MORPHOMETRIC RELATIONS OF FRESHWATER FISHES OF THE SUAZA RIVER ( HUILA DEPARTMENT , COLOMBIA )

The knowledge of freshwater fi shes in remote Andean mountains is scarce and restricted to changes of occurrence distribution along a specifi c gradient. The high number of endemic fi sh species requires further studies regarding their biology and their conservation status. The present study aims to estimate the length–weight, length– length, and length–girth relations for nine native freshwater fi sh species representing fi ve families (Loricariidae, Characidae, Heptapteridae, Crenuchidae, and Parodontidae): Chaetostoma thomsoni Regan, 1904; Lasiancistrus caucanus Eigenmann, 1912; Rineloricaria jubata (Boulenger, 1902); Bryconamericus huilae RománValencia, 2003; Gephyrocharax melanocheir Eigenmann, 1912; Pimelodella chagresi (Steindachner, 1876); Rhamdia guatemalensis (Quoy et Gaimard, 1824); Characidium fasciatum Reinhardt, 1867; Parodon suborbitalis Valenciennes, 1850. The fi shes were collected in the Suaza River (Huila, Colombia). These are the fi rst length–weight relations reported for all these species, mostly endemic to the Colombian Andes. The report also provides the new maximum size for four species.

In fi sheries biology, the knowledge of length-weight relations (LWRs) of fi shes is useful: • In determining weight and biomass when only length measurements are available; • For providing information on energy budget, reproductive patterns, and feeding conditions of individuals (Park and Huh 2015); and • For management and conservation of natural populations (Hossain et al. 2012).
Regarding other biometric relations, maximum girth of fi sh body is related to the effi ciency of gill and trammel nets, determining the size of the fi sh caught by different mesh sizes (Kurkilahti et al. 2002).High selection effectiveness would lower the pressure on undersized fi sh, and thereby increase the future size of an exploitable stock (Sistiaga et al. 2009).Relations between total and standard lengths of fi shes are also important in management for comparative growth studies (Sandoval-Huerta et al. 2015).These relations have been studied for the most common and exploited species, nevertheless few LWRs are known for fi sh without a prominent commercial interest that use to be restricted to remote areas (Gaspar et al. 2012).
Colombia is considered one of the world's top biodiversity hotspots, harboring probably more than 10% of global biodiversity (Arbeláez-Cortés 2013).The freshwater fi sh diversity in the Colombia Andes is very high with 188 reported species, the majority of which inhabit only the Colombian Andean basins (Maldonado-Ocampo et al. 2005).
However, our knowledge of the biology of these fi shes is very poor.In this context, we estimated biometric relations for nine captured species from an ecological survey in the Suaza River, located on the Colombian Andes, as a fi rst approach to study the biology of fi sh species in this remote area.This work represents the fi rst biological survey of the Suaza River, in Colombia.The river was studied from the source, in the Cueva de los Guacharos National Park, almost to the confl uence with the Magdalena River (1°38'-1°59′N, 76°06′-75°46′W).The river fl ows, from south to North, for 136 km through the cloud forest and Andean páramo ecosystems of the Huila Department.The altitude varies from more than 1880 to 720 m above sea level in the Magdalena River junction.
Fish sampling was conducted using a backpack electrofi shing unit (Hans Grassl model IG200/2D, 300-600 V, 0.2-2 A) at eleven sampling sites located along the river in January 2011.The collected fi shes were anesthetized and subsequently measured to the nearest 0.01 cm total length (TL), weighed with a digital scale to an accuracy of 0.05 g, and released after the survey, with the exception of some voucher specimens kept for posterior identifi cation.
In the laboratory, fi sh were preserved in alcohol (75%) and identifi ed to species using taxonomic keys and scientifi c literature (Maldonado-Ocampo et al. 2005, Mojica et al. 2005, Galvis et al. 2006).Once identifi ed, specimens were deposited in the fi sh collection of the Universidad de la Amazonia, Florencia, Colombia.
The parameters of LWRs were estimated using the allometric model where W is the total body weight (expressed in grams), TL is the total length (expressed in centimeters), a is the intercept and b the slope of the regression line.Lengthlength relations (LLRs) were also estimated by linear regression analysis: where SL is the standard length.Only extreme outliers attributed to data error were omitted from analyses (Froese 2006).The 95% confi dence limits (CL) of b were calculated (Zar 1999) to estimate differences between calculated slopes and those estimated in the future.
Maximum girth (G) of a fi sh was obtained from the maximum height and width, using the approximate formula proposed by Ramanujan in 1914 (cited after Arfken and Weber 2000).Maximum height and width were measured to the nearest 0.01 cm with a caliper.The perimeter of ellipse (where minor and major radii are the maximum height and width) is more accurate than the maximum circumference to estimate the true perimeter of the fi sh body.The length-girth relations (LGRs) between fi sh TL and maximum girth were estimated by linear regression analysis:

Statistical
analyses were performed using PAST software package (Hammer et al. 2001).
Descriptive statistics of the parameters studied and results of the regression analyses are shown in Table 1.
The knowledge on the biology of the presently studied fi sh species is scarce.Only one species (Bryconamericus huilae) has been evaluated and included in the IUCN Red List of Threatened Species as least concern (Jimenez-Segura 2016).From the nine studied species, four of them are endemic species of Colombian Andes, and future efforts should be focused on the study of freshwater fi shes of this singular region, seriously threatened for current and future exploitation projects (Finer et al. 2008, Finer andJenkins 2012).All LWRs for the studied species were not previously estimated according to FishBase database (Froese and Pauly 2016) and new maximum sizes for four out of the nine studied species were also provided (Table 1).Although at least 100 specimens are adequate to estimate LWRs, for rare species collected in remote areas, such as those of the present study, fewer measurements are acceptable (Froese et al. 2011).
The allometric coeffi cient b felt within the expected range of 2.5-3.5 as suggested by Carlander (1969) for fi sh's growth and Froese and Pauly (2016).High value of this coeffi cient for Characidium fasciatum was similar that estimated for the species Characidium zebra Eigenmann, 1909 in other Andean river (Gaspar et al. 2012) associated to the tendency to present heavier, compact bodies as they grow (Carlander 1969).
The parameters of the LWRs can vary signifi cantly according to sex and season, or due to other factors such as feeding rate, gonad maturity, growth phase, habitat, health and preservation techniques (Hossain et al. 2008), all of which were not accounted for in the present study.Nevertheless, these results will contribute to a better understanding of the poorly known Andean freshwater fi shes; and serve as baseline data for species without previous information on length-weight relations and for comparison with future studies of these fi shes.invaluable fi eld assistance and friendship.We thank the

Table 1
Descriptive statistics and estimated length-weight relation parameters for nine fi sh species captured in Suaza River, Colombian Andes W in g and TL in cm; n = sample size, Min TL = minimum total length, Max TL = maximum total length, a and b = equation parameters, CL = confi dence limits, r 2 = coeffi cient of determination; Bold type = Maximum total length reported data (according to FishBase).

Table 2
Total length-maximum girth (TL-G) and total length-standard length (TL-SL) relations for nine fi sh species captured in Suaza River, Colombian Andes n = sample size, a and b = parameters of the equation, CL = confi dence limits, r 2 = coeffi cient of determination.