The reproductive biology and size structure was determined for Lethrinops gossei, one of the most abundant fish species contributing 26% of the catch in the demersal trawl fishery of the Southeast Arm of Lake Malaŵi. Breeding seasonality was analyzed on the basis of monthly variation of gonado-somatic index and gonadal appearance, while sexual maturity was determined by fitting the logistic model to the proportions of sexually mature individuals. Length frequency data were used to characterize size structure of the population and to derive a selection ogive for the species based on the currently used cod-end mesh size of 38 mm in the fishery. Results suggest that even though mature individuals were encountered all year round, the main breeding events for L. gossei occur in March and September. It was also determined that the size at which 50% of female L. gossei reached maturity (L50%) was 137 mm total length, while 100% sexual maturity was reached at 170 mm total length. The size at first maturity was 122 mm total length, while the steepness parameter of the maturity ogive, (δ) was 7 mm. Distribution of sizes in the sample was roughly symmetric about a mean of 145 mm total length, while the size at 50% selection was 120 mm total length. These results are discussed in the context of their implications on the management of the demersal trawl fishery in the Southeast Arm currently dominated by L. gossei, given the misalignment of the closed season currently in October to December and the inadequacy of the cod-end mesh size of 38 mm. The main conclusion of this study was that the current cod-end mesh size in the fishery is too small and should be adjusted to prevent growth overfishing. A proposal is made for a closed season for the trawl fishery from March to April.

Introduction

Lake Malaŵi, set in the Great African Rift valley within tropical East-Central Africa, lies to the south of the equator between latitudes 9° 30′ to 14° 40′ S and longitudes 33° 36′ to 33° 50′ E. It is the third largest lake in Africa by volume and 9th largest in the world by surface area. The demersal trawl fishery of the Southeast Arm (SE Arm) of this lake has been one of the most important fisheries on the lake. While the commercial fishery in Malaŵi contributes only 10% to the total catch of the whole lake, the commercial catch in the SE Arm contributes 37% to the total catch from the lake (Banda et al., 1996; Government of Malawi, 2014; Banda et al., 1997). The highest number of commercial demersal trawlers is found in this part of the lake with the attendant management problems. Since the introduction of the demersal trawl fishery in the late nineties, a number of authors have reported species composition, catch per unit effort (CPUE) for the various species being exploited as well as changes in the size structure of the most important species. Turner (1977a,b, 1995) reported that only after 3 years of the start of exploitation of the demersal fisheries resources in the fishery, dramatic changes started to be seen with a trend towards a fishery dominated by small fish. The implication of these changes has been examined in terms of economic losses due to the decline of some commercially important species.

The study of reproductive biology of a species is an essential component in understanding its ecology, life history, and population dynamics. Reproductive strategies such as size or age-at-maturity and spawning seasonality are of particular interest to fisheries managers as they provide some of the inputs in developing management regulations (Welcomme, 2001). From a fisheries' perspective, reproduction is what perpetuates the benefits that come from the resource. Knowledge of reproductive traits of a fish species is of paramount importance as it allows the fisheries manager to put in place measures that are favorable for its successful reproduction.

For the fisheries manager, knowledge of reproductive traits of a fish stock is of paramount importance as it allows her/him to put in place measures that favor it. Most of the management measures in fisheries such as close-season, closed-area, minimum takeable sizes and cod-end mesh size as well as net mesh size regulations are all based on the manager's knowledge of the behavior and the biological characteristics of the species and their physical size ranges at various stages of their life cycle.

The importance of studying the biology of fish stocks can therefore not be overemphasized as the absence of this knowledge will render any measure taken to be either insufficient or counter to the biological needs of the fish stock that is the subject of the measure. For instance, a lack of knowledge of the reproductive cycle of a species will make it impossible to determine when to close the fishery in order to allow for a successful reproductive season. Lack of knowledge on the sizes at maturity for a species will make the manager unable to decide on the minimum takeable sizes and the cod-end mesh sizes of the trawls being used in the fishery. In addition, reproductive study of fishes is also an important aspect of fisheries management as it permits the manager to understand the life history and the species population dynamics that ensures successful reproduction and recruitment. Further, the study of reproduction biology will enable the fishery managers to identify some life cycle traits that require special environmental conditions for a prosperous reproduction. The objective of the study was to derive reproductive parameters of Lethrinops gossei such as breeding seasonality, size at maturity and size at selection, as well as size structure to use in formulating a fisheries management strategy for this species in the SE Arm of Lake Malaŵi.

Materials and methods

Breeding seasonality

The periods when the species breed in an annual cycle was investigated qualitatively, by examining gonads appearance to determine their developmental stage and by following the evolution of their gonado-somatic index (GSI). According to Barcellos et al. (2001), GSI has been proven to be a significant parameter for monitoring gametogenesis progression in teleost fish.

Monthly samples of fish were drawn from a main pooled sample collected on monthly basis from July 2011 to June 2012 from a research vessel of the Malaŵi Department of Fisheries fishing in the SE Arm of Lake Malaŵi. The sampling depth varied between 49 m to 84 m with 80% of the tows conducted in waters deeper than 60 m. In total and over the 12 months of a sampling program, 2400 specimens of L. gossei were collected for the study. The fish ranged in size between 100–203 mm in total length (TL). and were sexed. From the female fish, 50 individuals were drawn at random per month so that only the female gonads were used. This step was taken because in some cichlids the male gonads do not present a clear annual cycle of physical change and male individuals remain sexually active over longer periods than female fish King (1995). As a consequence, determination of the breeding season is more accurate when it is based on data from female fish only for cichlids.

To access the reproductive organs, the belly of the fish was slit open using a small pair of scissors to expose reproductive organs. The gonads were observed either directly by eye or under dissecting scope to determine at what stage they were in terms of development. Gonadal stage development was assessed using the criteria outlined in Table 1 (Weyl and Hecht, 1999).

The process described above produced 24 sets of data: 2 for each month for GSI and gonad appearance. These data were then analyzed graphically to depict for each month, the value of GSI and proportions of female fish in the various developmental stages 2–5.

The proportion of mature individuals was examined for its maximum to identify the peak breeding season.

To calculate the gonado-somatic index (GSI), the formulation proposed by (Wootton, 1990) was used in the form:
formula

where body weight is the total weight of the body excluding the gonads of the fish and gonad weight is the total weight of gonads.

Sexual maturity

To calculate the proportion, Pm of individuals which were sexually mature or reproductively active (developing, ripe and spent) at length and length at 50% maturity, cumulative frequencies of fish that were found to be mature or reproductively active were converted into proportions and plotted against their lengths in 10 mm class intervals.

The data were fitted into a logistic function in the form:
formula
where Pm is the proportion mature at length l, LM50 is the length at 50% sexual maturity, δ is the width or steepness parameter of the logistic ogive which is also interpreted as the inverse of the rate at which the fish mature.

An incongruity that was encountered was that not all matured individuals were in reproductive condition at the same time. The highest proportion of matured females was 82% in the month of March. According to King (1995), the occurrence of a situation where not all matured individuals are in reproductive condition at the same time leads to an over estimation of the value of Lm50 if the data are fitted to the logistic model directly. In order to circumvent this problem, data on the percent mature were multiplied by a correction factor of 1.22, obtained by taking the inverse of the decimal equivalent of 82%, which is 0.82. (King, 1995).

The model parameters δ and LM50 were estimated by plotting the observed data and fitting a least squares logistic curve using a programmed MS Excel spreadsheet that minimized the sum of squared deviations between observed and predicted data points. This was done using SOLVER (Microsoft Excel 2010) function over 1000 iterations.

The function used was that of a negative Binomial log-likelihood of the form:
formula
where Pl is the proportion mature, nl is the number of fish sampled in length class l and assessed for maturity, and yl is the number of fish in that sample that were mature.

Length at 50% maturity

The length at which the individuals attain 50% maturity, L50mat. was estimated as a reading made from a horizontal line cutting through the 50% graduation mark on the y-axis of the maturity ogive graph.

The mean size at first maturity

The mean size at first maturity Lfmat was estimated by the equation:
formula
where L1, L2, L3 . . ..L10 are lengths of the first 10 smallest specimens found to be mature but less than L50mat..

Size structure and length at 50% selection

The size structure of individuals in the population was studied using bargraph of length data collected from the integrated sample. A logistic Selection model was fitted to cumulative length-frequency data to characterize the recruitment of fish in the fishery. The model was of the form:
formula
where L is length and L50Sel. is the length at which 50% of the fish are retained and δ is the slope parameter.

Results

Breeding seasonality

Results from the examination of the development of both GSI and classification of gonads according to their appearance, showed that L. gossei bred throughout the year. Individuals with developed and ripe gonads were encountered throughout the year in varying proportions. Figure 1 is a bar graph plot of the proportions of various stages of gonads development overlaid by a curve of GSI evolution.

From Figure 1, the main breeding events for the species are in September and in March as shown by the highest proportion of ripe individuals followed by massive numbers of spent individuals. The development of GSI is also in agreement with a small peak and a big peak in September and March respectively. These two months are preceded by high proportions of individuals with developing gonads. The lowest proportion of ripe individuals was observed in the month of July.

Sexual maturity

Figure 2 shows the maturity ogive for L. gossei. The parameter of the logistic model of sexual maturity, δ was 7 mm, while the length at 50% maturity, Lm50 was 137 mm TL. Therefore, the Logistic model of maturity for this species became:
formula
with a coefficient of determination R,2 at 87%, the logistic model adequately described the maturity process.

Mean length at first maturity

Mean length at first maturity was calculated to be 122 mm TL.

Size structure of the population

Figure 3 below shows the distribution of lengths in the sample. The distribution of length frequencies is around a mean length of 145.3 mm TL. The distribution is approximately normal with a median of 145 mm TL.

Note the closeness of the median to the mean. The selection of the trawl net starts retaining fish at a minimum size of 85 mm.

Size selection of the trawl net

Figure 4 below shows the results of a logistic selection curve fitted to data on cumulative proportions of lengths retained by the net during fishing.

The size at which 50% of the fish is retained by the net LSel.50 was visually determined from the logistic curve and had a value of 120 mm TL. The net retained all fish of sizes 170 mm TL and above. Transformation of the equation to calculate δ gave a value of 7.

Discussion and conclusion

A number of researchers have worked on L. gossei from Lake Malaŵi. Recent studies on the species were completed by Singini, 2006 (unpublished MSc. Thesis, University of Malawi) and Duponchelle et al., 2000. With respect to breeding seasonality, the findings of this study agree with that reported by the former who reported that the species bred all year round with a peak season in March/April. Before these two authors, Lewis and Tweddle 1990 reported the breeding season for L. gossei to be March. Regarding maturity, this study found that the size at first maturity was 122 mm TL while that at 50% maturity was 137 mm TL. Singini reported a mean size at maturity of 136 mm TL. Both the finding of this study and that of Singini differ from that reported by Lewis and Tweddle in 1990, which was 159 mm TL.

Duponchelle and Panfili 1998, have reported that size at maturity often varies in successive years for the same population due to environmental conditions or over a longer period under fishing pressure. It is, however unlikely that the decline in the size at maturity observed in this study could be ascribed to fishing pressure as catches of L. gossei have been on the increase in recent years (Government of Malawi, 2014). Further to this, the picture portrayed by the length frequency bar graph, did not have the right-skewed appearance which is a classical sign of recruitment overfishing (Sparre and Venema, 1998). Considering that fishing effort and therefore fishing pressure has been stable in recent years (Government of Malawi, 2014) and that the main method of catching L. gossei has remained the traditional trawl, the logical explanation for the observation should be the recent changes in the aquatic environment brought about by climate change which has caused a drop in water level on the lake. The drop in volume of the lake and the resultant concentration of mineral salts in the waters has most likely led to increasingly high primary production which has reproductively benefited L. gossei more than the other species because of its existing higher standing spawner biomass. The attendant threat though is the finding in this study that shows that the size at 50% selection of the current cod-end of the trawl at 120 mm TL is less than both the mean size at first maturity which is 122 mm TL and the size at 50% maturity which 137 mm TL. This misalignment of these crucial fishery parameters, though small in magnitude could in the long run negatively affect the stock through recruitment overfishing.

This study concluded that stocks of L. gossei in the SE. Arm of Lake Malaŵi may still be resilient to the current level of fishing mortality but caution is needed to limit the fishing mortality preferably by changing the current cod-end mesh from the current 38 mm to 50 mm. Other than reducing the overall fishing effort exerted to the stocks, this measure would to some extent reduce the effects of clogging of the trawl net cod-end (Kanyerere, 1999) caused by the now popular all-day-long tows being practiced by many of the pair trawl operators in the SE. Arm of the lake.

Another observation that this study made was that since the current closed season in the months of October, November and December for the lake is intended for the large cichlids in shallow waters, it does not provide protection to this species which is found deeper in the lake but migrate during breeding to shallower areas accessed by small scale fishers. As it has been shown, the peak breeding season for this species is March. This period falls outside the traditional closed season. Since there has been some calls to institute a closed season for the trawl fishery on the lake and that L. gossei now accounts for a gigantic 26% of the trawl catch, it may be interesting to consider a closed season for the trawl fishery covering the period March-April since this is the period a number of workers have reported to be the breeding season for this species which is gradually becoming the mainstay of the trawl fishery.

Acknowledgements

We would sincerely like to thank the Lilongwe University of Agriculture and Natural Resources for providing facilities to complete the study. Many thanks go to my supervisors, who shared their experiences with me and made time for lengthy discussions on this subject.

Funding

We would sincerely like to thank the Regional Universities Forum (RUFORUM) for providing financial resources for the study.

References

Banda, M. C., Tomasson, T.,
1997
.
Demersal Fish Stocks in Southern Lake Malaŵi: Stock Assessment and Exploitation. Fisheries Bulletin No.35
.
Fisheries Department
,
Lilongwe
.
Banda, M. C., Tomasson, T., Tweddle, D.,
1996
. Assessment of the deep water trawl fisheries of the Southeast Arm of Lake Malaŵi using exploratory surveys and commercial catch data. In: I. G. Cowx (Ed.),
Stock Assessment in Inland Fisheries
, pp.
53
75
.
Fishing News Books
,
Osney Mead, Oxford, UK
.
Barcellos, L. J. G., Woehl, V. M., Wassermann, G. F., Quevado, R. M., Ittzes, I., Krieger, M. H.,
2001
.
Plasma levels of cortisol and glucose in response to capture and tank transference in Rhamdia quelen (Quoy and Gaimard), a South American catfish
.
Aquaculture Research 296
32
(
2
),
121
3
.
Duponchelle, F., Ribbink, A. J.,
2000
.
Fish Ecology Report
.
Lake Malaŵi/ Nyasa/Niassa Biodiversity Conservation Project
.
Duponchelle, F., Panfili, J.,
1998
.
Variation in age and size at maturity of female Nile tilapia, Oreochromis niloticus populations from man-made lakes of Cote d'Ivoire
.
Envir. Biol. Fish.
52
,
453
465
.
Government of Malawi, Fisheries Department, 2014. Catch Assessment Survey, p. 37. Lilongwe, Malawi.
Government of Malaŵi, Fisheries Department
,
2014
.
Catch Assessment Survey
, p.
48
.
Lilongwe Malaŵi
.
Kanyerere, G. Z.,
1999
.
Trawl selectivity and effects of clogging on selectivity in standard trawl surveys in Lake Malaŵi
.
Government of Malaŵi, Fisheries Department, Fisheries Bulletin 39
.
King, M.,
1995
.
Fisheries Biology, Assessment and Management
.
Fishing News Books
,
Osney Mead, Oxford, UK
.
Lewis, D. S. C., Tweddle, D.,
1990
. Breeding seasonality in some commercially important trawl-caught species in the southern Lake Malaŵi. In: T. J. Pitcher, C. R. Hollingworth (Eds.),
Collected Reports on Fisheries Research in Malaŵi: Occasional Paper
, Vol.
1
, pp.
23
35
.
ODA, Overseas Development Administration
,
London
.
Sparre, P., Venema, S. C.,
1998
.
Introduction to Tropical Fish stock Assessment. Part 1 Manual FAO Series Technical Paper No. 3061
,
Rome
.
Turner, G. F.,
1995
. Management, conservation and species changes of exploited fish stocks in Lake Malaŵi. In: T. J. Pitcher, P. J. B. Hart (Eds.),
The Impact of species changes in African Lakes
, pp.
365
395
.
Chapman &Hall, London
.
Turner, J. L.,
1977a
.
Some effects of demersal trawling in Lake Malawi (Lake Nyasa) from 1968 to 1974
.
J. Fish Biol.
10
,
261
71
.
Turner, J. L.,
1977b
.
Changes in the size structure of Cichlid populations of Lake Malaŵi resulting from bottom trawling
.
J. Fish Res Board Can.
34
,
232
238
.
Welcomme, R. L.,
2001
.
Inland fisheries: Biology and management
.
Blackwell Science
,
Oxford
.
Wootton, R. J.,
1990
.
Ecology of Teleost fishes
, 1st Ed.
Chapman & Hall
,
London
.

Note

Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uaem.