Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries

G Model

ARTICLE IN PRESS

FISH-4546; No. of Pages 7

Fisheries Research xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Fisheries Research journal homepage: www.elsevier.com/locate/fishres

Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries M. Elizabeth Conners a,∗ , Michael Levine b,1 a b

National Marine Fisheries Service, Alaska Fisheries Science Center, 7600 Sand Point Way NE, Seattle, WA 98115, United States Ocean Associates, Inc. 4007 N. Abingdon Street, Arlington, VA 22207, United States

a r t i c l e

i n f o

Article history: Received 24 May 2016 Received in revised form 9 September 2016 Accepted 12 September 2016 Handled by A.E. Punt Available online xxx Keywords: Octopus Enteroctopus dofleini Discard mortality Giant pacific octopus North pacific RAMP

a b s t r a c t Octopus are caught incidentally in several US federally-managed trawl, longline, and pot fisheries in Alaska. The majority caught are giant Pacific octopus Enteroctopus dofleini. Recent changes in fisheries management in Alaska have resulted in the creation of an octopus species complex with annual catch limits, leading to increased interest in management and catch accounting for this data-poor assemblage. This study characterized the incidental octopus catch in Alaska groundfish fisheries and the mortality rate of octopus caught and discarded at sea. Onboard fisheries observers collected data on octopus weight, sex, and condition at discard in a variety of Alaska groundfish fisheries from 2006 to 2011. A field study aboard a commercial pot-fishing vessel examined delayed mortality resulting from the capture process in giant Pacific octopus during routine pot fishing. Octopus incidental catch varied widely in size and condition at capture for various fishing gear types. Vessels fishing using pot gear captured larger octopus than vessels using longline or trawl gear. Initial condition at capture was best in pot gear, with over 90% of octopus discarded from pot vessels alive in excellent condition. Octopus taken in trawl gear had the highest immediate mortality rate, with 68–94% dead or injured at discard. Giant Pacific octopus held for 24–60 h following pot capture showed no signs of delayed mortality or decline in condition. These results suggest that assuming 100% mortality of discarded octopus may overestimate fishing impacts. © 2016 Published by Elsevier B.V.

1. Introduction Commercial fisheries capture a substantial amount of nontarget catch that is discarded, with a weight equal to approximately 28% of the total landed tonnage in the United States (Harrington et al., 2005). While catch of minor species may be unintended, this catch can be substantial enough to impact populations of the nontarget species (Zhou et al., 2011). Effective fisheries management must account for discarded incidental catch as well as retained catch. Catching, handling, and discarding practices can lead to immediate mortality of incidental catch on deck as well as mortality following discard, termed delayed mortality (Davis, 2002; Stoner, 2012). The most common conservative management approach is to assume 100% mortality of all incidental catch. However, this approach will overestimate the impact of the fishery if a species

∗ Corresponding author. E-mail address: [email protected] (M.E. Conners). 1 Current Address: NMFS Alaska Fisheries Science Center, 7600 Sand Point Way NE, Seattle WA 98115, United States.

frequently survives discard and does not experience substantial delayed mortality. Accounting for octopus discards has become increasingly important in the Alaska groundfish fishery, the largest fishery in the United States (National Marine Fisheries Service, 2015). Although there is currently no directed fishing for octopus of any species in this region, there is a large amount of incidental catch in the Bering Sea and Aleutian Islands (BSAI) and Gulf of Alaska (GOA) management regions (Conners et al., 2014; Conners and Conrath, 2015). The majority of the octopus bycatch is taken by pot fisheries for Pacific cod (Gadus microcephalus) in both regions. Annual catch from 2003 to 2015 has ranged from 72 t (t) to 587 t in the BSAI, and 149 t to 1298 t in the GOA (Table 1). Changes in the management of Alaska fisheries under the Magnuson-Stevens Act have led to annual catch limits for octopus in the BSAI and GOA since 2011. Stock assessment of this group is strongly hampered by a lack of information on abundance, distribution, and life history (Reuter et al., 2010). At the present time, data are not available or sufficient to support a model-based assessment for the octopus complex. Accurate catch accounting in the octopus species assemblage requires knowledge of the amount of octopus retained as well as the total mortality of discarded octopus. Total mortality is an estimate

http://dx.doi.org/10.1016/j.fishres.2016.09.010 0165-7836/© 2016 Published by Elsevier B.V.

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model

ARTICLE IN PRESS

FISH-4546; No. of Pages 7

M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

2

Table 1 Incidental catch of octopus (t, all species) in commercial groundfish fisheries for the Bering Sea and Aleutian Islands (BSAI) and Gulf of Alaska (GOA) fisheries management areas (NOAA Fisheries: www.alaskafisheries.noaa.gov). Total Catch (mt) Year

BSAI

GOA

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

269 338 338 351 181 212 72 177 587 86 223 422 444

212 283 149 166 266 339 321 330 927 415 442 1298 970

of the proportion of animals that die as a result of capture and discard, including delayed mortality that occurs after discard (Benoît et al., 2012). A range of capture stresses such as physical trauma due to fishing gear contact, temperature and pressure changes from capture depth to the surface, and exposure to air will influence the mortality of discarded catch (Davis, 2002). Delayed mortality is often estimated using containment experiments in which animals are held in tanks for a period of time following capture to observe mortality rates. This approach has been used with a variety of fish and invertebrate species in the North Pacific, including Pacific halibut (Hippoglossus stenolepis), snow crab (Chionoecetes opilio), and Tanner crab (C. bairdii) (Kaimmer and Trumble, 1998; Stoner et al., 2008). No containment experiments have been conducted for North Pacific cephalopod species, but the octopus Eledone cirrhosa demonstrated a survival rate of approximately 90% over 120–144 h following capture by beam trawl in the North Atlantic (Kaiser and Spencer, 1995). This study describes a fishery observer special project from 2006 to 2011, which collected information about the sizes, sexes, and condition of captured octopus across a range of fishery seasons, areas, and gear types. In addition, a small field containment study was conducted aboard a commercial pot fishing vessel to examine short-term delayed mortality in the fishery where most octopus are caught as bycatch. The goal was to estimate the proportion of octopus that die or show decreased condition during the first 24–48 h following capture. Together, the North Pacific fisheries observer special project and field containment study provide a “first look” at the extent of mortality of incidentally-caught octopus in Alaska groundfish fisheries. A larger study conducted at the Alaska Fisheries Science Center’s Kodiak Research Laboratory used on-shore containment to look at longer term delayed mortality (Conrath and Sisson, 2016). 2. Methods 2.1. Octopus incidental catch The incidental catch of octopus in Alaska is estimated by the National Marine Fisheries Service (NMFS) Alaska Regional Office catch accounting system (Table 1; www.alaskafisheries.noaa.gov). The NMFS directs the deployment of fisheries observers aboard commercial fishing vessels in all Alaska federal water groundfish fisheries to account for retained and incidental catch by recording information on vessel fishing location, effort, and catch composition and amount (Calahan, 2010). Observers are also deployed at shore-based processing plants receiving groundfish deliveries. While exact sampling protocols differ depending on the observed

vessel and gear type, observers generally randomly sampled 33% of the total vessel catch when aboard pot and longline vessels, and observed trawl vessel fractions typically varied from 2% to 29% (Calahan, 2010). Data from most vessel and plant observers are transmitted in real time to an in-season catch accounting system, and total catch by vessel class and regional management area is monitored throughout the year by the Alaska Regional Office. 2.2. Observer special project US federal fisheries observers in the BSAI and GOA collected supplemental data on octopus caught incidentally during fishing operations during 2006–2011. Observers on commercial fishing vessels collected data on all octopus found in catch species composition samples, and observers at processing plants collected octopus data from all octopus encountered during the delivery. Observers did not identify octopus to species. Data were collected in three of the four large marine ecosystems off Alaska (Fig. 1), including the Bering Sea (BS) and Aleutian Islands (AI), which are both within the BSAI management region. In 2006–2007, vessel-based observers noted the visual condition of octopus at the point of catch sample processing as either “alive”, “injured”, or “dead.” Vessel and plant observers also recorded the whole weight and sex of octopus. Sex was identified by looking for the specialized 3rd arm tip in males, and was left as “unidentified” where not determined. In 2008–2009, vessel and plant observers recorded the whole weight and sex of octopus but did not assess condition. In 2010–2011, the condition assessment was modified to more accurately reflect animal condition under normal crew handling procedures. Octopus condition was assessed at the point the vessel usually discarded incidental catch, and condition was not assessed if an octopus did not represent typical handling procedures. Condition was assessed using the viability codes “excellent,” “poor,” and “dead,” as determined using a key based on movement, external injuries, tissue strength, and color. A brief summary of this key is shown in Table 2. At processing plants, observers only recorded octopus sex and weight (noting if octopus was whole or gutted). A range of weighing scales were used by vessel- and plant-based observers, with accuracy from approximately ±0.1 kg–0.5 kg. It is important to note that this effort was not evenly distributed throughout the entire fleet and observer participation was voluntary. Observer special project data are considered an opportunistic look at incidentally caught octopus and not a randomly selected sample. 2.3. Discard mortality field study The experiment was conducted aboard the commercial fishing vessel F/V Aleutian Mariner in January 2013. The Aleutian Mariner is a 118 ft, house-forward vessel rigged for pot fishing, and fishes commercially in the BSAI for crab and Pacific cod. As with other pot boats, the Mariner routinely takes octopus as incidental catch in their Pacific cod fishery and agreed to participate in the project during their normal winter fishing season. Fishing was conducted in the southeast Bering Sea just north of the Alaska Peninsula, at depths ranging from 64 to 95 m (Fig. 1). Octopus were captured in standard 7 × 7 crab pots rigged for the Pacific cod fishery. When an octopus was caught, it was dropped onto a catch sorting table by the fishing crew, and moved from the sorting table to the point of discard at the outboard rail. Typical handling was rough, and octopus were separated from the table and transported as efficiently as possible. These practices are consistent with those observed by the authors on many other pot boats. Octopus sex was noted and weight was measured at the time of capture. Octopus were weighed using a 50 kg hanging scale (±0.5 kg accuracy, Salter, 235-6S-110 model). Octopus condition was then

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model FISH-4546; No. of Pages 7

ARTICLE IN PRESS M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

3

Fig. 1. Large marine ecosystems and fisheries management regions for federal groundfish fisheries off Alaska. Field study location in the southeast Bering Sea is indicated.

Table 2 Summary of three-stage criteria for assessing condition of octopus. Excellent

Poor

Dead

Octopus moves on its own before handling. Arms are very firm and suckers adhere strongly to surfaces. Skin is strongly colored (usually red-orange). No injuries to eyes or mantle, minor injuries to one or two legs only.

Octopus moves only in response to handling. Arms firm, but suckers adhere only weakly to surfaces. Skin color is less intense and changes more slowly. Injuries to several arms or minor injuries to mantle.

Octopus is dead or has serious injuries, including any visible injury to eyes or mantle injuries that penetrate to underlying muscle. Does not move in response to handling. Arms are flaccid, suckers do not adhere to surfaces. Color fading to grey or white.

assessed and graded as Excellent, Poor, or Dead using the same criteria as in the special project conducted by North Pacific fisheries observers in 2010–2011 (Table 2). To keep individuals separated and to prevent escape, each octopus was placed in a lidded tote (82 cm × 52 cm × 43 cm) with holes for water flow. Containers were secured with elastic bands and placed in an insulated tank that was provided with continuous flow-through seawater. Tank water temperature was monitored with a HOBO temperature data logger

(±0.1 ◦ C accuracy; Onset Computer Corporation). Each tank held a maximum of 4 totes. Octopus condition and octopus weight were then reassessed after a minimum holding time of 24 h. A subset of octopus were retained for periods greater than 24 h (ranging from 26 to 60 h total holding time) if space allowed, with condition and weight recorded at the end of the extended holding period. Octopus were not retained while the vessel was in port between fishing trips (Dutch Harbor, AK) due to the possibility of seawater contamination

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model FISH-4546; No. of Pages 7

ARTICLE IN PRESS M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

4

Fig. 2. Size frequency by weight of octopus observed from Bering Sea fisheries using a) bottom trawl nets, b) pelagic trawl nets, c) pot gear and d) longline gear.

with processing plant and vessel effluent. A brief experiment to test tolerance of captured octopus to air exposure was also conducted by leaving two octopus in totes on deck for one to two hours and reassessing their condition after this period.

Table 3 Characteristics of observer sampled octopus bycatch, from all years of the special project combined. Results are shown for each large marine ecosystem: GOA Gulf of Alaska, BS Bering Sea, and AI Aleutian Islands. Gear Type

2.4. Statistical analysis Linear models were used to examine octopus size differences between gear type and region and a generalized linear model (binomial family) was used to examine differences in sex ratio of the observed octopus between gear types, regions, and seasons. Individual octopus each represented a data point. Categorical blocking factors used for the analysis were fishing gear type (bottom trawl, pelagic trawl, pot, and longline gear); ecosystem (Bering Sea, Aleutian Islands, and Gulf of Alaska) and season (the pot fishery for Pacific cod occurs primarily in winter (January–March) and fall (September–November)). Weight data were checked for constant variance before being modeled, and residual and QQ plots from all models were checked to ensure that the data were consistent with model assumptions. Changes in weight of individual octopus during holding in the field study were checked using t-tests. To assure that the catch size of the experiment was adequate to detect mortality, we performed a statistical power analysis. Each octopus held for at least 24 h was coded as either 0 (no mortality or decline in condition) or 1 (observed mortality or decline). Using a binomial probability distribution, we calculated the probability of the observed result if true underlying delayed mortality rates were between 0 and 20%. All statistical analyses were done using R (v. 3.21). 3. Results 3.1. Characteristics of incidental catch Opportunistic collection by fisheries observers from fall 2006 through fall 2011 yielded 4151 recorded octopus in the BSAI and GOA regions, at capture depths ranging from 37 m to 971 m. The majority of octopus were encountered in the BSAI and GOA Alaska pot fisheries. The data record was not included in statistical analyses in cases where sex was unidentified or octopus weight was not recorded.

Bottom Trawl Pelagic Trawl Pots Longline

Sample Size

Average Weight (kg)

Sex Ratio (M:F)

GOA

BS

AI

GOA

BS

AI

GOA

BS

AI

28 57 691 78

125 368 867 274

118 9 97 128

7.3 10.7 14.4 8.2

10.6 3.7 14.4 4.7

6.8 7.5 8.4 5.9

1.1 0.2 2.1 0.5

0.8 0.6 0.9 0.5

0.8 1.0 0.6

The size of octopus captured varied strongly depending on vessel gear type (Table 3, Fig. 2). Patterns in size distribution for the various gear types were similar across all three ecosystems (BS, GOA, and AI). Pelagic trawl nets, which are exclusively used in the walleye pollock (Gadus chalcogrammus) fishery, take >70% small (≤2 kg) octopus. Longlines, used for a variety of fisheries, also tend to capture smaller octopus. Bottom trawl nets, predominantly used in various flatfish and rockfish fisheries as well as in the Pacific cod fishery, capture approximately 20% small (≤2 kg) octopus, but also capture a greater proportion of larger (≥10 kg) octopus than either pelagic trawl nets or longline gear. Pot gear, used to target Pacific cod, captured a greater proportion of large octopus than any other gear type. Comparatively few small octopus were encountered by observers in pot fisheries (Fig. 2). It is unclear whether this is because smaller octopus are not retained in the gear during retrieval or are simply not captured in the pots. The distribution of octopus weights had constant variance, so linear model analysis was used to examine size differences in the observer data. Gear type, octopus sex, and ecosystem all had strongly significant effects on octopus size, and both two and three-way interaction terms of these factors were significant (see Supplementary Table S1 in the online version at DOI: http://dx.doi. org/10.1016/j.fishres.2016.09.010). Pot gear caught the largest size octopus, and pelagic trawl and longline gear caught the smallest. Differences in mean octopus size by gear type were not consistent between ecosystems, reflecting regionally different targets of trawl and longline fisheries. Male octopus were on average slightly larger than female octopus (Fig. 3). Octopus caught in pot gear in the BS and GOA were of similar size, but the average octopus size for other

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model FISH-4546; No. of Pages 7

ARTICLE IN PRESS M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

5

Table 4 Results of observer special project, showing immediate mortality rates and octopus condition at catch sampling or discard by fishing gear type (includes BS, GOA, and AI ecosystem areas). Observer Special Project Data 2006–2007

Condition of Octopus in Catch Sample

Gear

Alive

Injured

Dead

Total

%Alive

Bottom Trawl Pelagic Trawl Pots Longline

41 9 481 180

24 16 61 53

62 112 3 16

127 137 545 249

32.3% 6.6% 88.3% 72.3%

2010–2011

Condition of Octopus at Discard

Gear

Excellent

Poor

Dead

Total

%Excellent

Bottom Trawl Pelagic Trawl Pots Longline

16 8 506 122

11 7 14 7

35 42 16 16

62 58 536 146

25.8% 13.8% 94.4% 83.6%

3.2. Observer condition coding

Fig. 3. Comparison of weight frequency by sex for octopus observed in pot gear fisheries in the a) Bering Sea and b) Gulf of Alaska.

gear types differed between the two ecosystems. Octopus captured in the AI in all gear types were significantly smaller than those in the other ecosystems. Gear type, sex, and ecosystem did not, however, account for the majority of the variation in octopus size: the overall R2 for the linear model, including all interaction terms, was 40%. The interaction terms in linear modeling of octopus size are largely due to differences in target fish and gear design between the ecosystems. Bottom trawls in the BS occur on largely soft bottom surfaces, while trawls in the AI and GOA require heavier gear and occur on harder, more complex, bottom surfaces (Stauffer, 2004). Bottom trawl fisheries in the AI (targeting mainly Atka mackerel Pleurogrammus monopterygius and Pacific cod) and GOA (targeting mainly rockfishes Sebastes sp. and flatfishes) tend to capture smaller octopus than those in the BS (targeting mainly flatfishes and Pacific cod). Similarly, pot fisheries in the BS and GOA capture much larger octopus than those in the AI. Nearly all pot fishing in the BS and GOA is for Pacific cod at relatively shallow depths, while pot fishing in the AI often targets sablefish (Anoplopoma fimbria) at much greater depths. Gear type also had a significant effect on the sex ratio of incidentally caught octopus; while most gear types caught more females than males, pot gear in all regions caught a much higher proportion of males (Table 3). Pelagic trawls and longline gear, which caught a higher proportion of small octopus, caught more females that pot and bottom trawl gears. Analysis also indicated a significant seasonal difference in sex ratios in both BS and GOA, with a higher proportion of males caught during the fall fishing season than during the winter (Supplementary Table S1 in the online version at DOI: http://dx.doi.org/10.1016/j.fishres.2016.09.010).

Observer condition coding of octopus was conducted in two different periods within the study, 2006–2007 and 2010–2011. The results of both periods suggest that immediate mortality of octopus bycatch is often considerably less than 100%, and that the condition of octopus at discard is highly gear-specific (Table 4). Bottom and pelagic trawl nets were associated with high rates of mortality in both the 2006–2007 and 2010–2011 study periods. Octopus caught by longline demonstrated considerably lower immediate mortality and higher condition coding than trawl-caught octopus. Pot-caught octopus had the lowest immediate mortality of all gear types (Table 4). The high proportion of octopus in excellent condition from pot gear is important, as this is the gear type that collects the majority of the bycatch for this assemblage. 3.3. Discard mortality field project The field study was conducted aboard the F/V Aleutian Mariner during the January 2013 Bering Sea Pacific cod season. Commercial cod pots were pulled, emptied, and reset in approximately the same locations twice per day. Fishing took place during three trips to the study area, with a total of 12 fishing days. Weather was better than average for winter Bering Sea fisheries, but air temperatures were generally below freezing. Water temperature in the holding tanks ranged from 3.2◦ to 4.3 ◦ C, averaging 3.61 ◦ C over the course of the study. Octopus were exposed to air on deck for as short a time as possible prior to tank placement (11 min ± 4 min) to minimize effects of air exposure. Forty-three octopus were captured during 12 days of fishing; a rate of approximately 1 per 50 pots pulled. Catch rates were highest at the start of each trip, when the gear had been soaking for one to two days. Octopus captured ranged in size from 5.5 kg to 22.0 kg, with males tending to be slightly larger (15.62 ± 4.72 kg) than females (12.97 ± 3.99 kg). The size frequency of octopus captured during the study period was consistent with pot-caught octopus in the observer special project. Examination of the gonads of retained octopus indicated that immature and mature animals of both sexes were captured. Despite fairly rough handling by the crew (who were instructed to treat octopus as they usually would during fishing), all of the octopus but one were in excellent condition at the normal point of discard. The only octopus rated “poor” displayed weaker muscle tone, slower movement, and lighter color than all other captured octopus. Thirty-six octopus were successfully held for at least 24 h in running seawater tanks. Most octopus had their condi-

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model

ARTICLE IN PRESS

FISH-4546; No. of Pages 7

M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

6

Table 5 Results of field study on octopus condition at discard and short-term delayed mortality. Octopus condition was recorded at capture (0 h) and after being held in flow-through seawater tanks for varying periods. Some octopus were observed at 24 h and again at a later release time. See Table 2 for condition codes. Holding Time (hrs)

Number of Octopus

0 24 25–36 37–48 48–60

36 21 3 2 1

Condition Code Excellent

Poor

Dead

36 21 2 2 1

1 0 1 0 0

0 0 0 0 0

tion assessed at 24 h and were in excellent condition at this point, including the one in initially poor condition. Some specimens were held for longer periods up to 60 h. Though no octopus declined from its initial condition, the octopus initially rated “poor” returned to poor condition the end of its extended holding period (38 h). All of the other octopus held for greater than 24 h were in excellent condition at the end of the holding period (Table 5). Two octopus were left in containers on deck to test tolerance to air exposure. These octopus did not show a decline from “excellent” to “poor” condition until more than an hour of air exposure. 4. Discussion Estimating the magnitude of discard mortality in a fishery requires knowing both the amount of catch discarded and the eventual mortality caused by the catching and handling process (Benoît et al., 2012). Although immediate mortality can be assessed by fisheries observers, assessing delayed mortality is more problematic. A number of approaches have been taken, including field containment experiments with various fish and crab species (Kaiser and Spencer, 1995; Kaimmer and Trumble, 1998; Stoner et al., 2008; Benoît et al., 2010) and lab experiments with fish (Davis and Ottmar, 2006; Davis, 2007). To our knowledge, only one experiment has held a species of octopus (Eledone cirrhosa) following capture (Kaiser and Spencer, 1995). The current study represents a first attempt to examine the immediate mortality of octopus due to capture and handling (the fisheries observer special project) and the short-term delayed mortality of captured octopus under the most common conditions (the field study). In some cases, there may be additional delayed mortality due to increased predation risk while a released animal is returning to its habitat. In the case of E. dofleini, this predation factor might need to be considered for small animals (<2 kg), that have several predators. Large octopus (10–20 kg) are much more predators than prey, however, being eaten only by marine mammals (Livingston et al., 2003). The fisheries observer special project highlights the importance of fishing gear type in the characteristics and discard condition of octopus bycatch: the size frequency, sex ratio, and discard condition of octopus catch were all gear dependent. Pot gear, which is responsible for the majority of octopus incidentally captured in Alaska, is selective for larger octopus and more male octopus than other gear types. The size of captured animals in pot gear suggests that the majority are E. dofleini and are sexually mature (Conrath and Conners, 2014). The reason for the sex selectivity of pot gear is unknown, but may be related to greater territorial movement in reproductive males, reduced feeding in mature females, or depth-based segregation related to reproduction. Both the observer data and the holding experiment indicated that immediate mortality is less than 10% in pot gear. These results are consistent with anecdotal observations by commercial fishermen and fisheries biologists in Alaska. This low mortality is probably a result of short gear retrieval times, rapid sorting of catch on deck, and minimal crushing injury in pot fishing. Low rates of

immediate mortality were also noted by observers in longline fisheries, where many octopus are observed holding onto bait and not “hooked” on gear. In contrast, bottom and pelagic trawl gear is associated with large rates of immediate mortality (68%–94% of octopus rated as dead or injured), most likely due to crushing injuries in the net codend and fish holds and long holding times during sorting and processing. In the field experiment, a total of 36 octopus were held successfully for at least 24 h. All of the octopus that were in excellent condition immediately after capture were still in excellent condition at discard 24 to 60 h later. Total sample size for assessing delayed mortality was small due to a lower than expected catch rate of octopus. Statistical power analysis showed that the probability of the observed result of no mortality out of 36 trials would be very small (p < 0.05) unless the true underlying mortality rate was 8% or larger. The probability of the observed result if the true mortality rate was 10% was 0.022, and the probability of the observed result with a true rate of 15% was 0.003. Thus while we may have missed a small delayed mortality, it is very unlikely that we missed any true short-term mortality over 10%. In the majority of studies of delayed mortality in fish, mortality is greatest in the first 1–5 days of the holding period (Benoît et al., 2012). The absence of observable decline in condition during the first 24–48 h after capture suggests that overall delayed mortality for octopus from this fishing gear is low. Separate long-term delayed mortality studies (Conrath and Sisson, 2016) collected octopus on commercial pot vessels in Kodiak, Alaska and held individuals for 21 days in a running seawater laboratory. This study showed no long-term delayed mortality of uninjured octopus (Excellent condition), and a 50% delayed mortality rate for visibly injured octopus (similar to Poor condition). Published studies of delayed mortality in marine species examine the degree of wounding, reflex or behavioral impairment, and physiological markers that vary with increasing stress as indicators of eventual mortality. The degree of wounding alone may not reflect the sum of physiological or environmental stresses on an organism, making it a poor predictor of eventual mortality (Davis and Ottmar, 2006; Davis, 2007; Stoner, 2012; Benoît et al., 2010). Similarly, attempts to link physiological markers with degree of stress have shown limited concordance with eventual mortality (Davis, 2002; Davis and Schreck, 2005; Lupes et al., 2006; Stoner, 2012). These authors developed a protocol called Reflex Action Mortality Predictor (RAMP). Reflex impairment has proven to be a reliable predictor of eventual mortality in a range of species including sablefish, northern rock sole (Lepidopsetta polyxystra), Pacific halibut, snow and Tanner crabs, and yellowtail flounder (Limanda ferruginea) (Kaimmer and Trumble, 1998; Davis and Ottmar, 2006; Davis, 2007; Stoner et al., 2008; Diamond and Campbell, 2009; Barkley and Cadrin, 2012; Stoner, 2012). The criteria used for this study were based on a similar evaluation of external injury and simple visual measures that reflect reflex impairment; looking at movement, muscle tone, and color of captured octopus was likely a reliable measure of octopus condition.

5. Conclusions The combination of observer data on condition and delayed mortality testing on octopus suggests that overall mortality of discarded giant Pacific octopus from pot and longline gear is very low. This is consistent with the published results of over 90% survival rate for Eledone cirrhosa (Kaiser and Spencer, 1995). At present, catch accounting for all species except Pacific halibut in Alaska assumes a discard mortality of 100%. This assumption may result in a substantial over-estimation of total fishery mortality of octopus captured, especially in the pot fishery that accounts for the major-

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010

G Model FISH-4546; No. of Pages 7

ARTICLE IN PRESS M.E. Conners, M. Levine / Fisheries Research xxx (2016) xxx–xxx

ity of octopus take. Under current regulations in Alaska, when any bycatch species or assemblage reaches its Annual Catch Limit, the target fishery taking that species must be closed. This resulted in an early closure of the BSAI pot cod fishery in fall 2011. Introducing a discard mortality factor into catch accounting for octopus would result in a smaller fraction of the pot gear bycatch being counted towards the Annual Catch Limit, and reduce the possibility of future closures of the economically important Pacific cod fishery. Acknowledgments The observer special project was conducted voluntarily by a number of North Pacific fisheries observers; we are grateful to these observers for taking extra time during their busy days to provide us with valuable data. Funding for the field study was provided by the AFSC’s Cooperative Research Program. Herb Murray and the crew of the Aleutian Mariner were extremely helpful and accommodating during the field project; their good humor with octopus wrangling was much appreciated. Christina Conrath (AFSC Kodiak) and David Scheel (Alaska Pacific University) consulted on the octopus viability key and provided experience and advice on handling and evaluating octopus. Managers Richard Marasco at Ocean Associates in Seattle and Kevin Kaldestadt of Aleutian Mariner LLC provided valuable flexibility in coordinating timing, travel, and budgets for the field project. Elizabeth Chilton, Christina Conrath, and two anonymous reviewers provided helpful comments on the draft manuscript. The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the National Marine Fisheries Service. Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. References Barkley, A., Cadrin, S., 2012. Discard mortality estimation of yellowtail flounder using reflex action mortality predictors. Trans. Am. Fish. Soc. 141, 638–644. Benoît, H., Hurlbut, T., Chassé, J., 2010. Assessing the factors influencing discard mortality of demersal fishes using a semi-quantitative indicator of survival potential. Fish. Res. 106, 436–447. Benoît, H., Hurlbut, T., Chassé, J., Jonsen, I., 2012. Estimating fishery-scale rates of discard mortality using conditional reasoning. Fish. Res. 125–126, 318–330. Calahan, J., 2010. At-sea monitoring of commercial north Pacific groundfish catches: a range of observer sampling challenges. NOAA, NMFS Alaska Fisheries Science Center. Q. Rep. J., 1–5. Conners, M.E., Conrath, C.A., 2015. GOA octopus complex. In: Stock assessment and fishery evaluation report for the groundfish resources of the Gulf of Alaska. North Pacific Fishery Management Council, 601W. 4th St., Anchorage, AK 99501. http://www.afsc.noaa.gov/refm/stocks/assessments.htm.

7

Conners, M.E., Conrath, C.A., Aydin, K.Y., 2014. BSAI octopus complex. In: Stock assessment and fishery evaluation report for the groundfish resources of the Bering Sea and Aleutian Islands. North Pacific Fishery Management Council, 601W. 4th St., Anchorage, AK 99501. http://www.afsc.noaa.gov/refm/stocks/ assessments.htm. Conrath, C.A., Conners, M.E., 2014. Aspects of the reproductive biology of the giant Pacific octopus (Enteroctopus dofleini) in the Gulf of Alaska. Fish. Bull. 112 (4), 253–260. Conrath, C.A., Sisson, N.B., 2016. Delayed discard mortality of the giant Pacific octopus in pot fisheries in the Gulf of Alaska. Fish. Res., submitted. Davis, M., Ottmar, M., 2006. Wounding and reflex impairment may be predictors for mortality in discarded or escaped fish. Fish. Res. 82, 1–6. Davis, M., Schreck, C., 2005. Responses by Pacific halibut to air exposure: lack of correspondence among plasma consitituents and mortality. Trans. Am. Fish. Soc. 134, 991–998. Davis, M., 2002. Key principles for understanding fish bycatch discard mortality. Can. J. Fish. Aquat. Sci. 59, 1834–1843. Davis, M., 2007. Simulated fishing experiments for predicting delayed mortality rates using reflex impairment in restrained fish. ICES J. Mar. Sci. 64, 1535–1542. Diamond, S., Campbell, M., 2009. Linking sink or swim indicators to delayed mortality in red snapper by using a condition index. Mar. Coast. Fish. 1, 107–120. Harrington, J., Myers, R., Rosenberg, A., 2005. Wasted fishery resources: discarded by-catch in the USA. Fish Fish. 6, 350–361. Kaimmer, S., Trumble, R., 1998. Injury, condition, and mortality of Pacific halibut bycatch following careful release by Pacific cod and sablefish longline fisheries. Fish. Res. 38, 131–144. Kaiser, M., Spencer, B., 1995. Survival of by-catch from a beam trawl. Mar. Ecol. Prog. Ser. 126, 31–38. Livingston, P.L., Aydin, K.Y., Boldt, J., Gaichas, S., Ianelli, J., Jurado-Molina, J., Ortiz, I., 2003. Ecosystem Assessment of the Bering Sea/Aleutian Islands and Gulf of Alaska Management Regions. In: Stock assessment and fishery evaluation report for the groundfish resources or the Bering Sea/Aleutian Islands regions. North. Pac. Fish. Mgmt. Council, Anchorage, AK. Lupes, S., Davis, M., Olla, B., Schreck, C., 2006. Capture-related stressors impair immune system function in sablefish. Trans. Am. Fish. Soc. 135, 129–138. National Marine Fisheries Service, 2015. Fisheries of the United States, 2014. U.S. Department of Commerce, NOAA Current Fishery Statistics No. 2014. https:// www.st.nmfs.noaa.gov/commercial-fisheries/fus/fus14/indexReuter. Reuter, R.F., Conners, M.E., Dicosmo, J., Gaichas, S., Ormseth, O., Tenbrink, T.T., 2010. Managing non-target, data-poor species using catch limits: lessons from the Alaskan groundfish fishery. Fish. Mgmt. Ecol. 17 (4), 323–335. Stauffer, G., 2004. NOAA Protocols for Groundfish Bottom Trawl Surveys of the Nation’s Fishery Resources. U.S. Dep. Comm., NOAA Tech. Memo. NMFS-F/SPO-65. Stoner, A., Rose, C., Munk, J., Hammond, C., Davis, M., 2008. An assessment of discard mortality for two Alaskan crab species, Tanner crab (Chionoecetes bairdi) and snow crab (C. opilio) based on reflex impairment. Fish. Bull. U.S. 106, 337–347. Stoner, A., 2012. Assessing stress and predicting mortality in economically significant crustaceans. Rev. Fish. Sci. 20, 111–135. Zhou, S., Smith, A., Fuller, M., 2011. Quantitative ecological risk assessment for fishing effects on diverse data-poor non-target species in a multi-sector and multi-gear fishery. Fish. Res. 112 (3), 168–178.

Please cite this article in press as: Conners, M.E., Levine, M., Characteristics and discard mortality of octopus bycatch in Alaska groundfish fisheries. Fish. Res. (2016), http://dx.doi.org/10.1016/j.fishres.2016.09.010