ICES Journal of
Marine Science
ICES Journal of Marine Science (2016), 73(2), 316– 328. doi:10.1093/icesjms/fsv215
Original Article
A comparison of cod life-history parameters inside and outside
of four year-round groundfish closed areas in New England, USA
Graham D. Sherwood 1* and Jonathan H. Grabowski 2
1
Gulf of Maine Research Institute, 350 Commercial Street, Portland, ME 04101, USA
Marine Science Center, Northeastern University, 430 Nahant Road, Nahant, MA 01908, USA
2
*Corresponding author: tel: +1 207 2281644; fax: +1 207 7726855; e-mail: [email protected]
Sherwood, G. D., and Grabowski, J. H. A comparison of cod life-history parameters inside and outside of four year-round
groundfish closed areas in New England, USA. – ICES Journal of Marine Science, 73: 316–328.
Received 19 March 2015; revised 19 October 2015; accepted 1 November 2015; advance access publication 27 November 2015.
From 1994 to 2002, five major year-round closed areas were established in the Gulf of Maine and on Georges Bank to promote recovery of groundfish
species, including Atlantic cod (Gadus morhua). Here, we present life-history data for cod sampled within and next to four of the five closed areas to
test the hypothesis that closed areas benefit cod. We found a positive effect of closure status on cod age, length and growth for three of the four
closed areas; contrary results for the fourth area (Jeffreys Ledge) may be due to recreational fishing pressure. Diet results were not consistent across
closed areas, but for the two areas farthest from shore (Cashes Ledge and Closed Area II), cod tended to have higher gut fullness and higher trophic
niche breadth (based on stable isotopes) inside vs. outside. Body shape analysis revealed a consistent effect of closure status on cod morphometry
with cod inside closed areas exhibiting less streamlined bodies. We discuss this apparent selection for more sedentary cod with lower productivity
potential but highlight the demographic result (i.e. higher age and greater size inside closed areas) as critical considerations for resource managers
that are designing or altering the configuration and/or scale of closed areas to protect and rebuild demersal fish species.
Keywords: age, Atlantic cod, closed areas, diet, Georges Bank, growth, Gulf of Maine, life-history variation, morphometrics, stable isotope.
Introduction
Gulf of Maine (GOM) and Georges Bank (GB) cod (Gadus morhua)
have been a mainstay for New England fishing communities for centuries with early European explorers and then settlers targeting this
species for both food and export (Kurlansky, 1997). In fact, human
exploitation of cod in this region predates the arrival of Europeans
with archaeological evidence of indigenous peoples fishing cod
going back millennia (Bourque et al., 2008). Over more recent
times (i.e. the last few decades), New England cod have undergone
major changes in both abundance and management. Following establishment of a US exclusive economic zone (EEZ) in 1976, cod
landings were quite high, although apparently not as high as in
the 1800s (Alexander et al., 2009). However, throughout the 1980s
and into the 1990s, cod fishing mortality was excessive, as evidenced
by a decline in the stock that clearly manifested by the mid-1990s,
which led to stricter effort controls taking effect in the fishery.
Among these input controls were limited days at sea, gear restrictions (e.g. minimum mesh size), and area closures.
Starting in 1994, five year-round closed areas were implemented
on GB and in the GOM (Figure 1) to replace or augment a series of
rolling closures which had been in effect in various forms since the
1970s (NEFMC, 2014). The year-round closures include Closed
Areas I and II (CAI and CAII) on GB and the Nantucket Lightship
Area (NLA) on Nantucket Shoals, which were established in their
current form in 1994, the Western GOM Closure Area or Jeffrey’s
Ledge (JL; also spans Stellwagen and Tillies Banks) which was
added in 1998, and the Cashes Ledge Closed Area (CL) in the
central GOM which was closed in 2002 (NEFMC, 2014). These
closed areas are not off-limits to all fishing activities, as is the case
for other marine protected areas (MPAs) around the world (e.g.
Edgar et al., 2007). Rather, they were intended to protect benthic
habitats and reduce groundfish mortality particularly for cod,
haddock, Melanogrammus aeglefinus, and yellowtail flounder,
Pleuronectes ferruginea. Certain commercial groundfishing technologies are excluded (i.e. primarily commercial bottom tending
gear such as trawling and gillnetting), while other activities are
# International Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
Comparison of cod life-history parameters inside and outside of four year-round groundfish
317
Figure 1. Map of GOM and GB showing location of four major closed areas and general sampling areas. CAI, Closed Area I; CAII, Closed Area II;
CL, Cashes Ledge; JL, Jeffrey’s Ledge. Bathymetry lines are 50 m contours. Ellipses show general area of sampling.
allowed, either throughout the areas (e.g. recreational fishing) or in
special access areas (e.g. scallop dredging and long-lining). As of
2015, the total amount of bottom currently off-limits to bottomtending gear, year-round, in the GOM and on GB is about
22 000 km2 (Murawski et al., 2005) or roughly the size of New
Hampshire.
In 2010, the management of groundfish in the US northeast
shifted from input controls to output controls, or quota management. At the time of this change, it was believed that New England
cod stocks were rebuilding with overfishing no longer occurring
(Mayo et al., 2009). In addition, since hard quotas were being
imposed with built-in uncertainty buffers, the continued necessity
of closed areas was questioned. Some stakeholders hold that
closed areas, for the purpose of rebuilding groundfish in New
England, have become obsolete. However, at the end of 2011, it
became apparent that the size of the GOM cod stock had been overestimated by the 2008 assessment (NEFSC, 2012), and that rebuilding was no longer attainable in the near term. The situation is not
much better for GB cod (NEFSC, 2013). As opposed to calls for
maintaining the current closed areas during this period of crisis,
there has been a growing pressure from the fishing industry to
reopen areas closed to groundfishing. This may be to lessen the financial burden of a 78% reduction (effective 2013) in quota for
GOM cod, but also because the process of reconfiguring closed
areas for different purposes other than groundfish rebuilding (e.g.
to protect benthic biodiversity) had already begun (NEFMC,
2014). More recently, an interim assessment for GOM cod determined that GOM cod spawning-stock biomass (SSB) is now at
318
just 3 –4% of the target level (Palmer, 2014), which has triggered a
shutdown of the groundfish fishery in most of the GOM (effective
November 2014). Over the backdrop of a deepening cod crisis in
the GOM/GB region and an ongoing discussion at the New
England Fishery Management Council level to alter the existing
boundaries of the closed areas in this region, it is essential that all evidence for their impact on cod be considered.
There is no consensus among managers and industry members
that closed areas have achieved the goal of reducing cod fishing mortality. However, there is some evidence that there have been ecological benefits that may translate into advantages for cod. For
example, studies have shown that benthic communities are more
diverse inside vs. outside closed areas (Collie et al., 1997; Fisher
and Frank, 2002; Hermsen et al., 2003). This may provide more
complete feeding opportunities for demersal fish like cod.
Murawski et al. (2005) examined catch patterns next to closed
areas for a variety of groundfish species and found that cod were
not the species most associated with closed areas (haddock was).
However, this result may not represent all cod. Sherwood and
Grabowski (2010) examined the life history of red cod at CL and
concluded that this colour morphotype represents a highly sedentary form of cod that may reside entirely within the closed area, primarily at Ammen Rock. Similarly, Lindholm et al. (2007) studied
residency behaviour of cod on Stellwagen Bank and found that a
proportion of acoustically tagged individuals were highly and consistently associated with specific bottom features. More broadly,
there is growing evidence throughout the North Atlantic that most
cod stocks and populations are made up of resident and migratory
components (Svedäng et al., 2007; Bergstad et al., 2008; Grabowski
et al., 2011; LeBris et al., 2013). Thus, while there may not be evidence for cod using closed areas more than open areas as a whole
in the GOM, there is a possibility that closed areas are favouring
one type of life history (i.e. resident) over another (i.e. migrant),
and that this difference has yet to be recognized.
The purpose of this study was to explore life-history parameters
of cod inside and outside of four year-round closed areas, all of
which are currently being considered for reconfiguration and/or
reopening (NEFMC, 2014). Specifically, we hypothesized that cod
growth, feeding success, condition and survival are higher inside
of closed areas than outside. We also considered the possibility
that closed areas may be promoting the recovery, or at the very
least, persistence of individuals with more sedentary behaviour
over those with migratory tendencies. We inferred migratory behaviour using body shape analysis (Sherwood and Grabowski, 2010; see
Morphometrics section for further explanation).
Methods
Sampling
Cod were collected by rod and reel and longline from inside and next
to four of the five year-round closed areas in the GOM and on GB
(Figure 1). Cod from “inside” closed areas came mostly from
areas well inside (.5 km), whereas cod from “outside” closed
areas were captured at distances .5 km from closed areas.
Immediately following capture, cod were measured (total length,
Ltot, nearest mm) and photographed for morphometric analysis
and then stored on ice for transport back to the lab where they
were kept frozen until they were dissected. Sampling occurred primarily from late spring to early autumn 2007–2009.
In the lab, cod were thawed and weighed (total weight, Wtot,
nearest g). The livers, stomachs, and gonads were removed and
G. D. Sherwood and J. H. Grabowski
weighed (Wliv, Wstom, Wgon, respectively, nearest g). All other
viscera were removed and gutted weight was recorded (Wgut,
nearest g). Sagittal otoliths were removed for age and growth analyses, and sex was recorded. Age was determined from cracked and
polished otoliths using an image analysis system. An 1 g muscle
tissue sample located directly anterior to the first dorsal fin was
obtained from each cod for stable isotope analysis.
Stomach content and stable isotope analyses
For each cod, individual dietary items were identified to species
(where possible), counted and measured (nearest mm), excess
water was removed, and items were weighed (nearest mg). To simplify analyses and consider only general trends in feeding, prey
items were partitioned into the following three broad prey groups:
benthos (various species of polychaetes, amphipods, molluscs, brittlestars, echinoderms, shrimp, lobsters, and crabs), pelagic fish
(Atlantic herring, Clupea harengus, redfish, Sebastes faciatus, silver
hake, Merluccius bilinearis, and pollock, Pollachius virens), and
demersal fish (haddock, sculpins, Myoxocephalus sp., cunner,
Tautogolabrus adspersus, and cod). Fish that were not identified to
species were assumed to be pelagic fish. Partial fullness index
(PFI) of the different prey groups was calculated for each cod to
remove the influence of length on stomach fullness (Bowering and
Lilly, 1992; Sherwood et al., 2007). PFI was calculated by dividing
the total weight (g) of prey group i in each cod by the total length
(Ltot, cm) of that fish cubed, and multiplying this proportion by
104. Total fullness index (TFI) was calculated in a similar manner
by substituting the weight of a prey group with the total weight of
all prey in the stomach (g). Finally, feeding intensity (FI) was calculated as the percentage of cod within each area (both inside and
outside) that had food in their stomachs (i.e. prey weight greater
than zero).
Stable isotope ratios of carbon (d13C) and nitrogen (d15N) in fish
muscle tissue are indicative of prey origin (pelagic vs. benthic) and
trophic level, respectively, and provide a time-averaged depiction of
diet (Sherwood and Rose, 2005). In previous studies, it was necessary to characterize the baseline for both d13C and d15N using
primary consumer stable isotope values to infer diet origin and
trophic level of higher consumers independent of baseline variations (e.g. Sherwood and Rose, 2005). Primary consumer stable
isotope values were not available for this study. However, given
that we were primarily interested in the variability of stable
isotope signatures (see Results) and that we employed a paired
design such that major baseline variation is less likely within a
pairing (i.e. inside vs. outside a closed area), we are confident in
the utility of our stable isotope results. Each muscle tissue sample
was dried in a drying oven at 608C for 48 h to constant weight,
homogenized to a fine powder using a mortar and pestle, placed
into 4 × 6 mm tin capsules, and weighed. Samples were then sent
to the Colorado Plateau Stable Isotope Laboratory (Northern
Arizona University, Flagstaff, AZ, USA) for analysis. Samples were
combusted to analyse the carbon and nitrogen stable isotope
ratios of CO2 and N2, respectively, using an elemental analyser followed by gas chromatograph separation interfaced via continuous
flow to an isotope ratio mass spectrometer. Stable carbon and nitrogen ratios (d) in this study are defined as the parts per thousand
deviations from the following standard materials: Pee Dee belemnite
limestone for d13C and N2 in air for d15N. Ten per cent of the samples
were analysed in duplicate; the average coefficient of variation for
duplicates was 0.0097 for d13C and 0.0135 for d15N.
Comparison of cod life-history parameters inside and outside of four year-round groundfish
Condition indices
Two indices of condition were considered: Fulton’s condition factor
(FCF) and liver somatic index (LSI). FCF was calculated as Wgut ×
21
L23
tot × 100 and LSI as Wliv × Wtot × 100. FCF is primarily an indicator of protein reserves (i.e. muscle mass), whereas LSI indicates
excess lipid stores in cod (Yaragina and Marshall, 2000). To
remove any seasonal variability in these indices, we calculated seasonally adjusted values. For example (for FCF), FCFadj ¼ FCF 2
FCFavg, where FCFavg is the average condition factor (all sizes and
areas combined) by month. FCF varied significantly by month
(analysis of variance (ANOVA), F5,626 ¼ 49.0, p , 0.0001). The
same calculation was performed for LSI, which also varied significantly by month (ANOVA, F5,564 ¼ 38.4, p , 0.0001). Both
FCFadj and LSIadj were significantly related to length: FCFadj ¼
0.029–0.00005 × L; r 2 ¼ 0.01; p , 0.05: LSIadj ¼2 1.73 + 0.003 ×
L; r 2 ¼ 0.07; p , 0.0001. The relationship for FCFadj vs. length was
deemed biologically insignificant given the low r 2 value and
length coefficient. On the other hand, the relationship between
LSIadj and length was strong enough to warrant correcting for
length before comparing among areas and status. Thus, the residuals
of this relationship were used for further testing. Gonadal somatic
index was not considered since very few cod were in spawning
condition.
Morphometrics
Body shape in fish is an indicator of swimming activity among
species (Sherwood and Rose, 2005). Within populations, body
shape may also vary according to migratory propensity such that
more migrant individuals have more streamlined body shapes
than more resident individuals (Morinville and Rasmussen, 2008;
Sherwood and Grabowski, 2010). We used body shape analysis
here to infer migrant vs. resident behaviour in cod. Cod morphometric analyses were conducted using the images collected at sea following the procedure outlined in Sherwood and Grabowski (2010).
In short, 11 homologous cod landmarks were digitized and the distances between landmarks were calculated for each fish using a boxtruss approach (Strauss and Bookstein, 1982; Cadrin, 2000). We
purposely excluded a landmark at the anterior margin of the first
anal fin to remove the possibility that variations in stomach fullness
would affect body shape results. Each of the resultant 17 distances
(Figure 5) was natural log-transformed to normalize distributions
and principal component analysis was then conducted on those
values. The first PC, which is an indicator of fish length (Cadrin,
2000), was used to standardize for differences in fish size by regressing it against each log-transformed box-truss dimension. Using the
residuals of each regression, discriminant function analysis (DFA)
was then performed to determine the percentages of cod that
could be categorized correctly by status (Solow, 1990). Loading
factors on the first discriminant function were also compared
among the status by Student’s t-test. Red cod which are a highly resident form of cod found in shallow habitats including at CL
(Sherwood and Grabowski, 2010) were excluded from all analyses.
Statistical analysis
Comparisons of life-history parameters were made within (inside
vs. outside) and among the four major closed areas. Hereafter, different closed areas are referred to as “area” and inside vs. outside is
referred to as “status”, both fixed variables. Age was compared
among area and status using ANOVA, whereby the effect of both
variables and their interaction on mean age was examined.
319
A Mann–Whitney test was also used to test for differences in
median age among status by area, and a chi-squared test was
employed to examine proportions of older cod (.5 years old, arbitrary) captured inside vs. outside of closed areas for all areas combined. The variation in cod length among areas and status and
their interaction was also examined by ANOVA and by Student’s
t-test (within areas). Length–weight relationships (LWRs) were
constructed using the formula: log W ¼ a + b × log L, where a is
the intercept and b is the slope or scaling coefficient. Analysis of
covariance (ANCOVA) with log W as the dependent variable,
status as a fixed factor (coded “0” for inside and “1” for outside),
log L as the covariate, and an interaction term between status and
log L were used to assess the difference in slopes among status and
within areas. von Bertalanffy growth function (VBGF) curves were
generated for each area– status combination (Chen et al., 1992).
Significant differences among these curves inside vs. outside for
each area separately and for all areas combined were tested by the
analysis of the residual sum of squares (ARSS, Chen et al., 1992).
Differences in FI among status were examined using chi-squared
tests (i.e. using the number of empty stomachs vs. number with
food) for all areas combined and by area. Total fullness index
(TFI) was compared among areas and status using ANOVA and
by status using Student’s t-tests for each area separately. Only TFI
values greater than zero were used. Before conducting these tests,
TFI was cube-root transformed to reduce heteroscedacity. This
analysis was repeated for each prey group using cube-root transformed PFI values. The relationship between prey size and predator
size (i.e. cod length) was explored by ANCOVA for all areas combined and within each area with status as a fixed factor (coding as
above) and length as the covariate; the interaction between length
and status was also examined. Given the dependency on length, residual prey size values were calculated and compared among status
by Student’s t-test for all areas combined and by area. Differences in
the variance of stable isotope values (d13C and d15N) inside and
outside of the closed areas for all areas combined were assessed
using Levene’s test. There was a significant effect of length on
d15N signatures (r 2 ¼ 0.34; p , 0.0001; N ¼ 310). Therefore, residual d15N values were compared among area and status by
ANOVA and within area by Student’s t-test. The same analyses
were performed for d13C, which was not related to length. Finally,
differences in FCFadj and residual LSIadj among areas and status
were examined using ANOVA and within areas using Student’s
t-test.
Results
Table 1is a summary of all results across all closed areas and all lifehistory parameters. The table shows mean values (+1 SE), sample
sizes, trends for differences between life-history parameters inside
vs. outside of closed areas (shading), as well as significant differences
(asterisks) where they existed. The results shown in the table are
further expanded below for each life-history parameter in the
order that they appear.
Both area and status influenced cod age distributions (Figure 2a).
Age varied significantly among areas (ANOVA: F3,679 ¼ 34.1;
p , 0.0001) and status (ANOVA: F1,679 ¼ 6.4; p , 0.05) and the
interaction between these two variables was marginally significant
(ANOVA: F3,679 ¼ 2.3; p ¼ 0.07). Median age was consistently
1 year older for cod captured inside closed areas compared with
outside for all areas, but this difference was significant for CAI
only (Mann –Whitney test: Z120 ¼ 2 3.92; p , 0.0001). Overall,
more old cod (age .5 years) were captured in closed areas (n ¼
320
Table 1. Summary of results for 11 life-history variables in cod sampled inside and outside of four major closed areas.
In
Life-history parameter
Mean age (years)
Mean length (mm)
Scaling coefficient (b)a
Assymptotic length (mm)
TFI (g cm23)b
Mean residual prey size
(mm)c
Mean residual d15N (‰)d
FI (%)e
Mean body conditionf
Mean liver conditionf
Mean body shapeg
Out
CAI
2.87 (0.15, 54)***
607.1 (14.9, 54)***
3.03 (0.06, 51)**
742.0 (37.1, 44)***
0.91 (0.03, 47)
20.05 (0.03, 93)
CAII
4.63 (0.18, 118)
692.9 (13.4, 118)
2.92 (0.04, 115)*
896.8 (34.0, 110)*
0.94 (0.02, 102)
20.01 (0.02, 283)
CL
3.05 (0.09, 217)
517.8 (7.8, 217)*
3.05 (0.04, 209)***
705.6 (28.7, 161)**
0.95 (0.02, 155)**
20.01 (0.03, 248)
JL
2.76 (0.09, 131)
473.1 (7.4, 131)
2.98 (0.08, 127)
565.7 (26.6, 109)
0.83 (0.02, 93)
0.02 (0.03, 186)
All
3.32 (0.07, 520)***
555.6 (6.2, 520)***
2.92 (0.02, 505)
838.4 (25.7, 424)***
0.91 (0.01, 397)*
20.01 (0.01, 810)
CAI
2.18 (0.07, 113)
466.6 (8.7, 113)
2.94 (0.06, 110)
648.0 (78.4, 76)
0.87 (0.02, 93)
0.07 (0.04, 73)*
CAII
4.18 (0.24, 22)
690.2 (15.9, 22)
2.74 (0.13, 21)
753.2 (49.0, 21)
0.88 (0.05, 20)
0.08 (0.06, 43)
CL
2.84 (0.12, 114)
485.1 (9.4, 114)
2.94 (0.06, 112)
610.2 (29.6, 107)
0.86 (0.02, 92)
0.01 (0.02, 193)
JL
3.07 (0.14, 67)*
520.0 (15.7, 67)**
3.38 (0.06, 65)
767.8 (182.5, 52)**
0.93 (0.03, 57)*
20.02 (0.01, 270)
All
2.75 (0.07, 316)
500.2 (6.6, 316)
2.99 (0.03, 311)
806.7 (21.4, 256)
0.88 (0.01, 262)
0.01 (0.01, 579)
20.36 (0.12, 18)
86.7 (53)
20.02 (0.01, 51)
21.38 (0.17, 50)
20.32 (0.17, 41)
0.23 (0.11, 19)*
86.4 (118)
0.00 (0.01, 117)
0.25 (0.08, 91)
20.02 (0.16, 57)
0.16 (0.06, 80)
77.8 (185)
0.01 (0.01, 193)*
20.44 (0.15, 178)
21.07 (0.11, 45)
20.15 (0.04, 80)
71.5 (130)
0.00 (0.01, 129)
0.32 (0.15, 121)
20.10 (0.10, 61)
20.01 (0.04, 197)
79.2 (486)
0.00 (0.00, 490)
20.19 (0.08, 440)
20.33 (0.07, 204)
0.01 (0.08, 16)*
83.7 (111)
0.03 (0.01, 110)**
0.80 (0.13, 108)***
0.76 (0.16, 40)***
20.16 (0.11, 20)
90.9 (22)
20.02 (0.01, 21)
0.20 (0.13, 19)
0.87 (0.19, 21)**
0.08 (0.06, 42)
80.0 (90)
20.01 (0.01, 89)
20.37 (0.17, 74)
0.21 (0.13, 45)***
0.01 (0.07, 36)
84.6 (65)*
0.02 (0.01, 67)
20.13 (0.20, 64)
0.22 (0.12, 44)*
0.01 (0.04, 114)
83.3 (288)
0.01 (0.00, 288)
0.20 (0.09, 265)**
0.45 (0.08, 150)***
G. D. Sherwood and J. H. Grabowski
Standard error and sample size are indicated in parentheses. Significant results (in vs. out by area and all areas combined) are indicated by asterisks and in bold. Shaded cells indicate values that were greater for each
comparison (significant or not).
a
From LWR (logW ¼ a + b.logL).
b
Power transformed (TFI0.16).
c
Residuals from log prey size (mm) vs. cod length (mm).
d
Residuals from d15N vs. cod length (mm).
e
Percentage of stomachs with food present (significance test: chi-squared).
f
Seasonally adjusted (see Methods).
g
Mean discriminant function 1 score from morphometric analysis; more positive values indicate more streamlined body form.
Significance levels: ***p , 0.0001; **p , 0.01; *p , 0.05.
Comparison of cod life-history parameters inside and outside of four year-round groundfish
47) compared with open areas (n ¼ 5; chi-squared test:
x21,680 = 18.8; p , 0.0001) and this difference was most pronounced for CAII and CL (Figure 2). Length also varied significantly
among areas (ANOVA: F3,679 ¼ 66.1; p , 0.0001) and status
(ANOVA: F1,679 ¼ 14.7; p , 0.0001) and the interaction between
these two terms was significant (ANOVA: F3,679 ¼ 10.2; p ,
0.0001). In this case, the mean length was significantly higher
inside than that outside for CAI (Student’s t-test: t120 ¼ 6.9; p ,
0.0001) and CL (Student’s t-test: t268 ¼ 2.6; p , 0.01). Although
not significant, the trend for the other two areas was higher mean
length inside CAII and higher mean length outside JL. Figure 2b
shows length distributions and medians.
LWRs were also affected by both area and closure status (Table 2).
Scaling coefficient or b values (i.e. the slope from equation: log
Figure 2. Age-frequency (a) and length-frequency (b) distributions
(bubble size represents percentage within grouping; smallest bubble is
,1% and largest bubble is 65%) for cod sampled in and out of four
major closed areas in the GOM and on GB. Percentages for lengthfrequency distributions are based on 10 cm length bins. Horizontal lines
are medians.
Table 2. LWRs (log W ¼ a + b × log L) for cod captured inside and
outside of 4 major closed areas as well as all areas combined.
Area
CAI
CAI
CAII
CAII
CL
CL
JL
JL
All
All
Status
In
Out
In
Out
In
Out
In
Out
In
Out
a
22.07
21.87
21.89
21.58
21.92
21.91
21.95
22.57
21.79
21.98
SE
0.10
0.08
0.06
0.24
0.05
0.10
0.13
0.09
0.04
0.05
b
3.04
2.93
2.92
2.74
2.96
2.94
2.97
3.34
2.88
2.99
SE
0.06
0.05
0.03
0.13
0.03
0.06
0.08
0.05
0.02
0.03
r2
0.98
0.96
0.98
0.96
0.96
0.95
0.92
0.98
0.96
0.96
N
56
127
122
21
401
113
135
76
717
340
321
W ¼ a + b × log L) tended to be higher inside than those outside
closed areas for CAI, CAII, and CL. However, b values were not significantly different among status within these three areas when the
interaction between the effect of status and log L was considered
(ANCOVA: p . 0.05). For the area JL, b was significantly higher
outside than that inside the closed area (ANCOVA: p , 0.05).
Somewhat counterintuitively, the interaction term was also significant for all data combined (ANCOVA: p , 0.05) such that b values
were higher outside than those inside closed areas. This result was
likely driven by the significant interaction effect for JL only and
the fact that the b value for outside JL was so high (Table 2).
Similar to the age distribution and LWR results, VBGF parameters
varied with area and closure status (Table 3). L1 was higher for cod
inside closed areas compared with outside for all areas except JL.
There was no consistent trend in k values among cod from inside
and outside of closed areas, although k values were generally higher
in the GOM (JL and CL) compared with those on GB (CAI and
CAII); the reverse was true for L1. VBGF curves were significantly different among cod captured inside vs. outside of closed areas for all
areas combined and within areas for CAI, CL, and JL. The difference
among curves was marginally significant for CAII (Table 3). Length at
ages 3 and 5 was higher inside vs. outside for all comparisons except
for JL (both ages) and CAII (age 3) (Table 3).
FI (i.e. per cent stomachs with food present) for all areas combined was marginally higher outside of closed areas than inside
(chi-squared test: x21,680 = 3.52; p ¼ 0.06). This result was driven
primarily by a significant difference in FI between inside and
outside for CL (chi-squared test: x21,268 = 4.71; p , 0.05). For the
cod that had food in their stomachs, TFI (cube root transformed)
was significantly higher inside than outside of closed areas
(Student’s t-test: t523; p , 0.01; all areas combined). Overall, TFI
varied significantly among areas and closure status, and the interaction was significant (Table 4); within areas, TFI was significantly
higher inside vs. outside for CL only (Student’s t-test: t192 ¼ 4.16;
p , 0.0001), while the trend was similar for CAI and CAII, but
not for JL. TFI was not significantly related to length.
PFIbenthos varied significantly among areas and closure status,
and the interaction was significant (Table 4, Figure 3). Within
areas, PFIbenthos was significantly higher inside for CL only
(Student’s t-test: t192 ¼ 4.90; p , 0.0001); the trend was similar
for CAI and CAII but not for JL (Figure 3). PFIpelagic did not vary significantly among areas and closure status (ANOVA), and PFIdemersal
varied only among closure status with higher values inside closed
areas. In this case, PFIdemersal was significantly higher inside for
CAI only (Student’s t-test: t99 ¼ 2.5; p , 0.05). While PFIdemersal
values were universally quite low compared with benthos and
pelagic fish, demersal fish were virtually absent from the diet of
cod outside of closed areas. Thus, while FI was generally higher
outside of closed areas, stomach fullness was generally higher
inside of closed areas where there was a greater representation of demersal fish.
For all areas combined, as well as within most areas, the prey size
was significantly related to cod length, status, and the interaction
between length and status (ANCOVA: see Table 5). In all cases
where the interaction term was significant (i.e. p , 0.05 for CAI,
CL and JL), the nature of the interaction was such that the slope
of the prey size –cod size relationship was higher inside closed
areas vs. outside. And although not significant, the slope of the
prey size– cod size relationship was higher inside than outside
CAII. None of the above results were affected by constraining the
data to cod lengths ,80 cm.
322
G. D. Sherwood and J. H. Grabowski
Table 3. VBGF parameters for cod from various area/status combinations.
Area
CAI
CAI
CAI
CAII
CAII
CAII
CL
CL
CL
JL
JL
JL
All
All
All
L1 (cm)
742.0
648.0
784.5
896.8
753.2
877.4
705.6
610.2
669.4
565.7
767.8
626.2
838.40
695.40
806.70
Status
In
Out
All
In
Out
All
In
Out
All
In
Out
All
In
Out
All
k
0.67
0.62
0.47
0.37
0.68
0.40
0.52
0.64
0.55
0.72
0.38
0.56
0.37
0.5
0.4
SE
37.1
78.4
66.3
34.0
49.0
29.1
28.7
29.6
21.9
26.6
182.5
36.1
25.7
31.6
21.4
SE
0.16
0.25
0.12
0.06
0.50
0.06
0.08
0.15
0.08
0.20
0.31
0.15
0.04
0.09
0.03
r2
0.79
0.34
0.58
0.72
0.41
0.69
0.64
0.46
0.57
0.34
0.57
0.41
0.65
0.5
0.61
N
44
76
120
110
21
131
161
107
268
109
52
161
424
256
680
RSS
120 901
430 558
738 454
633 519
61 013
732 669
844 138
585 879
1 515 869
479 286
345 543
902 530
3 298 596
1 762 379
5 253 140
F
12.89
d.f. denom
114
p
,0.0001
2.29
125
0.08
5.24
262
0.0016
4.87
155
0.0029
8.53
672
,0.0001
L3 (cm)
642.6
547.1
593.0
601.3
655.3
613.1
557.3
520.7
540.8
500.5
522.2
509.5
562.1
540.2
556.3
L5 (cm)
716.0
618.8
709.7
755.8
728.1
758.7
653.2
585.3
626.6
550.2
653.0
588.1
706.6
638.3
691.9
t0 was constrained to zero. Also shown are results from analysis of residual sum of squares (ARSS) and estimated length at ages 3 (L3) and 5 (L5).
RSS is residual sum of squares; L1 is in mm.
Table 4. ANOVA results for the effect of area, closure status and
their interaction on TFI, PFIbenthos, PFIpelagic and PFIdemersal (all cube
root transformed).
TFI
Effect
Area
Status
Area × status
F
3.56
5.3
3.45
p
,0.05
,0.05
,0.05
PFIbenthos
PFIpelagic
PFIdemersal
F
3.37
4.04
5.13
F p
N.S.
N.S.
N.S.
F
N.S.
5.37
N.S.
p
,0.05
,0.05
,0.01
p
,0.05
Denominator degrees of freedom in all cases was 525.
Cod from outside of closed areas (all areas combined) were
observed to occupy a more limited trophic niche space than cod
from inside closed areas. d15N and d13C values inside closed areas
were significantly less variable than outside (Levene’s test: p ,
0.05; Figure 4). Notably missing from outside of closed areas were
cod with high trophic positions (i.e. high d15N; a size effect), as
seen inside CAII and CL, and more pelagic (i.e. more negative)
carbon signatures, as seen inside CL (Figure 4). While some of the
more pelagic signal inside CL may be due to local baseline effects
(see Sherwood and Grabowski 2010), it is nonetheless noteworthy
that cod from inside CL displayed such a broad range in d13C
values (Figure 4), suggesting highly varied feeding opportunities.
Length-standardized d15N (i.e. residuals from d15N vs. L relationship) varied significantly among areas (ANOVA: F3,310 ¼ 5.33;
p , 0.01) and the interaction between area and status was significant (ANOVA: F3,310 ¼ 5.88; p , 0.01). Within areas, residual
d15N was significantly higher inside than outside CAII (Student’s
t-test: p , 0.05) and significantly higher outside than inside JL
(Student’s t-test: p , 0.05). Stable carbon isotope signatures
(d13C) were not related to length but varied significantly among
areas (ANOVA: F3,310 ¼ 7.90; p , 0.0001) and the interaction
between area and status was also significant (ANOVA: F3,310 ¼
8.56; p , 0.0001). Within areas, d13C was significantly higher (i.e.
more benthic) inside than that outside for CAI (Student’s t-test:
t34 ¼ 4.50; p , 0.0001) and significantly higher outside than
inside for CL (Student’s t-test: t122 ¼ 22.73; p , 0.01) and JL
(Student’s t-test: t116 ¼ 25.94; p , 0.0001).
Adjusted FCF (FCFadj) did not vary significantly among areas
nor status but was significantly related to the interaction between
these two variables (ANOVA: F3,625 ¼ 4.49; p , 0.01). Within
closed areas, FCFadj was significantly higher outside than inside
for CAI (Student’s t-test: t114 ¼ 2 2.47; p , 0.05) and higher
inside than outside for CL (Student’s t-test: t221 ¼ 2.01; p , 0.05).
For the other two areas, the trend was for higher FCFadj inside
than that outside for CAII (borderline significant, Student’s t-test:
t130 ¼ 1.84; p ¼ 0.07) and higher FCFadj outside than that inside
for JL (not significant). Adjusted LSI (LSIadj, length corrected, see
Methods) varied in both areas (ANOVA: F3,563 ¼ 4.88; p , 0.01)
and status (ANOVA: F1,563 ¼ 11.07; p , 0.001), and the interaction
between these two variables was significant (ANOVA: F3,563 ¼
25.54; p , 0.0001). Within closed areas, LSIadj was significantly
higher outside than that inside for CAI (Student’s t-test:
t113 ¼ 29.09; p , 0.0001) and significantly higher inside than that
outside for JL (Student’s t-test: t149 ¼ 2.62; p , 0.01). The trend
for the other two areas was for higher LSIadj inside than that
outside for CAII (p . 0.05) and higher LSIadj outside than that
inside for CL (p . 0.05).
DFA revealed that cod could be classified back to the correct
status (i.e. inside vs. outside) 65% of the time when all areas were
considered, and between 77 and 87% of the time within areas; the
average among areas was 81% (Table 6). We then explored the
mean loading values (from DFA) for cod by area/status
(Figure 5). Mean loading scores were consistently higher outside
than those inside closed areas. Morphometric features that were
higher outside included length measurements from the head
region; cod outside closed areas had longer heads. On the other
hand, variables that were higher inside closed areas were primarily
from the middle region of the body indicating deeper bodies.
Discussion
The purpose of this study was to compare a range of life-history variables (Table 1) for cod inside and outside of closed areas in New
England to provide evidence for or against the hypothesis that
closed areas have a net positive effect on cod. In this sense, a positive
effect can mean greater longevity (enhanced age structure), higher
growth, better feeding opportunities or higher condition. We also
considered the possibility that closed areas may be selecting for
more sedentary life styles (via body shape analysis; sensu Sherwood
and Grabowski, 2010), the first ever consideration, to our
Comparison of cod life-history parameters inside and outside of four year-round groundfish
323
Figure 3. Mean PFI of benthos, pelagic fish, and demersal fish by 5 cm length interval for cod inside (left panels) and outside (right panels) of four
major closed areas in the GOM and on GB. Numbers inside bars for CL “in” are PFI values (exceeded scale).
Table 5. ANCOVA results for the relationship between prey length
and predator (cod) length for all areas combined and within areas.
CAI
CAII
CL
JL
All
a
269.6
251.6
267.9
219.8
233.8
b
0.21
0.14
0.20
0.12
0.13
c
64.6
N.S.
40.6
42.6
44.0
b3 c
20.1
N.S.
20.09
20.09
20.08
r2
0.21
0.12
0.18
0.04
0.10
p
,0.0001
,0.0001
,0.0001
,0.001
,0.0001
N
166
326
441
456
1,389
Relationship is of the form Lprey ¼ a + b × Lcod + c × “status” + d × Lcod ×
“status”.
knowledge, of what may be considered area-based-managementinduced selection in cod. Closed areas had a fairly consistent effect
on cod life history variation (Table 1), particularly on age and
growth variables that were mostly enhanced inside closed areas. The
following is a detailed discussion of each of the variables and how
they responded to closure status in each of the closed areas as well
as overall.
Closed areas generally contained older cod than that which were
found in adjacent open areas. In fact, we found nearly 10 times more
cod aged .5 years inside closed areas than that which were found
outside. Additionally, the median age was consistently 1 year older
inside closed areas than that which were found outside (four out
of four comparisons). If we consider the per cent difference, the
median age was 50% higher inside closed areas than that outside
for three of the areas (3 vs. 2 years old; CAI, CL, JL) and 25%
higher for CAII (5 vs. 4 years old). Finally, there were some spatial
differences in closed area effects on age structure. The most complete age structures were seen at Closed Area II and CL (Figure 2),
the two areas farthest from shore, and likely less vulnerable to nonrestricted fishing activities (e.g. recreational fishing). In other words,
allowing activities such as recreational fishing inside closed areas
may erode some of the age benefits that might otherwise be fully realized if closed areas were completely closed to fishing.
The age benefit that we document here may be of great importance to the current management of cod in New England. The most
recent GOM cod assessment indicates that this stock is at 3 –4% of its
324
G. D. Sherwood and J. H. Grabowski
Table 6. Predicted vs. actual group membership for cod inside and
outside of 4 major closed areas in the GOM and GB from DFA
analysis (morphometrics).
Predicted
15
13
Figure 4. d N vs. d C for cod inside and outside of four major closed
areas in the GOM and on GB. Shaded area shows limited extent of
trophic niche space occupied by cod from outside of all areas
combined. Cod from inside closed areas occupy this same space but
also higher trophic positions (i.e. higher d15N; CAII and CL) and more
pelagic space (i.e. more negative d13C; CL).
Figure 5. Bottom: mean loading scores (DF1 from DFA) by area and
status for cod from inside and outside of four major closed areas in the
GOM and on GB. Error bars are standard error. Top: box-truss design
showing 17 measurements for morphometric analysis. Dotted lines
indicate distances that loaded positively on DF1 (i.e. were greater for
cod outside) and solid lines indicate measurements that loaded
negatively on DF1, or were greater for cod inside. See Table 6 for DFA
results. DF1 scores were significantly higher outside than inside (i.e.
more streamlined) for all areas except JL (Student’s t-test, p , 0.05).
target biomass level (Palmer, 2014). Additionally, the assessment
finds few fish over 5 years old and states that this truncation in the
age structure may compromise future recruitment success of the
stock (Palmer, 2014). There are many reasons why age truncation
Area
All
All
All
CAI
CAI
CAI
CAII
CAII
CAII
CL
CL
CL
JL
JL
JL
Actual count
In
Out
Total
In
Out
Total
In
Out
Total
In
Out
Total
In
Out
Total
In
133
52
Out
71
98
32
9
9
31
45
3
12
18
39
6
6
39
48
11
13
33
% Correct
65
65
65
78
78
78
79
86
81
87
87
87
79
75
77
may impact recruitment and rebuilding potential. It is well known
that fish populations that contain a wide range of ages and, in particular, large proportions of old females can be more resilient and
productive (Secor, 2000; Berkeley et al., 2004; Ottersen et al.,
2006; Fogarty and O’Brien, 2009; Secor et al., 2014). In cod, old/
large females help to ensure higher egg number and egg/larval viability (Kjesbu et al., 1996, Marteinsdottir and Steinarsson, 1998;
Trippel, 1998) and also tend to spawn more effectively (Hixon
et al., 2014). Specifically, older/larger females may release batches
over many weeks during a spawning season, thus making it more
likely for at least some of their eggs to encounter favourable conditions (i.e. sensu the match/mismatch hypothesis; Cushing, 1990),
whereas younger/smaller and inexperienced females may spawn
only one batch and potentially at the wrong time for egg survival
(Trippel, 1998; Wright and Trippel, 2009).
There are currently no measures in place explicitly to ensure
the survival of old/large females in the GOM/GB region or in any
other fished cod population in the North Atlantic. Gear restrictions
(e.g. minimum mesh sizes) and minimum size limits can protect
young fish, but there is little that can be done to avoid capturing
the largest individuals; to the contrary, these are often the most
sought after. Harvest slots (taking only medium sized fish), as is successfully used in many recreational fisheries (Gwinn et al., 2013),
would be virtually impossible to implement in what is primarily a
trawl fishery for cod, although this may be an effective management
option for the recreational cod fishery. Perhaps the only way to
ensure survival of old/large females is through area management,
of which closed areas are an integral part, along with rolling (spawning) closures (Berkeley et al., 2004). Our study indicates that closed
areas, in their current configuration, may be crucial in buffering
against further degradation in the age structure of GOM and
GB cod, a process that would almost certainly lead to delays in
rebuilding.
In general, our results suggest that the age/length/growth variables appeared to respond most consistently to closure status
(Table 1). Of the 16 possible comparisons (four variables × four
closed areas), 12 were significantly different among samples of
cod captured inside vs. outside of closed areas. The pattern of response to closed area status was opposite for JL cod compared
Comparison of cod life-history parameters inside and outside of four year-round groundfish
with cod from the other three areas, which all responded consistently. In particular, mean age, mean length, scaling coefficient (b), and
assymptotic length were all higher (significantly or not) inside
closed areas vs. outside for CAI, CAII, and CL, whereas the opposite
was true for JL. These results suggest that cod inside those closed
areas excluding JL (possibly because of higher recreational fishing
effort inside JL) may experience lower mortality leading to higher
age and length as well as higher growth resulting in enhanced b
values and assymptotic lengths.
Cod diet and trophic relationships responded less consistently to
closure status than age and growth variables. Although responses
varied, the general trend was for higher stomach fullness (CAI,
CAII, CL), smaller prey size (CAI, CAII, CL), higher d15N or
trophic position (CAII, CL) and lower FI (CAII, CL) inside vs.
outside closed areas. Lower FI is often related to higher prey size
(e.g. Albrey Arrington et al., 2002) which is consistent with our
d15N results (Madigan et al., 2012), but not the prey size findings.
Nonetheless, these results are consistent with cod inside closed
areas having access to a broader range of feeding opportunities.
This finding is further bolstered by the stable isotope results
(Figure 4), which indicated that cod inside closed areas occupied
broader tropic niches than those outside. Growing fish, including
cod, with access to a diverse prey base, including large prey like
other fish, tend to be more efficient than fish with access to more
simplified (i.e. low diversity) prey bases; this can translate into
higher condition and growth (Sherwood et al., 2002,2007). Thus,
our results revealing more diverse feeding opportunities inside
closed areas may explain why we also observed higher growth inside.
Our morphometric analyses indicated that the reclassification
rate was quite high (81% on average). Furthermore, there was a consistent functional aspect to the differences in body shape between
cod captured inside and outside of closed areas. Specifically, cod
from inside closed areas tended to have deeper bodies and shorter
heads compared with cod from outside closed areas that had
longer heads and narrower bodies (Figure 5). Sherwood and
Grabowski (2010) compared body shape between red (a highly resident morphotype) and normal cod (assumed to be more migratory)
and found similar differences in that the red cod had shorter heads
and deeper bodies than the normal cod. We interpret the body shape
results here as an indication of differences in migratory behaviour.
Cod inside closed areas have body shapes more consistent with sedentary behaviours, whereas cod outside closed areas are more fusiform, as would be expected of a more migratory life-history strategy.
Our finding that body shape differed inside vs. outside of closed
areas expands upon previous work on fisheries-induced selection
for specific traits. Specifically, Olsen et al. (2004) showed that
years of industrialized fishing resulted in maturation at younger
ages and smaller sizes in northern cod (Newfoundland; NAFO
Divisions 2J3KL). Fudge and Rose (2008) demonstrated that this
change in maturity corresponded to an increase in fecundity at
earlier ages. Meanwhile, Sinclair et al. found that cod fishing in
the Gulf of St Lawrence favoured fast growth in the 1970s and
then slow growth in the late 1980s and 1990s. Perhaps, these
changes in reproductive endpoints and growth rates also coincided
with changes in migratory propensity towards more resident/
benthic lifestyles (GDS, unpublished). These examples related the
effects of selective fishery removals on specific cod traits. In this
paper, we document what may be the first known instance of an
effect of spatial management on a cod life history trait–migration.
It is unclear whether this effect of closed areas on body shape, and
presumably migration, is negatively or positively affecting the
325
productivity of the stock. Robichaud and Rose (2004) reviewed migratory behaviour in 174 cod groups throughout the North Atlantic
and found that maximum historical biomass was highest for migratory groups. Whether selection for more sedentary cod in closed
areas will result in lower productivity in New England cod stocks
requires further investigation.
Collectively, our results suggest that closed areas are demographically and ecologically beneficial to cod in the GOM and on GB.
However, due to practical constraints, we sampled in relatively confined areas within and around the closed areas (Figure 1). Thus, it is
currently unknown whether the benefits we demonstrate are widespread throughout closed areas or restricted to key habitats (e.g.
Ammen Rock in CL or a habitat closure in the northern corner of
CAII). The spatial extent of closed areas necessary to sustain cod
has been considered previously. For northern cod, model simulations suggest that the early 1990’s stock collapse in Newfoundland
might have been prevented by prior establishment of closed areas
encompassing 80% of the fished area (Guénette et al., 2000). This
estimate, which would prove highly impractical to implement,
reflects the highly migratory nature of the northern cod before
stock collapse (Rose, 1993). Following collapse, migratory pathways
and behaviour was highly disrupted (Rose, 2003; Wroblewski et al.,
2005). In this case, Guénette et al. (2000) estimated that closing 20%
of fishable bottom would be required to promote stock recovery.
This estimate is more in line with the amount of bottom currently
covered by year-round closed areas in the GOM and on GB/
Nantucket Shoals (17%; NEFMC, 2014, estimated from Map
52, pg 166). Given the mostly resident nature of GOM cod
(Lindholm et al., 2007; Howell et al., 2008; Sherwood and
Grabowski, 2010), this amount may be sufficient for the region.
While our results demonstrated that the closed areas on GB and
in the GOM benefit cod, these closed areas have been found to have
ecological benefits for other species. For instance, sea scallop
biomass increased 14-fold within the closed areas on GB and southern New England from 1994 to 1998 immediately after their inception (Murawski et al., 2000). These authors also noted that closed
areas are an effective management tool to improve yield per
recruit by reducing mortality on sublegal individuals, thereby reducing growth overfishing. Given that sea scallops are far less mobile
than Atlantic cod, it is not surprising that the closed area benefits
for scallops were far more noticeable. In general, for more mobile
species, effective management of exploitation outside of closed
areas will be just as critical as the closed areas themselves
(Murawski et al., 2000). The combination of reduced fishing
effort coupled with closed areas protecting spawning aggregations
of haddock could explain the 10-fold increase in the GB haddock
stock between 1995 and 2005 (Brodziak et al., 2008). Collectively,
these studies suggest that closed areas can contribute to the recovery
of fishery species, even for mobile species like cod.
Closed areas have been used globally as an ecosystem management tool to rebuild fisheries in temperate marine ecosystems.
Several global syntheses of no-take reserves have found greater
biomass, density, species richness, and size of organisms within
them compared with adjacent open areas (Côte et al., 2001; Gell
and Roberts, 2003; Halpern and Warner, 2003; Lester et al., 2009).
Some have suggested that closed area effects are more idiosyncratic
and are dependent on factors such as their location, size, and age
(Jennings, 2000; Côté et al., 2001; Kaiser, 2005; Claudet et al.,
2008; Edgar et al., 2014). While Lester et al. (2009) found that
even small reserves can result in significant benefits irrespective of
location, with temperate reef reserves equalling or outperforming
326
tropical reef reserves, larger and older reserves generally result in
greater positive effects on species biomass and abundance
(Halpern and Warner, 2003; Claudet et al., 2010; White et al.,
2011; Edgar et al., 2014). In our study system, it is unclear
whether reserve size scales with benefits for cod and other species.
The efficacy of closed areas for rebuilding fishery species is likely
tightly coupled with their life histories. For instance, as species disperse over greater distances either as larvae or through adult movement, the likelihood of reserves having local benefits on biomass
decreases (Pelc et al., 2009; Claudet et al., 2010; White et al.,
2011). Given that cod have long-lived larvae that are potentially
broadly dispersed and are highly mobile as adults, it is somewhat
surprising that we found older cod inside the reserves than out.
This effect could be a function of the relatively large size of the closures in the GOM and GB, which collectively close over 22 000 km2
to commercial fishing for groundfish. Our findings also speak to the
importance of investigating more subtle population responses such
as growth and age distributions in addition to focusing on effects of
closures on the biomass and abundance of key species.
Future studies should be designed to examine the relationship
between the spatial scale of closures and their effects on the population dynamics and productivity of cod and other species more
closely. Furthermore, focusing on the benefits and costs (e.g. displacement of fishing effort to other areas or species) of closed
areas to fish populations needs to be weighed against those of
other species and ecosystem management tools such as reducing
effort and modifying gear to avoid overfished species. The
outcome of this research could help inform decisions regarding
changes to closed area boundaries. The New England Fisheries
Management Council (NEFMC) has completed an amendment
process to either modify or maintain New England closed areas
(as of summer 2015). The council has recommended status quo protections remain in place for CL and only minor reduction in the size
of JL. Major changes have been proposed for GB closed areas including reduction in CAII, opening of CA1 and addition of new closed
areas. However, as of final submission of this article these recommendations were not yet made rules by NOAA Fisheries.
It is unknown to what extent any changes in closed area boundaries or size may erode the benefits that may have taken some years to
accumulate. Russ and Alcala (2004) showed that closed area benefits
for predatory fish (biomass) may take 15 – 40 years to manifest.
Unfortunately, since we do not know the required size of New
England closed areas for promoting cod age structure, any reductions in their scale, through the current amendment process or
future actions, should be viewed as a risk that may impede rebuilding efforts, and should be carefully weighed against the potential
benefits to resource users.
Conclusions
This study demonstrates a net positive effect of New England closed
areas on cod biology. Age was consistently higher inside closed areas
and this finding alone may be enough to herald the benefit of closed
areas for promoting cod stock sustainability and recovery given the
well-recognized advantage of maintaining older spawners in fished
populations (Hixon et al., 2014; Secor et al., 2014). Closed areas also
appear to be selecting for more sedentary body types in cod which
suggests that they may be less effective for conserving individuals
with more migratory propensities. A shift in life-history strategy
towards more resident lifestyles may entail a loss in productivity
(sensu Robichaud and Rose, 2004). Additionally, the potential for
closed areas to create pockets of high density may make these
G. D. Sherwood and J. H. Grabowski
aggregations more prone to environmental disturbances and catastrophes (McGilliard et al., 2011). However, these effects are less
certain. What is certain is that New England cod stocks are in a
state of crisis (Palmer, 2014). Management should be concerned
not only with rebuilding SSB but also with conserving components
of the existing SSB that contribute to disproportionately higher reproductive output (i.e. old fish). Perhaps, the only way to do this is
through closed area management. Any alterations to the existing
closed areas in the GOM and on GB should be considered carefully
in light of what might further be lost (demographically) from
already highly degraded cod stocks.
Acknowledgements
We thank Curt Brown, Julien Gaudette, Aaron Lyons, Adam
Baukus, Nicole Stephens, Jessica Lueders-Dumont, Nicole Condon,
Josh Emmerson, Marissa McMahan, Spencer Blair-Glantz, Erin
Wilkinson, Pat Meyers, Laura Armstrong, Zach Whitener, Matt
Moretti, and Jane Johnson for assistance in the field and laboratory.
We thank John Shusta (FV Special J), Tim Caldwell (FV C.W.
Griswold), Bruce Kaminski (FV Never Enough), Kelo Pinkham
(FV Bad Penny), and Proctor Wells (FV Tenacious) for collecting
cod. Funding for this project was provided by the Northeast
Consortium (NA06NMF4720095) and the National Science
Foundation (OCE-01-22031).
References
Albrey Arrington, D., Winemiller, K. O., Loftus, W. F., and Akin, S. 2002.
How often do fishes “run on empty”? Ecology, 83: 2145– 2151.
Alexander, K. E., Leavenworth, W. B., Cournane, J., Cooper, A. B.,
Claesson, S., Brennan, S., Smith, G., et al. 2009. Gulf of Maine cod
in 1861: historical analysis of fishery logbooks, with ecosystem implications. Fish and Fisheries, 10: 428 – 449.
Bergstad, O. A., Jørgensen, T., Knutsen, J. A., and Berge, J. A. 2008. Site
fidelity of Atlantic cod Gadus morhua L. as deduced from telemetry
and stable isotope studies. Journal of Fish Biology, 72: 131– 142.
Berkeley, S. A., Hixon, M. A., Larson, R. J., and Love, M. S. 2004.
Fisheries sustainability via protection of age structure and spatial distribution of fish populations. Fisheries, 29: 23– 32.
Bourque, B. J., Johnson, B. J., and Steneck, R. S. 2008. Possible prehistoric fishing effects on coastal marine food webs in the Gulf of
Maine. In Human impacts on ancient marine ecosystems, pp.
165– 183. Ed. by T. C. Rick, and J. M. Erlandson. University of
California Press, Berkeley, CA.
Bowering, W. R., and Lilly, G. R. 1992. Greenland halibut (Reinhardtius
hippoglossoides) off southern Labrador and northeastern
Newfoundland (Northwest Atlantic) feed primarily on capelin
(Mallotus villosus). Netherlands Journal of Sea Research, 29:
211– 222.
Brodziak, J., Traver, M. L., and Col, L. A. 2008. The nascent recovery of
the Georges Bank haddock stock. Fisheries Research, 94: 123– 132.
Cadrin, S. X. 2000. Advances in morphometric identification of fishery
stocks. Reviews in Fish Biology and Fisheries, 10: 91 – 112.
Chen, Y., Jackson, D. A., and Harvey, H. H. 1992. A comparison of von
Bertalanffy and polynomial functions in modelling fish growth data.
Canadian Journal of Fisheries and Aquatic Sciences, 49: 1228– 1235.
Claudet, J., Osenberg, C. W., Benedetti-Cecchi, L., Domenici, P.,
Garcı́a-Charton, J.-A., Pérez-Ruzafa, A., Badalementi, F., et al.
2008. Marine reserves: size and age do matter. Ecology Letters, 11:
481– 489.
Claudet, J., Osenberg, C. W., Domenici, P., Badalementi, F., Milazzo, M.,
Falcón, J. M., Bertocci, I., et al. 2010. Marine reserves: fish life history
and ecological traits matter. Ecological Applications, 20: 830 – 839.
Comparison of cod life-history parameters inside and outside of four year-round groundfish
Collie, J. S., Escanero, G. A., and Valentine, P. C. 1997. Effects of bottom
fishing on the benthic megafauna of Georges Bank. Marine Ecology
Progress Series, 155: 159– 172.
Côté, I. M., Mosqueira, I., and Reynolds, J. D. 2001. Effects of marine
reserve characteristics on the protection of fish populations: a
meta-analysis. Journal of Fish Biology, 59: 178– 189.
Cushing, D. H. 1990. Plankton production and year-class strength in
fish populations: an update of the match/mismatch hypothesis. In
Advances in Marine Biology. Ed. by J. H. S. Blaxter, and A. J.
Southward. Academic Press Limited, San Diego, CA.
Edgar, G. J., Russ, G. R., and Babcock, R. C. 2007. Marine protected
areas. In Marine Ecology, pp. 533– 555. Ed. by S. D. Connell, and
B. M. Gillanders. Oxford University Press, South Melbourne, VIC,
Australia.
Edgar, G. J., Stuart-Smith, R. D., Willis, T. J., Kininmonth, S., Baker,
S. C., Banks, S., Barrett, N. S., et al. 2014. Global conservation outcomes depend on marine protected areas with five key features.
Nature, 506: 216 – 220.
Fisher, J. A. D., and Frank, K. T. 2002. Changes in finfish community
structure associated with an offshore fishery closed area on the
Scotian Shelf. Marine Ecology Progress Series, 240: 249 – 265.
Fogarty, M. J., and O’Brien, L. 2009. Recruitment in marine fish populations. In Fish Reproductive Biology: Implications for Assessment
and Management. Ed. by T. Jakobsen, M. Fogarty, B. A. Megrey,
and E. Moksness. Wiley-Blackwell.
Fudge, S. B., and Rose, G. A. 2008. Life history co-variation in a fishery
depleted Atlantic cod stock. Fisheries Research, 92: 107 – 113.
Gell, F. R., and Roberts, C. M. 2003. Benefits beyond boundaries: the
fishery effects of marine reserves. Trends in Ecology and Evolution,
18: 448 –455.
Grabowski, T. B., Thorsteinsson, V., McAdam, B. J., and Marteinsdóttir,
G. 2011. Evidence of segregated spawning in a single marine fish
stock: sympatric divergence of ecotypes in Icelandic cod? PLoS
One, 6: e17528.
Guénette, S., Pitcher, T. J., and Walters, C. J. 2000. The potential of
marine reserves for the management of northern cod in
Newfoundland. Bulletin of Marine Science, 66: 831 –852.
Gwinn, D. C., Allen, M. S., Johnston, F. D., Brown, P., Todd, C. R., and
Arlinghaus, R. 2013. Rethinking length-based fisheries regulations:
the value of protecting old and large fish with harvest slots. Fish
and Fisheries. doi:10.1111/faf.12053.
Halpern, B. S., and Warner, R. R. 2003. Matching marine reserve design
to reserve objectives. Proceedings of the Royal Society of London B:
Biological Sciences, 270: 1871– 1878.
Hermsen, J. M., Collie, J. S., and Valentine, P. C. 2003. Mobile fishing
gear reduces megafaunal production on Georges Bank. Marine
Ecology Progress Series, 260: 97– 108.
Hixon, M. A., Johnson, D. W., and Sogard, S. M. 2014. BOFFFFs: on the
importance of conserving old-growth age structure in fishery populations. ICES Journal of Marine Science, 71: 2171– 2185.
Howell, W. H., Morin, M., Rennels, N., and Goethal, D. 2008. Residency
of adult Atlantic cod (Gadus morhua) in the western Gulf of Maine.
Fisheries Research, 91: 123 – 132.
Jennings, S. 2000. Patterns and prediction of population recovery in
marine reserves. Reviews in Fish Biology and Fisheries, 10: 209 – 231.
Kaiser, M. J. 2005. Are marine protected areas a red herring or fisheries
panacea? Canadian Journal of Fisheries and Aquatic Sciences, 62:
1194– 1199.
Kjesbu, O. S., Solemdal, P., Bratland, P., and Fonn, M. 1996. Variation in
annual egg production in individual captive Atlantic cod (Gadus
morhua). Canadian Journal of Fisheries and Aquatic Sciences, 53:
610– 620.
Kurlansky, M. 1997. Cod: a Biography of the Fish that Changed the
World. Penguin Books, New York, NY. 294 pp.
LeBris, A., Fréchet, A., Galbraith, P. S., and Wroblewski, J. S. 2013.
Evidence for alternative migratory behaviours in the northern Gulf
327
of St Lawrence population of Atlantic cod (Gadus morhua L.).
ICES Journal of Marine Science, 70: 793 – 804.
Lester, S. E., Halpern, B. S., Grorud-Colvert, K., Lubchenco, J.,
Ruttenberg, B. I., Gaines, S. D., Airame, S., et al. 2009. Biological
effects within no-take marine reserves: a global synthesis. Marine
Ecology Progress Series, 384: 33– 46.
Lindholm, J., Auster, P. J., and Knight, A. 2007. Site fidelity and movement of adult Atlantic cod Gadus morhua at deep boulder reefs in the
western Gulf of Maine, USA. Marine Ecology Progress Series, 342:
239– 247.
Madigan, D. J., Carlisle, A. B., Dewar, H., Snodgrass, O. E., Litvin, S. Y.,
Micheli, F., and Block, B. A. 2012. Stable isotope analysis challenges
wasp-waist food web assumptions in an upwelling pelagic ecosystem. Scientific Reports, 2: 654.
Marteinsdottir, G., and Steinarsson, A. 1998. Maternal influence on the
size and viability of Iceland cod Gadus morhua eggs and larvae.
Journal of Fish Biology, 52: 1241 –1258.
McGilliard, C. R., Punt, A. E., and Hilborn, R. 2011. Spatial structure
induced by marine reserves shapes population responses to catastrophes in mathematical models. Ecological Applications, 24:
1399– 1409.
Mayo, R. K., Shepherd, G., O’Brien, L., Col, L. A., and Traver, M. 2009.
The 2008 assessment of the Gulf of Maine Atlantic cod (Gadus
morhua) stock. US Department of Commerce, Northeast Fisheries
Science Center Reference Document 09 – 03, 128 pp.
Morinville, G. R., and Rasmussen, J. B. 2008. Distinguishing between juvenile anadromous and resident brook trout (Salvelinus fontinalis)
using morphology. Environmental Biology of Fishes, 81: 171 – 184.
Murawski, S. A., Brown, R., Lai, H.-L., Rago, P. J., and Hendrickson, L.
2000. Large-scale closed areas as a fishery-management tool in temperate marine systems: The Georges Bank experience. Bulletin of
Marine Science, 66: 775– 798.
Murawski, S. A., Wigley, S. E., Fogarty, M. J., Rago, P. J., and Mountain,
D. G. 2005. Effort distribution and catch patterns adjacent to temperate MPAs. ICES Journal of Marine Science, 62: 1150– 1167.
New England Fishery Management Council (NEFMC). 2014. Omnibus
Essential Fish Habitat Amendment 2 Draft Environmental Impact
Statement (DEIS), Volumes I-V. http://www.nefmc.org/library/
omnibus-essential-fish-habitat-amendment-2-draft-environmentalimpact-statement-deis.
Northeast Fisheries Science Center (NEFSC). 2012. 53rd Northeast
Regional Stock Assessment Workshop (53rd SAW) Assessment
Report. US Department of Commerce, Northeast Fisheries Science
Center Reference Document 12-05, 559 pp.
Northeast Fisheries Science Center (NEFSC). 2013. 55th Northeast
Regional Stock Assessment Workshop (55th SAW) Assessment
Report. US Department of Commerce, Northeast Fisheries Science
Center Reference Document 13-11. 845 pp.
Olsen, E. M., Heino, M., Lilly, G. R., Morgan, M. J., Brattey, J., Emande,
B., and Dieckmann, U. 2004. Maturation trends indicative of rapid
evolution preceded the collapse of northern cod. Nature, 428:
932– 935.
Ottersen, G., Hjermann, D. O., and Stenseth, N. C. 2006. Changes in
spawning stock structure strengthen the link between climate and recruitment in a heavily fished cod (Gadus morhua) stock. Fisheries
Oceanography, 15: 230– 243.
Palmer, M. C. 2014. 2014 Assessment update report of the Gulf of Maine
Atlantic cod stock. United States Department of Commerce,
Northeast Fisheries Science Center Reference Document 14-14.
119 pp.
Pelc, R. A., Baskett, M. L., Tanci, T., et al. 2009. Quantifying larval export
from South African marine reserves. Marine Ecology Progress Series,
394: 65– 78.
Robichaud, D., and Rose, G. A. 2004. Migratory behaviour and range in
Atlantic cod: inference from a century of tagging. Fish and Fisheries,
5: 185 – 214.
328
Rose, G. A. 1993. Cod spawning on a migration highway in the northwest Atlantic. Nature, 366: 458– 461.
Rose, G. A. 2003. Monitoring coastal northern cod: towards an optimal
survey of Smith Sound, Newfoundland. ICES Journal of Marine
Science, 60: 453– 462.
Russ, G. R., and Alcala, A. C. 2004. Marine reserves: long-term protection is required for full recovery of predatory fish populations.
Oecologia, 138: 622– 627.
Secor, D. H. 2000. Longevity and resilience of Chesapeake Bay striped
bass. ICES Journal of Marine Science, 57: 808– 815.
Secor, D. H., Rooker, J. R., Gahagan, B. I., Siskey, M. R., and Wingate,
R. W. 2014. Depressed resilience of bluefin tuna in the western
Atlantic and age truncation. Conservation Biology, 29: 400 – 408.
Sherwood, G. D., and Grabowski, J. H. 2010. Exploring the life-history
implications of colour variation in offshore Gulf of Maine
cod (Gadus morhua). ICES Journal of Marine Science, 67:
1640– 1649.
Sherwood, G. D., Kovecses, J., Hontela, H., and Rasmussen, J. B. 2002.
Simplified food webs lead to energetic bottlenecks in polluted
lakes. Canadian Journal of Fisheries and Aquatic Sciences, 59:
1 – 5.
Sherwood, G. D., Rideout, R. M., Fudge, S. B., and Rose, G. A. 2007.
Influence of diet on growth, condition and reproductive capacity
in Newfoundland and Labrador cod (Gadus morhua): Insights
from stable carbon isotopes (d13C). Deep-Sea Research II, 54:
2794– 2809.
G. D. Sherwood and J. H. Grabowski
Sherwood, G. D., and Rose, G. A. 2005. Influence of swimming form on
otolith d13C in marine fish. Marine Ecology Progress Series, 258:
283– 289.
Solow, A. 1990. A randomization test for misclassification probability in
discriminant analysis. Ecology, 71: 2379– 2382.
Strauss, R. E., and Bookstein, F. L. 1982. The truss : body form reconstructions in morphometrics. Systematic Zoology, 31: 113– 135.
Svedäng, H., Righton, D., and Jonsson, P. 2007. Migratory behavior of
Atlantic cod Gadus morhua: natal homing is the prime stock separating mechanism. Marine Ecology Progress Series, 345: 1 –12.
Trippel, E. A. 1998. Egg size and viability and seasonal offspring production of young Atlantic cod. Transactions of the American Fisheries
Society, 127: 339– 359.
White, J. W., Botsford, L. W., Baskett, M. L., Barnett, L. A. K., Barr, R. J.,
and Hastings, A. 2011. Linking models with monitoring data for
assessing performance of no-take marine reserves. Frontiers in
Ecology and the Environment, 9: 390 – 399.
Wright, P. J., and Trippel, E. A. 2009. Fishery-induced demographic
changes in the timing of spawning: consequences for reproductive
success. Fish and Fisheries, 10: 283– 304.
Wroblewski, J., Neis, B., and Gosse, K. 2005. Inshore stocks of Atlantic
cod are important for rebuilding the east coast fishery. Coastal
Management, 33: 411– 432.
Yaragina, N. A., and Marshall, C. T. 2000. Trophic influences on interannual and seasonal variation in the liver condition index of
Northeast Arctic cod (Gadus morhua). ICES Journal of Marine
Science, 57: 42– 55.
Handling editor: David Secor