Bull. SouthernCaliforniaAcad. Sci. 113(1),2014,pp. 14—33 ©SouthernCaliforniaAcademyofSciences,2014 Status and Trends in the Southern California Spiny Lobster Fishery and Population: 1980-2011 Eric F. Miller MBC Applied Environmental Sciences 3000 Red Hill Ave., Costa Mesa, CA 92626, , [email protected] — Abstract. The California spiny lobster Panulirus interruptus fishery in southern ( ) California ranks amongst the State’s most economically important fisheries. An analysis of commercial harvest data confirms that the fishery was landing near- record catches in the late-2000s through early-2010s. Advances in recreational fishing technology likely tempered commercial fishery landings per unit effort. The commercial catch per trap pulled declined 15%, on average, in years after the introduction of a new rigid-style hoop net in the recreational fishery. Fishery- independent data sourced from power plant marine life monitoring recorded increased California spiny lobster abundance after 1989 with evidence of increased larval settlement beginning circa 1989. This timing was consistent with previously reported oceanographic changes in theNorth Pacific. Abundance indices forlobsters likely one-year’s growth away from recruiting into the fishery and the young-of-the- year, both derived from power plant abundance records, significantly predicted the commercial landings at index-appropriate temporal lags, i.e. one year for next year’s recruitment. Carapace lengths measured during power plant surveys in Santa Monica Bay, where commercial fishing was prohibited, significantly declined for the total sample and females only after the introduction ofthe new rigid hoop net. Male carapace lengths were not significantly different between the two periods. The power plant data confirms that, as of 2012, the fishery appears healthy but warns of the need to monitor sublegal individuals and their dependence on oceanographic conditions. These analyses also indicate the urgency of monitoring the recreational fishery harvest, especially the potential effects ofthe new rigid hoop net. Exploitation of spiny lobsters (Family Palinuridae) occurs worldwide. These fisheries rank among the most economically valuable in their respective regions (Phillips 2006; Lipcius and Eggleston 2008). Historically, substantial research was directed at understanding spiny lobster population variability (Polovina et al., 1995; Lipcius and Eggleston 2008) and predicting future fishery success by examining abundance indices of pre-recruitment size classes (Cruz et al., 1995; Cruz and Adriano 2001). These data commonly resulted in stock assessments and fishery management plans designed to maintain sustainable fisheries (Phillips et al., 2010). Despite its importance to California’s marine fisheries, California spiny lobster (Panulirus interruptus) historically received little research within California waters (Mitchell et al., 1969; Pringle 1986; Barsky 2001) in comparison to other spiny lobster fisheries worldwide. Recently, however, increased effort has been applied to understanding California spiny lobster biology, ecology, and fisheries (Arteaga-Rios et al., 2007; Mai and Hovel 2007; Parnellet al., 2007; Neilson and Buck 2008; Neilson et al., 2009; Koslow et al., 2012; Selkoe et al., 2010; Withy-Alien 2010; Miller et al., 201 la; Neilson 2011; Kay et al., 2012a; Kay et al., 2012b). 14 SOUTHERN CALIFORNIA SPINY LOBSTER FISHERYAND POPULATION DYNAMICS 15 Table 1. Summary of existing (as of December 2012) California spiny lobster recreational and commercial fishing regulations in California. Fishery Regulation Recreational Commercial Minimum Size 82.5 mm CL Seasonal Closure Mid-March through September Permit Needed Fishing License Limited-Entry Bag Limit 7 individuals None Capture Method Diving by hand. Hoop Net Trap Catch Reporting Report Card Logbook NA Mini-Season 5 days California spiny lobster (hereafter lobster) was commercially fished since the 1800s (Neilson 2011) with recreational or subsistence fishing likely as long. The commercial fishery peaked during the post-World War II years through the mid-1950s when the numbers ofimmaturelobsterstakenwasrecognizedand accounted forinnewregulations (Neilson 2011). Regulation changes between 1957 and 1976 reduced the catch ofsmaller, sublegal lobsters. Landings again increased in the late 1980s reaching a plateau of approximately 250 metric tons annually in the mid- to late-2000s (Neilson 2011; Koslow et al., 2012). Recreational harvesting of marine species was long considered inconsequential in comparisontocommerciallandings,butrecentevidencesuggestsotherwise(Egglestonetal., 2003; Colemanetal., 2004; BirkelandandDayton2005; Erismanet al.,2011). Thesouthern California recreational lobster fishery landings over time are unknown and likely varied in response to lobster population abundances, angler participation rates, and advances in fishing technology. Minimal information on historic recreational catch and participation exists. TheintroductionofalobsterreportcardissuedbytheCaliforniaDepartmentofFish andWildlife(DFW)in2008mayalleviatethisprobleminthefuture, butthelackofhistoric data hampers present analyses. This lack of information confounds understanding the population’shealth, intheabsenceoffishery-independentdata, consideringtherecreational catch was estimated at 30-60% ofthe commercial catch (Neilson 2011). The two lobster fisheries in California were governed by a suite ofsimilar regulations but also fishery-specific rules (Table 1). Equipment used in each fishery bore relevance to this discussion. Traps were the only technology available to commercial fishers. Regulations governing their design remained unchanged since 1976. Market demand resulted in the commercial fishery targeting smaller individuals by utilizing smaller entrancefunnels oneach trap resultinginminimalvariation in theharvested size sincethe 1980s (Neilson 2011; Barsky 2012; Healy 2012). In contrast, the recreational fishery targets all legal size classes, often prizing larger individuals (Neilson 2011). Recreational scuba divers once dominated the fishery with comparatively minor harvests by hoop net anglers (Neilson et al., 2009). This changed recently, coincidingwith the introduction ofa newrigid, conicalhoopnet (newhoopnet) in 2006 (Tackletour2006). More lobsterswere landed perhoop net set using the newhoopnet in comparison to the traditional hoop net (Neilson et al., 2009) thus raising concerns regarding their potential impact on the population if extensively adopted by the recreational fishery. Past experiences in other spiny lobster fisheries suggest this concern may be warranted. Recreational-only fishing periods used in Florida, similar to those used in California, resulted in significant 16 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES reductions in Caribbean spiny lobster (Panulirus argus) population density when pre- recreational season and pre-commercial season surveys were compared (Eggleston et al., 2003). Similar surveys have thus far not been completed in California. To enhance management strategies, California resourcemanagers recentlycompleted a stock assessment despite limited fishery-independent population data (Neilson 2011) a noted shortcoming in the assessment (Cope et al., 2011). The lack ofannual recruitment estimates and overall yearclass strength indices were highlighted as significant data gaps. Prior attempts to fill these gaps utilized historic plankton tows offshore off California (Pringle 1986; Koslow et al., 2012). Their results were informative of gross changes in population abundances, but inconclusive in explaining the interannual variation in lobster landings. Planktonic stage abundance indices often fail to predict future fishery patterns due to the variety of mortality sources acting on larval and pre-recruit stages (Houde 2008). Coastal power plant marine life entrapment monitoring programs were found to provide previously unused data supporting fisheries analyses (Field et al., 2010; Erisman et al., 2011; Miller et al., 2011b; Miller and McGowan 2013). Invertebrate records were never evaluated, but hold promise as a fishery-independent data source. Furthermore, three monitored power plants were situated in the Santa Monica Bay, California where commercial lobster fishing was prohibited, but recreational fishing was permitted. Using these data, this work aims to investigate important concepts regarding the California spiny lobster population that directly affects the fishery’s management. The primary purpose of this investigation is two-fold. First: fill knowledge gaps related to population abundance cycles over time; recruitment patterns and their relationship with the fishery landings; and interannual variability in larval settlement patterns. Second, use these complied data to test the hypothesis that the lobster population and commercial fishery have changed in measurable ways since the introduction of the new hoop net. Material and Methods Data Sources Total annual commercial landings, in metric tons (MT), were compiled from Perry et DFW al., (2010) for all fishing blocks in southern California. Fishing blocks represented a designated spatial grid of 10-minute latitude X 10-minute longitude numbered areas, except alongthecoastwhere thecoastline bounds the area and reduces thefishingblock’s overall size. Data was screened to remove those landings reported from fishing blocks encompassing bathymetry exceeding the lobster’s known maximum depth (73 m; Barsky 2001) and accounted for < 1% of the total cumulative landings, 1980-2008. Spatial distribution ofthe total harvest (1980-2008 cumulative) was visualized using ArcGIS 10 with five natural breaks segregating the data. Fishery-independent lobster data collected during power plant monitoring records as described by Miller and McGowan (2013) were compiled. Of the five power plants examined (Figure 1), three had intakes surrounded by soft-bottom sandy habitat where the intake structures themselves represented one of the few high-relief substrates in the area (Table 2). These were Scattergood Generating Station (SGS), El Segundo Generating Station (ESGS), and Huntington Beach Generating Station (HBGS). The primary Redondo Beach Generating Station (RBGS) intake was located near the King Harborbreakwall, awell documentedmature artificial reef(Stephenset al., 1994). Lastly, SOUTHERN CALIFORNIA SPINY LOBSTER FISHERY AND POPULATION DYNAMICS 17 Fig. 1. Totalmetrictons, basedonlandings, harvestedfromeachCalifornia Department ofFishand Gamefishingblock, 1980-2008commercialCaliforniaspinylobsterseasons(October-March).Theupper map depicts all Southern California Bight fishing blocks where lobster have reportedly been taken. The lowermap depicts only those fishing blocks contributing > 1% ofthe total commercial landings, 1980- 2008. Location ofthe fourpowerplants used in this analysis is presented. They are San Onofre Nuclear Generating Station (SONGS), Huntington Beach Generating Station (HBGS), Redondo Beach Generating Station (RBGS), El Segundo Generating Station (ESGS), and Scattergood Generating Station (SGS). Scattergood and El Segundo are less than 3 km apart from each other with Scattergood upcoast from El Segundo. San Onofre Nuclear Generating Station’s (SONGS) intakes were situated adjacent to or in a cobble stone reefwith a large kelp forest. At each power plant, the intake conduit emptied into a sedimentation basin (or forebay) within the power plant property. This reduces the overall water velocity and allows animals, such as fish and invertebrates, to take up residence within the forebay. Perpendicularto theconduit bulkhead lie steelmeshtraveling screens designed to prevent large material from passing further into the cooling water system. Traveling screen mesh was 10-mm squaremesh. At thediscretion ofpowerplant operators, aheat treatmentwas conducted to control biofouling growth within the cooling water system. This resulted in a forebay water temperature > 35°C for a minimum of one hour. During this time the traveling screens operated continuously removing all stressed and moribund marine life from the forebay. All material (marine life and debris) within the forebay became impinged upon the traveling screens, was carried out ofthe forebay, and washed offinto 18 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Table 2. Descriptiveparametersforeachpowerplantcoolingwater systemmonitored and supplying data on California spiny lobster abundance. Mean flowrefers to the volume ofcooling water circulated between heat treatments. Median carapace length (CL mm) and standard error for California spiny lobsters measured during surveys since 1993. Redondo Huntington San Parameter Scattergood El Segundo Beach Beach Onofre # ofIntakes 1 2 2 1 3 Intake Depth (m) 9.0 9.8 13.7 27.5 9.1 Riser Height (m) 3.2 3.0 3.0 2.4 2.9 Distance Offshore (m) 488 698 289 457 960 Habitat Surrounding Intake Sand Sand Reef Sand Reef, Kelp # ofSurveys 100 208 277 192 272 Mean Flow (106 m3) (SE) 273.8 (98.9) 66.1 (4.4) 87.6 (3.9) 59.7 (3.2) 168.5 (47.1) Years Surveyed 1993-11 1980-10 1980-06 1980-98, 2001-10 1980-93, 2006-07 Median CL (mm) 76 79 72 13 66 Standard Error 0.45 0.81 0.35 0.97 1.35 acollection basket. At the end ofthe heat treatment, few, ifany, animals remainedwithin theforebay. Lobsters impinged duringheat treatments were counted, batch-weighed, and measured to the nearest mm CL (beginning in 1993). The cooling water flow volumes between heat treatments were compiled from power plant records ofdaily cooling water circulated. Environmental data included the following indices: Pacific Decadal Oscillation (PDO; Mantua et al. 1997), North Pacific Gyre Oscillation (NPGO; Di Lorenzo et al. 2008), Multivariate ENSO Index (MEI; Wolter and Timilin 2012), and m daily seawater temperature at 5 (BST) recorded at the Scripps Institution of Oceanography (UCSD 2012). Data Analysis Fishing effort data was unavailable prior to 1997, therefore annual landings were presented unstandardized to effort. Both effort and landings data were provided by the DFW for the 2000-2010 fishing seasons (ca. October - March). Using these data, a catch per unit effort (CPUE; legal lobster count/trap pulled) was derived. Their use was limited due to the brevity ofthe CPUE series in comparison to the other data sets evaluated. The CPUE series was used to compare pre- and post-introduction of the new hoop net to investigate possible changes in the commercial fishery that could be attributed to the new recreational fishing technology. Comparisons were made using a Kruskall-Wallis test. Lobster abundance and biomass data collected during power plant monitoring were standardized to circulated coolingwatervolumes to account forvariation between plants and daily operations over time. The resulting entrapment rate (ER; count/106 m3 was ) used in subsequent analyses. Trends were tested for significance (meaningfulness) using a MS Excel add-in wherein the lobster time series was broken into intervals of varying length before analyzed using linear regression (Bryhn and Dimberg 2011). Individual weightswerenotrecorded duringsurveys; thereforeamean individualweightwasderived by dividing the aggregate biomass by the total count. This supplemented the size information collected after 1992. Median carapace lengths were calculated for lobsters measured at each power plant to indicate the size structure sampled. The annual median length (total, female, and male) from lobsters collected at SGS was examined for changes in the overall size structure with time at a location free ofcommercial fishing pressure. SOUTHERN CALIFORNIA SPINY LOBSTER FISHERY AND POPULATIONDYNAMICS 19 Sampling dates were matched to years corresponding to the fishing seasons and intervening period before the next season, i.e. October 1 - September 30. For example, datacollected fromOctober 1, 2009 - September 30, 2010wasincluded inthe2009 annual median length. This analysis was limited to SGS as it was the only power plant meeting two criteria: 1) located in the Santa Monica Bay where commercial fishing was disallowed, 2) robust sample size in years before and afterthe newhoopnetintroduction. The remaining power plants did not meet both of these criteria (Table 2). A Kolmogorov-Smirnov test was used to compare lengths from the two periods, pre- and post-new hoop net. Individuals between 72.5 and 81.5 mm CL were assumed to represent new fishery recruits the following year. Their abundance by heat treatment was standardized consistent with the ER technique to derive the next-year’s-fishery-index (NYFI; count/ 109 m3). Assuming commercial landings predominantly represent first-year recruits (Neilson2011), landingswereplottedas a function oftheNYFI afteradvancingtheindex one year. Individuals measuring less than 21 mm CL were assumed to be young-of-the- year. From their survey-specific abundance, a young-of-the-year index (YOYI, count/ 109 m3 was derived following the same methods described for the ER and NYFI. ) Environmental indices were compared with the YOYI to examine the potential relationship between the environment, as measured by the indices, and lobster settlement using a Spearman’s rank correlation. When identified, autocorrelation was addressed by adjusting the rcrit using the modified Chelton method (Pyper and Peterman 1998). All statistical analyses, other than the power plant trend analysis, were completed in R (R Development Core Team 2012). Results California Department of Fish and Wildlife records indicate lobster has been taken commercially acrossawideportion oftheCaliforniacoastline(Figure 1). Themajorityof MT fishing blocks contributed less than 22 of landed lobster during the 29 years reviewed. These areaswereexcluded and the remaininganalyses focused onthe43 fishing blocks supporting the bulk of the commercial fishery. Landings from the southern portion of the area were generally higher than those farther north (Figures 1 and 2a), including the peak along Point Loma in San Diego, California. Additional areas producing high biomass landings included rocky headlands at Dana Point and Palos Verdes along the mainland, and at San Clemente, San Nicholas, Santa Cruz, and Santa Rosa Islands. A shift in the early 1990s indicated increased landings from areas outside ofSan Diego County (fishing block series 800). The mean seasonal percentage of the total landings prior to 1994 in the southernmost fishing blocks accounted for 55% but declined to 45%, on average, through 2008 (Figure 2b). Duringthis 1994-2008 period, landings in the 700- series fishing blocks offshore ofLos Angeles and Orange Counties increased by only 2%, on average, from their 1980-1993 mean. Landings from the 600-series fishing blocks offshore ofSanta Barbara and Ventura Counties increased 8% during the latter 15 years in comparison to the 1980-1993 period. Commercial landings from these 43 blocks initially declined to a minimal level (<150 MT/annually) in the 1980s before increasing to relatively consistent annual landings of approximately 250 MT since 2000 (Figure 3a). Since 2000, the mean annual CPUE has remained > 0.36 legal lobsters/trap and ranged as high 0.54 legal lobsters/trap (Figure 3b). Prior to the introduction of the new hoop nets in the recreational fishery, 20 SOUTHERN CALIFORNIAACADEMY OF SCIENCES CO 100 H O d 03 landings of Percent 1980 1985 1990 1995 2000 2005 Year Fig. 2. a)TotalCaliforniaspinylobsterlandings(metrictons, MT)byfishingblockseriesandfishing season; b) percent oflandings by fishing block series. Block series: 600 = Santa Barbara and Ventura Counties, California; 700 = Los Angeles and Orange Counties; 800 = San Diego County. 0 SOUTHERN CALIFORNIA SPINY LOBSTER FISHERY AND POPULATION DYNAMICS 21 r 350 3000 - 300 2500 Metric 2000 - 250 1500 - 200 tons 1000 - 150 landed 500 - 100 0 - 50 - 0.5 - 0.4 Legal - 0.3 count/trap - 0.2 - 0.1 Fishing season 350 300 250 200 150 100 Fig. 3. a) Commercial California spinylobsterlandingsfromthosefishingblockscontributing > 1% ofthetotalcommerciallandings, 1980-2008. Thenext-year’s-fishery-index(NYFI; count/109m3)byyear derived from power plant entrapment surveys (1992-2011). b) Commercial fishery catch per unit effort (legalcount/trappulled)forthesamefishingblocksusedfora), c)Totallandings(1993-2010)plottedasa function of the NYFI after adjusting for a one-year lag. White circles indicate those seasons since the introduction ofa rigid-style hoop net in the recreational fishery (2007-2010). A sigmoidal function best describedthedistribution(r2 = 0.50).Dashedlinesindicate95%confidenceintervals,d)Subsamplingdata fgrreoamtefrigpurreedi3cctafboirlijtuyst(rN2Y=FI0.5<7).10D0a0/s1he9dmli3n,eosritnhdeicaastyem9p5to%tecotnhfriedsehnoclediinntdeircvaatlse.dinfigure 3cresultedin 22 SOUTHERN CALIFORNIAACADEMY OF SCIENCES the commercial CPUE averaged 0.46 legal lobster/trap. After the 2006 introduction, this declined to 0.39 legal lobster/trap, or a significant 15% reduction (KW, % = 3.98, df = 1, p = 0.046). Interannual CPUEvariabilitydeclined after the 2006 season as thecoefficient ofvariation declined from 1.00 prior to 2007 to 0.89 after 2006. Changes in the CPUE at the onset ofthis introduction were prominent across the Southern California Bight with mean monthly CPUE across all block series lower after the introduction, with one exception (Figure 4). The disparity between the two periods was generally higher during the first one-halfofthe season in all areas (Figure 4 b-d). In the San Diego area (fishing blocks numbered 801-899), the disparity was greatest during the first three months and waned thereafter, with post-introduction February CPUE slightly exceeding the pre- introduction mean. This was the only instance ofthe post-introduction period exceeding the pre-introduction mean. Trends in the fishery-independent population indices at each power plant were independent of each other with some exceptions (Figure 5). Entrapment rate trends at SGS significantlycorrelatedwith bothESGS (r = 0.58, p = 0.01, n = 18) and HBGS (r = 0.63, p < 0.01, n = 17). No correlation existed between the ESGS and HBGS trends. Trends at RBGS and SONGS were not correlated with any other power plant. Where data were available, an overall increasing trend in lobster population abundance (higher ER) was recorded beginning in the late-1980s, (Figure 5a). Monthly ER averaged over the entire series indicated general peak abundance in late-summer to early-fall period at each power plant (Figure 6). The ER at RBGS was the lone exception as it peaked in spring. Three power plants in Santa Monica Bay recorded data from an area free from commercial fishing pressure. At SGS, the farthest north ofthe three Santa Monica Bay power plants, lobster data was available beginning in 1993, from which a meaningful increase in ER was observed though 2011 (r2 = 0.51 P < 0.001), excluding a brief , depression in 2005 (Figure 5b). The SGS data were dominated by lobsters near the legal size limit (Table 1). Length measurements recorded at SGS were available from every year, 1998-2011. This was the only Santa Monica Bay power plant length series sufficiently spanning 2006 when the new hoop nets were introduced. Lobster carapace lengths pre-introduction were significantly larger than post-introduction across all samples (KS, D = 0.0998, p < 0.001; Figure 7a) and for females (KS, D = 0.1323, p < 0.001; Figure 7b), but not for males (KS, D = 0.0705, p = 0.10; Figure 7b). At nearby ESGS, the ER wasmorevariable than at SGS and extended to 1980 (Figure 5c). Lacking any significant trend, generallyhigher ERs were observed at ESGS in the last decade than during the preceding two. Most lobsters taken at ESGS were also near the fishery’s minimum size with the largest median CL recorded. Average weight oflobsters taken at ESGS was more variable prior to 1993, with a fairly stable mean weight of 500g during the 1993-2008 period before declining in later years. At the southeast edge ofthe Santa Monica Bay, RBGS also commonlyentrapped lobsterswith no significant trenddetected, but declining cooling water use by RBGS reduced the need for heat treatments (Figure 5d). Therefore, the usable time series ended in 2006. Entrapment rates at RBGS were also predominantly reflective of lobsters near recruiting to the fishery. Mean biomass of lobsters entrapped prior to 1990 were 80 g larger than after, suggesting an increased influence ofsmaller individuals since 1990. Huntington Beach Generating Station entraps few lobsters in comparison to the Santa Monica Bay power plants, and as with prior examples, a statistically insignificant abundance trend was observed (Figure 5e). In 1989, and several years after, the ER SOUTHERN CALIFORNIA SPINY LOBSTER FISHERYAND POPULATION DYNAMICS 23 a) all 0.8 Pre-rigid 0.6 ] Post-rigid 0.4 IlUL 0.2 i 0.0 b) 600-series (Santa Barbara Area) 0.8 0.6 Q_ CD 0.4 -I ilLi §0 0.2 H 1 0.0 CD c) 700-series (Los Angeles Area) 0.8 0.6 H 0.4 h 0.2 H ip io i fi 0.0 d) 800-^eries (San Diego Area) 0.8 H _L 0.6 0.4 -| LlLIL 0.2 (Lin 0.0 Oct Nov Dec Jan Feb Mar Month Fig. 4. MeanmonthlycommercialCaliforniaspinylobstercatchperuniteffort(± s.e.)before(2000- 2006)andafter(2007-2010)theintroductionoftherigid-stylehoopnetforrecreationalfishersa)acrossall fishing blocks in the 600-800 series, b) 600-series blocks, c) 700-series blocks, d) 800-series blocks. periodically increased substantially as a comparatively large number of lobsters were taken. Excluding 1989, the mean HBGS ER after 1989 was nearly four times greaterthan was recorded prior to 1989. Unlike the three Santa Monica Bay power plants, lobsters taken at HBGS were considerably smaller with a median CL of 13 mm (Table 2). This