Hydrobiologia 411: 65–70, 1999.

© 1999 Kluwer Academic Publishers. Printed in the Netherlands.

65

Reproduction of the annual fish Cynopoecilus melanotaenia (Regan, 1912)

in a temporary water body in Rio Grande do Sul, Brazil
(Cyprinodontiformes, Rivulidae)

Alexandre Arenzon, Alberto Carvalho Peret & Maria Beatriz Camino Bohrer
Ecology Center, Federal University of Rio Grande do Sul (UFRGS), CP. 15007, Porto Alegre 91501-970, Brazil
E-mail: alex@ecologia.ufrgs.br

Accepted 10 February 1999

Key words: annual fish, reproduction, Cynopoecilus melanotaenia, Brazil

Abstract
The reproductive cycle of the annual fish Cynopoecilus melanotaenia was studied in its natural environment, in
order to obtain relevant information about its biology. This data would be valuable for culturing optimization in
the laboratory. A total of 797 specimens of C. melanotaenia were collected on a monthly basis between April 1994
and March 1995 in a temporary water body, located in Rio Grande do Sul State, Brazil. The frequency distribution
of oocyte diameters indicates a continuous spawning life history style over a long period of time. The estimated
Lpm to females was 18.55 mm, when they are 54 days old.

Introduction such as the reproduction of the species. Information

Annual fish, are defined as a group of Cyprinodon-
tiformes fish that can be found in temporary ponds,
ditches and mudholes occurring in some parts of South
America and Africa that dry out seasonally (Myers,
1942). The complete drying out of that aquatic hab-
itat leads to the death of all adult and juvenile fish.
The population survives as buried eggs for up to 18
months according to Wourms (1972). When the next
rainy season comes, a new reproductive cycle begins,
ponds refill and embryos hatch. In addition, the larval
fish rapidly grow, become sexually mature and spawn
repeatedly over a long period (Myers, 1942, 1952;
Carvalho, 1957; Walford & Liu, 1965; Lacerda, 1969;
Costa, 1990; Arenzon, 1996).
Because of fast corporal growth, precocious sexual
maturity and long reproductive period, these annual
fish seem to present the basic characteristics to the spe-
cies to be used in toxicity tests. Some species can even
become sexually mature after 6–8 weeks of life (Vaz-
Ferreira et al., 1964; Walfourd & Liu, 1965; Weitzman
& Wourms, 1967; Arenzon, 1996).
The culture of test organisms in the laboratory im-
plies previous knowledge of biological characteristics
about spawning type, reproduction period and size
at the onset of maturity is extremely important. The
present study aims to obtain such information on the
annual fish C. melanotaenia.
Material and methods
From April 1994 to March 1995, specimens of C.
melanotaenia were collected in a temporary water
body, located in Tramandaí city, in the north of the
Coastal Plain of Rio Grande do Sul State, Brazil
(29 ◦ 580 4800 and 29 ◦ 580 5400 S; 50 ◦ 140 1200 and
50 ◦ 140 20" W). Specimens were collected monthly
using a hand net of 2 mm mesh. No specimens were
collected during December 1994, when the water body
was dry. The following data were obtained for each
specimen: total length in mm (Lt), sex and gonadal
developmental stage. In females, ovary macroscopic
characteristics, such as size, kind and oocyte size
where used for the identification of developmental
stage. The following stages were establishe: Initial
Development (D I), Intermediate Development (D II),
Advanced Development (D III) and Mature (D IV).
66
The spawning period was defined by monthly
analysis of the frequency distribution of females at
different gonadal developmental stages.
To assess gonadal maturity, oocyte total number
was counted and oocyte diameters measured in the D
III and D IV stages. Gonads were kept in Gilson’s
fluid until the complete detachment of oocytes from
connective tissue. After that, the oocytes were counted
and measured under a microscope using a calibrated
stage micrometer.
Female size at the onset of the reproductive stage
was estimated from the relative frequency distribution
of Mature stage (D IV) in each length class (2 mm in-
tervals). This size corresponds to the class where 50%
of specimens are in the Mature stage (Santos, 1978).
The line fitted to the data corresponds to: Frequency =
b
l - (eaLt ).
Results
A total of 797 specimens was collected (433 female,
356 male and 8 juveniles). The smallest individuals
whose sex could be determined, measured 10.5 mm
and 8.2 mm, male and female, respectively. Specimens
that could not be sexed were considered as juveniles.
C. melanotaenia has paired ovaries, of which an-
terior end is free, while the posterior one continues
into a little oviduct. Each ovary is covered with a thin
peritoneal covering. Ovaries are usually of the same
size in the initial gonadal development stages. How-
ever, in more advanced stages, the left and the right
side can have different sizes.
Considering the gonadal macroscopic observations
of females, the gonadal development stages were
described as following:
D I – Initial Development
The ovaries are small, tubiform and translucent, meas-
uring about 0.75 mm in length. The oocytes are spher-
ical, of relatively uniformed size (0.05–0.125 mm).
The number of oocytes is small. They do not occupy
all the interstitial gonadal space.
D II – Intermediate Development
The ovaries are larger and less transparent due to a
higher oocyte number. Oocytes occupy all intersti-
tial gonadal space. It is possible to observe a marked
difference in oocytes diameter (0.05–0.25 mm). The
ovary is still tubiform, measuring about 1.025 mm in
length.
D III – Advanced Development
At this stage a further increase in gonadal size ocurs.
It becomes more robust and compact, due to a higher
number of oocytes of intermediate size. The ovary
measures about 1.75 mm, and the oocytes have a
stronger difference in size (0.05–0.325 mm).
D IV – Mature
The ovary is turgid and with irregular surface because
of the presence of large mature oocytes. There is a
strong variability in oocyte number (from 49 to 219)
and size (0.05–1.375 mm). The ovary length is also
variable, depending on the number of oocytes. Once
in this stage, fish are considered adults. Two kinds
of oocytes are considered mature. They differ in the
presence or absence of a secondary chorionic layer,
called secondary envelope. The majority of the oo-
cytes do not show this secondary envelope yet. The
oocytes with this elaborated surface ornamentation are
present in an assorted number (1–15). It can have a
maximum diameter of 1.375 mm, minimum diameter
of 1.025 mm and an average diameter of 1.15 mm
(CV= 8.46%).
No other developmental stage was observed, as
well as any return to the last developed stages.
C. melanotaenia did not have a specific reproduc-
tion period. The mature stage was absent only during
December 1994 (dry period) and January 1995 (Fig-
ure 1). In January 1995, only specimens with total
length smaller than the size at the beginning of the
reproductive stage were collected. This information is
confirmed by observations in the laboratory, where we
kept the same males and females reproducing for more
than a year.
The monthly frequency distribution (%) of the dif-
ferent gonadal developmental stages related to female
length indicated that gonadal development follows the
specimens’ development (Figure 2). Thus, the Initial
Development stage (D I) occurs between a length of
8.20 and 14.20 mm, the Intermediate Development
stage (D II) between 10.20 and 20.20 mm, the Ad-
vanced Development stage (D III) between 14.20 and
24.20 and the Mature stage (D IV) between 16.20 and
34.20 mm.
The estimated size of the females at the beginning
of the reproductive stage was as 18.55 mm (Figure 3).
 67

Figure 1. Monthly gonadal developmental stages frequency distribution (%) of females of Cynopoecilus melanotaenia, collected from April
1994 to March 1995 in a temporary water body, located in the north of the Coastal Plain of Rio Grande do Sul State, Brazil.
68

Figure 2. Relative frequency distribution of the gonadal developmental stages in each total length class for females of Cynopoecilus melanot-
aenia collected from April 1994 to March 1995 in a temporary water body, located in the north of the Coastal Plain of Rio Grande do Sul State,
Brazil.

Figure 3. Relative frequency distribution of the Mature stage (D IV) in each total length class for females of Cynopoecilus melanotaenia
collected from April 1994 to March 1995 in temporary water body, located in the north of the Coastal Plain of Rio Grande do Sul State, Brazil.
Discussion
Our study showed a non-cyclic gonadal development
in C. melanotaenia, with very particular develop-
mental characteristics. Thus, it was necessary to create
a specific gonadal maturity scale, adapted to this
species, as suggested by Naumov (1959).
Two kinds of oocytes can be considered mature.
Both types present similar dimensions and differ only
in the presence or absence of a secondary chorionic
envelope. This secondary envelope is deposited on the
upper surface of the primary envelope and concludes
the oocyte development (Wourms, 1976). Wourms &
Sheldon (1976), using electron microscopy, sugges-
ted three theories about the possible function of this
secondary envelope. The eggs of annual fishes are
exposed to partial desiccation during the dry season
and the secondary egg envelopes function as chorionic
plastrons. Alternatively the spikes on the egg envel-
ope may serve to anchor it in place either when it is
deposited or when the habitat is refilled with water.
According to the third theory, a function of ‘shock
absorber’ and spatial barrier can be suggested.
A new theory is here proposed taking into con-
sideration our laboratory observation. When eggs are
in contact with the sediment of the water body, and
sediment adheres to the surface, among the spike-like
projections of the secondary envelope, this conveys
69
protected against desiccation. But, since the proposed
functions are not mutually exclusive, the secondary
envelope may serve all four purposes.
C. melanotaenia has internal fertilization (Meyer
& Lydeard, 1993). However, in oocytes that have sec-
ondary envelopes, it is not possible to know if it still is
an oocyte or a fertilized egg. The only exception to this
situation is when, in mature stage ovaries, we found
eggs at the beginning of embryonic development,
suggesting zygoparity as a mode of reproduction.
Length at the beginning of gonadal maturation
(Lpm) can indicate the gonadal development velocity
of the species.This process is extremely fast in C.
melanotaenia. Females can reach Lpm at 1.78 months
of age. Based on the observed data from the second
collection after the dry period (February 1995) 72%
of females were at the Mature stage. These females
showed a maximum age of 54 days or 1.8 months.
The average water temperature during this period was
25 ◦ C. Tanizaki (1988) reports that after the appear-
ance of a new generation of Cynolebias, it attains
sexual maturity after 60 days, and starts to spawn.
Vaz-Ferreira et al. (1964) reports that C. bellottii also
attains sexual maturity after 60 days. Wootton (1990)
presents the Cyprinodontiformes, among fish, as the
lower extreme to the necessary period of sexual mat-
uration, since they can become mature after only a few
weeks. This reproductive characteristic is as a strategy
70
to avoid local extinction, since the period between
hatching and death of the fish is very short under
natural conditions.
Thus, C. melanotaenia, and probably other species
of annual fish, show also other reproductive strategies
to ensure local survival. According to the Marza clas-
sification (Marza, 1938), and based on the data ob-
tained in this study, C. melanotaenia can be considered
to have an asynchronous ovary, since there is a ran-
dom mixture of oocytes at every conceivable stage.
This characteristic becomes clearer in more developed
stages.
In teleosts with asynchronous ovaries, recruitment
at all stages is continuous throughout the breeding sea-
son (Wallace & Selman, 1981). C. melanotaenia does
not seem to have a specific breeding period, and can
spawn over a prolonged period. The high frequency
of mature females at almost all collection dates con-
firms this. Gonadal development was observed to be
only interrupted by death of the fish on the dry period.
Based upon previous unpublished laboratory obser-
vations, we have found that the breeding period of
C. melanotaenia begins a few weeks after hatching
and terminates only when the fish become senile or
dies. This life history style can also be found in cyp-
rinodonts. Females of Oryzias latipes, a fresh-water
cyprinodont, exhibit a maturation cycle of 24 h and lay
a batch of eggs at dawn every day of the breeding sea-
son (Yamamoto, 1956). The asynchronous oocyte de-
velopment of C. melanotaenia guarantees continuous
egg production and spawning, across a long breeding
period.
Conclusion
The continuous egg production, the prolonged breed-
ing period and complex diapause process of C. melan-
otaenia, result in a ‘stock’ of eggs in the substratum.
These eggs, in different phases of embryonic devel-
opment, give rise to a stream of hatching embryos
throughout the inundation period. This adaptation pro-
tect the species of a single massive hatching during a
very short rainy period, which would not allow the fish
to mature and reproduce.
The strategies presented by C. melanotaenia fa-
vour its use as a laboratory test organism, since ones
can obtain new spawning stock in a very short period
of time. In addition, these strategies make it possible
to hatch the ‘stocked’ eggs to get newborn larval fish
whenever it is necessary.
References
Arenzon, A., 1996. Biologia e ecologia do peixe anual Cynopoecilus
melanotaenia (Regan, 1912) visando seu uso como organismo-
teste em testes de toxicidade (Cyprinodontiformes, Rivulidae).
M.Sc. thesis, Univ. Federal do Rio Grande do Sul, Brazil: 118
pp.
Carvalho, A. L., 1957. Notas para o conhecimento da biologia dos
peixes anuais. Rev. (brasil.) Biol. 17: 459–466.
Costa, W. J. M., 1990. Análise filogenética da família Rivulidae
(Cyprinodontiformes, Aplocheiloidei). Rev. (brasil.) Biol. 50:
65–82.
Lacerda, T. P., 1969. Estudos sobre os peixes anuais da região de
São Leopoldo. Dissertation, Univ. do Vale do Rio dos Sinos, São
Leopoldo, Brazil: 65 pp.
Marza, V. D., 1938. Histophysiologie de l’ovogenènese. Hermann,
Paris: 81 pp.
Meyer, A. & C. Lydeard, 1993. The evolution of copulatory organs,
internal fertilization, placentae and viviparity in killifishes (Cyp-
rinodontiformes) inferred from a DNA phylogeny og the tyrosine
kinase gene X-src. Proc. R. Soc. Lond. B. 254: 153–162.
Myers, G. S., 1942. Studies on South American fresh-water fishes.
Stanford Ichthyol. Bull. 2: 84–114.
Myers, G. S., 1952. Annual fishes. Aquar. J. 23: 125–141.
Naumov, V. M., 1959. The ovogenesis and ecology of the sexual
cycle of the murmansk herring Clupea harengus L. Spec. Sci.
Rep. Fishery U.S. Fish and Wild. Serv. 327: 203–62.
Santos, E. P. dos, 1978. Dinâmica de populações aplicada à pesca e
piscicultura. Hucitec-Edusp, São Paulo: 129 pp.
Tanizaki, K., 1988. Ecologia de peixes anuais raros do Estado do
Rio de Janeiro – gênero Cynolebias Steidachner, 1876 (Cyp-
rinodontiformes, Rivulidae). 1988. In Congresso Brasileiro de
Zoologia. Brazil: 15: 318 pp.
Vaz-Ferreira, R. Sierra de Soriano & B. Scaglia de Paulete, 1964.
Eco-etología de la reproducción en los peces del género Cy-
nolebias Steindachner, 1876. Arch. Soc. Biol. montevideo 26:
44–49.
Walford, R. L.& R. K. LIU, 1965. Husbandry, life-span and growth
rate of the annual fish, Cynolebias adloffi. Exp. Geront.1: 161–
171.
Wallace, R. B. & K. Selman, 1981. Celular and Dynamic Aspects
of Oocyte Growth in Teleosts. Am. Zool. 21: 325–343.
Weitzman, S. H. & J. P. Wourms, 1967. South American cyp-
rinodont fishes allied to Cynolebias with the description of a new
specie of Austrofundulus from Venezuela. Copeia 1: 89–100.
Wootton, R. J., 1990. Ecology of teleost fish. Chapman & hall,
Londres, 1990: 404 pp.
Wourms, J. P., 1976. Annual fish oogenesis. I. Diferentiation of the
mature oocyte and formation of the primary envelope. Dev. Biol.
50: 338–354.
Wourms, J. P. & H. Sheldon, 1976. Annual fish oogenesis. II.
Formation of the secondary envelope. Dev. Biol. 50: 355–366.
Yamamoto, T., 1956. The physiology of fertilization in the medaka
(Oryzias latipes). Exp. Cell Res. 10: 387–393.