Humpback shrimp biology in a central coast inlet, British Columbia, Canada.

ABSTRACT A population of humpback shrimp inhabiting Drury Inlet, British Columbia, was surveyed in November 2001 and March 2002 with trawl and trap fishing gear. Trawl catches were more representative of the shrimp population than trap catches in terms of age-structure and sex-stage proportions. Shrimp, as much as 81.4 tons, lived on a variety of benthic habitats, including deeper trawlable muddy and shallower untrawlable rocky areas. The shrimp population was comprised mainly of small age 1 shrimp (58% to 66%); the proportion that would be targeted by commercial fishers (i.e., larger older shrimp) was small (3% to 8%). Individual shrimp in the Drury Inlet population were generally smaller than shrimp reported from other areas of the BC coast. Females had not released their eggs in mid November and most (75%) were egg-bearing in mid March. Many more shrimp were in the transitional stage in March (24.9%) compared with November (0.4%). Shrimp numbers declined in trap catches at depths greater than 60 m. In trawlable areas, shrimp were abundant at depths 31 to 40 m and in rocky areas between 21 and 40 m. Egg-bearing females tended to be shallower in March just prior to egg hatch compared with when they were not carrying eggs in November. Males were collected from a broad depth range (11 to 80 m). Shrimp in the transitional stage were collected between 11 and 70 m, with more individuals being collected in shallower areas (11 to 20 m). The natural mortality rate of the population was high (mean M = 2.0). Mean fecundity was 905 [+ or -] 377 eggs per individual. We used a curvilinear model to describe the relationship between female size and fecundity. Shrimp with microsporidia infections were found in low prevalence (0.24%) in the sampled population. Humpback shrimp may experience competition for food resources from spiny pink shrimp, prawns, crangons, and eualids. Herring and shiner perch may prey on larval shrimp. Other species that may prey on juvenile and adult shrimp include eelpouts, english sole, sand sole, pricklebacks, giant wrymouths, staghorn sculpins, great sculpins, red rock crabs, and graceful crabs.

KEY WORDS: humpback shrimp, Pandalus, biology, British Columbia

INTRODUCTION

Humpback shrimp (Pandalus hypsinotus Brandt 1851) belong to the family Pandalidae and are found in Pacific Ocean waters 5 to 460 m deep (Butler 1980). In western regions of the Pacific Ocean, humpback shrimp are present from the western Bering Sea (Makarov 1941) to the Sea of Japan and Korea Strait (Vinogradov 1950). In eastern regions, humpback shrimp are present in the Aleutian Islands (Bergstrom 2000) and from Norton Sound, Alaska, to Puget Sound, Washington (Rathbun 1904). In British Columbia (BC), humpback shrimp are found in small isolated populations in protected inshore waters in inlets. These shrimp are protandric hermaphrodites; they generally begin their mature lives as small males. As they grow, they undergo a transition and change into females (Butler 1980).

In BC, until the mid 1990s, humpback shrimp were generally harvested as bycatch in commercial shrimp trawl and trap fisheries (Butler 1980, Boutillier & Nguyen 1999). However, as new markets developed, including the more lucrative live market, the harvesting of humpback shrimp has increased as commercial fishers began targeting isolated populations. Assessment and management of humpback shrimp fisheries are still in developmental phases and have not kept pace with the rapidly expanding targeted fisheries. A better understanding of humpback shrimp populations in BC waters will allow stock assessment biologists and fishery managers to deal more effectively with recruitment overfishing concerns.

In this study, our objective is to enhance the current understanding of humpback shrimp biology. We conducted two research cruises in a central coast inlet in BC and surveyed a population of humpback shrimp using trawl and trap fishing techniques at two points in time. We considered the selectivity of trawl and trap fishing gear, determined the biomass of shrimp in the inlet, and studied biological characters of the population including age, growth, sex, natural mortality, fecundity, diseases, and parasites. Competition and predation were inferred by the presence of by catch species.

METHODS

Study Site

The population of humpback shrimp chosen for this study inhabit Drury Inlet, located northeast from Port Hardy across Queen Charlotte Strait on BC's mainland coast (50[degrees]54' 14N, 127[degrees]01'64W) (Fig. 1). Drury Inlet is approximately 21 km long and 3.2 km wide, the upper portion of the inlet, Actaeon Sound, is 14 km long. Maximum depth is 101 m.

[FIGURE 1 OMITTED]

Sampling Methods

Each survey was conducted using trawl and trap fishing techniques as assessment sampling tools. Substrate type and accessibility dictated the sampling method used. Accordingly, Drury Inlet was divided into 2 areas, "trawlable" and "untrawlable" (Fig. 1). The trawlable area was defined as the portion of the inlet that was accessible to the trawl vessel and where depth exceeded 50 m. The untrawlable area was defined as near-shore areas from the shoreline to a depth of 50-m and portions of the inlet inaccessible to the trawl vessel. The two surveys, each 6 days in duration, occurred in mid November 2001 and mid March 2002.

Trawl Sampling

Trawl sampling was conducted from the 25-m Canadian Coast Guard Fisheries Research Vessel NEOCALIGUS. This vessel was equipped with a 17.7-m high-rise otter trawl and 1.7-m combination trawl doors. Trawl gear specifications include 17.7-m head and foot ropes, a 3.7-m rise, and a 3.8-cm poly mesh net with a 0.6-cm inch liner in the codend. A Nordmore separator grate (fish exclusion device) was used to reduce bycatch.

Tow locations were predetermined using a systematic sampling design. For the November 2001 and March 2002 surveys, 6 sites were trawled twice for a total of 12 tows each sampling period (Fig. 1). Tow times were generally 20 min in length except in cases when the Fishing Master determined that the trawl net may encounter untrawlable bottom. This occurred in four instances with the resulting tow times approximately 15 min.

Total catch from each trawl was weighed to the nearest 0.1 kg. A subsample of the catch from each trawl was randomly selected, and the humpback shrimp were separated and weighed. Catch weights were determined by applying the ratio in the subsample to the total weight in the trawl catch. Approximately 100 humpback shrimp were retained from each tow for length-frequency (LF) analysis.

Trap Sampling

Trap sampling occurred in both trawlable and untrawlable areas. Traps were deployed either from the NEOCALIGUS or from a 8-m skiff. All traps were plastic coated stainless steel, cone stacking, three ring frame traps with a 63.5-cm top diameter, 76.2-cm bottom diameter, and 30.5 cm high. All traps had three tunnels. Three different mesh sizes were used. The small mesh (SM) trap was fitted with 1.6-1.9 cm knotless black web, the medium mesh (MM) trap was fitted with 1.9-2.9 cm (about 1.3 x 1.3 cm) nylon web, and the large mesh (LM) trap was fitted with 3.6 (1.9 x 1.9 cm) nylon web. The SM trap is a shrimp research trap, the MM trap is a prawn research trap, and the LM trap is a commercially available prawn trap.

A total of 12 traps were fished on a single ground line. Each set consisted of 4 traps of each mesh size alternated along the length of the ground line. The trap type attached to the start of a ground line was assigned randomly to negate any bias associated with catchability of traps being located first and last on a ground line. Traps were baited with cat food grade canned tuna. Baits were replaced after each trap soak. Trap sets were soaked overnight for 16-25 h.

In November 2001 and March 2002, 29 trap strings were deployed each sampling period throughout Drury Inlet. Twelve trap strings were set at tow locations (6 sites, 2 replicates), 7 trap strings were set in untrawlable areas near tow locations, 5 trap strings were set in other areas of Drury Inlet, and 5 trap strings were set in the upper reaches of the inlet (Fig. 1).

For each trap catch humpback shrimp were sorted from other species, individually sexed and total numbers determined for each sex stage, and total shrimp weight determined. Shrimp for LF analyses were collected for each trap type on every string. For SM and MM trap models, which collected numerous humpback shrimp because of their smaller mesh sizes, approximately 35-40 shrimp were retained per string for LF analysis. For the LM trap model, all shrimp in each trap were retained for LF analysis.

Biomass Estimation

Geo-referenced modeling of the systematic survey data was used to estimate humpback shrimp biomass in Drury Inlet. For the biomass estimation procedure, we used the previously described "trawlable" and "untrawlable" area divisions (Fig. 1). Estimates of biomass were calculated independently for each of the two areas and then summed to provide an overall estimate of humpback shrimp biomass in Drury Inlet.

Within the trawlable area, spatial analysis was used to estimate humpback shrimp biomass based on catch densities determined from trawl and trap catches. The surface area of the inlet was divided into grid blocks each with an area of 0.25 square nautical miles. Catch densities from the tows were determined, and the weight density information from each sample tow was assigned to the grid block where the center point of the tow occurred.

Some sections within the trawlable area could not be sampled with trawl gear because the substrate was rocky or there was debris on the sea floor. Estimates of density in these sections were determined using traps. Trap catch rates were equated to trawl densities by placing control traps next to trawl locations. This trawl catch/trap catch relationship was used to estimate a weight density at each of the trap locations, and this density was applied to the appropriate grid block. This same method was also used to apply a weight density to grid blocks located just outside the trawlable area where trap catch data were available.

A sector geospatial interpolation was then used to calculate values for the unsampled grid blocks in the trawlable area. The sector interpolation examined an area within a circle with a radius of 10 grid blocks, with the target grid block (the block for which the value was being calculated) in the center of the circle. The circle was divided into six sectors, and the value for the target grid block was estimated using a distance-weighted average of the nearest sample in each of the sectors. Thus, samples closer to the target grid block had a greater influence on the interpolated value. Once blank grid blocks were filled with an interpolated value, the index of biomass for the entire trawlable area was calculated by summing the values of each grid block. This spatial analysis was done using the CompuGrid GIS software package.

Age Structure and Growth

Carapace lengths (CL) of humpback shrimp were measured using electronic calipers. Age composition of a sample was determined using length frequency (LF) modal analysis (Schnute & Fournier 1980). A minimum of 1,000 shrimp was included in each sample for LF analysis.

Sex

Humpback shrimp sex was determined by examining the endopods of the first and second pleopods (see Butler 1980). Sex was recorded as immature, male, transitional, female, egg-bearing female, or spent female (a female that has released her eggs).

Natural Mortality Rate Estimation

Three sources of data were used as indices of changes in abundance over the 4-mo period: (1) SM and MM trap catch per unit effort (CPUE) of ages 3+ humpback shrimp; (2) SM and MM trap CPUE of female humpback shrimp (including egg-bearing and spent individuals); and (3) the number of ages 3+ animals determined by the biomass estimation procedure. Gulland's (1983) estimation procedure for Z was applied using the following equation:

Z = -(12/(t2 - t1)) x ln (n2/n1)

where t1, t2 are sampling times in months; n1, n2 are trap CPUE or ages 3+ shrimp abundance estimates determined at two points in time. We assumed that Z = M because fishing mortality (F) in Drury Inlet was close to zero.

Fecundity

Female humpback shrimp carrying eggs were randomly collected from the retained trawl and trap samples in March 2002. Thirty humpback shrimp, stratified by size, were selected to determine fecundity estimates. The egg mass from each individual was slightly thawed and then teased apart with tweezers and the eggs visually counted.

Fecundity (number of eggs per individual) was modeled by the following equation (see Charnov 1979, Parsons & Tucker 1986, Hannah et al. 1995):

[F.sub.ec] = [al.sup.b]

where a = intercept with the abscissa; l = shrimp CL (mm); b = slope.

Diseases and Parasites

Humpback shrimp with the appearance of milky-white hemolymph visible between the head and abdomen region were examined from trap catches in trawlable areas in March 2002. Shrimp with abnormal white coloring were collected March 15 to 20, 2002 and preserved in Davidson Solution for analyses by the histology laboratory at the Pacific Biologic Station, Nanaimo, BC. Shrimp were examined for branchial isopods in November 2001 and March 2002.

Competition and Predation

Bycatch species collected by trawl and trap fishing gear were quantified to infer possible competition and predation processes.

RESULTS

Selectivity of Sampling Methods

Sex composition of humpback shrimp in trawl catches was similar during the two sampling periods (Table 1, Fig. 2). This consistency was not observed in trap catches. Even SM traps caught larger humpback shrimp than trawl gear did from the same habitat type (i.e., trawlable areas), especially in November 2001. At both sampling intervals, catches from trap gear showed higher proportions of female shrimp and lower proportions of male and transitional shrimp in comparison to trawl catches.

[FIGURE 2 OMITTED]

Biomass Estimation

Humpback shrimp biomass in the trawlable area of Drury Inlet was slightly lower in November 2001 (24.3 tons) compared with March 2002 (28.9 tons) (Table 2). However, biomass in the untrawlable area increased in March (52.5 tons) compared with November (19.5 tons). This resulted in a much higher humpback shrimp biomass estimate for the entire inlet in March (81.4 tons) compared with November (43.8 tons).

CPUE estimates derived using standardized SM trap catches were used to index changes in humpback shrimp abundance. CPUE estimates remained relatively unchanged in the trawlable area in November and March, but increased in rocky areas and the upper inlet in March (Table 3).

Decreases in the numbers of female and ages 3+ shrimp were observed between the November and March sampling periods (Table 4).

Age Structure and Growth

Biological data determined from humpback shrimp collected by trawl gear were considered to be the most representative of the population because trawl catches were less selective than trap catches. The age structure of the humpback shrimp population in Drury Inlet was similar in November 2001 and March 2002 (Fig. 3). The largest component of the population was age 1 individuals (58% in November, 66% in March); large-sized ages 3+ shrimp constituted only a small component of the population (8% in November, 3% in March; Table 5, Fig. 3).

[FIGURE 3 OMITTED]

The relationship between shrimp age and size is depicted using the von Bertalanffy growth model (Fig. 4, Table 5). CL of immature shrimp ranged from 9.4 to 12.6 mm; animals smaller than this will pass through the trawl net. Male shrimp were approximately 18.5-mm CL. Late-stage transitionals (23 mm) were more prevalent and larger in November than the early stage transitionals (19.1 mm) collected in March. The size of females was greater than 22.5 mm (Table 6). Humpback shrimp CL ranged from 9.4-30.7 mm.

[FIGURE 4 OMITTED]

Humpback shrimp increased in size between the sampling intervals (Table 7). Age 0 shrimp increased in size (0.38 mm) the most during the 4-mo period; ages 1 and 2 shrimp showed approximately the same amount of growth (0.11 and 0.10 mm respectively). Size change in ages 3+ shrimp was minimal (0.03 mm) between November and March.

The calculated length-weight relationship for humpback shrimp (n = 1,003) in Drury Inlet is:

Log total P. hypsinotus body weight = 2.7456 log CL - 2.859.

A humpback shrimp with a 25-mm CL would have a total body weight of 9.5 g.

Sex Stage Changes Over Time

Approximately 32% of the trawl sampled humpback shrimp population at both time periods was female (Table 1, Fig. 5). In November 2001, 100% of the female shrimp sampled (32.1% of the population) were not carrying eggs. Four months later in March, 5.7% of the females were not carrying eggs, 24% were egg-bearing, and 2.3% had already released their eggs. In November, 68% of the population were either males or transitionals, with most (67.6%) being males and only 0.4% transitionals. By March, 24.9% of the population was in the transitional state.

[FIGURE 5 OMITTED]

Sex Stages At Depth

During November 2001, humpback shrimp collected in standardized traps were most abundant between 41 and 60-m and less abundant in shallow water (Fig. 6). During March 2002, shrimp were abundant between 11 and 60-m. (Most of the high shrimp abundances in March were observed in the upper reaches of the inlet). At both time periods, shrimp abundance declined at depths >60-m. In trawlable areas, humpback shrimp were most abundant between 31 and 40-m; abundance generally declined at depths greater than this. In rocky areas, shrimp were more abundant at shallower depths 21 to 40-m. In the upper reaches, shrimp abundance was highest between 21 and 30-m and decreased at greater depths.

[FIGURE 6 OMITTED]

In November 2001, female shrimp were most abundant between 41 and 60-m (Fig. 7). The number of males increased slightly with depth, although the number collected in traps was low. In March 2002, most females were egg-bearing and, although they were collected between I I and 70-m, the highest numbers were caught at 21 to 30-m. More males were collected in March compared with November and their numbers appeared to decline at depth, with most males collected between 11 and 60-m. Transitional stage animals were present to 70-m depth, with more animals collected in shallower areas 11 to 20-m deep.

[FIGURE 7 OMITTED]

Female shrimp sampled from trawlable areas in November 2001 were most abundant at 31 to 40-m and generally less abundant as depth increased but still prevalent down to 80-m. In contrast, males increased in numbers as depth increased. In March 2002, shrimp in all sex stages were most abundant between 31 and 60-m. Shrimp abundance generally decreased at depths >60-m, except for transitional stage shrimp, which were collected consistently between 31 and 70-m.

Female shrimp sampled from rocky areas in November 2001 were most abundant between 21 to 40-m and 61 to 70-m. Males, although caught in small numbers, were more abundant as depth increased. In March 2002, shrimp were generally abundant at depths between 31 and 70-m, with more females and males collected between 31 and 40-m.

Natural Mortality Rate Estimation

Estimates of the Drury Inlet humpback shrimp natural mortality rate (M) ranged from 0.97 to 3.42, depending on the abundance index used (Table 8). The average mortality based on the five indices of abundance was 2.0. This natural mortality rate is equivalent to an annual survival rate of 14%.

Fecundity

The mean CL of female humpback shrimp examined (n = 30) was 23.7 [+ or -] 2.6 mm and the mean fecundity per individual was 905 [+ or -] 377 eggs. The number of eggs produced per individual ranged from 317 to 1,897. CL of females examined ranged from 19.5 to 27.9 mm. Larger females typically produced more eggs than smaller females.

By plotting In [F.sub.ec] as a function of In l we estimated the slope b using least square regression techniques.

In [F.sub.ec] = 2.5522 In l - 1.3399

where b = 2.5522; a = 0.26187. Figure 8 shows how the curvilinear model [F.sub.ec] = [al.sup.b] describes the actual number of eggs counted on individual female shrimp. Only 42% of the variation in the number of eggs produced by females is accounted for by size.

The mean number of eggs produced per individual in each age class was highest for age 3 shrimp (1,141 eggs) (Table 9). Approximately 90% of the age 3 shrimp carried eggs, compared with 61% of the age 2 and 1.3% of the age 1 shrimp. Thus, most eggs were produced by shrimp ages 2 and 3, with five to six times more eggs being produced by age 2 rather than age 3 females. Total egg production in Drury Inlet was estimated to be 1.69 x [10.sup.9] eggs.

Diseases and Parasites

In March 2002, humpback shrimp (n = 4,124) from 6 trap strings deployed in trawlable areas were examined for abnormal white coloring. Ten shrimp were white-colored (8 females, 2 males) comprising only 0.24% of the sampled population.

Examination of abnormal-looking white-colored shrimp (n = 8) for histologic diseases produced the following results:

* 5 shrimp had very heavy microsporidia infections in the musculature.

* 2 shrimp had very heavy systemic bacterial infections, which resembled tiny cocci-shape intracellular micro-organisms.

* 1 shrimp had necrosis and tissue degeneration in the center area of the hepatopancreas. In association with this area were abnormal looking cell nuclei somewhat characteristic of a viral infection.

No branchial isopods were observed in 9,352 shrimp examined (4,876 shrimp in November and 4,476 shrimp in March).

Competition and Predation

Ten shrimp, 28 fish, and 5 crab species were caught in Drury Inlet with trawl and trap fishing gear. However, most of these animals were caught infrequently or in low numbers. Organisms that constituted >1% of the total catches in terms of weight are displayed in Figure 9 and Figure 10. Abundant shrimp species in Drury Inlet were spiny pink shrimp (Pandalus borealis Kroyer 1838), prawns (Pandalus platyceros Brandt 1851), crangons (Family Crangonidae), and eualids (Eualus spp.). Fish and crabs collected were herring (Clupea pallasi Valenciennes 1847), shiner perch (Cymatogaster aggregata Gibbons 1854), eelpouts (Family Zoarcidae), english sole (Parophrys vetulus Girard 1854), sand sole (Psettichthys melanostictus Girard 1854), pricklebacks (Family Stichaeidae), giant wrymouths (Cryptacanthodes giganteus Kittlitz 1858), staghorn sculpins (Leptocottus armatus Girard 1854), great sculpins (Myoxocephalus polyacanthocephalus Pallus 1811), red rock (Cancer prodactus Randall 1839), and graceful (Cancer gracilis Dana 1852) crabs.

DISCUSSION

Humpback shrimp catches from trawl and trap fishing gear showed a different catch composition in terms of age and sex stage proportions. It has been shown for prawns that there is competition in traps whereby, during longer soak times, larger female shrimp chase small individuals out of traps, effectively biasing catches toward larger shrimp. This competition/predation probably explains some of the age-specific catchability that is reflected in the results for cohort analysis (Boutillier & Sloan 1988). The observed increase in the number of males in traps at greater depths may have resulted from decreased competition with females because these individuals seemed to prefer shallower areas. The proportions of the various sex stages were constant in trawl catches from both sampling periods, indicating that trawl catches probably represented the true sex stage proportions in the sampled population.

Estimating humpback shrimp biomass is complicated because of the broad area and diversity of habitats these animals occupy. Adult humpback shrimp are found at depths between 65 and 82-m on muddy and sandy substrates (Butler 1980, Balsiger 1981). In Drury Inlet, most shrimp were caught at depths <60-m. Shrimp were most abundant on trawlable grounds--an area where at least some of the substrate was comprised of mud because it was observed on traps retrieved to the surface and in the trawl net. In addition, there were sites in the upper reaches of the inlet (Actaeon Sound) where many humpback shrimp were collected and mud was brought to the surface on traps. Only a portion of the humpback shrimp population existed on trawlable grounds. Shrimp were also collected in shallower untrawlable rocky areas highlighting the fact that the use of relative indices of abundances from trap catches in trawlahle and untrawlable areas is critical in providing a broader scale biomass estimate of a humpback shrimp population.

The overall size of individual humpback shrimp in Drury Inlet is generally smaller than shrimp that have been observed in other areas on the BC coast and throughout the world. Humpback shrimp collected on the east coast of Vancouver Island and in Burrard Inlet (Butler 1964) were larger than those shrimp collected in Drury Inlet. Butler (1980) reported 39.6-mm CL as the maximum size of female humpback shrimp in BC. We compared our lengthweight relationship from Drury Inlet to the relationship Butler (1964) determined for humpback shrimp (n = 199) collected from the east coast of Vancouver Island and Burrard Inlet with trawl gear.

Log total P. hypsinotus body weight = 2.78578 log CL - 2.83882.

According to Butler (1964), a humpback shrimp with a 25-mm CL would have a total body weight of 11.4 g, almost 2 g heavier than a Drury Inlet shrimp of the same size. In Belize Inlet, another central coast inlet located north of Drury Inlet, the maximum CL of humpback shrimp collected was 38 mm (authors personal observations 2002) compared with 30.7 mm in Drury Inlet. It is unknown why humpback shrimp in Drury Inlet are smaller than shrimp located in other BC inlets, although environmental conditions such as food availability may be implicated. The variation in humpback shrimp size observed in populations in some inlets on the BC coast suggests that these populations may be distinct, even though they occur in the same general area. Kurata (198 l) reported female humpback shrimp with CL <--50 mm from the Sea of Japan.

Humpback shrimp generally mature as males, although some may mature early as females. We found only 1.3% of age 1 shrimp to be egg-bearing females. Males start their sex change from March to May, at approximately 2 y of age. Mature females are about 30 mo old (Butler 1980). The minimal size change in ages 3+ shrimp observed in Drury Inlet between November and March is to be expected because these animals are females, and they will not molt while they are carrying eggs. However, the somatic growth in males, measured during this time period, should not be considered representative of the growth over the entire year because somatic growth for shrimp is known to be greatest in the spring and summer, whereas in the fall and winter periods energy is generally directed to reproductive growth.

Natural mortality is greatest in egg-laying shrimp. Subsequent survival into the fourth year appears low (Butler 1980). We attributed decreases in the numbers of female and ages 3+ shrimp observed from the November to March sampling periods to natural mortality because fishing pressure in the inlet was low to none. Although the humpback shrimp natural mortality rate in Drury Inlet was determined over a 4-mo fall/winter time frame, which may be a period of higher senescence for female shrimp, natural mortality is generally caused by predation by fish and probably only to an infinitesimal degree by senescence (Bergstrom 2000). Other factors affecting shrimp mortality include larvae food supply and physical factors such as water temperature (Shumway et al. 1985).

We calculated the actual total mortality rate to be 86% (62% to 97%). In comparison, total mortality for humpback shrimp stocks in Alaska was estimated to be 61% (48% to 70%), largely due to predation and fishing (Alaska Department of Fish and Game 1985). Estimates of M for prawns in Howe Sound, BC averaged 0.88 (Boutillier & Bond 2000), which is lower than the average rate (M = 2.0) determined for humpback shrimp in Drury Inlet.

Our findings regarding the egg-bearing period in Drury Inlet were similar to those reported by Butler (1980) where female shrimp were generally not egged in November, but were still egg-bearing in March. Female humpback shrimp in BC become egg-bearing in late November to early December and remain that way until the eggs hatch in late March or April (Butler 1980). The egg-bearing period lasts about 4.5 mo around Vancouver Island (Butler 1964) or 6 mo in the Tatar Strait (Buyanovsky 1999). In Drury Inlet, egg-bearing females were in shallower waters in March compared with nonegged females in November. It may be advantageous for larvae to hatch in shallow water to avoid competition from larger, more mature shrimp, which seem to occupy deeper depths and avoid predators such as fish.

The relationship found in Drury Inlet confirms Kurata's (1981) findings in that fecundity increases with size in humpback shrimp. Many authors suggest that the relationship between shrimp female body length and fecundity is curvilinear rather than linear (Charnov 1979, Parsons & Tucker 1986, Hannah 1995). Charnov (1979) found the coefficient b (slope) to be around 3; this number was based on literature data of different Pandalus species. The value we determined for b in this study was lower (2.55). We found the relationship between CL and fecundity to be nearly linear, although our sample size was low. Some of the variability in relationships between female size and the number of eggs carried can probably be explained by females carrying egg batches in different stages of development and therefore exposed to different levels of egg mortality (Shumway et al. 1985, Bergstrom 2000). Pink shrimp (Pandalus borealis) lose eggs to cannibalistic neighboring shrimp. Eggs are also knocked off by the scavenging activity of the parent (Shumway et al. 1985).

Because humpback shrimp in Drury Inlet are relatively small-sized, the number of eggs produced by individual females should be lower than in other areas. In the Strait of Georgia, the mean egg count for 15 females, ranging in size from 24.7 to 31.2 mm CL, was 2,257 eggs (Butler 1980). From the Alaska coast, the mean estimate from partial counts for 5 females was 4,000 eggs (Hynes 1930). Off Hokkaido, egg counts for shrimp ranging in size from 10.2-15.0-mm CL was 264 to 3,000 eggs (Igarashi 1951). In the Sea of Japan, egg counts for females ranging in size from 30-50mm CL were 1,000 to 9,000 eggs (Kurata 1981).

Parasites found in adult humpback shrimp include protozoan microsporidia, rhizocephalan barnacles, and branchial isopods (Kurata 1964, Boutillier & Nguyen 1999). Microsporidiosis or cotton shrimp disease can infect most marine crustacea, including many species of shrimp and prawns. Microsporidia invade and replace host tissue such as muscle, heart, gonads, gills, and hepatopancreas making the meat soft, unpalatable, and unmarketable. Microsporidia infections are ubiquitous and occur in low prevalence in wild shrimp stocks on the west coast of Canada and may cause low level mortalities (Bower et al. 1994). Butler (1980) reported infections of pink shrimp (Pandalus borealis) with microsporidia.

Branchial isopods appear to infect sidestripe (Pandalopsis dispar) more often than humpback shrimp. The bopyrid isopod is a common parasite on Pandalus borealis (Shumway et al. 1985). Branchial infestation by Bopyroides hippolytes retards growth, accelerates sex change, and results in 0.3% to 0.6% mortality off Canada (Kurata 1964).

The disease conditions found in low prevalence in the Drury Inlet humpback shrimp population probably do not seriously impact the stock. However, one must be cognizant that in wild populations diseased individuals are usually rapidly removed from the population through predation, and most capture methods tend to select healthy specimens. Consequently, the true prevalence of diseases in the Drury Inlet population is unknown.

Spiny pink shrimp were abundant in Drury Inlet and might compete with humpback shrimp for food resources. However, Allen (1966) found that sympatric species often showed differences in food preferences and feeding habits. Closely related species with similar feeding habits may avoid competition by occurring on different substrata.

Shrimp zoeal larvae might be preyed on by herring (Black 1985) and possibly shiner perch; these fish species were more abundant in the inlet in March, at the time when shrimp eggs begin hatching. Demersal fishes are predators of humpback shrimp (Bergstrom 2000). There is evidence that eelpouts (Hart 1973), english sole (Hart 1973, Lassuy 1989), sand sole (Hart 1973, Butler 1980, Miller 1982), and staghorn sculpins (Hart 1973) prey on crustaceans like shrimp. Other fish species such as great sculpins, pricklebacks, and giant wrymouths probably also opportunistically feed on shrimp. Red rock crabs are scavengers and predators, feeding on snails, clams, echinoderms, and crustaceans (Carroll & Winn 1989).

Results from our study of humpback shrimp in Drury Inlet have implications for stock assessment and management of this shrimp species. Trawl gear should be used to collect shrimp to determine accurate unbiased age classes and proportions of sex stages. Both trawl and trap fishing gear are necessary to survey a population because shrimp live in a variety of habitats found on muddy trawlable and rocky untrawlable grounds. Because the overall size of individual humpback shrimp in different stocks may vary, one gear mesh size may not be appropriate for all stocks. Fishers will target larger sized female shrimp, which may constitute the smallest component of the stock. Fishery closures should consider that most females are still egg-bearing in March. Humpback shrimp may experience high natural mortality; consequently, the impact on populations from additional mortality due to fishing needs to be carefully considered. Fecundity data may be critical because the information can be used to develop fixed escapement models based on egg production estimates.

In conclusion, we have presented information regarding the biology of a humpback shrimp population inhabiting a small central coast inlet in BC. As commercial fishers increase their targeting on isolated stocks of humpback shrimp, concerns about recruitment overfishing become more justified. Increasing our understanding of the biology of humpback shrimp will aid in the creation of effective management programs that will help to protect stocks from overfishing.

TABLE 1.
Proportions of humpback shrimp sex stages collected on
trawlable grounds in Drury Inlet, November 2001 and March 2002.

                         Sex Stages Proportions (%)

Date        Fishing
             Gear       Immature       Male     Transitional

Nov 2001     Trawl         0.2         67.4          0.4
Mar 2002     Trawl         0.3         42.8         24.9
Nov 2001     Trap          0.1         10.3          0.7
Mar 2002     Trap          0.0         26.7         17.1

               Sex Stages Proportions (%)

  Date                   Female       Female
            Female     Egg-bearing    Spent          n

Nov 2001     32.1          0.0         0.0         1,235
Mar 2002      5.7         24.0         2.3         1,195
Nov 2001     88.9          0.1         0.0         6.199
Mar 2002     15.2         37.0         3.9         5,976

TABLE 2.
Humpback shrimp biomass (tons) in Drury Inlet,
November 2001 and March 2002.

                          Untrawlable
            Trawlable       Area
              (12.8         (27.2       Total
Date       [km.sup.2])   [km.sup.2])   Biomass

Nov 2001      24.3          19.5        43.8
Mar 2002      28.9          52.5        81.4

TABLE 3.
Comparison of the number of humpback shrimp collected per SM
trap in different habitat types in Drury Inlet, November 2001 and
March 2002.

                           Mean
                          Number      Mean     Mean
                          Shrimp    Soak Time  Depth
Location        Date     Per Trap   (Hrs:Min)   (m)

Trawlable-    Nov 2001      119       18:26      50
Main Inlet    Mar 2002      115       18:51      49
Rocky-        Nov 2001       44       21:33      45
Main Inlet    Mar 2002       86       19:55      47
Upper         Nov 2001       35       23:33      30
Inlet         Mar 2002       82       22:36      31
All Inlet     Nov 2001       76       20:30      45
              Mar 2002       95       20:02      45

TABLE 4.
Humpback shrimp densities by age (number of shrimp in
thousands) in Drury Inlet, November 2001 and March 2002

           Age 1           Age 2         Ages 3+

Date        No.     %      No.     %      No.     %

Nov 2001   4001.9  0.61   1880.2  0.29   675.6   0.10
Mar 2002   9679.9  0.73   3408.5  0.26   216.4   0.02

TABLE 5.
Age structure of humpback shrimp collected by trawl gear in Drury
Inlet, 2001 and 2002.

                      Age Structure
                                          Carapace Size (mm)
                  Carapace   Proportion
Date        Year  Size (mm)    at-age         Min   Max

Nov 2001     0      0-9.4       0.001          9.4  28.3
             1     9.5-19.8     0.58
             2    19.9-23.1     0.34
             3      23.2+       0.08
Mar 2002     0      0-13.8      0.003         10.2  28.4
             1    13.9-20.4     0.66
             2    20.5-23.5     0.31
             3    23.6+         0.03

                    von Bertalanffy Parameters

Date         [L.sub.[infinity]]    K     [T.sub.0]

Nov 2001           26.01          0.75     0.39

Mar 2002           26.71          0.68     0.22

TABLE 6.
Carapace lengths of humpback shrimp sex stages collected by
trawl gear in Drury Inlet, November 2001 and March 2002.

             Sex Stages Mean Carapace Lengths (mm)

                                                Female
                                                 Egg-    Female
Date      Immature  Male  Transitional  Female  bearing   Spent

Nov 2001    12.6    18.5      23.0       23.4      0.0      0.0
Mar 2002    10.5    18.6      19.1       22.5     23.2     22.5

TABLE 7.
Humpback shrimp carapace lengths (mm) by age in Drury Inlet,
November 2001 and March 2002.

Date               Age 0    Age 1    Age 2    Age 3+

Nov 2001             9.4     18.1     21.7      24.7
Mar 2002            10.9     18.6     22.0      24.8
Growth per month     0.38     0.11     0.10      0.03

TABLE 8.
Estimates of natural mortality (M), total mortality (A)
and survival (S) rates for humpback shrimp in Drury Inlet.

   Type          Gear      n1 (1)  n2 (1)  Months   M    A (2)  S (3)

By age         Trap SM       36    18         4    2.08  0.88   0.12
By age         Trap MM       33    17         4    1.99  0.86   0.14
By sex         Trap SM       65    39         4    1.53  0.78   0.22
By sex         Trap MM       47    34         4    0.97  0.62   0.38
By pop. size   Trawl/Trap  675.6   216.4      4    3.42  0.97   0.03

Average                                            2.00  0.86   0.14

(1) n1, n2 are trap CPUE or age 3 shrimp density estimates
determined at two points in time.

(2) A is the actual total mortality rate. A = 1 - [e.sup.-z]
where Z = M.

(3) S is the survivial rate. S = [e.sup.-z] where Z = M.

TABLE 9.

Humpback shrimp egg production in Drury Inlet, March 2002.

      Mean No.       % Shrimp
        Eggs         Gravid                  No. Shrimp      Total No.
        Per            or            No.       Gravid           Eggs
Age    Shrimp   n    Spent (1)    Shrimp      or Spent       Produced

  1       545    3      1.3      9,679,900     128,743       70,164,935
  2       676   13     60.6      3,408,500   2,066,574    1,397,004,024
  3     1,141   14     89.7        216,400     194,068      221,431,588
                                                          1,688,600,547
(1) Based on trawl data.

ACKNOWLEDGMENTS

The authors thank the crew of the CCGS NEOCALIGUS and Stock Assessment Division staff A. Phillips, B. Rusch, and S. Head for their help with biological sampling in Drury Inlet. H. Nguyen provided the shrimp biomass estimates. I. Murfitt helped produce the map in Figure 1. G. Meyer from the Histology Lab at the Pacific Biological Station analyzed the shrimp for parasites and diseases. D. Rutherford, D. Paltzat, and S. Shumway provided comments on the manuscript.

LITERATURE CITED

Alaska Department ofFish and Game. 1985. Commercial catches of shellfish by species, subarea and year, 1981-1983, 246 pp. Computer printout provided by Carmine DiCostanzo, Chief, Computer Services Section, Alaska Dep. Fish and Game, P.O. Box 3-2000, Juneau, AK 99802.

Allen, J. A. 1966. The dynamics and interrelationships of mixed populations of Caridea found off the north-east coast of England. In: H. Barnes, editor. Some contemporary studies in marine sciences. London: George Allen and Unwin. pp. 45-66.

Balsiger, J. W. 1981. A review of pandalid shrimp fisheries in the Northern Hemisphere. In: T. Frady, editor. Proceedings of the international pandalid shrimp symposium, February 13-15, 1979. Kodiak Alaska: University of Alaska. Fairbanks, Sea Grant College Program, Alaska Sea Grant Rep. 81-3. pp. 7-38.

Bergstrom, B. I. 2000. The biology of Pandalus. In: A. J. Southward, P. A. Tyler, C. M. Young & L. Fuiman, editors. Advances in marine biology. Vol. 38. London: Academic Press. pp. 57-228.

Black, E. A. 1985. The stomach contents of Pacific herring in the Queen Charlotte Straits 1978-1981. B. C. Ministry of Environment, Fisheries data report. No. 8. 142 pp.

Boutillier, J. A. & N. A. Sloan. 1988. Trap mesh selectivity in relation to the legal size regulation for prawn (Pandalus platyceros) in British Columbia. J. Shellfish Res. 7(3):427-431.

Boutillier, J. A. & H. Nguyen. 1999. Pandalus hypsinotus Humpback shrimp. A review of the biology and a recommended assessment framework for a directed fishery. Canadian Stock Assessment Secretariat Research Document 99/067. 22 pp.

Boutillier, J. A. & J. A. Bond. 2000. Using a fixed escapement strategy to control recruitment overfishing in the shrimp trap fishery in British Columbia. J. Northw. Atl. Fish. Sci 27:261 271.

Bower, S. M., S. E. McGladdery & I. M. Price. 1994. Synopsis of infectious diseases and parasites of commercially exploited shellfish. Annu. Rev. Fish Dis. 4:1-199.

Butler, T. H. 1964. Growth, reproduction and distribution of Pandalid shrimps in British Columbia. J. Fish. Res. Bd. Canada. 21(6):1403-1452.

Butler, T. H. 1980. Shrimps of the Pacific coast of Canada. Can. Bull. Fish. Aquat. Sci. 202:1-280.

Buyanovsky, A. I. 1999. Biology and distribution of humpback shrimp (Pandalus hypsinotus) in the Tatar Strait (the Sea of Japan). Scientific council meeting. Joint NAFO/ICES/PICES Symposium on Pandalid Shrimp Fisheries. NAFO SCR Doc. 99/79. 12 pp.

Carroll, J. C. & R. N. Winn. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Southwest)--brown rock crab, red rock crab, and yellow crab. US. Fish Wildl. Serv. Biol. Rep. 82(11.117). US. Army Corps of Engineers, TR EL-82-4:16 pp.

Charnov, E. L. 1979. Natural selection and sex change in pandalid shrimp: test of a life history theory. Am. Nat. 113:515-734.

Gulland, J. A. 1983. Fish stock assessment: a manual of basic methods. New York: Wiley. 223 pp.

Hannah, R. W. 1995. Variation in geographic stock area, catchability, and natural mortality of ocean shrimp (Pandalus jordani): some new evidence for a trophic interaction with Pacific hake (Merluccius productus). Can. J. Fish. Aquat. Sci 52:1018-1029.

Hannah, R. W., S. A. Jones & M. R. Long. 1995. Fecundity of the ocean shrimp (Pandalus jordani). Can. J. Fish. Aquat. Sci. 52:2098-2107.

Hart, J. L. 1973. Pacific fishes of Canada. Fisheries Res. Bd. Can. Bulletin 180. 740 pp.

Hynes, F. W. 1930. Shrimp fishery of southeast Alaska. Append. Rep. US. Comm. Fish. 1929. pp. 1-18.

Igarashi, T. 1951. Studies on Pandalus hypsinotus Brandt in Funka Bay, Hokkaido. Fac. Fish. Hokkaido Univ. Bull 2(1):1-9.

Lassuy, D. R. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)--English sole. US. Fish Wildl. Serv. Biol. Rep. 82(11.101). US. Army Corps of Engineers, TR EL-82-4. pp. 17.

Kurata, H. 1964. Larvae of decapod crustacea of Hokkaido. 3. Pandalidae. Bull. Hokkaido Reg. Fish. Res. Lab. 28:23-24.

Kurata, H. 1981. Pandalid shrimp fisheries of Japan. In: T. Frady, editor. Proceedings of the International Pandalid Shrimp Symposium. February 1979, Kodiak Alaska: University of Alaska. Fairbanks, Sea Grant Report 81-3. pp. 89-159.

Makarov, V. V. 1941. The decapod Crustacea of the Bering and Chukchi seas. Invest. Far East Seas USSR. 1:111-163.

Miller, B. S. 1982. Stomach contents of adult starry flounder and sand sole in East Sound, Orcas Island, Washington. J. Fish. Res. Board Can. 24:2515-2526.

Parsons, D. G. & G. E. Tucker. 1986. Fecundity of northern shrimp, Pandalus borealis (Crustacea, Decapoda) in areas of Northwest Atlantic. Fish. Bull. US. 84:549-558.

Rathbun, M. J. 1904. Decapod crustaceans of the northwest coast of North America. Harriman Alaska Exped. Ser. 10:1-210.

Schnute, J. & D. Fournier. 1980. A new approach to length-frequency analysis: growth structure. Can. J. Fish. Aquat. Sci. 37:1337-1351.

Shumway, S. E., H. C. Perkins, D. F. Schick & A. P. Stickney. 1985. Synopsis of biological data on the pink shrimp, Pandalus borealis Kroyer, 1838. NOAA Technical Report. NMFS 30:58.

Vinogradov, L. G. 1950. The classification of shrimps, crabs and other decapods from the Far East. Izv. Tikhookean. Nauchno Issled. Inst. Rybn. Khoz. Okeanogr. 33:179-358.

JASON S. DUNHAM, * KEN H. FONG AND JAMES A. BOUTILLIER

Fisheries and Oceans Canada, Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, British Columbia, V9T 6N7

* Corresponding author. E-mail: [e-mail]