 488
299-314.
489 490
Peck, M. A., Clemmesen, C. & Herrmann, J. -P. (2005). Ontogenic changes in the 491
allometric scaling of the mass and length relationship in Sprattus sprattus. Journal of 492
Fish Biology 66, 882-887. doi: 10.1111/j.1095-8649.2005.00651.x.
493 494
Pepin, P. & Penney, R. W. (1997). Patterns of prey size and taxonomic composition in 495
larval fish: Are there general size-dependent models? Journal of Fish Biology 51 496
(Supplement A), 84-100.
497 498
Pepin, P. & Penney, R. W. (2000). Feeding by a larval fish community: impact on 499
zooplankton. Marine Ecology Progress Series 204, 199-212.
500 501
Peters, R. H. (1983). The ecological implications of body size. pp. 10-23. Cambridge:
502
Cambridge University Press.
503 504
Porter, S. M., Brown, A. L. & Bailey, K. M. (2001). Estimating live standard length 505
of net caught walleye Pollock (Theragra chalcogramma) larvae using measurements 506
in addition to standard length. Fishery Bulletin 99, 691-696.
507 508
Robb, A. P. & Hislop, J. R. G. (1980). The food of five gadoid species during the 509
pelagic O-group phase in the northern North Sea. Journal of Fish Biology 16, 199- 510
217.
511 512
Roe, H. S. J. & Shale, D. M. (1979). A new multiple rectangular midwater trawl 513
(RMT 1+8M) and some modifications to the Institute of Oceanographic Sciences' 514
RMT 1+8. Marine Biology 50, 283-288.
515 516
Sabatés, A. & Saiz, E. (2000). Intra- and interspecific variability in prey size and 517
niche breadth of myctophiform fish larvae. Marine Ecology Progress Series 201, 261- 518
271.
519
520
Schael, D. M., Rudstam, L. G. & Post, J. R. (1991). Gape limitation and prey 521
selection in yellow perch (Perca flavescens), freshwater drum (Aplodinotus 522
grunniens), and black crappie (Pomoxis nigromaculatus). Canadian Journal of 523
Fisheries and Aquatic Sciences 48, 1919-1925.
524 525
Shirota, A. (1970). Studies on the mouth sizes of fish larvae. Bulletin of the Japanese 526
Society of Scientific Fisheries 36, 353-368.
527 528
Shirota, A. (1978). Studies on the mouth size of fish larvae - II. Specific 529
characteristics of the upper jaw length. Bulletin of the Japanese Society of Scientific 530
Fisheries 44, 1171-1177.
531 532
Timmermans, L. P. M. (1987). Early development and differentiation in fish. Sarsia 533
72, 329-422.
534 535
van Snik, G. M. J., van den Boogart, J. G. M. & Osse, J. W. M. (1997). Larval growth 536
patterns in Cyprinus carpio and Clarias gariepinus with attention to the finfold.
537
Journal of Fish Biology 50, 1339-1352.
538 539
Weikert, H. & John, H. -C. (1981). Experiences with a modified Bé multiple opening- 540
closing plankton net. Journal of Plankton Research 3, 167-177.
541 542
Wheeler, A. (1969). The fishes of the British Isles and North-West Europe. pp. 266- 543
278. London: MacMillan.
544
545
Wiebe, P. H., Burt, K. H., Boyd, S. H. & Morton, A. W. (1976). A multiple 546
opening/closing net and environmental sensing system for sampling zooplankton.
547
Journal of Marine Research 34, 313-326.
548 549
Wong, B. & Ward, F. J. (1972). Size selection of Daphnia pulicaria by yellow perch 550
(Perca flavescens) fry in West Blue Lake, Manitoba. Journal of the Fisheries 551
Research Board of Canada 29, 1761-1764.
552 553 554 555 556 557
TABLES
TABLE I. Developmental features and assigned scores for the determination of a developmental index for larval fish.
Developmental features based on observations from Balon (1975), Timmermans (1987), Kjørsvik et al. (1991) and Osse & van den Boogaart (1995), external features represented above the dashed line and internal features below.
TABLE II. Regression analysis of cod, haddock and whiting for i) natural log transformed gape size (SG) versus standard length (LS) data; ii) natural log transformed upper jaw length (LUJ) versus standard length (LS) data; iii) natural log transformed gape size (SG) versus upper jaw length (LUJ) data when carried out as a single and segmented regression.
Score Developmental Feature 1 Yolk-sac larvae with mouth closed 2 Yolk-sac larvae with mouth open 3 Exogenous feeding larvae – finfold complete 4 Exogenous feeding larvae – finrays visible 5 Exogenous feeding larvae – all fins present 6 Teeth not visible in mouth
7 Teeth visible in mouth 8 Incipient gut 9 Gut looped but stomach not visible
10 Stomach visible 11 Pyloric caeca visible Table
Species Regression type r2 F d.f. Regression equation
i) Cod Single 0.94 2377 140 ln(SG) = 1.019ln(LS)-2.281
Segmented 0.98 5350 138 ln(SG) = 1.376ln(LS)-2.938 Pre-inflection ln(SG) = 0.678ln(LS)-1.220 Post-inflection Haddock Single 0.94 3752 257 ln(SG) = 0.923ln(LS)-2.054
Segmented 0.98 13486 255 ln(SG) = 1.500ln(LS)-3.128 Pre-inflection ln(SG) = 0.654ln(LS)-1.164 Post-inflection Whiting Single 0.97 15610 552 ln(SG) = 0.924ln(LS)-2.005
Segmented 0.98 30285 550 ln(SG) = 1.331ln(LS)-2.824 Pre-inflection ln(SG) = 0.803ln(LS)-1.616 Post-inflection ii) Cod Single 0.96 3708 140 ln(SG) = 0.961ln(LUJ)+0.003
Segmented 0.97 5265 138 ln(SG) = 1.134ln(LUJ)+0.085 Pre-inflection ln(SG) = 0.773ln(LUJ)+0.160 Post-inflection Haddock Single 0.98 11080 261 ln(SG) = 0.911ln(LUJ)+0.032
Segmented 0.99 19512 259 ln(SG) = 1.100ln(LUJ)+0.043 Pre-inflection ln(SG) = 0.758ln(LUJ)+0.136 Post-inflection Whiting Single 0.98 21710 554 ln(SG) = 0.893ln(LUJ)+0.011
Segmented 0.98 27711 552 ln(SG) = 1.063ln(LUJ)+0.039 Pre-inflection ln(SG) = 0.807ln(LUJ)+0.103 Post-inflection iii) Cod Single 0.97 6705 187 ln(LUJ) = 1.070ln(LS)-2.410
Segmented 0.97 9163 185 ln(LUJ) = 1.198ln(LS)-2.628 Pre-inflection ln(LUJ) = 0.949ln(LS)-2.042 Post-inflection Haddock Single 0.97 11920 316 ln(LUJ) = 1.038ln(LS)-2.302
Segmented 0.99 31284 314 ln(LUJ) = 1.325ln(LS)-2.810 Pre-inflection ln(LUJ) = 0.872ln(LS)-1.751 Post-inflection Whiting Single 0.98 37740 628 ln(LUJ) = 1.040ln(LS)-2.277
Segmented 0.98 62172 626 ln(LUJ) = 1.221ln(LS)-2.627 Pre-inflection ln(LUJ) = 0.991ln(LS)-2.123 Post-inflection
The regression equation, coefficient of determination (r2), and associated ANOVA with F statistic (F) and degrees of freedom (d.f.) are given. All significance was at the p<<0.001 level.
TABLE III. Associated variables used for the calculation of larval gape size, when using i) standard length, ii) upper jaw length and iii) lower jaw length measurements.
variable Species Size range (mm)
a b
i) Cod <10.56 1.376 5.297 x 10-2 >10.56 0.768 2.952 x 10-1 Haddock <10.56 1.500 4.381 x 10-2 >10.56 0.654 3.122 x 10-1 Whiting <10.56 1.331 5.937 x 10-2 >10.56 0.803 1.987 x 10-1 ii) Cod <1.27 1.134 1.089
>1.27 0.773 1.174
Haddock <1.27 1.100 1.044
>1.27 0.758 1.146
Whiting <1.27 1.063 1.040
>1.27 0.807 1.109
iii) Cod <1.27 1.030 1.207
>1.27 0.672 1.439
Haddock <1.27 1.050 1.095
>1.27 0.782 1.246
Whiting <1.27 1.010 1.078
>1.27 0.804 1.213
1
FIGURE CAPTIONS 1
2
FIG. 1. Sample area in the western Irish Sea, for the collection of larval and juvenile cod, 3
haddock and whiting during April and May of 2001.
4 5
FIG. 2. Natural log plot of gape size and standard length for cod. A plot of the residuals 6
from a linear regression are shown in Insert A, and those for a segmented regression shown 7
in Insert B 8
9
FIG. 3. Natural log transformed standard length and gape size for a) cod, b) haddock and c) 10
whiting. The inflection point position at standard length (solid vertical line) ± 95% CI’s 11
(dashed vertical line) for cod (2.46 ± 0.16), haddock (2.32 ± 0.06) and whiting (2.29 ± 12
0.05) is shown.
13 14
FIG. 4. Natural log transformed standard length and upper jaw length for a) cod, b) 15
haddock and c) whiting. The inflection point position at standard length (solid vertical line) 16
± 95% CI’s (dashed vertical line) for cod (2.34 ± 0.29), haddock (2.34 ± 0.11) and whiting 17
(2.19 ± 0.10) is shown.
18 19
FIG. 5. Natural log transformed upper jaw length and gape size for a) cod, b) haddock and 20
c) whiting. The inflection point position at upper jaw length (solid vertical line) ± 95%
21
CI’s (dashed vertical line) for cod (0.21 ± 0.21), haddock (0.27 ± 0.12) and whiting (0.25 ± 22
0.12) is shown.
23 24 Figure Captions
2
FIG. 6. Relationship between developmental index and natural log transformed standard 25
length data for a) cod, b) haddock and c) whiting. Inflection points at standard length for 26
the change of larval stage are shown as solid vertical lines, ± 95% CI’s shown as dashed 27
vertical lines. Shaded area equates to the region of “intermediate” stage (circle symbols) 28
after the larval stage (cross symbols) and before the juvenile stage (triangular symbols).
29 30 31
Figure
Figure
Figure
Figure
Figure
Figure