Evidence of Considerable Shifts in Catch Composition in the Artisanal Spiny Lobster Fishery in Kenya

Simple Summary In Kenya, the artisanal lobster fishery is important socioeconomically for supporting local fishing communities and generating revenue for the government, yet detailed knowledge of many aspects of the fishery is lacking. In this study, the population structure and catch composition of spiny lobsters caught by divers were investigated, and the findings were compared to lobster survey data from the 1970s to identify potential changes in the artisanal landings that may have occurred over time. Abstract The artisanal lobster fishery in Kenya is small in world terms but important locally both in terms of supporting local fishing communities and generating revenue for the government. Despite its socioeconomic importance, detailed knowledge of many aspects of the fishery is lacking. The study reported herein aimed to investigate and provide information on the population structure and catch composition of spiny lobsters caught by artisanal fishers off six major landing sites along the coastline. A total of 2711 lobsters representing five palinurid species were collected during the study period (November 2000–March 2001). Among the five species, Panulirus longipes dominated the catches in Msambweni (75%) and Shimoni (58%), P. homarus in Mambrui (70%) and Kipini (72%), P. ornatus in Lamu (49%), and P. penicillatus in Kilifi (39%). P. versicolor was the rarest species observed in the catches across the six sites. The overall catch consisted of 33% P. ornatus, 32% P. homarus, 28% P. longipes, 6% P. penicillatus and 2% P. versicolor. Sitewise, Lamu contributed 31% of the overall catch, Kipini 23%, Shimoni 20%, Mambrui 13%, Msambweni 7%, and Kilifi 6%. A comparison of the results of this study and lobster abundance data from 1970s surveys revealed considerable shifts in the catch composition of artisanal landings over time. Future work on this fishery should concentrate on the lobster populations in the decades-old marine protected areas to obtain unfished reference data to assess the fishery and establish the underlying cause(s) of the observed shifts in catch composition.


Introduction
Kenya has both an artisanal and an offshore lobster fishery, of which the artisanal fishery is the most important in terms of production, employment creation, and revenue generation.The offshore production is based on deep-water lobster species (Metanephrops andamanicus, Puerulus angulatus, Scyllarides) caught as bycatch during semi-industrial prawn trawling off the north coast.The artisanal fishery, on the other hand, exploits five species of shallow-water tropical spiny lobsters belonging to the genus Panulirus, namely, Panulirus ornatus, P. longipes, P. homarus, P. versicolor and P. penicillatus [1].
Artisanal lobster fishing is undertaken along the entire coastline, with the major landing sites being Shimoni, Msambweni, Kilifi, Mambrui, Kipini, and Lamu.Spiny lobsters are harvested year-round, but the peak fishing season occurs during the northeast monsoon (November-April) when weather conditions are favorable.Even though artificial shelters have been shown to attract spiny lobsters at Gazi Bay on the south coast [2], Biology 2023, 12, 1477 2 of 15 breath-hold diving remains the most popular harvest method.Fishers capture lobsters in shallow waters (<10 m) during the daytime, rarely venturing beyond three nautical miles offshore due to the depth limits of breath-hold diving.Most divers access fishing grounds with dug-out canoes, outriggers, Arab dhows, and other small, non-mechanized crafts [3].Landed lobsters are either sold to local tourist hotels or exported live to overseas markets, primarily China.
Although the artisanal spiny lobster production is small in world terms, averaging only 100 metric tons annually, the fishery is socioeconomically important in supporting many fishers and their families and earning the government hard currency through the export of lobsters.Lobster fishing is popular among fisher folks because of the high price lobsters fetch compared to finfish and other crustaceans.Unfortunately, the ever-increasing demand for seafood has led to a tremendous increase in fishing effort in recent years, resulting in overfishing [4].Declines in catch rates and lobster sizes have been reported for some species [5][6][7].
Despite the socioeconomic importance of the resource and the apparent overfishing, the fishery is open access with a few applicable management controls, such as a minimum legal weight of retention (250 g) for all species, prohibitions on retaining berried females, and the use of scuba gear, spears, dynamite, and tangle nets.Four small marine protected areas (MPAs) are also in place where no fishing is permitted.However, these management controls are insufficient to ensure the long-term conservation and sustainable use of lobster resources, even if they are strictly enforced, since they are based mainly on limited scientific information.More comprehensive research is therefore needed to obtain the biological and population parameters in order to formulate stringent, science-based management regulations.Previous studies on Kenyan spiny lobsters have either dealt with a single species or populations in a small geographic area [3,4,6,8,9].Although both Mutegyera [1] and Mueni et al. [7] focused on the entire coastline, their findings varied substantially, especially regarding the relative abundance of lobster species.This study aimed to investigate and provide information on the species composition, relative abundance, spatial distribution, and population structure of spiny lobsters harvested along the Kenya coast.

Study Areas
The Kenyan coastline is approximately 640 km long, running in a south-westerly direction from the Kenya-Somalia border in the north to the Kenya-Tanzania border in the south (Figure 1).It is fringed by a continuous, well-developed reef that lies between 0.5 and 2 km offshore, except where major rivers (Tana and Sabaki) discharge into the Indian Ocean.The coral reefs, along with seagrasses and mangroves, provide an ideal habitat for spiny lobsters.Generally, the coastline is divided into the north and south coasts, with Mombasa Island serving as the reference point.
For this study, six sampling sites were selected based on their being major lobster landing centers: Shimoni and Msambweni on the south coast, Kilifi, Mambrui, Kipini, and Lamu on the north coast.

Sample Collection
This study was conducted between November 2000 and March 2001, coinciding with the lobster fishing season.A six-day trip was made to each of the selected six landing sites, where five divers were randomly selected from the respective local fishers and hired to assist with sample collection.In order for their catches to represent the typical landings of a given site, the divers were instructed to fish as they usually would and bring all their lobster catches into a makeshift laboratory at the landing site for analysis.Once brought to the laboratory, lobster samples were identified to the species level, sexed based on external sexual dimorphism, and measured to the nearest 0.1 mm (carapace length, CL) before being returned to the fishers.

Data Analysis
Lobster CL data were tested for normality (Kolmogorov-Smirnov test) and homogeneity of variance (Levene test) prior to formal analysis.For lobster size comparisons, CL data were grouped by landing site and compared for differences using the Kruskal-Wallis test.When the result was statistically significant (p < 0.05), a post hoc test was used to determine which comparisons had significant differences.The sex ratio was determined for each sampling site, and deviation from the expected ratio of 1:1 was tested using the Chi-square (χ 2 ) test.Panulirus penicillatus and P. versicolor were omitted from size and sex composition analyses due to their low proportions or absence from some landing sites.Species composition was calculated based on the number of each species within the catch.Relative abundance was estimated by dividing the number of lobsters in each species by the total number of lobsters in all species.All statistical analyses were performed using SPSS for Windows (version 22.0, IBM, Armonk, NY, USA), and p-values lower than 0.05 were considered statistically significant.

Limitations of This Study
The results provided in this study were derived from data collected approximately 22 years ago.Due to the cumulative effects of fishing on exploited stocks, these results may not reflect the current landings of the artisanal lobster fishery in Kenya.However, the findings are important since they can serve as a valuable reference for future assessments of the spiny lobsters in the region.

Species Distribution
In this study, five species of spiny lobsters belonging to the genus Panulirus were encountered in the landed catches at the six sampling sites (Table 1).These were Panulirus ornatus, P. longipes, P. homarus, P. penicillatus, and P. versicolor.Panulirus ornatus, P. longipes and P. homarus were the most widely distributed of the five species, occurring in all six sampling sites.Panulirus versicolor was found at four sites, while P. penicillatus was encountered at only three sites.Among the six sites, Shimoni, Kilifi, and Lamu had the highest number of spiny lobster species, with all five species present.These were followed by Msambweni and Kipini, with four species each.Mambrui had the least number of species, hosting only three.

Catch Composition
A total of 2711 lobsters were collected during the study period, of which 887 belonged to P. ornatus, 866 to P. homarus, 747 to P. longipes, 167 to P. penicillatus and 44 to P. versicolor. Figure 2 shows the composition of lobster species by landing site.P. ornatus was the most abundant species in Lamu, accounting for 49.3% of the sampled catch.It also constituted a substantial proportion of the catches landed at the other five sites.P. homarus was most prevalent in the estuarine sites of Mambrui (70%) and Kipini (72%) but least common on the south coast sites of Msambweni and Shimoni, contributing just 1% of the catch.P. penicillatus appeared to be the most harvested species in Kilifi.In Msambweni and Shimoni, P. longipes accounted for 75% and 58% of the catch, respectively, making it the most important species on the south coast.P. versicolor was the rarest species observed in the catches across the sites.Pooling all the catches together (Figure 3), P. ornatus had the highest relative abundance (33%), followed by P. homarus (32%), P. longipes (28%), and P. penicillatus (6%), while P. versicolor had the lowest relative abundance of just 2%.Sitewise, Lamu contributed most of the pooled catch (31%), followed by Kipini (23%) and Shimoni (20%).The contribution of the other three sites ranged from 6% to 13% (Figure 4).The northern sites of Lamu, Kipini, and Mambrui together accounted for 67% of the lobster landings.
the south coast sites of Msambweni and Shimoni, contributing just 1% of the catch.P. penicillatus appeared to be the most harvested species in Kilifi.In Msambweni and Shimoni, P. longipes accounted for 75% and 58% of the catch, respectively, making it the most important species on the south coast.P. versicolor was the rarest species observed in the catches across the sites.Pooling all the catches together (Figure 3), P. ornatus had the highest relative abundance (33%), followed by P. homarus (32%), P. longipes (28%), and P. penicillatus (6%), while P. versicolor had the lowest relative abundance of just 2%.Sitewise, Lamu contributed most of the pooled catch (31%), followed by Kipini (23%) and Shimoni (20%).The contribution of the other three sites ranged from 6% to 13% (Figure 4).The northern sites of Lamu, Kipini, and Mambrui together accounted for 67% of the lobster landings.

Size and Sex Composition
The descriptive statistics of the carapace length of males and females of the three most abundant species (i.e., P. ornatus, P. homarus and P. longipes) collected from the six sites are presented in Table 2.The overall mean and standard deviation of carapace length were 82 mm ± 22 mm SD for P. ornatus, 69 mm ± 12 mm SD for P. homarus and 70 mm ± 11 mm SD for P. longipes.In terms of the mean size of lobsters by landing sites, lobster samples from Kipini had the largest mean size, while those from Mambrui had the smallest mean size.Carapace length frequency distributions constructed for males and females exhibited a unimodal distribution for all three species over the range of individuals measured (Figures 5-7), except for P. ornatus males and females from Shimoni and Kilifi and P. longipes females from Kilifi, which displayed a bimodal distribution.Analysis of variance (Kruskal-Wallis) test showed significant differences in the lobster size distributions among the landing sites for all three species (Kruskal-Wallis test, p < 0.05).Subsequent Dunn's post hoc tests also revealed significant differences in most paired site comparisons (Table 3).
The sex ratio of P. ornatus, P. homarus and P. longipes determined by site did not depart from the expected male-to-female ratio of 1:1 for all sites (Table 4), except Shimoni and Kipini, where males of P. ornatus and P. longipes significantly outnumbered the females in the population.In terms of pooled data, P. ornatus samples were male-biased, with the sex ratio significantly deviating from a 1:1 ratio.In contrast, male-to-female ratios of P. homarus and P. longipes in the pooled catch were near parity.The sex ratio of P. ornatus, P. homarus and P. longipes determined by site did not depart from the expected male-to-female ratio of 1:1 for all sites (Table 4), except Shimoni and Kipini, where males of P. ornatus and P. longipes significantly outnumbered the females in the population.In terms of pooled data, P. ornatus samples were male-biased, with the sex ratio significantly deviating from a 1:1 ratio.In contrast, male-to-female ratios of P. homarus and P. longipes in the pooled catch were near parity.lobsters started in the mid-1950s.Brusher [10] found that 50% of the fishery comprised undersized lobsters.Mutegyera [1] also noted that 89% of all P. ornatus females landed at Kizingitini (Lamu) were below the mean size of the berried females (98.9 mm CL), indicating that a significant percentage of the lobsters were harvested before they had a chance to reproduce.Okechi [24] estimated the mean sizes of male and female P. ornatus at first capture at 77.5 mm CL and 67.5 mm CL, respectively, suggesting the capture and retention of immature lobsters.More recently, Mueni et al. [7] reported that P. ornatus, P. homarus and P. versicolor were harvested before they reached the minimum legal weight of retention (250 g) based on the length at first capture they estimated for the three species.In the present study, more than half of all P. ornatus collected were smaller than the size (84 mm), at which 50% of females reached size at first maturity [7], confirming the continued harvest of undersized lobsters.During a 2001 survey [5], the most frequent complaints of local fishers along the coastline were the scarcity of lobsters in coastal waters and the declining sizes of lobsters caught.

Conclusions
When compared to the data from lobster surveys conducted in the 1970s, the results of this study show a significant change in the catch composition of Kenya's artisanal lobster fishery over time.In particular, the results show that the traditionally dominant species, P. ornatus, has lost its position as the most harvested species along the coastline and now (at the time of this study) ranks marginally higher in terms of abundance than P. homarus and P. longipes.An interesting question that comes to mind is why only P. ornatus has been negatively affected and not the other four species with which it cooccurs in nearshore habitats under the same environmental conditions.Unlike the affected P. ornatus, for example, P. homarus and P. longipes have thrived and become as important in the catch as P. ornatus over time, with a combined contribution of 60%, up from 10% in the 1970s.Although several factors are known to influence catch composition and abundance of lobsters, the available data are insufficient to draw any conclusion on the specific factor(s) responsible for the observed shifts in lobster catch composition.Additional research is therefore required to determine the underlying causes of this anomaly, which appears to have varying impacts on the lobster species.In the short term, research on the fishery should focus on the decades-old marine protected areas in order to obtain unfished reference data to empirically assess the status of the fishery.In the long term, however, continuous monitoring of the fishery is required to identify trends and potential issues, such as overfishing or population shifts, for timely regulatory interventions.Alternatively, researchers can use length-based stock assessment models such as length-based spawning potential ratio (LBSPR), length-based indicators (LBI) or length-based Bayesian biomass approach (LBB) to assess the status of the lobster stocks and establish reference points for their management.Length-based models are a viable alternative to conventional catchbased methods for assessing fisheries that lack substantial datasets (i.e., time series of total catch, absolute or relative abundance, fishing effort, life-history parameters, etc.), given their robustness to produce less biased and reliable estimates as well as the simplicity and costeffectiveness of length data collection.Moreover, the application of length-based models in lobster management can support efforts to assess stock health, evaluate management effectiveness, and explore different length-based management scenarios.

Figure 1 .
Figure 1.Map of the Kenyan coast showing sampling sites and other lobster landing centers.

Figure 1 .
Figure 1.Map of the Kenyan coast showing sampling sites and other lobster landing centers.

Figure 2 .Figure 2 .
Figure 2. Species composition of spiny lobsters caught off the six sampling sites during the study period (November 2000-March 2001).

Figure 3 .
Figure 3. Catch composition of spiny lobsters in the pooled data from all sampling sites.

Figure 4 .
Figure 4. Catch contributions of the respective sampling sites to the pooled spiny lobster data.

Figure 3 .Figure 3 .
Figure 3. Catch composition of spiny lobsters in the pooled data from all sampling sites.

Figure 4 .
Figure 4. Catch contributions of the respective sampling sites to the pooled spiny lobster data.

Figure 4 .
Figure 4. Catch contributions of the respective sampling sites to the pooled spiny lobster data.

Figure 5 .
Figure 5. Size frequency distributions of Panulirus homarus from three sampling sites during the study period (November 2000-March 2001).

Figure 5 .
Figure 5. Size frequency distributions of Panulirus homarus from three sampling sites during the study period (November 2000-March 2001).

Figure 6 .
Figure 6.Size frequency distributions of Panulirus longipes from six sampling sites during the study period (November 2000-March 2001).

Figure 6 .
Figure 6.Size frequency distributions of Panulirus longipes from six sampling sites during the study period (November 2000-March 2001).

Figure 7 .
Figure 7. Size frequency distributions of Panulirus ornatus from six sampling sites during the study period (November 2000-March 2001).

Figure 7 .
Figure 7. Size frequency distributions of Panulirus ornatus from six sampling sites during the study period (November 2000-March 2001).

Table 1 .
Distribution of spiny lobster species along the Kenya coast based on samples caught off the six sampling sites during the study period (November 2000-March 2001) *.
* Symbol (x) indicates the presence of a species in a particular site.

Table 2 .
Descriptive statistics of the carapace length of male and female spiny lobsters caught off the six sampling sites during the study period (November 2000-March 2001).

Table 3 .
Kruskal-Wallis and Dunn's post hoc test results for the carapace length of spiny lobsters caught off the six sampling sites during the study period (November 2000-March 2001).Statistically significant (p < 0.05) values in bold.