Lough and O'Brien: Life-stage recruitment models for Gadus morhua and Melanogrammus aeglefinus on Georges Bank 
127 
Table 2 
Yearly spawning egg and hatching abundances and life-stage mortality rates used in the haddock (Melanogrammus aeglefinus ) 
age-1 recruitment hindcasts by program surveys. Marine Resource Monitoring, Assessment, and Prediction (MARMAP) surveys 
occurred from 1977 and 1987, and the U.S. Global Ocean Ecosystems Dynamics (GLOBEC) surveys occurred from 1995 to 1999. 
Survey egg and larval abundances and mortality rates are from Lough et al. (2006) and Mountain and Kane (2010). Pelagic 
juvenile mortality was specified as 6%/d. Demersal juvenile mortality rates were estimated from Equation 7. 
Survey 
and year 
Spawning 
abundance 
(no.xl012) 
Egg 
mortality 
(%/d) 
Hatching 
abundance 
(no.xl012) 
Larval 
mortality 
(%/d) 
Pelagic 
juvenile 
mortality 
(%/d) 
Demersal 
juvenile 
mortality 
(%/d) 
MARMAP 
1977 
No egg data 
7.017 
10.1 
6.0 
3.00 
1978 
No egg data 
12279.967 
29.9 
6.0 
2.27 
1979 
7.988 
10.3 
2.032 
14.0 
6.0 
2.19 
1980 
80.803 
17.2 
2.071 
11.5 
6.0 
2.12 
1981 
13.317 
9.3 
2.192 
11.3 
6.0 
2.54 
1982 
4.004 
16.5 
0.090 
17.8 
6.0 
2.07 
1983 
13.488 
9.7 
2.397 
19.1 
6.0 
1.91 
1984 
1.651 
0.0 
2.394 
16.6 
6.0 
2.26 
1985 
3.025 
3.7 
1.534 
9.2 
6.0 
1.89 
1986 
2.492 
5.3 
0.978 
10.5 
6.0 
2.57 
1987 
2.224 
0.8 
2.078 
11.0 
6.0 
2.04 
GLOBEC 
1995 
7.274 
12.0 
0.580 
10.3 
6.0 
1.78 
1996 
10.866 
11.3 
1.030 
11.5 
6.0 
1.83 
1997 
12.286 
13.4 
0.700 
9.1 
6.0 
1.92 
1998 
9.160 
7.8 
1.920 
4.7 
6.0 
1.81 
1999 
16.512 
9.9 
1.740 
6.1 
6.0 
1.97 
sampled bimonthly with about 30 standard stations 
during the MARMAP period (Fig. 2). The GLOBEC 
sampling period provided the best coverage of the spawn- 
ing seasons because surveys were conducted at about 40 
standard stations monthly from January through June 
throughout the Georges Bank region (Fig. 2B). Other 
years outside the MARMAP and GLOBEC time series 
were surveyed only once or rarely twice during the 
spawning season so that we have the least confidence 
in those years for estimating egg abundance during the 
spawning season. 
Sampling methods and data procedures for the ich- 
thyoplankton time series are described in full detail 
elsewhere (Lough et al., 2006; Mountain et al., 2008; 
Mountain and Kane, 2010). Fish eggs were identified, 
staged, and counted from quantitative 61-cm bongo- 
net hauls. Counts of staged eggs were normalized to 
the number of eggs per 10 m 2 per day (no./[10 m 2 xd]). 
Egg densities were averaged by survey, and survey 
means were expanded by the number of days repre- 
sented by each survey and summed over the season’s 
beginning and end times to estimate the total number 
of eggs spawned (Fig. 3). Only late-stage cod and had- 
dock eggs can be visually identified, and therefore the 
abundances of earlier stage eggs were approximated 
by the proportion of the late-stage eggs. Differential 
egg mortality between the two species may have intro- 
duced a systematic bias in the yearly egg production 
estimates — a bias that varies with their relative bio- 
mass. This bias is discussed in Lough et al. (2008) but 
cannot be readily corrected. 
Egg stage 
Another method for estimating egg mortality was the 
use of a simple wind-driven transport model developed 
by Mountain et al. (2008) for the GLOBEC years, which 
was redone for the time series 1979-2005 (Mountain 3 ). 
Particles simulating the egg distributions were allowed 
to drift in the modeled flow for 17 days (egg duration) 
and were considered lost to Georges Bank if they 
crossed the 200-m isobath. The percentage transport 
loss of particles was determined from mid-February to 
mid-April for cod and from mid-March to mid-May for 
haddock and converted to a daily mortality rate (%/d). 
The observed survey egg mortality rates from the 
GLOBEC and MARMAP years were regressed against 
3 Mountain, D. 2008. Unpubl. data. Northeast Fisheries 
Science Center, Woods Hole, MA 02543-1097. 
