Reproductive traits of Gymnogeophagus labiatus ( Teleostei , Cichlidae ) , in an upper stretch of Sinos river , Caraá , Brazil

1. Programa de Pós-Graduação em Ecologia, Instituto Nacional de Pesquisas da Amazônia, Av. André Araújo, 2936, 69011-970, Manaus, AM, Brazil. (juliatovarv@gmail.com) 2. Programa de Pós-Graduação em Biologia Animal, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, Prédio 43435, 91501-970, Porto Alegre, RS, Brazil. (biovrl@hotmail.com) 3. Setor de Ictiologia, Museu de Ciências Naturais, Fundação Zoobotânica do Rio Grande do Sul, Rua Dr. Salvador França, 1427, 90690-000, Porto Alegre, RS, Brazil. (marco-azevedo@fzb.rs.gov.br)

sizes 15, 30, 40 mm) set for 24 h.Fish were fixed in 10% formalin and voucher specimens catalogued in the Coleção de Peixes do Museu de Ciências Naturais da Fundação Zoobotânica do RS (MCN 18800, 18801).Temperature was measured in the field; photoperiod and rainfall data were provided by the 8 th District of Meteorology, in Porto Alegre, RS.Standard length (SL, mm), total weight (Wt, g), and gonad, stomach, and liver weights (Wg, Ws, Wl, g) were recorded for each specimen to the nearest 0.01 mm and 0.0001 g.We used these data to calculate the gonadosomatic (GSI), stomach repletion (SRI), and hepatosomatic (HSI) indices according to these equations: GSI = (Wg.100)/Wt(Vazzoler, 1996); SRI = (Ws.100)/Wt(santos, 1978); and HSI = (Wl.100)/Wt (santos, 1978).Gonadal maturation stages were identified through macroscopic analysis following Vazzoler (1996).Reproductive period was estimated by monthly variation in GSI and the frequency of ripe individuals.Spearman rank correlation was used to test for significance between biotic indices (GSI, SRI, and HSI) and abiotic data (temperature, photoperiod, and rainfall).Size at maturation was made using the length of the smallest ripe individual, male and female, as an estimate of the minimum size in which the species might be ready to reproduction (sato & Godinho, 1988).Sex ratio was estimated for the total sample as well as for monthly samples and standard-length classes (15-mm intervals).Fecundity and oocyte development were estimated by counting and measuring oocytes from 10 ripe females under a stereomicroscope with a ruled scaled ocular.Relative fecundity is given through the number of oocytes per milligram of female's total body weight, as suggested by adebisi (1987).

RESULTS
Along the study period, water temperature in Sinos river ranged from 12.8ºC, in July, to 26.8ºC, in January.These months corresponds to winter and summer in southern hemisphere.The photoperiod ranged from 615 min (June) to 844 min (December).As expected for temperate South America, the monthly rainfall data showed high variation along the year without well defined dry or wet seasons.The highest rainfall value was in July (233.5 mm) and the lowest (74.1 mm) in April (Tab.I).
Results from the analysis of 174 males (ranging from 24.58 to 156.38 mm SL) and 132 females (ranging from 17.9 to 83.59 mm SL) of G. labiatus showed a peak in mean GSI values in September along with a peak in the values of temperature as well as with the increase in photoperiod (Figs 1, 2; Tab.I).Ripe individuals were found from August to November, December, March and April (males) and from September to February (females) (Figs 3, 4), which defines a seasonal reproduction for this species, although high values of GSI can be found all along the year to males.
The Spearman rank correlation showed a significantly negative correlation between GSI mean values of males and photoperiod (r = -0.6105;P = 0.0350) and a very significant negative correlation between GSI mean values of males and temperature (r = -0.7206;P = 0.0082).For females, a significantly positive correlation was found only between GSI mean values and photoperiod (r = 0.6364; P = 0.0402).The monthly variation in the mean values of SRI, HSI and GSI for males (Fig. 5) showed similar behavior of the curves of SRI and HSI between some months.A significantly positive correlation between mean values of GSI and HSI was observed for males (r = 0.6316; P = 0.0276).The curves of SRI, HSI and GSI of females also showed a similar behavior but no significant correlations were detected (P > 0.05) (Fig. 6).
The standard length of the smallest ripe male was 104.74 mm, and the smallest ripe female was 55.00 mm.
The overall sex ratio of this G. labiatus population was highly skewed with 1.32 males: 1 female.Monthly sex ratios were near parity throughout the year, except for March and December where males were more abundant.There was a large difference in the size distribution by gender.Females were more abundant at the 47├62 mm and 62├77mm SL intervals, with more than twice the number of males in the later.Males were more abundant in the 17├32 mm, 32├47 mm e 77├92 mm SL intervals.Males reached the larger sizes so that the largest standard length intervals (above 92 mm) were composed only by males (Tab.II).
Fecundity showed a linear relationship (R 2 = 0.4192) with total weight.Absolute fecundity ranged from 64 to 174 oocytes, with a mean of 113.4 (± 31.24sd) and relative fecundity ranged from 0.0073 to 0.0176 oocytes per mg, with a mean of 0.0125 (± 0.0026 sd).The analysis of oocyte frequency by oocyte-diameter class showed the presence of small-diameter oocytes (previtellogenic) between 0.03 and 0.36 mm at high frequencies.Larger oocytes were also found but at lower frequencies (Fig. 7).
As long as ripe females are found from September to February while males can be found with high values of GSI along the year, we suppose that females determine the reproductive period of G. labiatus.We conclude that this species have a long and seasonal reproductive period along six months, comprising the spring and summer in the southern hemisphere with a positive correlation with photoperiod.santos & Fontoura (2000) report similar results to a population of Geophagus brasiliensis (Quoy & Gaymard, 1824), which reproduces between September and April.Although the peaks of reproduction of males and females are coincident in this population, occurring in November and February, GSI curve of males showed small peaks indicating gonadal development in other months such as observed for G. labiatus in the present study.Females require a higher energetic investment to the development and maturation of their gametes rather than males (ForsGren et al., 2002), which might explain the occurrence of mature males along the year while females concentrate their efforts to reproduction in a period of the year where environmental conditions are more favorable to the survival of offspring.This is corroborated by the correlations found between GSI and environmental factors.As long as males do not show a marked seasonal behavior in the maturation of their gonads, a negative correlation appeared.Females, on the other hand, showed a marked seasonality in gonadal maturation with high and low values of GSI in spring and summer and autumn and winter, respectively with a positive correlation to photoperiod which is low in winter and rises in spring and summer.
Sex-ratio analysis showed that males predominated over females in the population as a whole.They also grew larger so that in the largest standard-length intervals, there were only males in our samples, showing a marked sexual dimorphism by size.Males often reach larger sizes than females in other cichlid species which agrees with our results, but in those species the number of females is usually higher than the number of males what differs from our findings (hartz et al., 1999 Reproductive traits of Gymnogeophagus labiatus... Iheringia, Série Zoologia, Porto Alegre, 101(3): 200-206, 30 de setembro de 2011 Kullander, 2003).According to Fryer & iles (1972), in cichlid populations from African lakes, males are larger than females because their growth rate is higher.Although, according to hartz et al. (1999), males of cichlids are larger and colorful due to their mating behavior, which may include territorial defense, competition for females, and also, the lower energetic cost for the development of sperm compared with oocytes.
Fishes that provide parental care usually produce a low number of oocytes, presumably because protection of eggs and larvae demand higher energetic cost.On the other hand, it guarantees a higher survival of progeny.Many Characiformes that show parental care have low fecundities, often between 20 to 400 oocytes (Vazzoler & menezes, 1992).Data on fecundity of neotropical cichlid species are scarce and usually do not take into account relative fecundity making standardized comparisons quite difficult.This feature, along with the fish length, although a raw approximation, might allow a comparison between species rather than the use of absolute fecundity.Studies show that some cichlid species may have low absolute fecundity, such as Apistogramma hoignei Meinken, 1965 (51 oocytes), Crenicichla geayi Pellegrin, 1903 (230) (Winemiller, 1989), and Gymnogeophagus lacustris Reis & Malabarba, 1988(131 to 253) (hartz et al., 1999).Similarly, we found that G. labiatus has a low mean of 113.4 oocytes.On the other hand, other cichlids have higher values of fecundity, such as Aequidens pulcher (Gill, 1858) (861), Astronotus ocellatus (Agassiz, 1831) (2301), Caquetaia kraussii (Steindachner, 1879) (3702), Cichlasoma orinocense Kullander, 1983 (1287) (Winemiller, 1989), Geophagus brasiliensis (341 to 3191) (santos & Fontoura, 2000;mazzoni & iGlesiasrios, 2002), Cichla monoculus (Spix & Agassiz, 1831) (3100 to 17987) (chellaPPa et al., 2003;Gomiero & braGa, 2004), C. kelberi (Kullander & Ferreira, 2006) (6072) (souza et al., 2008) and C. ocellaris (Bloch & Schneider,1801) (13769) (Gomiero & braGa, 2004).Gymnogeophagus labiatus is more closely related and have also similar size with species like G. brasiliensis and G. lacustris; the other species mentioned are smaller, or bigger, or not closely related to G. labiatus.Although all those species display brood protection, the type of parental care may affect the amount of oocytes produced.For example, mouthbrooding species, like G. labiatus, have their fecundity limited by buccal cavity volume (hartz et al., 1999) while the substrate spawner G. brasiliensis do not show the same limitation (GoodWin et al., 1998).GoodWin et al. (1998) suggests that the evolution of traits such as the loss of adhesive threads on eggs (Peters & berns, 1982), reduced fecundity, and increased egg size with higher juvenile survival (noaKes & balon, 1982) have accompanied the evolution of mouthbrooding.According to laGler et al. (1977), for a given species, fecundity is inversely proportional to the level of parental care.The species size, the oocyte diameter of each species, and its reproductive strategy may all play a role in specific fecundity.
The pattern of oocyte development in G. labiatus was similar to that found in species whose oocyte development is asynchronous, as proposed by Vazzoler (1996) suggesting a multiple spawning to this species.Winemiller (1989) has analyzed aspects of reproduction in six cichlid species and all of them showed a long reproductive period (six to twelve months) with two or more spawning events per year.Those species were characterized as acyclic spawners.In the same way, Satanoperca pappaterra (Heckel, 1840) (Vazzoler, 1996), Gymnogeophagus lacustris (hartz et al., 1999), species of Cichla (chellaPPa et al., 2003;Gomiero & braGa, 2004;souza et al., 2008) and Geophagus brasiliensis (barbieri et al., 1981), were characterized as multiple spawners, although the later was defined as total spawner by santos & Fontoura (2000).
According to Winemiller (1989), fishes that exhibit a set of life-history traits such as an intermediate or long generation time, high investment in individual offspring, late maturation, non-seasonal reproduction, and long reproductive period may be characterized as equilibrium strategists, which is similar to the K-strategy as proposed by PianKa (1970).As a group, cichlids tend to exhibit traits related to equilibrium strategy (Winemiller, 1989).Our data for G. labiatus, such as low fecundity, asynchronous oocyte development, multiple spawning, and mouthbrooding support that life-history model for cichlids.