Trophic and reproductive ecology of a species of Hemibrycon (Pisces: Characidae) in Tinajas creek, Quindío River drainage, upper Cauca basin, Colombia

Mail: croman@uniquindio.edu.co; albotero33@yahoo.com  Abstract: We studied trophic and reproductive ecology of a new endemic species of characid fish Hemibrycon sp., from Tinajas Creek, Quindío River drainage, upper Cauca basin, Colombia. The diet consisted primarily of Diptera (Tipulidae and Chironomidae), Hymenoptera (Formicidae) and Coleoptera (Dytiscidae), but it also eats algae (Clorophyta) and seeds. Alochtonous food items are important factor for this species. The species reproduces in both the wet and dry seasons. During September-October they store lipids in their celomic cavity. Fecundity is low 05 oocytes) and the sex ratio is 1:2.28, with a predominance of males. Hemibrycon sp. is syntopic with: (4  Cetopsorhandia boquillae, Astyanax aurocaudatus, Trichomycterus caliense, and Poecilia caucae. Physical and chemical of data of their habitat are included.

Collecting took place during both the dry (February-March, July-August) and wet seasons pril-June, September-December). (A Fish were captured using a drag net, a seine and preserved in ice, in situ; then were carried to the Laboratory of Ichthyology, Department de Biology, Universidad del Quin-dio, Armenia, Colombia (IUQ), where they are now deposited. 120 specimens were examined, 51.5 mm of mean standard length (25.5-75.5 mm SL). Fish were measured, weighed and cut along their major axis to remode their digestive tracts and gonads. 116 stomach contents were analyzed, using occurrence index, numeric index (Hynes, 1950;Hyslop, 1980), volumetric analysis (Pedley & Jones, 1978;Capitoli, 1992), and an importance index (Oda & Parrish, 1981), were I= ccurrence % * Volume %)/100. (O Then Principal Components statistical analysis was applied, considering volume, quantity and frequency as variables and different items found on the stomachs of the individuals. To standardizer frequency, volume and quantity values this procedure was followed: for each variable its mean value and its standard deviation were calculated, then mean value was taken from each individual value and the result was divided by the standard deviation, all of this was done to make it possible to compare between different ized fishes. s The analysis of feeding activity (R) was based on the quotient of stomach content (Ghazai et al, 1991), where R= We/Wt * 100 (We: Stomach weight (g) and Wt: Individual weight (g). Using field information about diet from earlier work, Roman-Valencia (pers. com.), Román-Valencia & Ruiz (2005), Román-Valencia et al. (2003) and of course those from present work, trophic diversity was calculated, by Shannon-Winner index cited in Ferris Salas (1994), between Hemibrycon sp., & Bryconamericus caucanus, Cetopsorhamdia boquillae, Hyphessobrycon poecilioides, Roeboides dayi and Astyanax aurocaudatus. The units with the diversity is measured come from the infor-mation theory and its depend on the type of used logarithms, in our case bits for garithm base 2 (Magurran, 1989). lo Morisita overlapping diet index was applied (CH) cited by Granado (1996) and Spearman Correlation Range (Rs) (Fritz 1974) to compare between the diet Hemibrycon sp., Astyanax aurocaudatus (Roman-Valencia & Ruiz, 2005) nd the one of the Cetopsorhandia boquillae a (Román-Valencia & Giraldo pers. com.), to calculate previous indexes those items that could not e quantified as discrete units were not used. b To calculate the Gonadosomatic Index (RGS) (Vazzoler, 1996) the following equation was used: RGS= Wo/Wc*100, where, Wc=Wt-Wo, and Wo=gonad weigh (g), Wt=Individual weigh with internal organs (g) and Wc=body weigh (g) without internal organs. Size at maturity was determined using a graphic statistical method and the sex ratio was tested by Chi-Cuadrado (Sokal & Rohlf, 1995) and the Spearman Correlation Range was used to determine diet season and variation as it was proposed by Ortaz (2001). Statisgraphics plus 5.0, Statistix 7.0 and Krebs 0.1 (Windows) were used for statistical procesing. s Physical and chemical variables were recorde thus: Water color and substrate by direct observation, dissolved oxygen, and saturation of oxygen by electrode method, pH with electrometric method (pH meter), width and depth of the creek with measuring tape (decameter), to establish the variation along time of these variables a Variation Coefficient (% CV) was pplied. a Finally, two water samples were taken, during dry season (March) and the other in wet season (May). To be analyzed for Biochemical Oxygen Demand (DBO), Dissolved Oxygen using Winkler Method, Alkalinity, hardness, dissolved and Chlorates by APHA Standard Methods (1992).

H abitat
Hemibrycon sp. inhabit Tinajas Creek, which has an average width of -1.50 m, and 0.4 m average depth. its riversides are usually covered by a secondary forest layer composed of Caña Brava (Poaceae) and Heliconias (Heliconia sp), Guadua angustifolia and agricultural systems like: coffee (Coffea arabiga) and plantain (Musa paradisiaca). The substrate texture of the river basin is a mixture of sand, stones and decomposed organic matter. Water is crystalline but in the wet season it turns brown, superficial temperature is 18.2 °C, with low variation coefficient (17.9-19.7 °C, %CV de 7.17), and atmospheric temperature is 18.52 °C (17.5-18.5 °C; %CV 2.57). The latter presented almost no change during the study. Dissolved oxygen presented a high and stable value during the observation period, 7,18 mg/l medium value (6.13-8.2 mg/l %CV 2.57), and the mean oxygen saturation percentage was 90.71 % (77.8-107.5 %). pH values near 7 (6.7-7.85 %CV 7.2) were found (Table 1). DBO and DQO presented almost no variation between wet and dry seasons ( Table  2). But other items did change alkalinity (24 mg/ l CaCO3 in dry and 42 mg/l CaCO3 in wet), total solids (71 mg/l in ray, and 98 mg/l in wet), dissolved solids (67 mg/l in dry, 94 mg/l in wet); and chlorides (37 mg/l Cl in dry 200 mg/l Cl in wet) ( Table 2). Augmentation during wet season on past items could be related with erosion and anthropic processes occurring at this basin; it can be said that valves presented are characteristically for this kind of habitat of Neotropical ighlands. h Hemibrycon sp. Is syntopic with: Cetopsorhamdia boquillae, Astyanax aurocaudatus, Trichomycterus caliense, and Poecilia caucana.

T rophic ecology
Hemibrycon sp. possess a stomach that is longer (mean 10.3 mm) than wide (mean 5.34 mm), located in the anterior part of the celomic cavity surrounded by other organs. Four to nine (mean 5) pyloric cecae are located at the front art of the stomach. p Hemibrycon sp. stomach contents showed a wide diet spectrum of 52 items (Table 3). The most consumed items were: diptera (Chironomidae, Branchycera, Ceratopogonidae, Nematocera), insect pieces, seeds, plant material and Hymenoptera of family Formicidae. The numeric analysis (Table 3) which indicates the importance of each item selected, gave these results: diptera (Tipulidae, Chironomidae y Branchycera), insects A significative and positive correlation was found between the standard length and the intestine length (r= 0.69, li = 5.35392 + 0.629008*ls, n=75). A non-significant and positive correlation between the weight of the stomach and the standard length (r= 0.48, n=80); and the relationship between standard length and the total weight was (r= 0.67, wt = -4.38232 + 0.137985*ls n=75).

ti
Feeding activity is diurnal asinferred from the digestion condition of the food items examined. Hemibrycon sp. Feeds on organisms that are found at the bottom of the water column, as well as adult insects of alochtonous origin that fall inside this environment. Important differences were not observed between the wet and dry seasons changes diet (Rs=0.8 p=0.0001), which suggests that resources remain mostly unchaned in the environment.

R eproductive characteristics
g The alimentary spectrum of fish that share the same habitat with the studied species and other characids, presented a high trophic diver-sity with ndency to euophagy: Hemibrycon sp. te 4.31 bits, Hemibrycon boquiae 3.56 bits, Bryconamericus caucanus 3.24 bits, Hyphessobrycon poecilioides 2.84 bits, Roeboides dayi 3.51 bits and Astyanax aurocaudatus 3.1 bits. The data by the Morisita index (CH) when comparing the ali-mentary preferences of the species with the ones of sintopic species, the data were low as well as the Sperman quotient of correlation, which also were low (Rs) and it registered a non significative correlation of the variables: between Hemi-brycon sp. -Astyanax aurocaudatus (CH=0.0006 y Rs 0.54 p=0.01) and Hemibrycon sp.
Previous analysis of the reproductive aspects of the Hemibrycon sp. indicate that the Gonodosomatic index (RGS) obtained its maximum values with females in June-July and December-February, with spawning occurring in March and August (Fig. 1). Males the maximum valvues of RGS were collected during Mayo-Julio and November-January. The minimums occurred in February-April and the July-August (Fig. 1). In September and October specimens were found with empty gonads and high amounts of lipids i side the body cavity. n The minimum size of maturity was 40 mm SL for females and 30 mm SL for males, 50 % of the population reaches full gonad deve-lopment for females when they are 53 mm standard length and for males 50 mm (Fig. 1). The mean oocyte count was 445 (range 168-741) with a mean diameter 0.74 mm (0.85-0.5 mm). During this study females were more numerous, and comprised 68.8% of the samples, significantly greater with a 1:2.28 (X2 =O.0098, GL 1 á 0.005). No sexual dimorphism was observed. Species are not pieces of a machine with fixed properties, nor constants populations rather, the same species responds differenty in different places or moments to the challenges of survival (Margalef, 2002). Most tropical fish do not usually show trophic specializations and often change their diet as their biotype changes or according with the fluctuation of the seasons (Lowe-McConnell, 1987). This is the case for Hemibrycon sp., given that it does not only feed on aquatic insects but also on alochtonous material (Ara-nnea, Himenoptera, vegetal part and beetles Coleoptera) that fall occasionally into the water, during the dry and wet seasons. This response is similar to that found by Román-Valencia et al., (pers. com.) for Bryconamericus caucanus nd p Unidentified nematodes were found in the stomachs of Hemibrycon sp.; as well as Bryconamericus caucanus (Román-Valencia & Muñoz, 2001), Argopleura magdalenensis (Román-Valencia & Perdomo, 2004) and Hemibrycon boquiae (Roman-Valencia et al., per. com.), the presence of apparently undigested nematodes inside the fishe's gut means that these nematodes were endoparasites of the Hemibrycon sp., these endoparasites are of indirect cycle in adult fishes ranado, 1996) (Kaiser, 1973;Htun -Kan, 1978). Hemibrycon sp. reproduce when the dry season is ending (August) and again when the wet season is beginning (February-April). Similar results were found on other upper Cauca characids as: Roeboides dayi (Román-Valencia et al., 2003), Creagrutus brevipinnis (Román-Valencia, 1998) and Bry-conamericus caucanus (Román-Valencia & Muñoz, 2001). Confirm what was said by Lowe-to Intestinal length relative to body length is yet another indication of diet: herbivores tend to have a relative longer intestine, than carnivores (Kramer & Bryant, 1995). With the ratio in carnivores less than one (Bussing, 1993). This is confirmed for Hemibrycon sp., and allows us to classify it as a carnivore and insectivore, taking into account that its diet contains mainly insects and algae, the vegetable material (plants, fruits and seeds) fruits are only 4 of the 52 items (Table 3), compared to 45 insect items. This feature has McConnell (1987), these features support that the few estimations of reproductive cycles at small basins show that there different reproductive strategies coexist, that includes spawns at dry season or at wet season or throughout the year. September-October individuals presented fattissue and not gonads, so they when preparing emselves for the next reproductive event. th The Shannon-Wiener diversity index gives an approximation of trophic specialization values reflect europhagic behaviour of predators (Ferriz Salas, 1994), it works for Hemibrycon sp., & Hemibrycon boquiae, Bryconamericus caucanus, and Astyanax aurocaudatus; similar results were obtained for Argopleura magdalenensis (Román-Valencia & Perdomo, 2004). This trophic diversification let species be not -specialized and opportunistic and this behavior reduced inter and intraspecific competition, and favors differential usage of resources (Ferriz & Salas, 1994). For the sake of last considerations is why it can be said that Hemibrycon sp. is an opportunistic and u-rophagic species. e Hemibrycon sp. present little or no competition with Astyanax aurocaudatus or Cetopsorhamdia boquillae as concerns feeding resources, because the Morisita index (CH), resulted in a non-significant, less than 0.6 (sensus Zaret & Rand, 1971, in Herrera & López, 1997; the same was shown by Correlation Spearman range (Rs); to conclude that food is a mechanism of ecological segregation for syntopic species that could compete for resources inside the environment (Hemibrycon sp., A. aurocaudatus and C. boqui-llae), even though their diets are composed of high numbers of macroinvertebrates, differences in the exploitation method because A. auro-caudatus eats more alternative items like alochtonous seeds and fruits (Román-Valencia & Ruiz, 2004), C. boquillae, feeds at late afternoon until night and prefers Trichoptera, Ephe-meroptera, and Odonata (Román-Valencia & Giraldo, pers. com.), and Hemibrycon sp. prefers diptera and insect of alochtonous origin such as Hymenoptera and they search and catch their prey during daylight also, they have a wider alimentary spectrum.