investigating key biological parameters of nile tilapia

Hydroécol. Appl. (2021) Tome 21, pp. 157–179© EDF, 2021https://doi.org/10.1051/hydro/2020001

https://www.hydroecologie.org

Investigating key biological parameters of Nile tilapia(Oreochromis niloticus L.) in a large Asian reservoirto better develop sustainable fisheries

Étude des principaux traits biologiques du Tilapia du Nil(Oreochromis niloticus L.) d’un grand réservoir asiatiquedans la perspective d’améliorer le développement de lapêche durable

D. Beaune1,2,*, J. Guillard2, M. Cottet3, K. Kue3, R. Lae4, V. Chanudet5,S. Descloux6, A. Tessier2,7

1 Biogéosciences, UMR 6282 CNRS, Université Bourgogne Franche-Comté, 6, Boulevard Gabriel,21000 Dijon, [email protected]

2 UMR CARRTEL. (INRA –USMB), Centre Alpin de Recherche sur les Réseaux Trophiques et lesEcosystèmes Limniques, 75, bis Av. de Corzent, 74203 Thonon les Bains Cedex, France

3 Nam Theun 2 Power Company Limited (NTPC), Environment & Social Division, EnvironmentDepartment, Gnommalath Office, PO Box 5862, Vientiane, Laos

4 IRD, Lemar UMR 6539, IUEM Technopôle Brest-Iroise, rue Dumont d’Urville, 29280 Plouzané, France5 EDF, Hydro Engineering Center, Savoie Technolac, 73370 Le Bourget du Lac, France6 EDF-Dismantling and Environmental Services Division, 154, avenue Thiers, 69006 Lyon, France7 MAREPOLIS, Université de Perpignan, 58, avenue Paul Allduy, 66860 Perpignan, France

Abstract – The Nile tilapia (Oreochromis niloticus) was introduced in Southeast Asia, suchas in Thailand, in the mid-1960s for aquaculture purposes (Pullin et al., 1997; De Silva et al.,2004). The species was later promoted for aquaculture development in the early 1990s inLao PDR (Garaway et al., 2000). In Lao People’s Democratic Republic (PDR), like in othercountries, this exotic species is well established due to its self-reproduction (De Silva et al.,2004; CABI, 2018) facilitated by its particular life history traits and plasticity (Ishikawa et al.,2013). The species was then suspected to be introduced in the seventies in Laos andsubsequently in the Nam Theun watershed where the NT2 Reservoir was impounded in2008. Population parameters of this alien fish were investigated using the FiSAT II softwarewith the most popular methods of bioparameters assessment to check their interchange-ability within a same stock and the same year 2016. The length-at-age data analysis (usingotoliths analysis, N= 258) gave slightly different results with the length frequency distributionanalysis of fish landings (11 820 individuals). Furthermore, experimental fishing dataprovided irrelevant parameters due to insufficient representativeness of the sample size.The growth parameter K was estimated to be 0.23 year�1, with asymptotic length

mailto:[email protected]

https://doi.org/10.1051/hydro/2020001

https://www.hydroecologie.org

158 D. Beaune et al.

L∞=52.5 cm based on the length frequency distribution analysis with the fish landing data.According to these landing data, the total, natural and fishing mortality were Z=1.41 year�1,M=0.30 year�1 and F=1.11 year�1. The exploitation rate E=0.79 is over the Emax = 0.594obtained by relative yield and biomass per recruit. This estimated stock of 165 tons (morethan 700000 tilapias) was characterized by high mortality (no population growth). Theseresults showed that the population is overfished with too many juveniles caught(L50 = 210.4mm; 50% mature stages at 295mm). To maximize the yield per recruit,increase the biomass and sustain this fishery, enlarging the gillnet mesh size of the gill-net isrecommended. This example highlights the variability of the parameters calculated fromdifferent methods and thus weaken worldwide and even inter-site comparisons. Despite thisissue, the Growth Performances Indices (w

0) gathered into the literature can serve as

baseline and confirmed the wide phenotypic plasticity of the species due to environmentalfactors. Analyses revealed difference between fast growing domesticated fish rose foraquaculture with w

0higher to the tilapia growing in natural and challenging environments.

Key words – cichlidae, FiSAT, fishing management, growth rate, otolith, reservoir

Résumé – Le Tilapia du Nil (Oreochromis niloticus) a été introduit en Asie du sud-ouest,notamment en Thaïlande, dans le milieu des années 1960, à des fins d’aquaculture. Cen’est que dans les années 1990 que cette espèce a été introduite au Laos pour ce mêmeobjectif. Cette espèce exotique, de par sa capacité de reproduction naturelle induite parses traits d’histoire de vie et sa plasticité phénotypique, est bien établie au Laos, ainsi quedans d’autres pays asiatiques. Cette espèce présente dans le bassin versant de NamTheun aurait ainsi colonisé le réservoir de NamTheun 2 (NT2) lors de sa mise en eau en2008, avec également des actions sporadiques d’empoissonnements. Les paramètresde la population du réservoir de NT2 ont été étudiés en recourant à FiSAT II avec lesméthodes d’évaluation des bio-paramètres les plus populaires pour vérifier leurinterchangeabilité dans un même stock et pour une même année (2016). L’analysede la relation taille-âge, via l’analyse de 258 otolithes, a donné des résultats légèrementdifférents de l’analyse des fréquences de tailles issues des débarquements réalisées sur11 820 individus. Les données issues des pêches expérimentales ont quant-à-ellesfournis des résultats non pertinents de par le faible effectif des échantillons. Lacroissance (k) a été estimée à 0,23 cm an�1 avec une longueur asymptotique (L∞) de52,5 cm. Selon les données de débarquements la mortalité totale, naturelle et par pêcheétaient respectivement de 1,41, 0,30 et 1,11 an�1. Le taux d’exploitation (0,79) étaitsupérieur au taux d’exploitation maximal (0,594) obtenu par le rendement relatif et labiomasse par recrue. Le stock estimé de 165 tonnes (plus de 700 000 individus) étaitcaractérisé par une mortalité élevée. Ces résultats ont montré que la population étaitsurexploitée et que de nombreux juvéniles étaient capturés (L50 = 210,4mm; 50%matureà 295mm). Pour maximiser le rendement par recrue, augmenter la biomasse, et soutenirla pêche, il est recommandé d’augmenter la taille des mailles des filets de pêche. Cetexemple met en évidence la variabilité des paramètres calculés à partir de différentesméthodes ce qui sous-entend que les comparaisons avec les autres populations dans lemonde, mais également inter-sites sont difficiles. Toutefois les indices de performancede croissance (w

0) relevés dans la littérature peuvent servir de référentiel et confirmer la

Nile tilapia in a Lao reservoir 159

grande plasticité phénotypique de cette espèce selon les facteurs environnementaux.Les analyses ont révélé que les populations aquacoles présentaient une croissance plusrapide que celle des populations vivant dans les environnements naturels contraignants.

Mots-clés – cichlidae, FiSAT, gestion des pêcheries, taux de croissance, otolithe,réservoir

1 Introduction

Tilapia (including all species) is thesecond most important group offarmed fish after carps, and the mostwidely grown of any farmed fish in theworld (FAO, 2009). The most desirableNile tilapia (Oreochromis niloticus(Linnaeus, 1758); [Perciformes: Cichli-dae]), because of its advantageousbiological traits (feeding habits andfast growth) was deliberately intro-duced in tropical countries during the1960s up to the 1980s, promoted byinternational aid and developmentagencies (FAO, 2009). Consequentlythis alien fish originating from EastAfrica, colonized natural freshwatersystems and is currently well estab-lished in 75 countries (Africa: 16,Asia: 25, Europe: 6, North America:18, Oceania: 3, South America: 7;(FishBase, 2018).

This species is invasive, and be-came the dominant species in terms ofdensity and biomass in various tropicalcolonized areas (Zambrano et al.,2006; Weyl, 2008; Zengeya et al.,2013). In an ecosystem, Nile tilapiapopulation can affect the nutrient cy-cling (e.g. indirectly contribute to theoccurrence of cyanobacteria bloomsand water eutrophication (Figueredoand Giani, 2005), trigger the extinctionofnativeandendemicspecies (Canonico

et al., 2005; Casal, 2006) and probablyaffectotherecologicalprocesses(Canonicoet al., 2005; Cucherousset and Olden,2011). So in freshwater tropical colonizedareas, the Nile tilapia is now the mainfishing resource of protein for humanpopulation, and eventually providesfood security (FAO, 2009). An ade-quate fishing management is thenrequired to avoid overexploitationand preserve sustainable fisheries. Inthis context, knowledge on the keybiological parameters (growth, length,mortality) has to be investigated tobetter understand the local populationbiology and then to better developsustainable fisheries.

One of the most popular methods forestimating biological parameters is touse length frequency data that are fastand cheap to gather (Sparre, 1998) orthe length-at-age analysis based onscales, opercular bones or otoliths agereadings. Both analyses can be per-formed with the FAO-ICLARM (FishCenter) Stock Assessment Tool (FiSATII) which is broadly used for tropical fishstock assessment (> 1700 citations)and contains both routines for theanalyses (Gayanilo et al., 2005). Inthis paper the widely used ’lengthfrequency distribution analysis’ wascompared to the parameters obtainedwith ’analysis of length-at-age’ from fishagedwith otoliths using FiSAT II. These

Table 1. Biological parameters from three methods (2 LFA and 1otoliths) of the stockO. niloticus stockusing FiSAT II. The parameters are: predicted asymptotic length (L∞, Standard (SL) and Total Length(TL)), growth coefficient K, growth performance index w’, and hypothetical length at which age is zero t0,the exploitation rate E, total coefficient of mortality Z, natural mortality M, and fishing mortality F.ID = insufficient data.Tableau 1. Paramètres biologiques pour trois méthode (2 LFA et 1 otolithe) du stock d’O. niloticus viaFiSAT II. Les paramètres sont : longueur asymptotique prédite (L∞, standard (SL) et longueur totale(TL)), coefficient de croissance K, indice de performance de croissance w’ et longueur hypothétiquepour laquelle l’âge est égal à zéro t0, le taux d’exploitation E, coefficient total de mortalité Z, mortaliténaturelle M et mortalité par pêche F. ID = données insuffisantes.

Methods SL∞ (cm) TL∞ (cm) K (year�1) (w’) t0 (years) E Z/M/F (year�1)

Analysis of Length-at-agedata (all fish) (N =258,incl. 58 not sexed)

45.47 55.61 0.14 2.636 �0.28

Analysis of Length-at-agedata (♂Males) (93)

43.81 53.58 0.14 2.604 �0.47

Analysis of Length-at-agedata (♀Females) (107)

31.75 38.83 0.23 2.540 �1.17

Length FrequencyDistribution Analysisfrom fish landings (11 820)

52.5 64.21 0.23 2.977 0.79 1.41/0.30/1.11

Length Frequency DistributionAnalysis from experimentalfishing (95)

24.1 29.47 0.14 2.085 ID ID


three methods are the most commonin the abundant O. niloticus literature(see ref. of Tab. 1).

The study case population is fromLao People’s Democratic Republic, inthe second-largest reservoir in thecountry, the NamTheun2 (NT2). Thisreservoir was impounded in April 2008and after a classical trophic upsurge,the system seems to have reachedecological stabilization in 2011 (Cottetet al., 2016). The Nile tilapia representsthe major part of the catch by the localfishery and is of significant economicimportance (52% of relative biomasscontribution of total catch, (Cottet andVisser, 2017). Biological parameters ofthis fish species constitute essentialdata to assess population dynamic andthe exploitation level for fisheries man-agement. This is especially needed in

this oligotrophic reservoir with relativelylow fish density and where the fishingactivity settled from less than ten years.

Objectives of this study are toprovide: (1) First analyses of detaileddata on a tilapia population in Lao PDR;(2) comparison of popular bio parame-ter assessment methods in a context ofimproving fisheries management andfurthermore; (3) comparison of otherworld tilapia population data.

2 Materials and methods

2.1 Study site

The NamTheun 2 (NT2) hydropowerreservoir is located in the center of LaoPDR (Khammouane Province) on twosub catchments of the Mekong River(Descloux et al., 2016). The reservoir


impoundment started in April 2008 andcommercial operation began inApril 2010. The reservoir has a totalvolume of 3.9 billionm3, covering anarea of 489 km2 at full supply level and86 km2 at the (theoretical) minimumoperating level. The project design ledto divert the waters from the NamTheun watershed to the XeBangfaiwatershed (Descloux et al., 2016). Atotal of 33 fish species belonging to10 families were identified in the reser-voir (between 2008 and 2013) (Cottetet al., 2016). Cyprinidae dominated fishpopulations with a total of 24 species. Atotal number of 2182 households livingwas estimated in 16 hamlets along thereservoir in2014;85%of themareactivein fishing activities with a fluctuatingaverage of [2.5–8 kg HouseHold�1

day�1] (Cottet and Visser, 2017).

2.2 Fish sampling

Data were collected with two sam-pling methods: fish landing and scien-tific experimental fishing.

Each of the three villages of Done,Nongbouakam and Nakai Tai (stars ofthe Fig. 1), was sampled for fiveconsecutive days per month betweenJanuary to December 2016. Each day,the first ten fishermen reaching thelanding site were individually inter-viewed about their fishing gears (type,mesh size, number of hooks, etc.) andtheir catches regarding O.niloticuswere all recorded. For each fisherman,when the number of individuals washigher than 30, only 30 individuals wererandomly selected to be further mea-sured and weighted. A total of11 820 individuals were recorded for

the year, representing about 48% ofthe yield (mixed techniques used from246different fishers, including gillnetswith mesh size ranging between 10 and200mm, hook lines and set lines).

Second data set was obtained fromscientific experimental fishing at fivestations (Fig. 1). A total of 95 fish werecaught from theseexperimentalmonthlygillnet surveys in 2016 (gillnets:25� 2m; mesh size: 10, 15, 20, 25,30, 35, 40, 50, 60 and 70mm) repre-senting 24nights of sampling effort. Thegillnet mesh sizes were selected tocapture a wide range of fish.

2.3 Ethical statement

The experimental fishing were per-formed by the NamTheun2 PowerCompany who has the national autho-rization for fish population monitoringand research in rivers and reservoir bythe Lao Government namely the Minis-try of Agriculture and Forestry repre-sentative on Province level and theNakai Natural Protected area (namelythe Watershed Management Protec-tion Secretariat). This authorizationincludes the traditional fishing methodwith gillnets used in this study. Thisauthorization started in 2008 to allowfish population monitoring in the damarea and as an obligation of monitoringbetween the company and the Govern-ment of Laos in the ConcessionAgreement –Environment and Biodi-versity Obligation. The sampling didnot catch endangered or protected fishspecies, the targeted species areexploited species and the experimentalfishing followed established guidelinesfor treatment of animals in research.

Fig. 1. Map of the NamTheun 2Reservoir in Lao PDRat its higher level and localization of sampling siteby experimental gillnet fishing (black dots) and the three sampled villages by fish landings (black stars).

Fig. 1.Carte du réservoir de Nam Theun 2 au Laos à son niveaumaximal de remplissage et localisationdes sites d’échantillonnage par pêche expérimentale (point noir) et des villages échantillonnés par lesuivi des débarquements (étoile noire).


2.4 Biological parameters

For both sampling methods, thestandard length (SL±1mm) and sexof each individual were recorded.Lengths were assigned to a class sizeof 10mm.

For the analysis of length-at-age,258Nile tilapia individuals were ran-domly selected from the fish landing in2016, and from experimental fishing.Age was determined for each individualby counting annuli on the otoliths aftervalidation of only one annulus forma-tion per year (Tessier et al., 2017).

For the length frequency distributionanalysis, data were first obtained fromfish landing.

This analysis was performed with theElectronic Length Frequency Analysis(ELEFAN I) package incorporated inFiSAT II (Sparre, 1998; Gayanilo et al.,2005). ELEFAN is a very popularmethod frequently used in length fre-quency analysis of fish. It is a non-parametric, ad hoc method, and notdependent on cohort distributions di-rectly. Lengths of each cohort are fixedsuch as in the von Bertalanffy growthmodel (1) which makes a strong


assumption about growth, with theformula containing the asymptoticlength (L∞) or the length the speciescan reach, the length (L) at age classtime t, the instantaneous growth coeffi-cient K and t0 the point at which the vonBertalanffy curve intersects the x-axis:

Lt ¼ L∞ 1� expð�K t�t0ð Þ� �

: ð1Þ

The growth performance index (w’),also found as Ф’ or ø’ in literature, isquantified using (Pauly and Munro,1984) formula (2):

w0 ¼ log10K þ 2log10L∞; ð2Þ

with L∞ expressed in Total Length (TL)here, as required in FiSAT II routine forcomparability. Total mortality coefficient(Z) is estimated from fish landings andexperimental fishing using the length-converted catch curve method (Pauly,1983) from the pooled length frequencydata for the study period (the curve=Z).The natural mortality coefficient (M) isestimated following Pauly’s empiricalformula (Pauly, 1980) (3):

ln Mð Þ ¼ �0:0152� 0:279ln L∞ð Þþ 0:6543ln Kð Þ þ 0:463ln T°Cð Þ;

ð3Þ

linking natural mortality with the vonBertalanffy parameters (K and L∞) andthe mean annual temperature (T °C) ofthe surfacewater inwhich thefish live (inthis case) temperatures of the surfacewaterwere recorded for several stationsevery month by the Environment andSocial Division, Environment Depart-ment of the NamTheun2 Power Com-panyLimitedwhich is themanagerof the

reservoir. The average surface temper-ature for 2016 was 26.5 °C. Fishingmortality (F) was computed from therelationship F=Z�M, while the exploi-tation rate (E) was calculated from (4):

E ¼ FZ¼ F

F þMð Þ : ð4Þ

The recruitment model with pulsewas obtained by projecting the lengthfrequency data back on the time axisusing the growth parameters deter-mined above. The normal distributionof the recruitment model was deter-mined by NORMSEP (Gayanilo et al.,2005).

2.5 Virtual population analysis,fish stock and fishing management

The probabilities of capture areinferred from the length-convertedcatch curves, by backward extrapola-tion of the catch curve and comparisonof the numbers actually caught withthose that “ought” to have been caught(Sparre, 1998; Gayanilo et al., 2005).The standard length for which 25, 50and 75% of the fish are caught (L25,L50,L75) were estimated with running ave-rages technique (Pauly, 1983) withFiSAT II. A virtual population analysis(VPA) estimated the tilapia stock (bio-mass, population size (N)) at the NT2Reservoir from fish landing data andparameters determined above. For thebiomass estimation, the standardlength-weight relationship for theseTilapia was calculated with (5):

log10W ¼ log10aþ blog10L; ð5Þ


where b= [3.23323–3.27171] anda= [0.01699–0.01897] (IC95), N=653fish from 6 to 410mm; length-weightrelationship R2 = 0.9943; (unpub. spe-cies results, method based on (Tessieret al., 2017)).

Based on the previous parametersand frequencies, FiSAT II can still beused to predict the relative yield andbiomass per recruit (Y’/R and B’/R) ofO.niloticus to the fisheries. Plots of Y’/Rvs. E (=F/Z) andofB’/Rvs. E, fromwhichEmax (exploitation rate which producesmaximum yield), E0.1(exploitation rate atwhich the marginal increase of relativeyield-per-recruit is 10%of its unexploitedbiomassE=0) andE0.5 (valueofEunderwhich the stock has been reduced to50%of its unexploited biomass) are alsocomputed through the first derivative oftheBevertonandHolt function (BevertonandHolt,1966).Theseareusedastargetreference points for best fishing man-agement.

Yield contours are plotted to assessthe impact on yield of changes inexploitation rate E and critical lengthratio Lc/L∞. (Lc being mean length offish at first capture, = L50). The resultingisopleths map can be used to deter-mine the best E and Lc/L∞ to reach themaximal Y’/R and B’/R at the currentsituation.

Additional length is calculated asadditional help fishing managementdecision: Lopt which is the lengthwhere the biomass of an unfishedcohort reaches its maximum (Beverton,1992; Froese et al., 2008) (6):

Lopt ¼ L∞3

3þ MK

� � : ð6Þ

The standard length at which 50% ofNile tilapias are still immature and theother half starts to be mature wasestimated based on the 258 fish mea-sured above (SL±1mm). These lengthswere compared with Lc. Maturitystage was described by a 1 or a 0observing: (1) gonads presence (þ ovulefor females=stage3 for thisspecies)and(þ semen dripping from the gonads formales=stage4) or (0) no visible gonads.

A logistic regressionwas fitted inRwiththe general linear model procedure usingthe fishR package. The Thompson andBell yield and stock prediction routine inFiSAT II was used for prediction ofharvested fish and biomass changeswithan example of management recommen-dation (Thompson and Bell, 1934). Thismodel combines features of Beverton andHolt’sY’/Rmodelwith thoseofVPA,whichit inverts. The current fishing rate and Lcwere comparedwith increased Lc and Loptdetermined bellow only, without modifica-tion of F.

2.6 Lao reservoir and worldwide w’comparison

Biological parameters from other tila-pia populations of various ecologicalcontexts (naturaland farmed)havebeengathered from available literature andanalysed for comparisons with the NT2tilapia population. According to (Moreauet al., 1986), the growth performanceindex (w’) values should be equal orroughly similar between different stocksin a normal distribution. We tested thisassumption and hypothesised that thispan-tropical species showed a pheno-typic plasticitywith awide spectrumofw’


irrespectively to continents or watersystems, but fish from aquacultureraised in optimal settings should havea better w’ than in natural and challen-ging conditions (Pauly et al., 1988). Wealso checked (with objective (2)) ifbiological parameter variability couldalsobedue tovariability inmethodology.

To test the normal distribution ofworldwide O.niloticus w’, Shapiro-Wilkand graphical tests were performed onTable 1 values. These datawere collectedfrom published works based on biblio-graphic search, with the keywords “Oreo-chromisþ niloticus” in Web of Science(from 1970 to August2018) and fromrelevant works found in the referencessections. Continental effects on popula-tions, and other potential effects such ashydrosystems (reservoir, lake,cage,pond,etc.)weretestedwithanalysisofvariances,Kruskal-Wallis and Pairwise comparisonsusing Wilcoxon rank sum tests. Thesetests were realized with R software (RDevelopment Core Team, 2011). Only theaverage was considered for cases withmultiple values at a same site.

Because the reported fish lengthswere standard length (SL) or total length(TL) in the literature, we established theO. niloticus TL/SL relationship based on322fish caught and measured into theNT2Reservoir (the 258 aged Nile tilapiaindividuals randomly selected from thefish landing in 2016 þ64 not aged). Therelationship was (R2=0.99): TL=1.223�SL, and it allowed recalculate compa-rable growth performance indices (w’).

3 Results

3.1 Biological parameters

At theNT2Reservoir,ageof the258fishgiven by otolith readings ranged from 0 to

8yearsold (Max. ♀=8years, Max. ♂=7years), with standard length ranging from31 to 360mm. The L∞ mean (±SE) was454.7±123.8mm (♀=317.5±124.1mm;♂=438.1± 284.9mm).The mean growthcoefficient (K) was 0.14±0.06yr�1

(♀=0.23±0.34yr�1;♂=0.14±0.17yr�1)and the growth performance indices (w’)was 2.636 (♀=2.540; ♂=2.604; Tab. 1).For length frequencydistributionanalyses,data ranged from 65 to 495mm for fishlanding and from 37 to 303mm forexperimental fishing. The L∞ means, (K)and (w’) were respectively (no SE avail-able) L∞=525.0mm, (K)=0.230yr�1,(w’)=2.977forfishlandingandL∞=241.5mm, (K)=0.140yr�1, (Tab.2). (w’)=2.085for experimental fishing.

The mean estimated total mortalitycoefficient (Z) was 1.41± 0.23 year�1,IC95 = [1.33–1.50]; natural mortality co-efficient (M) = 0.30 year�1 (at 26.5 °C)fishing mortality coefficient (F) = 1.11year�1 and exploitation rate (E) = 0.79from the fish landing data (Fig. 2 left).The experimental fishing did not allowrelevant results (e.g. relative age of fishat 22 years old, Z<M, etc.).

The recruitment model showed con-tinuous recruitment of O.niloticus allalong the year with slight peaks inJanuary and in March–April (11.0 and18.4%, respectively).

3.2 Virtual population analysisand fishing management

With current Fish mortality, F of 1.11,the optimal Fopt should be 0.5M=0.15and the limit Flim should be 1/3M=0.10.The mean length at the entry into thefishery at L25, L50, and L75 were 202.43,210.38 and 219.28mm respectively

Tab

le2.

Biologica

lparam

etersof

Nile

tilap

ia(O

.nilo

ticus

)in

vario

usloca

tions

from

literature.

The

parametersare:

pred

ictedas

ymptotic

leng

th(L∞,

Stand

ardan

dTotal

Leng

th,italicsva

lues

arereca

lculated

values

),grow

thco

efficien

tK,g

rowth

performan

ceinde

xw’,an

dhy

pothetical

leng

that

which

ageis

zero

t 0,theex

ploitatio

nrate

E,totalc

oefficien

tof

mortalityZ,na

turalm

ortalityM,an

dfish

ingmortalityF.

Tab

leau

2.Param

ètresbiolog

ique

sdu

Tila

piadu

Nil(O

.nilo

ticus

)da

nsdive

rses

loca

lités

etissu

sde

lalittérature.Le

spa

ramètresso

nt:long

ueur

asym

ptotique

préd

ite(L∞,lon

gueu

rstand

ardet

long

ueur

totale,les

valeursen

italiq

ueso

ntde

sva

leursreca

lculée

s),coe

fficien

tdecroiss

ance

K,ind

ice

depe

rforman

cede

croiss

ance

w’,long

ueur

hypo

thétique

pour

laqu

elle

l’âge

estég

alàzé

rot0,taux

d’ex

ploitatio

nE,co

efficien

tde

mortalitétotalZ

,na

turelm

ortalitéM

etmortalitépa

rpê

cheF.

Con

tinen

tCou

ntry

System

Site

SL∞

(cm)

TL∞

(cm)

K(yea

r�1)

w’

t 0 (yrs)

EZ/M

/F(yea

r�1)

Referen

ces

Africa

Ben

inLa

keToh

o34

.341

.50.3

2.71

3�0

.75

0.3

1.10

/0.74/0.36

(Mon

tcho

etal.,20

15)

Africa

Burkina

Fas

aRes

ervo

irBorom

o17

.621

.30.52

22.37

4(Baijotet

al.,19

96)

Africa

Burkina

Fas

aRes

ervo

irTap

oa29

.836

0.39

2.70

4(Baijotet

al.,19

96)

Africa

Côted’Ivoire

Lake

Aya

meI

25.4

30.8

0.22

2.32

0(O

uatta

raet

al.,20

09)

Africa

Côted’Ivoire

Lake

Aya

meI

29.2

35.3

0.4

2.69

80.3

1.42

/0.94/0.48

(Tah

etal.,20

09)

Africa

DRCon

goLa

keAlbert

40.3

48.8

0.5

3.07

6(M

orea

u,19

79)

Africa

Gha

naRes

ervo

irBon

tang

a23

.628

.90.58

2.68

40.3

1.90

/1.35/0.55

(Kwarfo-Ape

gyah

etal.,20

09)

Africa

Ken

yaLa

keNya

nzaGulf

48.6

58.8

0.59

3.30

9�0

.64

0.5

2.12

/1.00/1.12

(Njiruet

al.,20

07)

Africa

Ken

yaLa

keNya

nzaGulf

50.7

61.3

0.39

3.16

60.6

1.71

/0.72/0.99

(Dac

he,19

94)

Africa

Ken

yaLa

keVictoria

38.2

46.2

0.69

3.16

80.5

2.18

/1.14/1.05

(Yon

goan

dOuta,

2016

)Africa

Mad

agas

car

Lake

Kainji

58.2

70.4

0.4

3.29

7(M

orea

u,19

79)

Africa

Mad

agas

car

Lake

Itassi

59.1

71.5

0.14

2.85

5(M

orea

u,19

79)

Africa

Nigeria

Lake

Kainji

44.0

53.2

0.29

2.91

4(D

uFeu

,20

03)

Africa

Nigeria

Lake

Gbe

dike

re36

.243

.8(Ade

yemie

tal.,20

09)

Africa

Nigeria

Res

ervo

irOpa

46.9

56.7

0.26

2.92

2(King,

1997

)Africa

Tch

adLa

keTch

ad37

.345

.20.31

82.81

2�0

.11

(Blach

eet

al.,19

64)

Africa

Uga

nda

Lake

Nab

ugab

o38

.346

.30.41

2.94

5�0

.06

(Bwan

ikaet

al.,20

07)

Africa

Uga

nda

Lake

Wam

ala

24.81

30.0

(Bwan

ikaet

al.,20

07)

Africa

Burkina

Fas

oRes

ervo

irBolmigou

19.6

23.7

0.28

62.20

6(Baijotet

al.,19

96)

Africa

Burkina

Fas

oRes

ervo

irKok

olo

19.3

23.3

0.19

62.02

7(Baijotet

al.,19

96)

Africa

Burkina

Fas

oRes

ervo

irRam

iteng

a15

.018

.20.58

2.28

4(Baijotet

al.,19

96)

Africa

Burkina

Fas

oRes

ervo

irSorou

18.2

220.7

2.53

0(Baijotet

al.,19

96)

Africa

Burkina

Fas

oRes

ervo

irTan

gugiga

14.5

17.6

0.46

22.15

6(Baijotet

al.,19

96)

Africa

Burkina

Fas

oRes

ervo

irTan

guiga

14.5

17.6

0.46

2.15

4(FishB

ase,

2018

)Africa

Egy

ptLa

keBurullos

28.6

34.6

0.21

2.40

1�0

.42

(San

gak,

2010

)Africa

Egy

ptLa

keMariut

33.6

40.6

0.35

62.76

9�0

.12

(Jen

sen,

1957

)Africa

Egy

ptLa

keMariut

35.2

42.6

0.46

2.92

10.54

1(El-Z

arka

etal.,19

70)

Africa

Egy

ptLa

keNas

ser

52.1

63.7

0.26

3.02

4(M

orea

uet

al.,19

86)

Africa

Egy

ptLa

keNas

ser

47.8

57.8

0.51

63.23

7(M

orea

u,19

79)


Tab

le2.

(con

tinue

d).

Con

tinen

tCou

ntry

System

Site

SL∞

(cm)

TL∞

(cm)

K(yea

r�1)

w’

t 0 (yrs)

EZ/M

/F(yea

r�1)

Referen

ces

Africa

Egy

ptLa

keNas

ser

53.8

65.1

0.21

62.96

2(M

orea

u,19

79)

Africa

Egy

ptLa

keEdk

u28

.534

.50.2

2.37

7�0

.53

0.4

1.39

/0.61/0.78

(Abd

-Alla

andTalaa

t,20

00)

Africa

Egy

ptCan

alEl-B

ahrEl-F

arao

uny

30.8

37.3

0.29

42.61

1�0

.09

0.4

1.15

/0.65/0.49

(El-K

ashe

ifet

al.,20

15)

Africa

Egy

ptRiver

Nile

39.8

48.1

0.14

72.53

2�0

.22

0.6

0.97

/0.40/0.56

(Has

sanan

dEl-K

ashe

if,20

13)

Africa

Egy

ptRiver

Nile,Dam

ietta

bran

ch30

.036

.30.12

2.19

9�1

.12

0.4

0.64

/0.39/0.25

(Authm

anet

al.,20

09)

Africa

Egy

ptRiver

Nile,Ros

etta

bran

ch23

.528

.50.39

2.50

1�0

.82

0.5

1.62

/0.80/0.82

(Mah

mou

dan

dMaz

rouh

,20

08)

Africa

Egy

ptLa

keNoz

haHyd

rodrom

e31

.538

.10.21

12.48

5�0

.43

0.4

0.87

/0.52/0.35

(Mah

mou

det

al.,20

13)

Africa

Gha

naLa

keVolta

33.5

41.0

0.35

2.76

9(D

eGraaf

andOfori-Dan

son,

1997

)Africa

Gha

naLa

keVolta

33.5

41.0

0.55

2.96

5�0

.38

0.3

1.21

/090

/0.31

(Ofori-Dan

son,

2005

)Africa

Ken

yaLa

keNya

nzaGulf,Victoria

53.4

64.6

0.25

3.01

80.3

0.82

/0.54/0.28

(Getab

u,19

92)

Africa

Nigeria

Lake

Gbe

dike

re36

.243

.80.32

2.78

8(Ade

yemie

tal.,20

09)

Africa

Uga

nda

Lake

Albert

39.0

47.7

0.5

3.05

6(Sse

nton

go,19

71)

America

Brazil

Res

ervo

irBarra

Bon

ita33

.641

.10.63

3.02

70.4

2.81

/1.61/1.20

(Cos

taNov

aesan

dCarva

lho,

2012

)America

Mex

ico

Lake

Coa

tetelco

17.88

21.9

0.34

2.21

1�1

.54

(Góm

ez-M

árqu

ezet

al.,20

08)

Asia

Ban

glad

esh

Res

ervo

irKap

tai

46.0

55.6

0.39

3.08

10.4

1.13

/0.80/0.59

(Ahm

edet

al.,20

03)

Asia

Irak

River

Tigris

24.3

29.4

0.10

81.96

9(Atte

eet

al.,20

17)

Asia

Lao

Res

ervo

irNam

The

un2

52.5

64.21

0.23

2.97

70.5

1.41

/0.30/1.11

Thisstud

y(LFA

metho

d)Asia

Lao

Res

ervo

irNam

The

un2

45.5

55.6

0.14

2.63

6�0

.28

Thisstud

y(otoliths

metho

d)Asia

Pak

istan

Res

ervo

irCha

shma

18.8

22.78

0.07

1.56

0�0

.16

0.9

1.11

/0.16/0.95

(Meh

aket

al.,20

17)

Asia

SriLa

nka

Res

ervo

irAda

walaw

e36

.444

.00.49

2.97

7../1.01

/..(M

orea

uet

al.,20

08)

Asia

SriLa

nka

Res

ervo

irKau

dulla

45.1

54.5

0.35

3.01

70.6

1.95

/0.75/1.07

(Amaras

ingh

ean

dde

Silva,

1992

)Asia

SriLa

nka

Res

ervo

irMinne

riya

36.0

43.5

0.32

2.78

2../0.78

/..(M

orea

uet

al.,20

08)

Asia

SriLa

nka

Res

ervo

irMinne

riya

45.1

54.5

0.43

3.10

60.7

3.57

/0.87/2.61

(Amaras

ingh

ean

dde

Silva,

1992

)Asia

SriLa

nka

Res

ervo

irTab

bowa

42.0

50.7

0.64

3.21

60.4

1.96

/1.17/0.79

(Amaras

ingh

e,20

02)

Asia

SriLa

nka

Res

ervo

irVictoria

31.6

38.2

0.6

2.94

2../1.19

/..(M

orea

uet

al.,20

08)


Fig. 2. Length-converted catch curve (left) of O. niloticus (N =11820) estimating annual instantaneoustotal mortality (Z) = 1.41 year�1; Natural mortality (at 26.5 °C) (M) = 0.30 year�1; Fishing mortality(F) = 1.11 year�1; Exploitation rate (E) = 0.79. NT2 Reservoir, Lao, PDR (year 2016). Black dots of theleft figure were used for Z calculation (i.e. the slope) using least squares line regression (R2 =0.98). Thewhite dots represent the fish that would have been caught if fully recruited. The yellow dots on the left arethe not fully recruited points discarded from the calculation; The ratio: expected and actual catches,gives probability of capture for each class size(right).

Fig. 2. À gauche : Courbe des captures converties en tailles d’O. niloticus (N= 11820) estimant lamortalité instantanée annuelle (Z) = 1,41 an�1 ; naturelle (à 26,5 °C) (M) = 0,30 an�1 ; par pêche(F) = 1,11 an�1 ; le taux d’exploitation (E) = 0,79 pour NamTheun2 en 2016. Les ronds noirscorrespondent aux données utilisées pour calculer Z en utilisant la régression linéaire des moindrescarrées (R2 = 0.98). Les ronds blancs représentent les poissons qui auraient été capturés si lerecrutement avait été total. Les ronds jaunes sont des données correspondant à un recrutement nontotal et écartés du calcul. À droite : Le ratio : les captures prévues et les captures réelles donnent laprobabilité de capture pour chaque taille de classe.


(Fig. 2 right). The mean length of fish atfirst capture (Lc) was equivalent toL50 = 210.38mm. Immature fish werefound up to 360mm but some maturefish started at 112mm. The proportionof mature fish was 50% for a standardlength of 295mm (Fig. 3). From theFiSAT II virtual population analysis, thestock of the reservoir was estimated tobe 165 tones with about 744 000Niletilapias (Fig. 4). The exploitation ratewas E=0.79. This rate was >E10 =0.508, >E50 = 0.336 and >Emax =0.594 obtained by relative yield andbiomass per recruit (Fig. 5). The yield-per-recruit isopleths diagram of thevarious length at entry for O.niloticus

into the fishery based on differentvalues of E and a constant value ofM/K=1.304 and L∞=525mm showedthe discontinued curves indicating therange producing the maximum yield-per-recruit (Fig. 6). The maximumtheoretical value of relative yield-per-recruit at the meeting point of theeumetric yield curve (i.e. stable popu-lation) with the maximum sustainableyield (MSY) (so-called potential yield-per-recruit) was Y’/R=0.062 withE=0.7 and Lc/L∞=0.6, that to sayLc=315mm. The current level ofexploitation (Y’/R=0.05) is on the rightside of the eumetric curve indicatingpotential overfishing with too many

0 100 200 300 400 500

Standard Length (mm)

Pro

porti

on M

atur

e

0.0

0.5

1.0

01

Fig. 3. Fitted logistic regression for proportion of mature Nile tilapia (O. niloticus) by standard lengthwith L50 = 295mm shown in blue (N=258) in NT2Reservoir, Lao, PDR. The current mean length of fishat first capture was Lc=210.38mm.

Fig. 3.Régression logistique ajustée pour la part de Tilapia du nil (O. niloticus) mature selon la longueurstandard avec L50 = 295mm indiqué en pointillé bleu (N=258) pour NamTheun2, Laos. La longueurmoyenne actuelle des poissons à la première capture (Lc) était de 210,38mm.


short fish caught (Fig. 6). Accordingly tothe figures, a reduction of E (moving tothe left) and/or a reduction of thetargeted class size (with bigger hooksand larger gill net mesh (moving to thetop)), would lead to an increase inrelative yield.

Value of Lopt=365.9mm is consid-ered as the size of exploitation at whichO. niloticus yield attains its maximumpotential in this reservoir.

A Thompson and Bell Yield-StockPrediction with a L50 of 315mm wascompared to the current prediction withL50=210.38mm. For the same fishingeffort of F=1.11, but with an increase ofmesh size, the yield was near optimalcapability and with þ125% than thecurrent level. On the other hand, thebiomasshasmore than doubled (207%).

3.3 Lao reservoir and worldwide w’comparison

Biological parameters of severalNile tilapia populations from variouscontinents and hydrosystems aresummarized in Table 2. The rangeof parameter values was wide forpopulations from the different ecosys-tems (lakes, reservoirs, rivers): L∞=[SL: 145.5–591.3 cm], K = [0.07–0.70], w’ = [1.56–3.31]. Globally thew’ distribution was normal(W= 0.94605, p-value = 0.0001584).The average w’ was significantlyhigher in fish farms = 3.387; IC95 =[3.332–3.443] (ponds, cages, tanks,Tab. 3) than in ecosystem 2.735;IC95 = [2.624–2.846] (lakes, reser-voirs, rivers) (W= 173, p< 0.0001).

Fig. 4. Length-Structured (TL) Virtual Population Analysis (VPA) of O.niloticus (from fish landings) inNT2Reservoir, Lao, PDR (year 2016).

Fig. 4. Analyse de la structure de taille (TL) d’une population virtuelle (VPA) d’O. niloticus (issus desdébarquements de pêche) pour le réservoir de NamTheun2, au Laos, en 2016.

Fig. 5. Relative yield and biomass per recruit (selection ogive) ofO. niloticus. NT2Reservoir, Lao, PDR(year 2016). E10 = 0.508, E50 = 0.336 and Emax = 0.594 (yellow dotted line). Current E=0.79.

Fig. 5. Rendement relatif et biomasse par recrut (sélection ogive) d’O.niloticus pour le réservoir de NamTheun 2, auLaos, en2016.E10 =0.508,E50 =0.336andEmax=0.594 (lignepointillée jaune). ActuellementE=0,79.


Fig. 6. 3Dmap of relative yield per recruit (Y’/R) (left), and relative biomass per recruit (B’/R) (right) as afunction of relative size at first capture (Lc/L∞) and exploitation rate (E=F/Z Knife-edge with fixed M/K=1.304. The relative yield and relative biomass per recruit increased with the vertical axis (isoplethcolor going to orange). For the Y’/R, if the situation point is at the left of the slope (eumetric yield curve)the scenario is in underfishing situation, at the right of the slope: in overfishing. The situation in 2016 inthe NT2Reservoir for Nile tilapia was E=79 and Lc/L∞=0.4. A reduction in exploitation rate and in Lcwith mesh size/hooks increase lead to an increase in relative yield (and overall biomass) and would bethe fishing management recommendation.

Fig. 6. Représentation 3D du rendement relatif par recrue (Y’/R) (à gauche) et de la biomasse relative parrecrue (B’/R) (àdroite) en fonctionde la taille relativeà lapremièrecapture (Lc/L∞) et du tauxd’exploitation(E=F/Z). Tranche d’âge avecM/K fixé à 1,304. Le rendement relatif et la biomasse relative par recrue ontaugmentéavec l’axevertical (la couleurdes isoplèdesallantà l’orange).Pour leY’/R, si lepoint desituationest à gauche de la pente (courbe de rendement eumétrique), le scénario est en situation de sous-pêche, àdroitede lapente : ensurpêche.Lasituationen2016duréservoir deNamTheun2pour le tilapiaduNilétaitE=79 et Lc / L∞=0,4. Une réduction du taux d’exploitation et de Lc avec une augmentation de la taille dumaillage ou recourt aux hameçons conduit à une augmentation du rendement relatif (et de la biomasseglobale) et constituerait donc une recommandation de gestion pour cette pêche.


There were differences within theecosystem and aquaculture groups(Wilcoxon rank sum tests with P< 0.05, Fig. 7). For ecosystems, thepopulations were not different be-tween Africa and Asia (W= 222, p-value = 0.797).

4 Discussion

The NT2Reservoir was impoundedin 2008 and tends to reach an ecologi-cal equilibrium, despite significant ex-ploitation of the ichtyo resources

(Cottet et al., 2016; Cottet and Visser,2017). More than 2000households relyon the O. niloticus for food. represent-ing half of their fisheries. Consequently,knowledge on biological parameters ofO. niloticus is useful for fishing man-agement in this recent Asian reservoirand more largely for the South EastAsian region. Globally, several meth-ods exist to help fish stock manage-ment, most of them usable in variousroutines with FiSAT II (Sparre, 1998).This study showed that the results canbe slightly different between length-at-age analyses versus length frequency

Table 3. Biological parameters of Nile tilapia (O. niloticus) in aquaculture (Pauly et al., 1988). Withpredicted asymptotic length (L∞, Standard and Total Length), growth coefficient K, and growthperformance index w’.Tableau 3. Paramètres biologiques du Tilapia du Nil (O.niloticus) en aquaculture (Pauly et al., 1988).Les paramètres sont : les longueurs asymptotiques prédite (L∞ pour la longueur standard et la longueurtotale), le coefficient de croissance K et l’indice de performance de croissance w’.

Continent Country Hydrosystem SL∞ (cm) TL∞ (cm) K (year�1) (w’)

Africa Burkina Faso Pond 14.6 17.9 2.385 2.881Africa Burkina Faso Pond 22.3 27.3 0.88 2.816Africa Cameroon Pond 21.9 26.8 4.329 3.492Africa CAR Tank 15.1 18.5 7.305 3.396Africa CAR Tank 15.6 19.1 3.003 3.039Africa CAR Tank 17.3 21.2 2.997 3.128Africa CAR Tank 24.9 30.5 1.572 3.164Africa CAR Tank 25.9 31.7 1.447 3.162Africa Côte d’Ivoire Cage 19.3 23.6 4.737 3.421Africa Côte d’Ivoire Cage 23.9 29.2 1.881 3.206Africa Côte d’Ivoire Pond 19.1 23.4 6.604 3.557Africa Côte d’Ivoire Pond 23.4 28.6 3.69 3.480Africa Sierre Leone Pond 18.9 23.1 2.95 3.198America Puerto Rico Pond 15.3 18.7 5.342 3.272America USA Pond 14.2 17.4 7.291 3.342America USA Pond 20.8 25.4 6.58 3.629America USA Pond 22.5 27.5 10.458 3.899America USA Pond 22.8 27.9 4.179 3.512America USA Pond 22.8 27.9 8.028 3.795Asia Philippines Cage 12.9 15.8 13.933 3.540Asia Philippines Cage 13.2 16.1 6.544 3.232Asia Philippines Cage 15.4 18.8 7.843 3.444Asia Philippines Cage 22.4 27.4 2.915 3.340Asia Philippines Pond 12.4 15.2 7.155 3.216Asia Philippines Pond 12.7 15.5 12.348 3.474Asia Philippines Pond 13.5 16.5 9.22 3.400Asia Philippines Pond 14.9 18.2 5.685 3.276Asia Philippines Pond 15.3 18.7 10.377 3.560Asia Philippines Pond 15.6 19.1 10.742 3.592Asia Philippines Pond 16.3 19.9 9.352 3.570Asia Philippines Pond 17.4 21.3 10.631 3.683Asia Philippines Pond 20.6 25.2 5.409 3.536Asia Philippines Pond 23.6 28.9 6.879 3.758Asia Philippines Pond 23.8 29.1 3.854 3.514Asia Thailand Pond 12.9 15.8 5.476 3.134Asia Thailand Pond 13.7 16.8 9.712 3.436Asia Thailand Pond 15.8 19.3 8.889 3.521Asia Thailand Pond 16.6 20.3 10.942 3.654Asia Thailand Pond 16.9 20.7 7.651 3.514Asia Thailand Tank 10.3 12.6 19.644 3.494Asia Thailand Tank 10.7 13.1 7.263 3.095Asia Thailand Tank 11.1 13.6 13 3.379Asia Thailand Tank 11.4 13.9 6.423 3.096Asia Thailand Tank 11.8 14.4 4.15 2.937Asia Thailand Tank 12 14.7 9.871 3.328Asia Thailand Tank 12.3 15.0 6.534 3.170Asia Thailand Tank 13.3 16.3 7.093 3.273Asia Thailand Tank 13.3 16.3 7.628 3.305Asia Thailand Tank 13.7 16.8 4.75 3.125


Table 3. (continued).

Continent Country Hydrosystem SL∞ (cm) TL∞ (cm) K (year�1) (w’)

Asia Thailand Tank 15.7 19.2 7.157 3.421Asia Thailand Tank 15.8 19.3 7.109 3.424Asia Thailand Tank 18.3 22.4 5.881 3.469Europe Belgium Pond 19 23.2 4.789 3.413Europe Belgium Pond 21.5 26.3 2.265 3.195Europe Belgium Pond 22.2 27.2 4.271 3.498Europe Belgium Pond 23.1 28.3 2.54 3.307Europe Belgium Pond 40.8 49.9 0.716 3.251Europe Belgium Tank 13.2 16.1 13.207 3.537Europe Belgium Tank 17.3 21.2 12.654 3.753Europe Belgium Tank 18.6 22.7 4.221 3.339Europe Belgium Tank 18.9 23.1 10.665 3.756Europe Belgium Tank 22.5 27.5 2.984 3.354Europe Belgium Tank 23.3 28.5 5.419 3.643Europe Belgium Tank 29.9 36.6 2.076 3.443


distribution growth analyses, with thesame fish stock, sampled at the sameyear. The sampling method can affectbiological parameters results and thuspotentially changing other parameterssuch as mortality M or even Bevertonand Holt Y/R analyses (Knife-edge)which is calculated mainly with K andL∞.

Otolith readings are a robust methodbut the lower sample size of this studyprevented recording fish of extremelarge sizes (for L∞ determination). Forinstance, a 8 years old female wasaged in this study, which is the oldestage published for Nile tilapia in theliterature (FishBase, 2018), but had aSL of only 280mm (in the literature).Thus indicating a probable margin ofmore growing. The L∞ is = 454 versus525mm for the length at age and thelength frequency analyses respectively.However, K was 0.14 for males and forall sex pooled, but 0.23 for femalesonly, equal to 0.23 found from thelength frequency analyses. Theseresults complicated the inter andintra-population comparisons regularly

found in literature (Tab. 2). Further-more, site comparisons have to becautious because some parametersbased on experimental fishing with arefrom low sample size or short period ofsampling. However, in this study case,insufficient sampling showed irrele-vant results for this type of analysisdespite costly sampling efforts(24 nights, 3 staff/boat for this proto-col). The main conclusion for fishlandings or experimental samplingchoice of protocol is that bulky databased on fishermen are more robusteven if the catches are not standard-ized with a fishing protocol, because ofthe high data number reaching repre-sentativeness. Furthermore, this meth-od is logistically easier. This method isthus recommended according to thescientific questions and needs. Thisexample recalls best practices forusing the ELEFAN approach (Breyand Pauly, 1986; Pauly and Morgan,1987; Taylor and Mildenberger, 2017):the length frequency data should (i)contain relatively high counts that arerepresentative of the length distribution

Fig. 7. The growth performance indices (w’) of Nile tilapia (O. niloticus) in various hydrosystems.Groups significantly different are represented by letters (Wilcoxon rank sum tests, P< 0.05). Beloweach box, the effective of bibliographic resources indicates the number of used. N are indicated belowthe boxes. The star represents the NT2Reservoir, Lao PDR.

Fig. 7. Les indices de performance de croissance (w’) du Tilapia du Nil (O. niloticus) dans divers hydro-systèmes. Les groupes significativement différents sont identifiés par une lettre (test de Wilcoxon,P< 0,05. L’étoile représente NamTheun2 au Laos.


of the population or catches; (ii) besampled at regular intervals throughoutthe year with the same sampling effortand (iii) cover a substantial degree ofsmall class sizes, as their relativelyhigher numbers and faster growth willaid in the fitting of the growth curve. Ageneral rule-of-thumb could be that thesmallest class size should start at least25% of L∞. Our experimental fishingstudy case failed for condition (i).

This finding highlights the need tocompare the biological parameters of afish population with cautious. Indeed,various methods seem to provideslightly different results. The speciesrange for K is [0.14–0.41 years�1] onFishbase (FishBase, 2018), but seemsto be wider on Table 1. w’ can be used

to compare populations as long as similarunits and definitions are used (seeroutine’s requests in FiSAT II) andcautions have to be taken for ichthyo-logists for comparisons using external w’data.However, themeaningful lowgrowthperformance indices w’ can be noticed forpopulationssufferingchallengingenviron-ment (e.g. oligotrophic waters, low tem-perature, etc. and high w’ for populationunder relatively severe predation/fishingpressure (Amarasinghe, 2002; Njiruet al., 2007; Moreau et al., 2008; CostaNovaes and Carvalho, 2012; Yongo andOuta, 2016).

Based on these data, O. niloticus ofNT2Reservoir seems to be overfished.The current E is 0.79 which is superiorto Emax = 0.594, leading to an over


exploitation of the population with re-ducedbiomass,densityandultimately toreduced yield per recruit relatively to itsoptimal potential (0.05 insteadof 0.062).Reducing the fishingmortality from 1.11to 0.45 would increase the populationbiomass by 195% and thus the yield by114%. Furthermore, the fish start to becaught at 210mm through the currentgillnet mesh size. Many of these fish didnot reproduced more than once or areeven immatureandwill never participateto the population renewal. Using biggermesh size would lead to an increase inthe yield. The social, economic andtechnical implications of changes incurrent fishing practices must not beoverlooked.Adedicated feasibility studyshould probably be done and a peda-gogic and technical support should beprovided to local populations. Otherecological effects of the O.niloticusdeserve to be studied.

ACKNOWLEDGMENTS

This work was funded by the EDF-CIH (Centre d’Ingénierie Hydraulique–France), by the NamTheun2 PowerCompany (NTPC–Lao PDR) through aproject led by UMR CARRTEL (INRA–

USMB, France). The fieldwork wasperformed thanks to the NamTheun2Power Company, whose shareholdersare Électricité de France, Lao HoldingState Enterprise and Electricity Gener-ating Public Company Limited ofThailand. The authors would like tothank all villagers who participated inthe fish landing survey, the staff ofthe NTPC, especially the EnvironmentDepartment, including Khampong

THAN-Onkeo, Sengthien Vongvilay-suk, Seesouk XAYABOUNPHENG,and Sengsavanh PHOMPHILATH, fortheir technical and human support inthe laboratory in Lao PDR. We wouldalso like to thank Michel COLON andJean-Christophe HUSTACHE of theUMR CARRTEL for their help in thefield and the laboratory in Lao PDR.

REFERENCES

Abd-Alla A. & Talaat, K.M., 2000. Growthand dynamics of tilapias in Edku Lake,Egypt, Bull. Nat. Inst. Oceanogr. Fish.26: 183–196.

Adeyemi S.O., Bankole N.O., Adikwu I.A. &Akombu P.M., 2009. Age, growth andmortality of some commercially impor-tant fish species in Gbedikere Lake,Kogi State Nigeria, Int. J. Lakes Rivers 2(1): 45–51.

Ahmed K.K.U., Amin S.M.N., Haldar G.C. &Dewan S., 2003. Population dynamicsand stock assessment of Oreochromisniloticus (Linnaeus) in the Kaptai Res-ervoir, Bangladesh, Indian. J. Fish. 50(1): 47–52.

Amarasinghe U.S., 2002. The fishery andpopulation dynamics of Oreochromis-mossambicus and Oreochromis niloti-cus (Osteichthyes, Cichlidae) in ashallow irrigation reservoir in Sri Lanka,Asian Fish. Sci. 15: 7–20.

Amarasinghe U.S., & de Silva S.S., 1992.Population dynamics of Oreochromismossambicus and O. niloticus (Cichli-dae) in two reservoirs in Sri Lanka,Asian Fish. Sci. 5: 37–61.

Attee R.S., Alheni A., Jassem K., Abd A. &Khalifa S.Z., 2017. Description of thegrowth of Nile tilapia (Linnaeus, 1758)Oreochromis niloticus in the Tigris Riversouth of Baghdad, Int. J. ChemTechRes. 10(9): 9.


Authman M.M.N., El-Kasheif M.A. & Shal-loof K.A.S., 2009. Evaluation and man-agement of the fisheries of Tilapiaspecies in Damietta Branch of the RiverNile, Egypt, World J. Fish. Mar. Sci. 1:167–184.

Baijot E., Barry I., Bouda S., Kabore K.H. &Traore A.C., 1996. Comparative growthof Oreochromis niloticus and Sarother-odon galilaeus in small artificial lakes inBurkina Faso and the larger lakes ofAfrica. In: The third International Sym-posium on Tilapia Aquaculture,ICLARM Conf. Proc., pp. 28–32.

Beverton R. & Holt S., 1966. Manual ofmethods for fish stock assessment-Part2-tables of yield functions, FAO Fish.Tech. Paper 38: 67.

Beverton R.J.H., 1992. Patterns of repro-ductive strategy parameters in somemarine teleost fishes, J. Fish Biol. 41:137–160.

Blache J., Milton F., Stauch A., Iltis A. &Loubens G., 1964. Les poissons desbassins du Tchad et du bassin adjacentdu Mayo Kebbi : étude systématique etbiologique.

Brey T. & Pauly D., 1986. Electronic lengthfrequency analysis: a revised andexpanded user’s guide to ELEFAN 0,1 and 2.

Bwanika G.N., Murie D.J. & Chapman L.J.,2007. Comparative age and growth ofNile tilapia (Oreochromis niloticus L.) inlakes Nabugabo and Wamala, Uganda,Hydrobiologia 589(1): 287–301.

CABI, 2018. Oreochromis niloticus. Inva-sive Species Compendium. Walling-ford, UK: CAB International. www.cabi.org/isc.

Canonico G.C., Arthington A., McCrary J.K.& Thieme M.L., 2005. The effects ofintroduced tilapias on native biodiversity,Aquat. Conserv.: Mar. Freshw. Ecosyst.15(5): 463–483.

Casal C.M.V., 2006. Global documentationof fish introductions: the growing crisisand recommendations for action, Biol.Invasions 8(1): 3–11.

Costa Novaes J.L., & Carvalho E.D., 2012.Reproduction, food dynamics and exploi-tation level of Oreochromis niloticus(Perciformes: Cichlidae) from artisanalfisheries inBarraBonitaReservoir, Brazil,Revista de Biol. Trop. 60(2): 721–734.

Cottet M. & Visser T.A.M., 2017. Fish catchand fishing practices in the Nam Theun2 Reservoir and watershed (Lao PDR),Lakes Reserv.: Res. Manag.

Cottet M., Descloux S., Guédant P., CerdanP. & Vigouroux R., 2016. Fish popula-tion dynamic in the newly impoundedNam Theun 2 Reservoir (Lao PDR),Hydroécologie Appliquée 19: 321–355.

Cucherousset J. & Olden J.D., 2011.Ecological impacts of nonnative fresh-water fishes, Fisheries 36(5): 215–230.

Dache S.A.O., Observations on the fisheries,growth and mortality of Oreochromisniloticus (tilapia) in Nyanza Gulf of LakeVictoria. In: Proceedings of the secondEECregional seminaron the recent trendsof research on Lake Victoria fisheries,ICIPE Press, Kenya, 1994, pp.56–65.

De Graaf G.J. & Ofori-Danson P.K., 1997.Catch and fish stock assessment instratum VII of Lake Volta. FAO.

De Silva S.S., Subasinghr R.P., Bartley D.M. & Lowther A., 2004. Tilapias as alienaquatics in Asia and the Pacific: areview. Rome: FAO.

Descloux S., Guedant P., PhommachanhD. & Luthi R., 2016. Main features of theNam Theun 2 hydroelectric project (LaoPDR) and the associated environmentalmonitoring programmes, Hydroécolo-gie Appliquée 19: 5–25.

Du Feu T.A., 2003. Population parametersfor the six commercial species in LakeKainji, Nigeria using length frequency

http://www.cabi.org/isc

http://www.cabi.org/isc


datasampled fromartisanalfishcatches,ACP-UE Fish. Res. Rep. 14: 15.

El-Zarka S., Shaheen A.H. & Aleem A.A.,1970. Tilapia fisheries in Lake Mariut.Age and growth of Tilapia nilotica (L.) inthe Lake, Bull. Inst. Oceanogr. Fish.Egypt 1: 149–182.

FAO, 2009. Oreochromis niloticus. In:Cultured aquatic species fact sheets.CD-ROM. multilingual). http://www.fao.org/tempref/FI/CDrom/aquaculture/(Accessed: 20/07/2018).

Figueredo C.C. &Giani A., 2005. Ecologicalinteractions between Nile tilapia (Oreo-chromis niloticus L.) and the phyto-planktonic community of the FurnasReservoir (Brazil), Freshw. Biol. 50(8): 1391–1403.

FishBase, 2018. FishBase. World WideWeb electronic publication. www.fishbase.org (Accessed: 24/07/2018).

Froese R., Stern-Pirlot A., Winker H. &Gascuel D., 2008. Size matters: howsingle-species management can contri-bute to ecosystem-based fisheries man-agement, Fish. Res. 92(2-3): 231–241.

Garaway C., Lorenzen K. & Chamsingh B.,2000. Developing fisheries enhance-ment in small waterbodies: lessons fromLao PDR and Northeast Thailand. In:Reservoir and culture-based fisheries:Biology and management. Bangkok,Thailande: ACIAR, pp. 227–234.

Gayanilo F.C., Sparre P. & Pauly D., 2005.FAO-ICLARM stock assessment toolsII: Revised version: User’s guide. FAOComputerized Information Series: Fish-eries. Rome, FAO: Food and Agricul-ture Organization of the United Nations.

Getabu A., 1992. Growth parameters andtotal mortality in Oreochromis niloticus(Linnaeus) from Nyanza Gulf, LakeVictoria, Hydrobiologia 232(1): 91–97.

Gómez-Márquez J.L., Peña-Mendoza B.,Salgado-Ugarte I.H. & Arredondo-Fig-ueroa J.L., 2008. Age and growth of thetilapia, Oreochromis niloticus (Perci-formes: Cichlidae) from a tropical shal-low lake in Mexico, Revista de Biol.Trop. 56(2): 875–884.

Hassan A.A. & El-Kasheif M.A., 2013. Agegrowth and mortality of the cichlid fishOreochromis Niloticus (L.) from theRiver Nile at Beni Suef GovernorateEgypt, Egypt. J. Aquat. Biol. Fish. 287(1828): 1–24.

IshikawaT.,ShimoseT.&TachiharaK.,2013.Life history of an invasive andunexploitedpopulation of Nile tilapia (Oreochromisniloticus) and geographical variationacross its native and non-native ranges,Environ. Biol. Fish. 96: 603–616.

JensenK.W., 1957.Determinationof ageandgrowth of Tilapia nilotica L., T. galilaeaArt., T. zillii Ger., and Lates niloticus C. etV. by means of their scales, K. Nor.Vidensk. Selsk. Forh. 30: 150–157.

King R.P., 1997. Growth performance ofNigerian fish stocks, Naga, ICLARM Q.20(2): 31–35.

Kwarfo-Apegyah K., Ofori-Danson P.K. &Nunoo F.K.E., 2009. Exploitation ratesand management implications for thefisheries of Bontanga Reservoir in theNorthern region of Ghana, West Afr. J.Appl. Ecol. 14(1).

Mahmoud H.H. & Mazrouh M.M., 2008.Biology and fisheries management oftilapia species in Rosetta branch of theNile River, Egypt, Egypt. J. Aquat. Res.34(3): 272–285.

Mahmoud H.H., Ezzat A.A., Ali T.E.-S. & ElSamman A., 2013. Fisheries manage-ment of cichlid fishes in Nozha Hydro-drome, Alexandria, Egypt, Egypt. J.Aquat. Res. 39(4): 283–289.

http://www.fao.org/tempref/FI/CDrom/aquaculture/

http://www.fao.org/tempref/FI/CDrom/aquaculture/

http://www.fishbase.org

http://www.fishbase.org


Mehak A., Mu Y., Mohsin M., Noman M. &Nazir K., 2017. Population dynamics ofNile Tilapia (Oreochromis niloticus) atChashma Barrage, Pakistan, Indian J.Geo-Mari. Sci. 46(01): 206–210.

Montcho S.A., Agadjihouèdé H., Montcho-wui E., Lalèyè P.A. & Moreau J., 2015.Population parameters of Oreochromisniloticus (Cichlidae) recently introducedin lake Toho (Benin, West Africa), Int. J.Fish. Aquat. Stud. 2(3).

Moreau J., 1979. Biologie et évolution despeuplements deCichlides (Pisces) intro-duits dans les lacs malgaches d’altitude.Thèse doctorat d’État, Institut NationalPolytechnique de Toulouse, Toulouse,France.

Moreau J., Bambino C. & Pauly D., 1986.Indices of overall growth performance of100 tilapia (Cichlidae) populations[Africa; Asia]. In: The First AsianFisheries Forum, Asian Fisheries Soci-ety, Manila, Philippines, pp. 201–206.

Moreau J., AmarasingheU.&AjithKuramaP.,2008. Population dynamics of non-exploited and under-exploited fish speciesin three reservoirs of Sri Lanka. In: Aquaticecosystems and development: Compara-tive asian perspectives, biology of inlandwaters. Leiden: Backhuys Publishers.

Njiru M., Okeyo-Owuor J.B., Muchiri M.,Cowx I.G. & Van der Knaap M., 2007.Changes in population characteristicsand diet of Nile tilapia Oreochromisniloticus (L.) from Nyanza Gulf of LakeVictoria, Kenya: what are the manage-ment options?, Aquat. Ecosyst. HealthManag. 10(4): 434–442.

Ofori-Danson P.K., 2005. An assessment ofthe purse-seine (winch-net) fishery inLake Volta, Ghana, Lakes Reserv.:Res. Manag. 10(3): 191–197.

Ouattara N.I., Iftime A. & Mester L.E., 2009.Age et croissance de deux espèces de

cichlidae (pisces): Oreochromis niloti-cus (Linnaeus, 1758) et Sarotherodonmelanotheron rüppell, 1852 du lac debarrage d’Ayamé (Côte d’Ivoire, Afriquede l’ouest),Travauxdumuséumnationald’histoirenaturelle «GrigoreAntipa »52:313–324.

Pauly D., 1980. On the interrelationshipsbetween natural mortality, growth pa-rameters, and mean environmentaltemperature in 175 fish stocks, ICESJ. Mar. Sci. 39(2): 175–192.

Pauly D., 1983. Length-converted catchcurves: a powerful tool for fisheriesresearch in the tropics (part 1), Fishbyte1(2) 9–13.

Pauly, D. & Morgan, G.R., 1987. Length-based methods in fisheries research.In : ICLARM Conference Proceedings13, ICLARM, Manila, Philippines, andKuwait Institute for Scientific Research,Safat, Kuwait.

Pauly, D. & Munro J.L., 1984. Once more onthe comparison of growth in fish andinvertebrates, Fishbyte (FAO Philippines)2(1): 21.

Pauly D., Moreau J. & Prein M., A compari-son of overall growth performance oftilapia in open waters and aquaculture.In : The Second International Sympo-sium on Tilapia in Aquaculture, ICLARMConference Proceedings, ICLARMContribution 322, ICLARM, Manilla,Philippines, 1988, pp. 469–479.

Pullin R.S., Palmares M.L., Casal C.V., DeyM.M. & Pauly D., Environmentalimpacts of tilapias. In : Proceedings ofthe Fourth International Symposium onTilapia in Aquaculture, USA, Fitzsim-mons, K.M., 1997, pp. 554–572.

R Development Core Team, 2011. R: Alanguage and environment for statisticalcomputing. Available from http://www.R-project.org.

http://www.R-project.org

http://www.R-project.org


Sangak Y.K., 2010. Distribution and abun-dance of fishes with special reference toTilapia species in Lake Burullos. Ph.D.,Faculty of Science, Zagazig University,Egypt.

Sparre P., 1998. Introduction to tropical fishstock assessment. Part 1. Manual, FAOFish. Tech. Paper 306, 1–407.

Ssentongo G.W. 1971. Yield equations andindices for tropical freshwater fishpopulations. Master Sc., University ofBritish Columbia.

Tah L., Da Costa K.S., Douba V.N., KouassiJ.N. & Moreau J., 2009. Preliminaryassessment of major Tilapiine fishesexploited by artisanal gillnet fishery inLake Ayame I (Bia Basin, Cote d’Ivoire),J. Fish. Int. 4(4): 83–90.

Taylor M.H. & Mildenberger T.K., 2017.Extending electronic length frequencyanalysis in R, Fish. Manag. Ecol. 24(4):330–338.

Tessier A., Guillard J., Chanudet V. & CottetM., 2017. Length-weight relationshipsof eight Asian freshwater fish species inNam Theun 2 Reservoir (Laos PDR),J. Appl. Ichthyol. 33(3): 658–660.

ThompsonW.F. &Bell F.H., 1934. Biologicalstatistics of the pacific halibut fishery 2.

Effect of changes in intensity upon totalyield and yield per unit of gear, Rep. Int.Fish Commission 8: 49.

Weyl O.L.F., 2008. Rapid invasion of asubtropical lake fishery in centralMozambique by Nile tilapia, Oreochro-mis niloticus (Pisces: Cichlidae), Aquat.Conserv.: Mar. Freshw. Ecosyst. 18(6):839–851.

Yongo E. & Outa N., 2016. Growth andpopulation parameters of Nile tilapia,Oreochromis niloticus (L.) in the openwaters of Lake Victoria, Kenya, LakesReserv.: Res. Manag. 21(4): 375–379.

Zambrano L., Martínez-Meyer E., MenezesN. & Peterson A.T., 2006. Invasivepotential of common carp (Cyprinuscarpio) and Nile tilapia (Oreochromisniloticus) in American freshwater sys-tems, Can. J. Fish. Aquat. Sci. 63(9):1903–1910.

Zengeya T.A., Robertson M.P., Booth A.J.& Chimimba C.T., 2013. Ecologicalniche modeling of the invasive potentialof Nile tilapia Oreochromis niloticus inAfrican river systems: concerns andimplications for the conservation ofindigenous congenerics,Biol. Invasions15(7): 1507–1521.

investigating key biological parameters of nile tilapia

Documents