Improving outplanting designs for northern abalone (Haliotis kamtschatkana): the addition of complex substrate increases survival.
KEY WORDS: crabs, density dependence, emigration, pinto abalone, predation, sea stars, Haliotis kamtschatkana
The northern, or pinto, abalone (Haliotis kamtschatkana Jonas, 1845) is an endangered marine gastropod that occurs in rocky subtidal areas along the Pacific Coast of North America (Committee on the Status of Endangered Wildlife in Canada (COSEWIC) 2009). The northern abalone fishery was successful commercially until 1978, when declines of up to 83% in local abundance led to the banning of all commercial, recreational, and traditional abalone harvesting in British Columbia in 1990 (Campbell 2000). Despite this prolonged closure, northern abalone populations have not recovered and the species was listed as Endangered by COSEWIC in 2009. Major threats to the persistence and recovery of this species include poaching and low recruitment (Fisheries and Oceans Canada 2007). The northern abalone is now so rare in the southern part of its range (Lessard et al. 2007, Egli & Lessard 2011, Lessard & Egli 2011) that males and females are unlikely to be in close enough proximity for broadcast spawning to be successful (Campbell et al. 2003, Seamone & Boulding 2011). This situation suggests that some form of outplanting of captive-bred individuals will be necessary until these southern population are rehabilitated, and has resulted in several related projects that explore how best to accomplish this (Hansen 2011, Read et al. 2012), including the current study.
The release of captive-bred individuals into the wild in an attempt to restore depleted populations is a common method for species recovery (Stoner & Davis 1994, Booth & Cox 2003, Carlsson et al. 2008). However, the survival rates of outplanted hatchery-reared abalone have been highly variable, and predation has been shown to be a major cause of this variation (Shepherd et al. 2000, Tegner 2000). Predation rates may be higher on outplanted individuals because abalone grown in hatcheries lack the natural epiphytes that camouflage wild abalone shells (Schiel & Welden 1987, Schiel 1992). Adult northern abalone are preyed on by the sunflower star (Pycnopodia helianthoides (Brandt, 1835)), the red rock crab (Cancer productus Randall, 1839), the Pacific giant octopus (Enteroctopus dofleini (Wulker, 1910)), and the sea otter (Enhydra lutris Linnaeus, 1758) (Sloan & Breen 1988, Shepherd & Breen 1992, Watson 2000). Laboratory studies by Griffiths and Gosselin (2008) have identified additional predators on juvenile abalone less than 25 mm in shell length (SL) including the sharpnose crab (Scyra acutifrons Dana, 1851), the blackclaw crestleg crab (Lophopanopeus bellus (Stimpson, 1860)), and the pile perch (Rhacochilus vacca (Girard, 1855)).
Several field experiments have used complex habitats in an attempt to reduce predation on outplanted abalone. Dixon et al. (2006) assembled boulders from nearby areas into 1- and 2-layer habitats into which they outplanted small juvenile abalone (Haliotis laevigata). They observed that survival was greater in the 2-layer habitat and they attributed the increased survival to predator refuges provided by the added cryptic habitat. A positive correlation between survival of outplanted juvenile Haliotis midae and habitat containing stacked boulders has also been demonstrated in South Africa by de Waal and Cook (2001). Artificial habitats made of concrete bricks or polyethylene pipes have also been used as outplanting modules for abalone, and survival in these habitats has been relatively high compared with controls (McCormick et al. 1994). On the other hand, James et al. (2007) found no differences in the survival rates of abalone of 10-30 mm in SL outplanted into complex and noncomplex artificial habitats at densities of 2040 abalone/[m.sup.2].
In an effort to rebuild the local abalone population, the Bamfield Huu-ay-aht Community Abalone Project (BHCAP) outplanted 150,000 1-mo-old juveniles (SL, 1 mm) into Barkley Sound in plastic milk crates filled with concrete brick pieces and covered with plastic mesh with openings of 3.18 x 3.18 mm (Oceanlink 2008). The mesh size of these cages would have excluded the predators adult crabs (Cancer productus) and large sea stars (Pycnopodia helianthoides); however, small crabs and sea stars would have been able to pass through the mesh and feed on the abalone (E.G. Boulding, pers. obs.). Nevertheless, microsatellite genotyping of biopsies from wild-captured emergent abalone suggests that some of postlarval and juvenile abalone outplanted by BHCAP survived (Read et al. 2012).
The objective of this study was to understand the mechanisms responsible for high predation rates on higher than ambient densities of outplanted juvenile Haliotis kamtschatkana. The first experiment tested whether removing small predators from cages that exclude large predators increased the survival of outplanted abalone. We hypothesized that juvenile abalone outplanted into cages have poor survival rates because they experience higher than usual predation rates by small predators that take refuge inside the cage. Therefore, we predicted that outplanted juvenile survival would be greater in cages in which predators were removed than in similar cages in which predators were present. The second experiment tested whether adding cobbles and boulders to outplanting plots increased the survival of outplanted abalone. We hypothesized that adding complex substrate to a plot would increase the survival rate of outplanted juveniles because the cobbles and boulders would provide cryptic refuges, thereby reducing mortality from predators. We predicted that survival would be greater in plots with more and smaller crevices compared with plots with less cryptic habitat and larger crevices.
MATERIALS AND METHODS
Small-Predator Removal Cage Experiment
To test experimentally whether the survival of caged outplanted juveniles can be increased by removing small predators from the cages, the number of live juveniles in artificial habitats modules (cages) were compared between small-predator removal and small-predator present treatments, each replicated 10 times. Ten cages were assigned randomly, but permanently, to the small-predator removal treatment and 10 were assigned randomly as controls. The cage design (details in DeFreitas (2003)) consisted of modified commercial crab traps (l m in diameter, 0.3 m tall, with a stainless steel mesh size of 66 x 91 mm) filled with 24 uniform pieces of concrete brick measuring 397 x 97 x 97 mm each (cage weight, approximately 120 kg, see photo in Appendix IA). The stacked bricks provided each cage with approximately 3.5 [m.sup.2] of surface area (DeFreitas 2003). These "habitat module" cages have been used to facilitate population surveys by sampling cryptic juvenile stages (SL, <50 mm (Sloan & Breen 1988)) without disturbing natural boulder habitats (Stoner & Davis 1994, DeFreitas 2003, Dixon et al. 2006). The cage excluded larger predators, but the moderately large mesh holes (approximately 66 x 91 mm) allowed the outplanted juvenile abalone as well as small predators to enter or exit the cage (K. Read, unpubl. data). On July 15, 2008, the cages were placed 5-10 m apart and at depths ranging from 5-7 m in suitable abalone habitat (low to medium wave exposure and the presence of boulders and cobble as well as kelp (Lessard & Campbell 2007)). In total, 20 cages were placed at 2 sites, 10 cages at each site, with the sites approximately 100 m apart. Cages were left on the substrate for 2 wk prior to abalone outplanting to allow a diatom film to form on the bricks.
On August 6, 2008, 1,000 juvenile abalone (mean SL, 22.2 [+ or -] 0.7 mm) were obtained from the BHCAP hatchery where they had been produced by spawning wild broodstock in 2005. The abalone had been held in tanks with low structural complexity that contained nothing but the abalone and the bull kelp on which they fed. This may have influenced their subsequent behavior (Straus & Friedman 2009). Juveniles were sedated by bubbling carbon dioxide through the tanks, and SL was measured using vernier calipers. Ziploc bags containing 50 abalone were transported to the site in coolers within 1 h to minimize handling stress. Scuba divers first removed all predators (Cancer productus, Pycnopodia helianthoides, Scyra acutifrons, and Lophopanopeus bellus) and competitors (sea urchins) from inside the cage and then emptied a bag of 50 abalone into each cage. The cages were then monitored briefly to ensure that all outplanted abalone had attached successfully to the artificial substrate.
The first survey performed 1 wk after outplanting was also the point at which the small predator removal treatment was first applied, and therefore no differences in abalone survival, mortality, or predator abundance was expected at that time. The second survey conducted 2 wk after outplanting was the first time that differences in the small-predator removal treatment and the control were expected. A third survey was conducted on November 8 and December 15, 2008 (approximately 20 wk after outplanting), and a fourth survey was conducted on June 17, 2009 (approximately 44 wk after outplanting). We made final observations and took photos of the cages on August 24, 2009.
During each survey, all live abalone were removed from the cage, counted, then returned to the cage or substrate position from which they were collected. We counted live abalone and empty or broken abalone shells that were present within the cages or within a 2-m radius around the cage. During the second, third, and fourth surveys, all predators in cages assigned to the small-predator removal treatment were removed from the cage, counted, and measured, then transported to a location at least 50 m away from any cages. Predators found in the control cages were also counted and measured, but were then returned to the cage. The divers also counted and classified all reported abalone predators that were present in a 2-m radius around the cages. We were able to find only a small percentage of the shells of dead outplanted abalone and were consequently not able to distinguish completely survival from emigration. Therefore, we used the slope of the regression of the log-transformed number of live abalone present in the cage during each survey versus time to estimate the instantaneous rate of loss per day.
Live abalone and predator abundance data generated from the surveys were heteroscedastic and not normally distributed; therefore, the data were log transformed. Repeated-measures ANOVA (Sokal & Rohlf 1995, Barnard et al. 2007) was carried out using IBM SPSS Statistics 19. For the predator data, neither a [log.sub.10](y + 1) transformation nor a [square [square root of y + 5] transformation (Sokal & Rohlf 1995) resulted in the data meeting the homoscedastic assumptions of parametric statistics. Therefore, nonparametric Mann-Whitney U-tests were also used to compare the total number of predators and the number of specific predators (Pycnopodia helianthoides, Cancer productus, and Scyra acutifrons) between the predator removal cages and the control cages. We did not find any of the small crab Lophopanopeus bellus in the cages during any of the surveys. The sea otter was still very rare in Barkley Sound and therefore was not a significant predator on abalone in this area.
Complex Substrate Addition Experiment
To test whether the size and amount of cryptic habitat influenced the survival of outplanted juveniles, 30 circular fenced plots were deployed at the same 2 sites (see photographs in Appendix IB-D) in a fully crossed, 2-factorial design. The plots (diameter, 0.89 m; height, 0.15 m) were constructed using plastic mesh netting (mesh size, 3.18 x 3.18 mm) obtained from Dynamic Aqua-Supply Ltd. (Surrey, BC). The plots were spaced at 8-m intervals along a lead-line, approximately 3 m away from the rocky substratum, in the sand, with the goal of retaining more abalone on these "islands" (Sloan & Breen 1988). Because of space constraints, a randomized block design (Krebs 1989) was used in which 15 plots were each placed in 2 sites, approximately 100 m apart. The plots were secured to the substrate using rebar driven into the sand. To control for spatial heterogeneity in the substratum, we added a base layer of gravel approximately 0.1 m thick to all plots.
To test whether the presence and/or amount of cryptic habitat increased outplanted abalone survival (as assayed by retention of live abalone within the plot and a 4-m radius around the plot), the 30 plots were randomly assigned a substrate type treatment (cobbles (diameter, 150 [+ or -] 20 mm) or boulders (diameter, 300 [+ or -] 20 mm)) and a mound height treatment (high, 0.9 m; or low, 0.3 m) in a fully crossed 2-factor design with 2 levels each (factors are substrate type (cobble and boulder) and mound height (low and high)). The high substrate treatments consisted of either 94 cobbles (C94) or 14 boulders (B14) that were stacked 0.90 m high; the low treatments consisted of either 61 cobbles (C61) or 9 boulders (B9) stacked 0.30 m high. Control treatments contained only the gravel (G) layer (Appendix IB-F).
The amount of added cryptic habitat varied among treatments, with cobble plots containing the most crevice volume (1.23-2.87 [m.sup.3]), boulder plots containing less (0.099-0.792 [m.sup.3]), and controls having none because the abalone were too large to hide in the gravel. Consequently, we predicted that mean survival of outplanted abalone would be greatest in the cobble treatment, moderate in the boulder treatment, and lowest in the gravel treatment because of the decreasing structural complexity.
After the different substrate treatments were assembled inside the fenced plots, they were left for 2 wk to allow a biofilm to accumulate. Outplanting was done on July 11, 2009. Nine hundred abalone (average SL, 51.5 mm) were obtained from the BHCAP hatchery where they had been held since they were settled in 2005. It was not necessary to tag the outplanted juveniles because the exposed nacre and unique coloring of the periostracum of their shell, and its lack of epiphytes and epifauna, allowed us to distinguish them easily from wild abalone. Abalone were remove unharmed from the tank walls by exposing them to the predatory sunflower starfish (Pycnopodia helianthoides) to elicit an escape response. The moving abalone were then gently collected and packed in Ziploc bags (30 per bag) and transported in coolers to the outplanting site within 2 h. Dive teams released the abalone into the survey plots, placing each abalone either inside rock crevices or directly onto gravel and monitoring them until they were securely attached.
Abalone and predator surveys were completed each day after outplanting for 6 days, and then after 13 days and 27 days. The surveys consisted of searching each plot for surviving abalone, empty shells or fragments, and predators, which were then counted and recorded. The area surrounding each plot out to a radius of 4 m was searched following the same survey protocol, and any dead abalone found were considered to have originated from the nearest plot. Numbers of live abalone and predators were compared among substrate treatments considering the total area surveyed (inside plot + surrounding area) and the area inside the plots only. The number of dead abalone and cumulative dead abalone were compared for the total search area only. To minimize disturbance, which may attract additional predators (Boulding & Hay 1984), plots were left intact during most surveys; however, to gain more accurate estimates, plots were completely disassembled during the 4-day, 13-day, and 27-day (intensive) surveys.
The number of live abalone, number of dead abalone shells, and number of predators were heteroscedastic with nonnormal distributions, so a log(y + 1) was applied to meet the assumptions of a parametric ANOVA (Sokal & Rohlf 1995). The data were then analyzed using repeated-measures ANOVA using IBM SPSS Statistics 19. Site was initially included as a factor in this model but was removed from all later models because preliminary analyses revealed that site was not a significant factor (site and its interactions had a P value > 0.25). Both the initial repeated-measures ANOVA and the nonparametric Friedman test showed an effect of mound height but no effect of the factor substrate type on abalone survival. Therefore, the factors mound height and substrate type were combined into a new complex substrate addition treatment, called substrate added, and the data were reanalyzed with repeated-measures ANOVA with 2 levels: substrate added and control. Last, the slope of the regression of the log-transformed number of live abalone present in the plot during each survey versus time was used to estimate the instantaneous rate of rate of loss per day.
A nonparametric Friedman test was also used to confirm the results of the repeated-measures ANOVA, because even the transformed data did not completely satisfy all the assumptions of a parametric test (Siegal 1957). The Friedman test (Sokal & Rohlf 1995) ranked the median value and compared the mean rank of blocks (survey times) among the 5 substrate treatments (high cobble, low cobble, high boulder, low boulder, and control), and was conducted in PASW 17 (SPSS). An SPSS script containing a nonparametric post hoc test designed by Gnambs (2010) was then used to carry out a multiple comparison test to determine the significance of substrate treatments. This post hoc test provided the differences in mean ranks between treatments as well as 2 estimates of the critical rank difference (Schaich & Hammerle 1984, Conover 1980, as cited in Gnambs 2010). We chose to use the more conservative estimate of the rank difference (Schaich & Hammerle 1984).
Small-Predator Removal Cage Experiment
Abalone remained in the cages for 1 wk before the first survey when the initial predator removal was applied to the treatment cages. As expected, data from this first survey did not show significant differences among cages for any of the 4 response variables (live abalone, dead abalone shells, grouped predators, and individual predators). At the time of the second survey (2 wk after outplanting) and the third survey (20 wk after outplanting), there were more surviving abalone found in the small-predator removal cages than in the control cages (Fig. 1A), and the difference was borderline significant (repeated-measures ANOVA, P = 0.069). No difference was detected among the treatment and control plots 44 wk after outplanting.
The small-predator removal treatment was effective for slow-moving predators. At the 2-wk survey point, the sea star Pycnopodia helianthoides was completely absent from all cages in the small-predator removal treatment but was still found in the control cages (Fig. 1B; Mann-Whitney U-test, P < 0.011). In contrast, the small-predator removal treatment had no effect on the number of small crabs found during the 2-wk survey (means for Scyra acutifrons: 1.6 crabs per predator removal cage and 1.3 crabs in the control cage; means for Cancer productus: 0.2 crab per predator removal cage and 0.3 crab in the control cage).
No small-predator removal treatment was applied for 18 wk between the 2-wk survey and the 20-wk survey. Few live abalone (n = 41) remained in the cages at the 20-wk survey. In addition, there were no significant differences among the small-predator removal cages and the control cages for the response variables grouped predators and individual predators. The same was true at the 40-wk survey, when a grand total of 2 live abalone remained in all the cages.
The instantaneous rate of loss of live abalone from the cages per day was greater (slope = -0.1217) during the first 2 surveys, when the density of live abalone inside the cages was greater than later during the experiment (slope = -0.0179), when the density dropped below 1 abalone per cage (Fig. 1C).
Complex Substrate Addition Experiment
Adding substrate of any size (cobble vs. boulders) increased the survival of outplanted abalone (Fig. 2A, B). It also increased the abundance of predators in and around the experimental plots relative to the control, gravel-only plots (Fig. 2C, D). The nonparametric Friedman test gave similar results to the repeated-measures ANOVA. The mean number of live abalone in the total area surveyed was significantly lower in the control treatments than in the substrate-added treatments (repeated-measures ANOVA, P < 0.01; Friedman test, P = 0.036). The mean number of live abalone inside the plots was also significantly lower in the control plots than in the substrate-added plots (repeated-measures ANOVA, P < 0.001; Friedman test, P = 0.014). Surprisingly, the abundance of predators was significantly greater inside the substrate-added plots than in the control plots (Fig. 2D; repeated-measures ANOVA, P < 0.001; Friedman test, P = 0.001). There was no significant difference between substrate-added plots and the controls in the mean number of dead abalone shells (Fig. 2E) in either the parametric (P = 0.445) and the nonparametric (P = 0.363) tests. Repeated-measures ANOVA revealed that adding complex substrate also increased significantly the mean predator abundance in the total area (Fig. 2C; repeated-measures ANOVA, P = 0.001), but the Friedman test found only a borderline effect (P = 0.06). Also, repeated-measures ANOVA showed no effect of complex substrate addition on cumulative mortality among the substrate-added treatments and the controls, whereas the Friedman test showed that the low boulder treatment had the lowest mean cumulative mortality of all treatments (Fig. 2F; P < 0.001). About half of the outplanted abalone shells that were recovered had been broken.
Reanalysis of the data with only 2 levels--substrate-added and control--confirmed that the more rapid decrease in abundance in the control plots was attributable in part to higher emigration rates. More live abalone were found outside the control plots than inside them, whereas the reverse was true for the substrate-added plots (Fig. 3A). Comparison of semilog regressions forced though the y-intercept showed that the instantaneous rate of loss for the first 5 surveys (Fig. 3B) was significantly higher in the control plots (slope = -1.7501) than in the substrate-added plots (slope = -0.6463; t-test, df = 8, P < 0.05). The instantaneous rate of loss from the substrate-added plots became less (slope = -0.0262) after the mean density of abalone dropped below 1 abalone per plot (Fig. 3B).
Surprisingly, the number of sunflower sea stars (Pycnopodia helianthoides, primarily with a diameter < 10 cm) was significantly higher in the substrate-added plots than in the control plots (Fig. 4A; day 4 survey: t-test unequal variances, t = -5.13, df = 17, P < 0.001). The number of sharpnose crabs (Scyra acutifrons) was also significantly greater in the substrate-added plots than in the control plots (Fig. 4B; day 4 survey: t = -2.36, df = 17, P = 0.030). The same was true of the total number of predators (Fig. 4C). In contrast, the densities of the larger and more motile red rock crab were not significantly different between the 2 treatments even when the aggregate sum inside and within a 4-m radius outside the plots (0.009 crabs/[m.sup.2]) was used. There was no significant difference between treatments in the mean cumulative number of dead abalone shells that were found inside and within a 4-m radius of the plots (Fig. 4D). Instead of the expected negative correlation, the day 4 survey data actually showed a weak trend toward a positive correlation between the number of predators within a plot and the number of live abalone remaining in a plot (r = 0.296, df = 28, P = 0.120).
We also examined whether placing plots inside the bay (site 1) or outside of the bay (site 2) had an effect on abalone survival and mortality as well as predator abundance. The effect of site placement did not influence any of the dependent variables significantly except for total predator abundance (Fig. 5). Mean predator abundance was significantly higher inside the bay for the total area in and around the plots (repeated-measures ANOVA, P < 0.037).
Our outplanted hatchery-raised abalone experienced very high rates of predation despite efforts to protect them by putting them in cages. The high predation rates by visual predators could be attributable to the abalone's poorly camouflaged shells (Schiel & Welden 1987, Schiel 1992) that often had shiny, exposed nacre (K. Read, pers. obs.). It is more difficult to explain the high predation rates by predators that use solely chemoreception, such as the sunflower sea star, Pycnopodia helianthoides, unless the abalone have also failed to develop strong behavioral predator avoidance mechanisms (Straus & Friedman 2009, Hansen 2011). Our prediction that outplanted juvenile survival would be higher in cages where predators were removed was somewhat supported by the results. The borderline significant increase in live abalone abundance in our predator removal cages suggests that the juvenile abalone either survived better or were less likely to leave the cages if all the sea stars are removed. The significantly lower sea star densities in our experiment suggest that removing slow-moving predators such as sea stars from cages may be effective if done often enough. We saw a reduction 1 wk after the first predator removal treatment was applied, but did not see a reduction after an 18-wk interval or after a subsequent 14-wk interval. Removing predators from seeding sites has been shown to reduce abalone mortality effectively over the short term (Tegner 2000), but there is no evidence to demonstrate its effectiveness in the long term.
Weekly cleaning of predator exclusion cages by divers may provide a means to increase the survival of outplanted abalone; however, more cost-effective methods should be investigated. A larger outplanted shell size can increase predator handling time and reduce abalone mortality (Andresen & van der Meer 2010). If slightly larger juveniles had been outplanted into cages with smaller mesh size, then we might have been able to increase survival for a modest increase in rearing costs. Unfortunately, smaller mesh would become fouled more quickly than the large mesh currently used on this cage module design, and might also impede water flow and food transport through the cages. An additional problem is that the cage modules seem to be behaving as artificial reefs, in that they continue to attract predators even though they no longer contain any abalone (E. G. Boulding, pers. obs.).
Our prediction that abalone survival would be greater in plots with more and smaller crevices compared with plots with less cryptic habitat and larger crevices was supported when we compared the complex substrate addition treatment with the gravel-only control. The retention of outplanted abalone in the plots was significantly improved by the addition of complex substrate. However, retention was not affected by the type (cobbles (diameter, 150 [+ or -] 20 mm) or boulders (diameter, 300 [+ or -] 20 mm)) or depth of complex substrate (0.3 m or 0.9 m), even though these factors should affect the size and availability of crevices, respectively. This contrasts with the results of de Waal and Cook (2001), who found that outplanted abalone survived better in habitats with stacked boulders of less than 300 mm in diameter than in habitats with stacked boulders of 300-500 mm in diameter. This difference between studies could be accounted for by our provision of adequate cryptic habitat in all complex substrate addition treatments by adding a minimum of 0.3-m depth of added cobbles or boulders. Conversely, in our gravel-only control plots, where cryptic habitat was absent, both abalone retention and the observed numbers of sea star and sharpnose crab predators were significantly reduced. This suggests that the presence of refuges may reduce predation in this system (Drolet et al. 2004). The addition of cryptic habitat to natural sites has been reported previously to increase survival rates in outplanted abalone (Booth & Cox 2003, Dixon et al. 2006).
Emigration from the control plots, rather than mortality resulting from predation, may be the primary cause of the rapid decrease in the number of abalone found in control plots. One day after outplanting, the mean number of live abalone remaining in control plots was significantly less than the number in the substrate-added plots. Furthermore, a greater proportion of outplanted abalone from control plots was found in the poor-quality sand habitat between the plots and the cobble habitat 2-4 m away (K. Read, pers. obs.), presumably moving toward what is likely a higher quality habitat (see Lessard and Campbell (2007)), with more refuge from predators and more food (Day et al. 2004). Abalone retention might have been improved by outplanting our abalone in release modules and then waiting for 24 h before releasing them (Sweijd et al. 1998).
The large number of broken shell fragments that we recovered suggested that predation on our outplanted abalone was high. Hansen (2011), who also outplanted individuals from this hatchery-reared abalone population, found that only 40% of the juveniles survived the first 24 h, and her recovery of broken shells (Emmett & Jamieson 1989), as well as predator surveys, suggested that density-dependent predation by the red rock crab Cancer productus was responsible. Another important shell-breaking predator on juvenile abalone is the pile perch Rhacochilus vacca, which reaches local densities of 0.119/[m.sup.2] in nearby Bamfield Inlet (Boulding et al. 2001). An 82-mm-long Rhacochilus vacca can consume abalone of SL 25 mm (Griffiths & Gosselin 2008), so could prey on abalone that emigrated from our large predator exclusion cages. It seems plausible that the largest (300-350 mm) Rhacochilus vacca (Boulding et al. 2001) could consume the abalone of SL 50 mm used in our complex substrate experiment.
Our day 4 survey of our complex substrate addition plots revealed significantly higher densities of both predators and prey in substrate-added plots than in control plots. Hansen (2011) also found that the red rock crab Cancer productus showed significant aggregation at all outplanting sites relative to control sites, and that the sunflower sea star Pycnopodia helianthoides showed aggregation at one of her outplanting sites. This high aggregation of predators at the outplanting sites likely explains the low survival rate of released juvenile abalone (Dixon et al. 2006). Small crab and sea star predators may initially be attracted to outplanting sites because they detect a high density of live abalone using their distance chemoreception (Boulding & Hay 1984), but may remain in or near a cage, pipe, or mesh fence long after the abalone have dispersed or been eaten because it offers them refuge from their own predators. This density-dependent response by predators may explain the low recovery rate of outplanted abalone (e.g., Rogers-Bennett & Pearse 1998), and outplanting at lower densities may increase survival and retention of individuals in future restoration attempts (Goodsell et al. 2006).
The results of the repeated-measures ANOVA for our complex substrate addition experiment showed that the abundance of predators was significantly greater inside the bay, perhaps because of the lower wave exposure there. Tomascik and Holmes (2003) demonstrated a negative correlation between exposure index and predator density in another area of Barkley Sound. Selecting appropriate habitat, where predator abundance is low and refuges are present, can increase the survival of outplanted abalone (de Waal & Cook 2001) and is an important factor to consider prior to outplanting.
This study demonstrates that adding cryptic habitat provides refuge from large, motile predators, but mortality rates of the hatchery-reared abalone were still very high. Both of the large predator exclusion cages and the fenced plots with added boulders or cobbles actually had higher densities of small predators inside them than were found in the surrounding control habitat, suggesting they were acting as artificial reefs. Manual removal of small predators from outplanting habitats was extremely time-consuming, but was a successful method to decrease the density of a major predatory sea star. The finding that the northern abalone emigrates from habitats with low substrate complexity could potentially affect outplanting success and should be considered in future seeding efforts. Overall, the instantaneous loss rates of outplanted abalone were high and might have been reduced by outplanting at lower densities that were more similar to current wild abalone densities in healthy populations (Goodsell et al. 2006). Future outplanting should be done in areas with sufficient cryptic habitat to increase abalone retention and with low numbers of natural predators.
In addition, further understanding of the changing local kelp ecosystem (Watson 2000) is also important for the continued survival of northern abalone populations (Lessard et al. 2007). Poaching remains a significant source of local abalone mortality in more remote areas (Campbell 2000) and will continue to undermine efforts to rehabilitate the fishery unless it is addressed (Hauck & Sweijd 1999). Working with locals to help them understand the impacts of poaching on natural recruitment is essential.
APPENDIX 1: ENCLOSURE DESIGNS
Small-Predator Removal Experiment
The small-predator removal experiment included an enclosure design 1 m diameter and 0.3 m tall (Panel A). The predator exclusion cages had mesh openings 66 x 91 mm.
Complex Substrate Addition Experiment
Panels B-F shows 5 substrate and mound height treatments inside 0.9-m-diameter plastic mesh fences with openings of 3.18 x 3.18 mm: control (B), high cobble (C), low cobble (D), low boulder (E), and high boulder (F).
We acknowledge the support of the Bamfield Marine Sciences Centre, Bamfield Huu-Ay-Aht Abalone Project (BHCAP), and Fisheries and Oceans Canada (DFO)'s Pacific Biological Station. Funding was provided from Natural Sciences and Engineering Research Council (Canada) Strategic Project grant STPSC 357084-07 (L. A. Gosselin, E. G. Boulding, C. D. G. Harley), DFO, and the University of Guelph. We thank T. Bird, D. Brouwer, D. Bureau, S. Gray, C. Hansen, K. Harrer, E. Herder, B. Seamone, I. Reding, and G. Wittig for scuba support; D. Richards and J. Richards for technical support; I. Smith for editing the figures; and J. Bogart, R. Danzmann, B. Robinson, and members of the Bogart/Boulding/Fu/Heyland lab group for reading the manuscript.
Andresen, H. & J. van der Meer. 2010. Brown shrimp (Crangon crangon, L.) functional response to density of different sized juvenile bivalves Macoma balthica (L.). J. Exp. Mar. Biol. Ecol. 390:31-38.
Barnard, C., F. Gilbert & P. McGregor. 2007. Asking questions in biology: a guide to hypothesis testing, experimental design and presentation in practical, work and research projects. Harlow, UK: Pearson Education. 243 pp.
Booth, J. D. & O. Cox. 2003. Marine fisheries enhancement in New Zealand: our perspective. N. Z. J. Mar. Fresh 37:673-690.
Boulding, E. G. & T. K. Hay. 1984. Crab response to prey density can result in density-dependent mortality of clams. Can. J. Fish. Aquat. Sci. 41:521-525.
Boulding, E. G., D. Pakes & S. Kamel. 2001. Predation by the pile perch, Rhacochilus vacca, on aggregations of the gastropod Littorina sitkana. J. Shellfish Res. 20:403-409.
Campbell, A. 2000. Review of northern abalone, Haliotis kamtschatkana, stock status in British Columbia. Can. Spec. Publ. Fish. Aquat. Sci. 130:41-50.
Campbell, A., J. Lessard & G. S. Jamieson. 2003. Fecundity and seasonal reproduction of northern abalone, Haliotis kamtschatkana, in Barkley Sound, Canada. J. Shellfish Res. 22:811-818.
Carlsson, J., R. B. Carnegie, J. F. Cordes, M. P. Hare, A. P. Leggett & K. S. Reece. 2008. Evaluating recruitment contribution of a selectively bred aquaculture line of the oyster, Crassostrea virginica used in restoration efforts. J. Shellfish Res. 27:1117-1124.
Conover, W. J. 1980. Practical nonparametric statistics. 2nd ed. New York: John Wiley and Sons. 493 pp.
COSEWIC. 2009. COSEWIC assessment and update status report on the northern abalone Haliotis kamtschatkana in Canada. Ottawa: Committee on the Status of Endangered Wildlife in Canada. 55 pp. Available at http://www.registrelep-sararegistry.gc.ca/document/ default_e.cfm?documentID=1810.
Day, R., P. Gilmour & S. Huchette. 2004. Effects of density and food supply on postlarval abalone: behaviour growth and mortality. J. Shellfish Res. 23:1009-1018.
DeFreitas, B. 2003. Estimating juvenile northern abalone, Haliotis kamtschatkana, abundance using artificial habitats. Can. Tech. Rep. Fish. Aquat. Sci. 2482:69-70.
de Waal, S. & P. Cook. 2001. Quantifying the physical and biological attributes of successful ocean seeding sites for farm-reared juvenile abalone (Haliotis midae). J. Shellfish Res. 20:857-861.
Dixon, C. D., R. W. Day, S. M. H. Huchette & S. A. Shepherd. 2006. Successful seeding of hatchery-produced juvenile greenlip abalone to restore wild stocks. Fish. Res. 78:179-185.
Drolet, D., J. H. Himmelman & R. Rochette. 2004. Use of refuges by the ophiuroid Ophiopholis aculeate: contrasting effects of substratum complexity on predation risk from two predators. Mar. Ecol. Prog. Ser. 284:173-183.
Egli, T. P. & J. Lessard. 2011. Survey of northern abalone, Haliotis kamtschatkana, population in the Strait of Georgia, British Columbia, October 2009. 15 pp. Available at http://www.dfo-mpo.gc.ca/librariesbibliotheques/manu-eng.htm.
Emmett, B. & G. S. Jamieson. 1989. An experimental transplant of northern abalone, Haliotis kamtschatkana, in Barkley Sound, British Columbia. Fish Bull. 87:95-104.
Fisheries and Oceans Canada. 2007. Recovery strategy for the northern abalone (Haliotis kamtschatkana) in Canada. Vancouver: Fisheries and Oceans Canada. 42 pp. Accessed April 2012. Available at http://www. sararegistry.gc.ca/document/default_e.cfm?documentiD=1342.
Gnambs, T. 2010. Post-hoc comparisons for Friedman test. Available at http://timo.gnambs.at/en/scripts/friedmanposthoc.
Goodsell, P. J., A. J. Underwood, M. G. Chapman & M. P. Heasman. 2006. Seeding small numbers of cultured black-lip abalone (Haliotis rubra Leach) to match natural densities of wild populations. Mar. Freshw. Res. 57:747-756.
Griffiths, A. G. & L. A. Gosselin. 2008. Ontogenetic shift in susceptibility to predators in juvenile northern abalone, Haliotis kamtschatkana. J. Exp. Mar. Biol. Ecol. 360:85-93.
Hansen, S. C. 2011. Restoring endangered northern abalone (Haliotis kamtschatkana) populations in British Columbia, Canada, using hatchery-raised individuals. MS thesis, Thompson Rivers University. 165 pp.
Hauck, M. & N. A. Sweijd. 1999. A case study of abalone poaching in South Africa and its impact on fisheries management. ICES J. Mar. Sci. 56:1024-1032.
James, D. S., R. W. Day & S. A. Shepherd. 2007. Experimental abalone ranching in Port Phillip Bay, Victoria. J. Shellfish Res. 26:687-695.
Krebs, C. J. 1989. Ecological methodology. New York: Harper & Row. 654 pp.
Lessard, J. & A. Campbell. 2007. Describing northern abalone, Haliotis kamtschatkana, habitat: focusing rebuilding efforts in British Columbia, Canada. J. Shellfish Res. 26:677-686.
Lessard, J., A. Campbell, Z. Zhang, L. MacDougall & S. Hankewich. 2007. Recovery potential assessment for the northern abalone (Haliotis kamtschatkana) in Canada. Canadian Science Advisory Secretariat research document 2007/061. 107 pp. Available at http:// www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2007/ 2007_061-eng.htm.
Lessard, J. & T. P. Egli. 2011. Survey of northern abalone, Haliotis kamtschatkana, population in Queen Charlotte Strait, British Columbia, May 2009. 21 pp. Available at http://www.dfo-mpo. gc.ca/libraries-bibliotheques/manu-eng.htm.
McCormick, T. B., K. Herbinson, T. S. Mill & J. Altick. 1994. A review of abalone seeding, possible significance and a new seeding device. Bull. Mar. Sci. 55:680-693.
Oceanlink. 2008. Bamfield Huu-ay-aht Community Abalone Project (BHCAP). Accessed December 5, 2011. Available at http://www. oceanlink.info/Conservation/abalone/BHCAP/outplant3.html.
Ochwala-Doyle, F., C. A. Gray, N. R. Loneragan & M. D. Taylor. 2010. Using experimental ecology to understand stock enhancement: comparisons of habitat-related predation on wild and hatchery-reared Penaeus plebjus Hess. J. Exp. Mar. Biol. Ecol. 390:65-71.
Read, K. D., M. A. Lemay, S. Acheson & E. G. Boulding. 2012. Using molecular pedigree reconstruction to evaluate the long-term survival of outplanted hatchery-raised larval and juvenile northern Abalone (Haliotis kamtschatkana). Conserv. Genet. 13:801-810.
Rogers-Bennett, L. & J. S. Pearse. 1998. Experimental seeding of hatchery-reared juvenile red abalone in northern California. J. Shellfish Res. 17:877-880.
Seamone, B. & E. G. Boulding. 2011. Aggregation of the northern abalone, Haliatis kamtschatkana, with respect to sex and spawning condition. J. Shellfish Res. 30:881-888.
Schaich, E. & A. Hamerle. 1984. Verteilungsfreie statistische Prufverfahren. Berlin: Springer. 366 pp.
Schiel, D. R. 1992. The enhancement of paua (Haliotis iris Martyn) populations in New Zealand. In: S. A. Shepherd, M. J. Tegner & S. A. Guzman-del Proo, editors. Abalone of the world: biology, fisheries and culture, 1st edition. Oxford: Blackwell Scientific. pp. 474-484.
Schiel, D. R. & B. C. Welden. 1987. Responses to predators of cultured and wild red abalone, Haliotis rufescens, in laboratory experiments. Aquaculture 60:173-188.
Shepherd, S. A. 2000. Studies on southern Australian abalone (genus Haliotis) XII: long term recruitment and mortality dynamics. Aust. J. Mar. Freshw. Res. 41:475-492.
Shepherd, S. A. & P. A. Breen. 1992. Mortality in abalone: its estimation, variability and causes. In: S. A. Shepherd, M. J. Tegner & S. A. Guzman-del Proo, editors. Abalone of the world: biology, fisheries and culture, 1st edition. Oxford: Blackwell Scientific. pp. 276-304.
Shepherd, S. A., P. A. Preece & R. W. G. White. 2000. Tired nature's sweet restorer? Ecology of abalone (Haliotis spp.) stock enhancement in Australia. In: A. Campbell, editor. Workshop on rebuilding abalone stocks in British Columbia. Ottawa: NRC Research Press. pp. 84-97.
Siegal, S. 1957. Nonparametric statistics. Am. Stat. 11:13-19.
Sloan, N. A. & P. A. Breen. 1988. Northern abalone, Haliotis kamtschatkana, in British Columbia: fisheries and synopsis of life history information. Can. Spec. Publ. Fish. Aquat. Sci. 103. 46 pp.
Sokal, R. R. & F. J. Rohlf. 1995. Biometry. New York: W.H. Freeman. 887 pp.
Stoner, A. W. & M. Davis. 1994. Experimental outplanting of juvenile queen conch, Strombus gigas: comparison of wild and hatchery-reared stock. Fish. B NOAA 92:390-411.
Straus, K. M. & C. S. Friedman. 2009. Restoration aquaculture of the pinto abalone (Haliotis kamtsehatkana kamtschatkana Jonas): impacts of rearing method on behavior, growth, and survivorship. Mar. Freshw. Res. 60:1021-1028.
Sweijd, N. A., P. A. Cook, Q. Snethlage & D. Harvey. 1998. Experimental abalone (Haliotis midae) seeding in South Africa. J. Shellfish Res. 17:897-904.
Tegner, M. J. 2000. Abalone (Haliotis spp.) enhancement in California: what we've learned and where we go from here. Can. Spec. Publ. Fish. Aquat. Sci. 130:61-7l.
Tomascik, T. & H. Holmes. 2003. Distribution and abundance of Haliotis kamtschatkana in relation to habitat, competitors and predators in the Broken Group Islands, Pacific Rim National Park Reserve of Canada. J. Shellfish Res. 22:831-838.
Watson, J. 2000. The effects of sea otters (Enhydra lutris) on abalone (Haliotis spp.) populations. Can. Spec. Publ. Fish. Aquat. Sci. 130: 123-132.
KAITLYN D. READ, (1) JOANNE LESSARD (2) AND ELIZABETH G. BOULDING (1*)
(1) Department of Integrative Biology, University of Guelph, Ontario, N1G 2W1, Canada; (2) Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, BC, V9T 6N7, Canada
* Corresponding author. E-mail email@example.com
|Printer friendly Cite/link Email Feedback|
|Author:||Read, Kaitlyn D.; Lessard, Joanne; Boulding, Elizabeth G.|
|Publication:||Journal of Shellfish Research|
|Date:||Apr 1, 2013|
|Previous Article:||Reproductive potential can predict recruitment rates in abalone.|
|Next Article:||Assessing the recovery of pink abalone (Haliotis corrugata) by incorporating aggregation into a matrix model.|