ICES Journal of Marine Science, 59: 239–253. 2002
doi:10.1006/jmsc.2001.1131, available online at http://www.idealibrary.com on
Does the North Atlantic current affect spatial distribution of
whiting? Testing environmental hypotheses using statistical and
GIS techniques
X. Zheng, G. J. Pierce, D. G. Reid, and I. T. Jolliffe
Zheng, X., Pierce, G. J., Reid, D. G., and Jolliffe, I. T. 2002. Does the north Atlantic
current affect spatial distribution of whiting? Testing environmental hypotheses using
statistical and GIS techniques. – ICES Journal of Marine Science, 59: 239–253.
This paper describes spatial relationships between whiting, Merlangius merlangus
(Linnaeus, 1758), abundance in the northern North Sea and contemporaneous
measures of environmental conditions: sea surface temperature (SST), sea bottom
temperature (SBT), and depth, with particular reference to the processes underlying
patterns in SST. Generalised additive models (GAMs) were used to provide quantitative descriptions of the relationships between local abundance and environmental
conditions. GIS (geographic information system) techniques were used to provide
qualitative description of spatial patterns and to confirm the results revealed from
GAMs.
GAMs fitted to both long-term averaged and individual years’ data revealed marked
seasonal changes in the spatial relationships between whiting abundance and environmental variables. The GAM results were supported by GIS analysis. In winter and
spring (December–April) in the northern North Sea, the spatial pattern of SST
apparently has an important influence on the spatial distribution of whiting at the same
time. Where the water is relatively warm whiting abundance is relatively high, probably
reflecting the indirect influence of North Atlantic waters entering the northern North
Sea. However, there are no consistent optimum SST bands for whiting. These positive
relationships between abundance and SST disappear in summer.
2002 International Council for the Exploration of the Sea. Published by Elsevier Science Ltd.
All rights reserved.
Keywords: depth, GAM, GIS, Northern Atlantic current, sea surface temperature,
spatial patterns, whiting.
Received 13 November 2000; accepted 13 July 2001.
X. Zheng and G. J. Pierce: Department of Zoology, University of Aberdeen, Tillydrone
Avenue, Aberdeen, AB24 2TZ, Scotland, UK; tel: 01224 273273; e-mail:
. D. G. Reid: Fisheries Research Services, Marine Laboratory
Aberdeen, PO Box 101, Victoria Road, Aberdeen, AB11 9DB, Scotland, UK; tel:
01224 876 544; e-mail: . I. T. Jolliffe: Department of
Mathematics Science, University of Aberdeen, Aberdeen, Scotland, UK; e-mail:
i.jolliff
[email protected]
Introduction
Of all environmental factors in the ocean water temperature is often assumed to have the largest effect on fish
and fisheries. Water temperature not only directly influences the physiological capability of an organism but
also affects various physical and chemical processes such
as the solubility of dissolved gases and viscosity of sea
water. Both of these vary inversely with temperature
(Vernberg and Vernberg, 1972). Brett (1970) summarised much of the existing data on the thermal limits
of embryonic and post-embryonic fish from different
1054–3139/02/040239+15 $35.00/0
latitudes. He showed that for embryonic fish in latitudes
between 50N and 60N the survival range of water
temperature is above zero and under 16C, while for
post-embryonic fish in the same latitudes the range
is a little wider. In several stocks of Atlantic cod
higher water temperature has been associated with
higher growth rates (Pederson and Jobling, 1989;
Brander, 1994). Serchuk et al. (1994) found that the
larval metabolic requirements of cod increased with
temperature.
Temperature
also
affects
fish
distribution.
Temperature-related displacements of cod have been
2002 International Council for the Exploration of the Sea. Published by Elsevier Science Ltd. All rights reserved.
240
X. Zheng et al.
0
0
50 100 150
4 gears
January
100
50
50 100 150
4 gears
February
150
100
50
0
0
50 100 150
4 gears
May
All gears
150
100
50
0
150
100
50
0
0
50 100 150
4 gears
September
All gears
150
100
50
50 100 150
4 gears
June
50 100 150
4 gears
October
50
0
50 100 150
4 gears
April
0
50 100 150
4 gears
August
0
50 100 150
4 gears
December
150
100
50
0
0
50 100 150
4 gears
July
150
100
50
0
0
100
50 100 150
4 gears
March
0
0
All gears
0
150
0
0
All gears
150
50
0
0
All gears
All gears
0
All gears
50
100
All gears
50
100
150
All gears
100
150
All gears
150
All gears
All gears
Figure 1. Map of the study area showing the location of the land masses and indicating sea depth (m). The study area consists of
a set of adjacent of ICES statistical rectangles (*) and was chosen on the basis of the analysis of the spatial relationships between
whiting abundance and sea surface temperature (Zheng et al., 2001).
150
100
50
0
0
50 100 150
4 gears
November
Figure 2. Spatial comparison of whiting abundance (LPUE) derived from all gears and the four main trawling gears (see the text)
for demersal fish by ICES rectangles. Data are long-term averages.
Does the North Atlantic current affect spatial distribution of whiting
241
Table 1. The formula, degrees of freedom used, residual degrees of freedom, and pseudo-coefficient of
determination, r2, of the best GAMs fitted to the data in different seasons (data are long-term
averages). The predictors and smoothers were automatically selected by the GAM stepwise procedures
from the pre-specified predictors and smoothers (smooth spline smoother and locally weighted
regression smoother), respectively. Pre-specified predictors are SST, SBT, and depth.
Formula
Degrees of
freedom used
in model fit
Residual
degrees of
freedom
r2
LpSST+s(SBT, 3)+s(depth, 3)
Lps(SST, 4)+SBT+s(depth, 4)
LpSST+s(depth, 3)
Lps(SST, 3)+s(depth, 3)
Lps(SST, 3)+s(SBT, 3)+s(depth, 3)
LpSST+s(SBT, 4)+depth
Lps(SST, 3)+s(SBT, 4)+depth
Lps(SST, 3)+s(SBT, 3)+depth
Lps(SST, 4)+s(SBT, 4)+depth
Lplo(depth, span=0.5)
LpSBT+s(depth, 3)
LpSST+lo(depth, span=0.75)
8
10
5
7
10
7
9
8
10
6
5
5
44
42
34
47
42
46
32
40
36
45
43
38
0.59
0.78
0.66
0.68
0.55
0.56
0.72
0.60
0.67
0.56
0.49
0.69
Month
January
February
March
April
May
June
July
August
September
October
November
December
Lp is LPUE; SST is sea surface temperature. s(x) is a smooth spline smoother and lo(y) is a locally
weighted regression smoother, where x represents the degrees of freedom for the smoother and y
represents the span argument in the locally weighted regression smoother. The locally weighted
regression smoother is used to specify the percentage of the observations each local neighbourhood
should contain for the smoother.
40
20
10
0
–20
s(Depth, 3)
s(SBT, 3)
SST
20
–10
–20
–40
–30
–40
0
–60
7.0
7.4
7.8
SST
January
7.5
8.2
40
s(Depth, 4)
0
0
–20
–20
350
20
20
SBT
s(SST, 4)
20
150
250
Depth
January
50
8.0
8.5
S (...truncated)
