Age and growth of Oreochromis niloticus ( Perciformes : Cichlidae ) in Mexico

Age and growth of Oreochromis niloticus from Lagoon of Coatetelco, Morelos State, Mexico were studied from January through December, 1993. Scales of 318 specimens were collected. Modal length at capture was 10. 5 -11. 5 cm standard length. Scales rings were forrned during December. Back-calculated lengths-at-age showed no significant differences by sexo Four check marks were recorded. According to the growth curve parameters for population, the fish grow at a low rate (k=O.07) until they achieve a size (Loo) of 2 9. 19 cm. Length-frequency analysis (Bhattacharya's Gaussian component deterrnination procedure) do not differ significantly (t-student, p

The family Cichlidae, with about 700 species, is the predominant perch-like fish occuring naturalIy in Africa and Madagascar, and also in the Central and South America Tropical Zone (Fryer &. Hes 1972;Morales 1991).They have been introduced into a large number of tropical and sub-tropical countries around the world in the last four or five decades, either accidentally or deliberately.
The tilapias are important food fish in aquaculture and they are normally reared in open ponds in the tropics (Soderberg 1990).The Nile Tilapia Oreochromis niloticus (Linnaeus, 1757), is considered a suitable species for rural fish farming, taking into account its fast growth rate, adaptability to a wide range of culture conditions andits high capacity to transform farm waste into protein.
The tilapias has proven to bring both social and econoIÍlic benefits to the rural people living close to aquatic systems under, which show exploitation and management fisheries (Arredondo 1983).
The annual production of tilapias in Mexico was 92 891 t in 1994 (Anonymous 1995).This figure represent about 7.9 % of tota! production from fish caught in lakes and from determine the rate at which a rational exploitation should be carry out (Oduleye 1982).
The present study deals with the deterrnjnation of age and growth rate in O. niloticus from Coatetelco Lagoon, Marelos State, Mexico, by means of scales reading and length frequency distribution analysis.

MATERIALS AND METHODS
Samples from the commercial catch at Coatetelco Lagoon, located in Morelos State (18°45' N, 99°20' W, altitudc 1 100 m), were taken approximately at monthly periods from January lo December 1993.Fish were caught by static gill-net with mesh size 5.08 cm.The specimens were collected by the same fisherman in the same place and day.A total of 318 fish were analyzed: 157 females and 161 males.Fish were measured for total length (TL) and standard length (SL) to the nearest mm and weighed (total weight) to the nearest g.The sex was determined by examination of internal organs.Ten 10 fifteen scales were removed from the left side of the body in the region between the middle of the pectoral fin and the lateral line (Tesch 1968, Ruiz-Dura et al. 1970).111 ey were washed, mounted on glass sUdes and examined by a sEde projector.The rings that continued around the entire circumference of the scale were counted and the total number of these rings was used to estimate the age of each specimen.Counts for each specimen were perfoffiled twice, approximately at periods two months apart, wíthout knowledge of the length of the specimen.The scale radius (R) was defined as the maximum distance from the focus to the distal edge of the scale.The distance from the focus to each annulus was measured along this axis.The relationship between SL and R was linear but no! directly proportionaJ, havíng a positive intercept on the ordinate.SL-R regressions were tested for differences in slopes between sexes using analysis of covariance (ANCOVA).Back calculated length-at-age of individual fish were estimated ti-om the standard length-scale radius regressions using the Fraser-Lee method (Tesch 1968).Mean weighted back-calculated lengths at-age were tested for differences between sexes using Student's t-test.AIso the length frequency distribution analysis was employed through the modificatíon of Bhattacharya's procedure (Pauly and Caddy 1985).
The validation of ageing and the periodicity of the annulus formatíon were verified through marginal growth examination and focus-to-annulus distance (mean and range).Marginal growth was examined in all scale aged during the whole sampling period.The percentages of fish having edge growth were compared by month of capture (Yamaguchí et al. 1990).
The mean values resulted from length frequency distribution analyses and those obtained by hard-part readings, they were used lo estimate the growth parameters of the von Bertalanffy growth function (vBGF) which was fiHed to mean weighted back-calculated lengths-at-age using a non-linear regression program available in StataCorp (1997), to obtain the nominal vaJue for Loo and k.However, the vBGF was used according to Barlow (1992).Boxplot scatter plots (Salgado Ugarte 1992), residual analyses (Curts 1984) and Student t-test were used.

RESULTS
TI1e total sample size was 318 fishes (Table 1).Mean standard body length of males was 8.9 to 14.8 cm (females 9.0 lo 15.8 cm) and theÍr weight 25.7 lo 106.5 g (females 25.8 to 100.6 g).The standard length-frequency distribution of tilapias in the lagoon is shown in figure 1. Modal length at capture was 10.5-11.5 cm standard length and it observed a bias  to values more th� 12.0 cm.The total length was highly co r related with standard length: Si.. = -0.2�9+ 0.794 TL, r 2 = 94.6%,n = 318.

Age determination by length�frequency distribution analyses:
The length-frequency distribution analyses for pooled data were smoothed by running average over five (Pauly 1987) to clarify the distribution and to obtain more efficiently the component means.Component characterization obtained through the Bhattacharya's procedure were carr ied out using sexes combined data, because the mean values for males and females did not differ significantly (t-student, p<0.05).Seven size groups were estimated.
Age determination by scales: Scales were collected from 318 fish.Seventy-three percent (110 for females, 122 for males) of the scales were readable and used for study of age and growth.Check marks were distintic in all scales fields.Data indicated that two annulus were formed per year.Observation of the position of the last complete ring scales of 232 specimens, showed that most of the fish caught during December-January (71.8%) and June (63.3%),had a complete ring close to the edge.However, the marginal growth index (Fig. 2) suggested , that annuli are formed between December and January.The mean focus-to-annulus distance estimated for each individual, was consistent for each annulus (Table 2).Modal age at capture was represented mainly by age two (57.32%), and a lowest percentage (1.72%)by age zero.
The regressions of SL (cm) on scales radius (R) were estimated separately for separated and combined sexes.The slopes of the SL-R regressions did not differ significantIy between sexes (ANCOVA, p<0.05).Therefore, the body-&cale relationship used for back-calculate length-at-age for combined data was: SL :::: 6.0931 + 1.2617 R, n=232 fish, r2:65.2% Interceptsfrom this regressions were used to back-calculate length-at-age which are summarized in Table 3.The greatest growth in length occurred in the frrst year of life, and growth increment declined steadily after the frrst year reaching a low figure around 9.81 mm SL.
Length at age estimated hard part readings by means of boxplot (Fig. 3) allowed to establish at the median as the central tendency value for being resistent.In this case, both the mean and median were similar value and mean value was used.The(efore, the mean values age from I to IV of the hard part readings were similar with the mean values age from I to IV of the length frequency analysis (t-student, p<0.05) through Bhattacharya's procedure.The von Bertalanffy growth function parameters obtained both through Bhattacharya's procedure and hard part readings were analyzed, and the results did not show statistical differences (t-student, p <0.05), and its express ion according to Barlow (1992) is (Fig.

DISCUSSION
The length-frequency distribution of O. níloticus shows a young looking population, because they were caught with static gíll-net size 5.08 cm.We obtained small specimens (from 8.8 to 16.0 cm standard length), in comparison with Oreochromis aureus reported at others reservoirs in Mexico, caught with gíll-net size from 7.62 to 10.16  cm, which ranging from 15.0 to 35.0 cm total length (Bernal 1984, Morales 1992, Guzmán 1994), because fishermen did not allow to use other gill-net size.We are convinced that specimens with a larger average total body length could be obtained in the Lagoon of Coatetelco, because this aquatic system has a high nutritional potential and it is considered as eutrophic (Granados 1990).On the other hand, if fishermen do not change their attitude respect the use of other fishing gear, regardless the lagoon is characterized as an eutrophic system, the fishing effort and the capture could decreasing in time.Therefore, it is necessary to realize research with different fishing gear to determine which one requires less energy per unit of effort, selectivity of catch as to the size and age, and increased survival of the catch.
The formatiQn of.rings in the scales of tilapia were observed in December-January (winter), which corresponds with the rninimum value of marginal growth.This event may be attributed to lower temperatures (21 °C) prevailing in the lagoon during this season.However, the percentage of fish caught during December-January and June had a complete ring close to the edge, and therefore the growth rings found (mainly June) on the•scales were probably due to growth interruption by the gonad maturation and the spawning activity (although spawn all the year around), because O. niloticus is a mouth brooder and it is possible that in the females, this breeding habit would prevent feeding and thus lead to growth ring formation.Yamaguchi et al. (1990) found that the formation 'of rings occurred every January when they used the rninimum value of marginal growth.
Respect to the formation of marks in the scales, Fagade (1974) in Tilapia malanotheron of Lagos Lagoon found that the complete growth rings on the opercular bone are formed probably by the spawning activity during the height of the rainy season and incómplete growth rings formed during the dry season.Garrod (1959) has shown that in the mouth breeding T. esculenta caught in L. Victoria, the growth rings are formed in the scales as a result of spawning, and the same pattern was reported by Alejo et al. (1989) in O. mossambicus, and Guzmán (1994) in O. aureus from Mexico.lt is probably that the frequency of spawning may be influenced by the abundance and season availability of food and by other environmental factors in different localities (Babiker andIbrahim 1979, Oduleye 1982).
Our data show that two growth band are formed, and this together with the high correlation found between standard length and scales radius, demonstrate the validity of using scales for estimating fue • age and past growth history of tiHl.pía, as presented by Yamaguchi et al. (1990).
Age determination by the two methods used in this work, allowed• to compare the average values attained to estimate the growth rate of tilapia, because the scale size could be different in fishes of the same length and also within the specific place where they are taken
-Figueroa and Tejeda-Salinas 1989), including releases into Coatetelco Lagoon in 1990.Studies in this aquatic system are very important, because they are the basis to Figueroa and Guzmán-Arroyo (1986),

Fig
Fig. l.Length-frequency distribution of O. niloticus at Coatetelco lagoon, Morelos State, Mexico.

Fig. 4 .
Fig. 4. Relationship between age and standard body length by O. niloticus.

(
Carlander 1987).Therefore, the length frequency distribution for trus species during the study was based on mainly by the number of fishennen that realized this activity (40 individuals), and the gill-net mesh-size that fishennen society used.The mean values from 1 to IV age group by the two methods used were similar, and most of the total sample (133 specimens) was represented by fish with two growth checks.In another fraction the fish of age O (4 specimens) and IV (16 specimens) were less represented, due to gill-net mesh size, in the range of 8.0-16.0cm.Making thus, the market price in Coatetelco is low.With reference to growth parameters, Morales (1991) for O. aureus and Yamaguchi et al. (1990) for O. niloticus reported similar values of body length (25.0 cm) at two years oId.Moreau (1985) who worked with T. rendalli found 28.25 cm body length with 0.06 growth coefficient.It is probably that the similarity with our results could be due to the temperature of water (24-30 oC), even though the eutropruc grade and catch of fish are very different.Studies in several other fish species of Oreochromis (Guerra and Peña 1985, Hernández 1987, Jaramillo and Sánchez 1991) from temporary ponds in Morelos State during one year, had been carried out and similar results in standard body length were obtained.This rapid linear growth in the first year of life may represent an evolved adaptation to reduce its vulnerability to the environment.Flores (1994) found that O. niloticus in ponds at Mexico City, attains 20.7 cm asymptotic length standard (Loo) with 0.2038 growth coefficient.Although this asymptotic value is smaller than the one obtained in this study, the mean values used by him were similar to those used to estimate the growth parameters of the von Bertalanffy growth function (VBGF).Compared to the standard body length obtained in this study,Fryer and Hes (1972)   mentíoned that T. nilotica in Lake Albert reaches a length of 50 cm, and does not breed until it has attained a length of about 28 cm.However, trus same species in the lagoon called Buhuka (600 m and 2 m deep), adjacent lake Albert, did not exceed a length of 17 cm, and were breeding at a length of 10 cm.This population was considered with "dwarf' or pygmy Ti/apia.AIso, cited that under natural conditions the growth achieved in the first year of life by several species of Tilapia is from about 9 to 12 cm.This represents adaptations to adverse conditions for growth as a result of selection for low growth rates and relatively earIy maturátion.This last observation becomes important because the Coatetelco lagoon presented a collapse in 1986 by a crack at ground, and. the water was absorbed.In 1990, after solving the problem, the system was flooded and O. niloticus were íntroduced.Therefore, factors such as temperature, pH, total hardness, depth leve!, nutrients and productivity could be very different and the growth rate of the fish would be affected.It is important to mention that the informatíon presented in ibis work requíre additional analyses, besides the use of other bone structures (opercular bones, otolith, vertebrae) to compare results and to obtain better estimates of growth parameters.

TABLE 2
Range, mean distance and standard deviation of seales radius at eaeh annulus in O. niloticus

TABLE 3
Back-calculated standard length (cm) at age for O. niloticus Back calculated 1ength at age estimated by scale readings.