ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on March 30, 2007
ICES Journal of Marine Science: Journal du Conseil 2007 64(5):920-928; doi:10.1093/icesjms/fsm028
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Home range and elevated egg densities within an inshore spawning ground of coastal cod
1 Centre for Ecological and Evolutionary Synthesis (CEES), Department of Biology, University of Oslo, PO Box 1066 Blindern, N-0316 Oslo, Norway
2 Institute of Marine Research, Flødevigen 4817 His, Norway
Correspondence to N. C. Stenseth: tel: +47 22 85 45 84; fax: +47 22 85 40 01; e-mail: n.c.stenseth{at}bio.uio.no
Espeland, S. H., Gundersen, A. F., Olsen, E. M., Knutsen, H., Gjøsæter, J., and Stenseth, N. C. 2007. Home range and elevated egg densities within an inshore spawning ground of coastal cod. – ICES Journal of Marine Science, 64: 920–928.Atlantic cod (Gadus morhua) along the Norwegian Skagerrak coast are structured into genetically distinct local populations. Mechanisms contributing to this genetic structure may include spawning site fidelity of adult cod as well as retention of pelagic early life stages close to the spawning grounds. Spawning in sheltered inshore localities is likely to favour retention of eggs and larvae, the opposite situation to offshore spawning. A combined study was made of area utilization by adult cod and the distribution of cod eggs within an inshore locality of the Norwegian Skagerrak coast. The behaviour of adult cod was studied using acoustic telemetry and kernel modelling, and eggs were sampled throughout the spawning season. Generalized additive models were applied to test hypotheses about the spatial dynamics of the eggs, and the best model described a central spawning area that retained its integrity through time. Adult cod were confined to small parts of the study area and remained there throughout the spawning season. The average home range of the adult cod was 27 ha. Overall, the study demonstrated two mechanisms by which coastal (i.e. inshore) cod maintain their population structure: spawning site fidelity and the spatial dynamics of their eggs.
Keywords: Atlantic cod, egg retention, Gadus morhua, home range, local populations, spawning, stock structure
Received 17 July 2006; accepted 10 February 2007; advance access publication 30 March 2007.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
Marine species display a great variation in population structure (Ward, 2006). Hence, knowledge of population structure is important both for understanding biological processes and for implementing sustainable management. Different mechanisms may either enhance or erode population structure. These mechanisms may include lifelong behavioural effects such as year-round site fidelity or accurate homing to the natal spawning area (Robichaud and Rose, 2004), but also short-term mechanisms operating within a specific spawning season. Retention of planktonic life stages close to the spawning area may promote population structure. Advection of pelagic early life stages (eggs and larvae) by ocean currents will tend to break down local population structure, if the dispersers survive and spawn in non-natal areas (Waples, 1998). Spawning in inshore areas sheltered from strong currents could reduce such passive drift of eggs and larvae and help to maintain local self-reproducing units (Hastings and Botsford, 2006).
The Norwegian Skagerrak coast holds a complex of genetically distinct populations of Atlantic cod (Gadus morhua), with small but significant neutral genetic differentiation at a scale of only a few tens of kilometres (Knutsen et al., 2003, 2004). Many of these populations also differ in adaptive traits (Olsen et al., 2004). It is already known that the cod spawn inshore, perhaps to minimize the advection of eggs. The spawning grounds are typically separated by just a few tens of kilometres (Knutsen et al., 2007). On a year-round basis, there is evidence that the Skagerrak coastal cod are fairly stationary, most (>90%) marked fish being recaptured less than 20 km from their point of release (Danielssen and Gjøsaeter, 1994) and indicating a high degree of site fidelity between spawning events. However, with the distance between spawning areas being short, cod would also need to show good site fidelity throughout the spawning season.
We used hydro-acoustic equipment to monitor movements of mature fish on a fine spatial and temporal scale. Such methodology has been widely used to track various gadoid species (Hawkins et al., 1974; Clark and Green, 1990; Robichaud and Rose, 2001; Cote et al., 2004). Next, we use generalized additive models (GAMs) to describe the spatio-temporal dynamics of cod eggs in the same region, so adding to current knowledge by estimating the home range of adult cod throughout the season, and at the same time investigating the spatio-temporal dynamics of their eggs.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The study was conducted in the strait between the island of Tromøya and the mainland (58°29'N 8°51'E) on the Norwegian Skagerrak coast (Figure 1). The strait is open to the sea at both ends, and the distance to open water is 9 km (in a southwest direction) and 7 km (to the northeast). The average water depth is 30 m, but deeper areas extend down to 50–60 m. Within the study area, there are sills at both ends 25 m deep. There is considerable water replacement of surface water of the strait, but the water is stagnant at deeper levels (Dahl and Danielssen, 1986). Coastal cod from the Tromøy strait are genetically different from cod in neighbouring fjords, and form part of a network of local populations typically separated by only a few tens of kilometres (Knutsen et al., 2003). The study took place during 64 d from 8 February to 12 April 2001, of which ultrasonic tracking was conducted from 8 February to 22 March and sampling of eggs from 23 February to 12 April. Eggs were sampled 18 times during the 48-d period, with vertical tows made with a WP2 net of diameter 0.46 m and mesh size 320 µm. The tows were usually made from 30 m deep to the surface, meaning that in shallow water the tows covered the whole water column. The tows were conducted on 53 stations arranged in a square grid covering the target area. Because of limited daylight, not all stations were sampled in a single day, so sampling effort was divided between two sets of stations (Figure 1). Every second station in the grid (1–27) belonged to set A and the rest (30–56) to set B, and both sets covered the entire geographical extent of the target area. Eggs were preserved in 4% formalin. On 28 March, four hauls were made from the bottom to the surface and three from 10 m to the surface at stations 1 and 21. The results of these hauls indicated that most eggs (>88%) were collected deeper than 10 m.
|
The number of eggs in each sample was enumerated, but because of the large number of eggs found, the diameters of only 20% from each sample were measured and identified; the raised numbers of cod eggs in the subsamples were used in subsequent analyses. The eggs were identified to species by combining diameter and pigmentation pattern of eggs in late stages of development. The cod eggs were at a variety of developmental stages, but these were not recorded for each egg individually. In the later part of the study period, several larvae were also caught. The diameters of eggs were measured taking into account a shrinkage of formalin-preserved eggs of
7% (Hiemstra, 1962). Eggs with an original diameter from 1.11 (Hiemstra, 1962) to 1.67 mm (Ehrenbaum, 1905) were defined as cod eggs. In all, 3343 eggs were measured and 2377 were identified as cod eggs. The number of cod eggs from each subsample was divided by the station depth and plotted against depth, latitude and longitude of the station, and day sampled (Figure 2). Those eggs not identified by this method as cod had size distribution peaks of 1.02, 1.87, and 2.3 mm and were presumed to be of Platichthys flesus, Trisopterus minutus, Merlangius merlangus, Microstomus kitt, and Pleuronectes platessa. It is easy to misclassify early stages of cod and haddock eggs (Hurley and Campana, 1989; Lough et al., 1996; Bradbury et al., 2000), but information from local fishers is that the area does not hold adult haddock.
|
Ten cod were tagged with ultrasonic transmitters and tracked using a stationary positioning system (VRAP, Vemco Ltd, Shad Bay, Nova Scotia, Canada) capable of discriminating up to ten different transmitters. The system consisted of three hydrophone buoys arranged in a triangle 600–900 m apart. Positions were calculated based on the relative delay of pulse arrivals at the hydrophones and could be calculated up to 1 km outside the buoy triangle, depending on ambient conditions. Environmental noise and echoes may contribute to error in position calculation, so a stationary transmitter was used to assess the precision of the positioning system. For a transmitter held stationary within the triangle for 1 h, all calculated positions fell within a 5 m2 area. When the transmitter was held 370 m outside the triangle, the positions calculated had a wider margin of error (34 m2). Therefore, home ranges extending far beyond the triangle may be somewhat overestimated.
A local fisher deploying eel pots caught all ten cod used for this study, over a period of time and in several locations within the study area. Cod ranged in size from 55 to 79 cm, and based on this relatively large body size, all were assumed to be sexually mature (Olsen et al., 2004). Cod were tagged with surgically implanted ultrasonic transmitters (Vemco, continuous V16-6L). measuring 90 x 16 mm, weighing 37 g (14 g in water), and constituting 0.8–2% of the fish body weight. Fish were anaesthetized in a 7:20 000 solution of quinaldine, and after surgery, kept in tanks for 2–3 d to ensure that they had recovered fully. They were then released. Recent studies have shown that such transmitters can be used effectively on marine fish, inserted gastrically (Lucas and Johnstone, 1990; Winger and Walsh, 2001; Winger et al., 2002) or surgically (Comeau et al., 2002; Bridger and Booth, 2003; Cote et al., 2004). Our surgical procedures were also tested for any effect on behaviour prior to the study (Espeland, 2002); no behavioural effects were found.
During tracking, formation and break-up of ice occasionally moved the hydrophone buoys. Positions recorded during such periods, as well as immediately after release (1–4 d) and obviously false positions, were excluded from the analyses. Rocks and rough ground can easily reduce the accuracy of a position fix, so the number of fixes for each fish varied (22–1679 for nine fish, and one fix for one fish). The three buoys had to be hit by the same signal for a position to be calculated. The number of positive signal detections by each buoy was compared with the total number of possible detections to give the ratio of time a transmitter was detected (Table 1). Manual tracking was conducted on several occasions to locate fish that could not be detected with the VRAP system, intensively throughout June and occasionally thereafter (when tracking with the VRAP system had ended).
|
Studies of home range and area utilization are well developed within the field of terrestrial biology (Mohr, 1947; Jennrich and Turner, 1969), and development of various statistical tools has facilitated the concept of utilization distribution (UD). It can be defined as "the two-dimensional relative frequency distribution for the points of location of an animal over a period of time" (Winkel, 1975). Methods to estimate the UD of an animal include a non-parametric method of kernel density estimation, introduced by Worton (1989), a method consisting of placing a kernel (a symmetrical density function) above each position in the sample. Then, by superimposing a grid on the data, the sum of the kernel density distributions can be identified at each grid intersection, and the overall frequency distribution of the sample can be deduced.
A kernel UD was estimated for each fish (Kenward and Hodder, 1996), using the formula below to give the frequency distribution:
|
| (1) |
where n is the number of observations and x and y the coordinates of the position fixes. E(x) and E(y) are the expected average positions calculated as arithmetic means of the sample. The kernel K is a unimodal symmetrical bivariate probability density function, and h is the smoothing parameter (describing the weight of the tails of kernel K). In this study, a fixed kernel distribution was estimated for each individual range with the smoothing parameter set at href = n–1/6 
, where is the standard deviation of the coordinates of the positions. We did not choose h based on least square cross validation (LSCV) because it may create an over-fragmented distribution (Blundell et al., 2001; Hemson et al., 2005).
Different values of 
may be used to describe different percentages of the mass of the distribution. = 0.90 was considered as containing the home range area. Autocorrelation can cause an underestimation of home range (Swihart and Slade, 1985), but reducing the number of positions to achieve independence may cause bias because of small samples (Hansteen et al., 1997; Rooney et al., 1998; De Solla et al., 1999). Therefore, we did not attempt to reduce the autocorrelation. Here, we included all the distributions we calculated, but those based on samples with fewer than 30–50 position fixes may be biased compared with true distributions (Seaman and Powell, 1996; Seaman et al., 1999).
The spatial and temporal distribution of cod eggs was modelled with GAMs (Hastie and Tibshirani, 1990). The response variable (y) was the natural logarithm of egg density, standardized by volume filtered (eggs per metre hauled). Hauls containing zero eggs and hauls not sampled because of ice presence (100) were excluded from the analyses. The models were also tested using null data with no major differences. Predictor variables were the bottom depth at each sampling station (d), day (t), longitude (x), and latitude (z). Model selection was based on the generalized cross validation (GCV) criterion (Wood, 2000). Residual plots were used for checking the fit of the models.
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Home ranges (
= 0.90) of mature fish were estimated from the positions observed for nine fish (Table 1, Figure 3). The average home range size was 27.04 ha (3.05–77.33 ha). One fish did not yield sufficient positions to calculate a UD, but manual tracking confirmed that the fish was present throughout the study period. After removal of the buoys in late March, three fish (numbers 2, 3, and 6) disappeared altogether (i.e. moved > 10 km from the release point), and two (8 and 9) moved intermediate distances (5 and 7.5 km); the other four stayed in the study area. Three stopped moving totally during summer (1, 4, and 7) and were therefore classified as dead. Two of the fish originally released (5 and 9) were taken by fishers during summer, and the other two fish remained in the area until the next spawning season (February/March 2002).
|
Our general model [formulation (2) below] describing the spatial and temporal distribution of cod eggs had bottom depth and day number as additive effects, and latitude and longitude modelled as a smoothed surface (Figure 4):
|
| (2) |
the error term. The model had a GCV of 0.38, and was superior to more constrained models that exclude one or more covariates (GCV > 0.41). Egg density increased throughout the spawning season, peaked, then declined to a more moderate level (Figure 4). The effect of bottom depth on egg density was less clear, but egg densities tended to be high in the shallower areas (Figure 4). Densities peaked in the central northern region of the study area and dropped towards the east and were moderate south and southwest of the northern peak (Figures 4 and 5).
|
|
Our general starting model (2) considered the temporal dynamics to have the same effect anywhere, indicating no shift in the distribution of cod eggs. To test the stability of the density patterns through time, we built a more complex model (3), in which day (t) was considered to interact with depth (d), latitude (z), and longitude (x):
|
| (3) |
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Several mechanisms may work to enhance or erode small-scale population structure in marine species. Here, we highlight two mechanisms working within the spawning season that may help to explain why local populations of coastal cod are able to persist along the Norwegian Skagerrak coast (Knutsen et al., 2003). First, the inshore distribution of eggs suggests inshore spawning activity, and the stability of the distribution throughout the spawning season is indicative of inshore retention of the early life stages. Second, we found a high degree of site fidelity of adult cod in the area during the spawning season, evidenced by acoustic tracking.
On the spawning ground, the greatest egg densities were in a central basin, with no significant shift in location of egg densities through time. We found evidence for a temporal peak in egg densities, likely to describe the peak of the spawning season (Kjesbu, 1989). Absence of a clear drop in egg density after the peak suggests that cod eggs are being retained within the basin. These results are consistent with previous findings (Knutsen et al., 2007). In fjords, where physical structures such as a sill influence the concentration of eggs and the topography of the basin, the same structures may also influence water circulation and prevent dispersal of eggs and larvae.
There was some evidence for a weak shift in egg densities northwards through the strait over time. This temporal effect could be a weak current taking the eggs north, or perhaps a shift in the main spawning area. From comparing home ranges of cod with predictions of their egg density, we conclude that mature cod utilize areas slightly south of the highest densities. This can be an effect of the same weak current taking the eggs from the major spawning site north to the main hatching basin, or that the tagged cod were slightly unrepresentative of the population, implying that several untagged cod spawned slightly north of the study area.
The sample of mature fish that we tracked had a high degree of site fidelity. Not only did they confine themselves to the strait, but also to certain areas within the strait, utilizing areas much smaller than the total area available (160 ha). The home ranges were close to areas with high egg densities, although just south of the "hotspot". Manual tracking confirmed that the fish stayed in the strait throughout the spawning season. Through spawning years, fish need to show either year-round site fidelity or accurate homing to their natal spawning area. However, our few adult fish precludes us drawing strong conclusions about homing tendencies in the population.
A sedentary behaviour of spawning fish and an indication of local retention of eggs is clear from our analysis. Local populations of coastal cod may spawn inshore to prevent advection of eggs and larva (Wroblewski et al., 2005), and the results of this study demonstrate the effects of both mechanisms on a much smaller scale then previously documented (Bradbury et al., 2000, 2003; Green and Wroblewski, 2000), in accordance with the observed geographic scale of earlier studies on population structure.
|
Acknowledgements |
|---|
We acknowledge the support we received from the Institute of Marine Research during the various stages of fieldwork. Notable support was received form Svein Erik Enersen and Henry Knutsen, as well as from Terje Jørgensen and Jan Tore Øvredal. Harry Andreassen and Asbjørn Vøllestad both provided valuable input to the project during its various stages. Comments provided by David Robichaud and two anonymous reviewers helped us sharpen our argument. We acknowledge the financial support received from the Institute of Marine Research and the University of Oslo (through the Centre for Ecological and Evolutionary Synthesis, CEES). The work forms part of the project "Dynamics and genetics of oceanic—coastal cod population complexes", funded through the Norwegian Research Council, NRC.
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Blundell G. M., Maier J. A. K., Debevec E. M. Linear home ranges: effects of smoothing, sample size, and autocorrelation on kernel estimates. Ecological Monographs (2001) 71:469–489.[CrossRef][Web of Science]
Bradbury I. R., Snelgrove P. V. R., Fraser S. Transport and development of eggs and larvae of Atlantic cod, Gadus morhua, in relation to spawning time and location in coastal Newfoundland. Canadian Journal of Fisheries and Aquatic Sciences (2000) 57:1761–1772.
Bradbury I. R., Snelgrove P. V. R., Pepin P. Passive and active behavioural contributions to patchiness and spatial pattern during the early life history of marine fishes. Marine Ecology Progress Series (2003) 257:233–245.[Web of Science]
Bridger C. J., Booth R. K. The effects of biotelemetry transmitter presence and attachment procedures on fish physiology and behavior. Reviews in Fisheries Science (2003) 11:13–34.[CrossRef][Web of Science]
Clark D. S., Green J. M. Activity and movement patterns of juvenile Atlantic cod, Gadus morhua, in Conception Bay, Newfoundland, as determined by sonic telemetry. Canadian Journal of Zoology (1990) 68:1434–1442.
Comeau L. A., Campana S. E., Castonguay M. Automated monitoring of a large-scale cod (Gadus morhua) migration in the open sea. Canadian Journal of Fisheries and Aquatic Sciences (2002) 59:1845–1850.
Cote D., Moulton S., Frampton P. C. B., Scruton D. A., Mckinley R. S. Habitat use and early winter movements by juvenile Atlantic cod in a coastal area of Newfoundland. Journal of Fish Biology (2004) 64:665–679.[CrossRef][Web of Science]
Dahl F. E., Danielssen D. S. Resipientundersøkelser i Arendalsområdet i perioden 1975–1979. Flødevigen meldinger (1986) 5:1–65. (in Norwegian).
Danielssen D. S., Gjøsaeter J. Release of 0-group cod, Gadus morhua L. on the southern coast of Norway in the years 1986–1989. Aquaculture and Fisheries Management (1994) 25:129–142.
De Solla S. R., Bonduriansky R., Brooks R. J. Eliminating autocorrelation reduces biological relevance of home range estimates. Journal of Animal Ecology (1999) 68:221–234.[CrossRef][Web of Science]
Ehrenbaum E. Eier und Larven von Fishen. In: Nordisches Plankton 1 (1905) Kiel: Lipsius & Tischer.
Espeland S. H. Long term movement patterns and area utilization by coastal cod (Gadus morhua L.). In: MS thesis (2002) Norway: University of Oslo.
Green J. M., Wroblewski J. S. Movement patterns of Atlantic cod in Gilbert Bay, Labrador: evidence for bay residency and spawning site fidelity. Journal of the Marine Biological Association of the UK (2000) 80:1077–1085.[CrossRef]
Hansteen T. L., Andreassen H. P., Ims R. A. Effects of spatiotemporal scale on autocorrelation and home range estimators. Journal of Wildlife Management (1997) 61:280–290.[CrossRef][Web of Science]
Hastie T. J., Tibshirani R. J. Generalized Additive Models. (1990) London: Chapman and Hall.
Hastings A., Botsford L. W. Persistence of spatial populations depends on returning home. (2006) 103. Proceedings of the National Academy of Sciences of the USA. 6067–6072.
Hawkins A. D., MacLennan D. N., Urquhart G. G., Robb C. Tracking cod Gadus morhua L. in a Scottish sea loch. Journal of Fish Biology (1974) 6:225–236.[CrossRef][Web of Science]
Hemson G., Johnson P., South A., Kenward R., Ripley R., Macdonald D. Are kernels the mustard? Data from global positioning system (GPS) collars suggests problems for kernel home-range analyses with least-squares cross-validation. Journal of Animal Ecology (2005) 74:455–463.[Web of Science]
Hiemstra W. H. A correlation table as an aid for identifying pelagic fish eggs in plankton samples. Journal du Conseil Permanent International pour l'Exploration de la Mer (1962) 27:100–108.
Hurley P. C. F., Campana S. E. Distribution and abundance of haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua) eggs and larvae in the waters off southwest Nova Scotia. Canadian Journal of Fisheries and Aquatic Sciences (1989) 46:103–112.
Jennrich R. I., Turner F. B. Measurement of non-circular home range. Journal of Theoretical Biology (1969) 22:227–237.[CrossRef][Web of Science][Medline]
Kenward R. E., Hodder K. H. Ranges V. An analysis system for biological location data. (1996) Wareham, UK: ITE.
Kjesbu O. S. The spawning activity of cod, Gadus morhua L. Journal of Fish Biology (1989) 34:195–206.[CrossRef][Web of Science]
Knutsen H., Jorde P. E., André C., Stenseth N. C. Fine-scaled geographical population structuring in a highly mobile marine species: the Atlantic cod. Molecular Ecology (2003) 12:385–394.[CrossRef][Medline]
Knutsen H., André C., Jorde P. E., Skogen M. D., Thuroczy E., Stenseth N. C. Transport of North Sea cod larva into the Skagerrak coastal population. (2004) 271. Proceedings of the Royal Society of London, Series B (Biological Series). 1337–1344.
Knutsen H., Olsen E. M., Ciannelli L., Espeland S. H., Knutsen J. A., Simonsen J. H., Skreslet S., Stenseth N. C. Marine Ecology Progress Series (2007) 333:249–255.[Web of Science]
Lough R. G., Caldarone E. M., Rotunno T. K., Broughton E. A., Burns B. R., Buckley L. J. Vertical distribution of cod and haddock eggs and larvae, feeding and condition in stratified and mixed waters on southern Georges Bank. Deep-Sea Research Part II (1996) May 1992;43:1875–1904.[CrossRef]
Lucas M. C., Johnstone A. D. F. Observations on the retention of intragastric transmitters, and their effects on food-consumption, in cod, Gadus morhua L. Journal of Fish Biology (1990) 37:647–649.[Web of Science]
Mohr C. O. Table of equivalent populations of North American small mammals. The American Midland Naturalist (1947) 37:223–249.[CrossRef]
Olsen E. M., Knutsen H., Gjøsaeter J., Jorde P. E., Knutsen J. A., Stenseth N. C. Life-history variation among local populations of Atlantic cod from the Norwegian Skagerrak coast. Journal of Fish Biology (2004) 64:1725–1730.[CrossRef][Web of Science]
Robichaud D., Rose G. A. Multiyear homing of Atlantic cod to a spawning ground. Canadian Journal of Fisheries and Aquatic Sciences (2001) 58:2325–2329.
Robichaud D., Rose G. A. Migratory behaviour and range in Atlantic cod: inference from a century of tagging. Fish and Fisheries (2004) 5:185–214.[CrossRef][Web of Science]
Rooney S. M., Wolfe A., Hayden T. J. Autocorrelated data in telemetry studies: time to independence and the problem of behavioural effects. Mammal Review (1998) 28:89–98.[CrossRef][Web of Science]
Seaman D. E., Powell R. A. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology (1996) 77:2075–2085.[CrossRef][Web of Science]
Seaman D. E., Millspaugh J. J., Kernohan B. J., Brundige G. C., Raedeke K. J., Gitzen R. A. Effects of sample size on kernel home range estimates. Journal of Wildlife Management (1999) 63:739–747.[CrossRef][Web of Science]
Swihart R. K., Slade N. A. Testing for independence of observations in animal movements. Ecology (1985) 66:1176–1184.[CrossRef][Web of Science]
Waples R. S. Separating the wheat from the chaff: patterns of genetic differentiation in high gene flow species. Journal of Heredity (1998) 89:438–450.
Ward R. D. The importance of identifying spatial population structure in restocking and stock enhancement programmes. Fisheries Research (2006) 80:9–18.[CrossRef][Web of Science]
Winger P. D., Walsh S. J. Tagging of Atlantic cod (Gadus morhua) with intragastric transmitters: effects of forced insertion and voluntary ingestion on retention, food consumption and survival. Journal of Applied Ichthyology (2001) 17:234–239.[CrossRef][Web of Science]
Winger P. D., McCallum B. R., Walsh S. J., Brown J. A. Taking the bait: in situ voluntary ingestion of acoustic transmitters by Atlantic cod (Gadus morhua). Hydrobiologia (2002) 483:287–292.[CrossRef][Web of Science]
Winkel W. V. Comparison of several probabilistic home range models. Journal of Wildlife Management (1975) 39:118–123.[CrossRef][Web of Science]
Wood S. N. Modelling and smoothing parameter estimation with multiple quadratic penalties. Journal of the Royal Statistical Society, Series B—Statistical Methodology (2000) 62:413–428.[CrossRef]
Worton B. J. Kernel methods for estimating the utilization distribution in home range studies. Ecology (1989) 70:164–168.[CrossRef][Web of Science]
Wroblewski J., Neis B., Gosse K. Inshore stocks of Atlantic cod are important for rebuilding the east coast fishery. Coastal Management (2005) 33:411–432.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||