In LobsterScience/bio.lobster: Lobster Stock Assessment Tools for DFO Maritimes

Analysis and Response

Indicators of Stock Status

Primary indicators are used to define stock status in relation to the reference points, and secondary indicators are those in which time series trends are displayed but are not associated with reference points.

The data sources available for constructing indicators for LFA 33 are mainly fishery-dependent. Commercial logbooks report information on date, location (grid), effort, and estimated catch. At-sea samples are collected during normal commercial fishing operations and provide information on bycatch and Lobster caught, including carapace size, sex, egg presence, and stage; shell hardness; occurrence of culls and v-notches; and the number of traps, location, and depth. The Fishermen and Scientist Research Society (FSRS) are contracted to conduct a recruitment trap project involving volunteer fishermen who record data on Lobsters that are captured in standardized traps.

Primary Indicators

Primary indicators are the focus for defining stock status, by describing the time series trends relative to reference points. The primary indicator for describing stock status is the unmodelled commercial Catch Per Unit Effort (CPUE). Exploitation estimated using the Continuous Change In Ratio (CCIR) method from recruitment trap data is used as the primary indicator of fishing pressure that is independent of landings reported in the logs. The recruitment trap data for the 2018–19 fishing season were not available for this Science Response report.

Catch Per Unit Effort

The time series of commercial catch rates is made up of two data sources. The first was the voluntary log books, which began in the 1980s and continued until 2013 in LFA 33. Mandatory logs have been in place in LFA 33 since the mid 2000s and provide a more complete data set with which to evaluate changes in catch rates (Tremblay et al. 2012). In years where both voluntary and mandatory logs were available, the magnitude and trends over time were similar (Tremblay et al. 2013), so both logbook types were included together. In the current analysis, the two commercial catch rate series are treated as a single continuous time series beginning in 1990, when there was increased participation in the voluntary logbook program. The combined catch rate data series from 1990–2016 was used to define the Upper Stock Reference (USR) and Limit Reference Point (LRP). The median of this time series was used as the proxy for Biomass at Maximum Sustainable Yield (BMSY), 0.35 kg/Trap Haul (TH). Following the recommendations of DFO (2009), the USR and LRP were set to 80% and 40% of the BMSY proxy. The value used to compare the commercial catch rates to the USR and LRP is the 3-year running median, as this dampens the impact of any anomalous years that may occur due to factors outside of changes in abundance.

logs=lobster.db("process.logs")
        CPUE.data<-CPUEModelData(p,redo=F)
        cpueData = CPUEplot(CPUE.data,lfa= p$lfas,yrs=1982:2020,graphic='R', plot=F)$annual.data

```{r CPUE1, include=TRUE, echo=FALSE, warning=FALSE, message=FALSE, comment = FALSE, error = FALSE, fig.cap="Time series of commercial catch rates (black points), along with the 3-year running median (blue Line). Red Triangle indicates an incomplete data point as all commercial logbooks for this season not yet received. The horizontal lines represent the upper stock (dashed green line) and limit reference point (dotted red line)."}

33

crd = subset(cpueData,LFA==33,c("YEAR","CPUE")) mu=median(crd$CPUE[crd$YEAR %in% c(1990:2016)]) usr = mu * 0.8 lrp = mu * 0.4

crd = merge(data.frame(YEAR=min(crd$YEAR):max(crd$YEAR)),crd,all.x=T)

par(mar=c(1.5,5.5,2.0,3.0)) xlim=c(1990,max(crd$YEAR)) plot(crd[,1],crd[,2],xlab=' ',ylab='CPUE (kg/TH)',type='p',pch=16,xlim=xlim, ylim=c(0, 1.05*max(crd$CPUE))) points(max(crd$YEAR), crd$CPUE[which(crd$YEAR==max(crd$YEAR))], pch=17, col='red', cex=1.5) running.median = with(rmed(crd[,1],crd[,2]),data.frame(YEAR=yr,running.median=x)) crd=merge(crd,running.median,all=T) lines(crd[,1],crd$running.median,col='blue',lty=1,lwd=2) abline(h=usr,col='green',lwd=2,lty=2) abline(h=lrp,col='red',lwd=2,lty=3)

creates variables referenced in text

current.cpue=round(crd$CPUE[which(crd$YEAR==max(crd$YEAR))],2) print.usr=round(usr,2) print.lrp=round(lrp, 2) ratio=round(current.cpue/usr, 1)

The trend in CPUE indicates that a significant increase in the stock biomass has occurred in the
last ten years (Figure 2). For most of the time series, CPUE has fluctuated just above the USR,
substantially increasing after 2008 so that it is currently more than triple the USR. The 3-year
running median value for CPUE for the 2019-2020 fishing season is 'r current.rm' kg/TH, which is well
above the USR ('r print.usr' kg/TH) and LRP ('r print.lrp' kg/TH).


## Continuous Change In Ratio (CCIR)

The CCIR method is used as an indicator of fishing pressure. It is based on the recruitment trap data,
so reflects trends in exploitation in the inshore portion of the LFA, where the majority of the fishery occurs.
The CCIR method provide estimates of population parameters based on the changes in observed proportions of
components within the population. Estimating exploitation using CCIR relies on defining and monitoring two
(or more) components of the population, consisting of a reference (non-exploited) component and an exploited
component. The premise of this method is that the proportion of reference individuals within the population
will increase with the cumulative removals from the exploitable component (Claytor and Allard 2003). The
strength of this approach is that it does not rely directly on fishery-dependant landings data, so the CPUE
indicator and CCIR exploitation estimates are based on independent time series.
The implicit assumptions of the CCIR include that, over the sampling period, the population is closed,
the ratio of catchability of the two components is constant, the ratio of the catchability of the monitoring
traps and the commercial traps is constant, and the monitoring effort is directly proportional to harvesting
effort. The recruitment trap catch data provide information on the changes in the pre-exploitable reference
group (sub-legal-sized Lobsters) relative to the exploitable group (legal-sized Lobsters) needed to estimate
exploitation. The Removal Reference (RR) was defined as the 75th quantile of the posterior distribution of
the maximum modeled CCIR exploitation rate. Given that regional Lobster stocks are currently in a highly
productive state and population growth has not decreased under the range of estimated exploitation, it is
reasonable to assume the RR is less than the fishing mortality corresponding to maximum sustainable yield, FMSY.
The time series of exploitation estimates is shown in Figure 3. For the first half of this time series,
exploitation estimates were fairly high—just below the RR. Since 2013, exploitation has declined to about
two-thirds of the level of the RR. In the last two years, exploitation has increased but the running median
remains near two-thirds of the RR. The 3-year running median value of CCIR exploitation for the 2019-20 fishing
season is x.xx, which is below the RR (0.xx).

## Secondary Indicators

Secondary indicators represent important time-series trends that are tracked individually, but no reference points
are defined. The secondary indicators for LFA 33 are landings and total effort (trap hauls), as well as the
recruitment trap legal and sub-legal catch rate series.

### Landings and Effort
Levels of commercial landings are related to population abundance, as fishery controls are input-based (effort
controls) rather than output-based (e.g., total allowable catch). Changes in levels of fishing effort,
catchability (including the effects of environment, gear efficiency), Lobster size distribution, and the spatial
overlap between distribution of Lobster and effort will impact landings, thereby weakening the relationship with
abundance.

Fishing effort can be used as a proxy for fishing pressure. It is an important indicator for fishery performance,
as increases in landings may be due to increases in commercial sized biomass, or increased fishing effort, or
both. Fishing effort, number of trap hauls, in the Lobster fishery is controlled by fishing season length, trap
limits, and a limited number of fishing licences. Consequently, there is a maximum fishing effort that can be
deployed. This maximum is never met as factors such as weather conditions, seasonally variable catch rates, and
fishing partnerships all limit the total number of trap hauls.

Generally, the trend in landings is similar to the trend in the primary indicator, CPUE, as effort has remained
fairly consistent over the time series (Figure 4). There has been a significant increase in landings over the last
ten years that corresponds with an increase in CPUE. There are some annual fluctuations in effort, with a slight
increasing trend over time.

```r

    land = lobster.db('annual.landings')
    logs=lobster.db("process.logs")
        CPUE.data<-CPUEModelData(p,redo=F)
    land =land[order(land$YR),]


#Landings limits
        xlim<-c(1982,p$current.assessment.year+1)


####LFA 33####
        d1 = data.frame(YEAR = land$YR, LANDINGS = land[,"LFA33"])
        d2 = subset(cpueData,LFA==33)
        d2 = merge(data.frame(LFA=d2$LFA[1],YEAR=min(d2$YEAR):max(d2$YEAR)),d2,all.x=T)
        data=fishData = merge(d2,d1)
        data$EFFORT2=fishData$EFFORT2 = fishData$LANDINGS * 1000 / fishData$CPUE


        par(mar=c(1.5,5.5,2.0,3.0))
        plot(data$YEAR,data$LANDINGS,ylab='Landings(t)',type='h',xlim=xlim, xlab=" ", ylim=c(0,max(data$LANDINGS)*1.2),pch=15,col='gray73',lwd=4,lend=3)
        lines(data$YEAR[nrow(data)],data$LANDINGS[nrow(data)],type='h',density = 30,pch=21,col='steelblue4',lwd=4, lend=3)
        par(new=T)

            plot(data$YEAR,data$EFFORT2/1000,ylab='',xlab='', type='b', pch=16, axes=F,xlim=xlim,ylim=c(0,max(data$EFFORT2/1000,na.rm=T)))
            points(data$YEAR[nrow(data)],data$EFFORT2[nrow(data)]/1000, type='b', pch=24,size = 2,bg='black')
            axis(4)

The 2019 landings for LFAs 27-32 are preliminary as there remains outstanding logbooks, and 2019 landings in LFA 27 do not include information from the Gulf. Landings in LFA 27 are on track with recent years, with 2018 being an all-time high for the area. In LFA 28, even with incomplete data the landings for 2019 already exceed recent years. Landings in LFA 29 and 30 remain comparable to previous years with 2018 being an all-time high for both areas (Figure 4). Landings in LFA 31A and 31B show all-time high in time series and LFA 32 are comparable to most recent years (Figure 5).

Recruitment Trap Catch Rates

The recruitment trap survey provides the best information on the abundance of under-sized Lobsters. It is also the only source of data on abundance for LFA 33 that is collected in a standardized manner. The catches of legal- (≥82.5 mm) and sub-legal-sized (70–82.5 mm) Lobsters were modelled with a Bayesian approach in order to characterize the credible intervals of the predicted time series used as the indicator. The numbers of legal- and sub-legal-sized Lobsters were assumed to follow a negative binomial distribution with the log number of traps used as an offset. For sub-legal-sized Lobsters, the predictors included temperature, the number of legal-sized Lobsters caught, and year. For legal-sized Lobsters, the predictors were temperature, the day of the season, and year. All of these effects were significant. Temperature is assumed to affect catch rates of all Lobsters, while larger Lobsters (legal-sized) are assumed to reduce entrance of smaller Lobsters (sub-legal-sized) into traps. The resultant models were used to predict the number of Lobsters (for each size class) per trap for each year at a common temperature, date, and number of legal-sized Lobsters per trap. The results from the recruitment trap models showing the median number of legal- and sub-legal-sized Lobsters per trap with their 95% credible intervals are presented in Figure 5. Both legal and sub-legal size classes show a gradual increasing trend, which is not as dramatic as the increase in landings and CPUE over the last ten years. It is important to note, however, that the recruitment traps are mainly located close to shore, where smaller Lobsters are and do not cover the entire range of fishing locations in LFA 33.

Bycatch

Bycatch estimates are calculated using effort from logbook data for all LFAs except LFA 27. To estimate bycatch in LFA 27, effort was estimated from the combined Gulf Logbooks and Maritimes logbooks using the median CPUE and landings from the Gulf logs and logbook effort from the Maritimes Region. 2019 Gulf Landings in LFA 27 are calculated by adding estimated slip landings based on previous years to the logbook landings. 2019 Bycatch estimates for LFAs 27-32 are preliminary due to log records being incomplete with no bycatch estimate for LFA 29, 30 and 32 due to low sampling numbers or no data available.

knitr::include_graphics("LFA27Bycatch.png")

knitr::include_graphics("LFA31ABycatch.png")

knitr::include_graphics("LFA31BBycatch.png")

LobsterScience/bio.lobster documentation built on Feb. 14, 2025, 3:28 p.m.