A cave population of Isbrueckerichthys alipionis ( Gosline , 1947 ) in the Upper Ribeira karst area , southeastern Brazil ( Siluriformes : Loricariidae )

A cave population of the armored catfish Isbrueckerichthys alipionis is reported from the Santana Cave, in the rio Betari watershed, Upper Ribeira karst area, Iporanga, São Paulo State, southeastern Brazil. The cave population was compared to an epigean population of I. alipionis and no significant differences where found in morphology or degree of pigmentation. As the cave population is known for at least 30 years and is apparently isolated from epigean streams, it is classified as troglophilic. The discovery of this troglophilic species in the Santana Cave is an additional strong argument for the conservation of that cave.

The Brazilian subterranean ichthyofauna distinguishes worldwide by its diversity, with more than 20 species known to exhibit some degree of eye and pigmentation reduction (classical troglomorphisms), indicating a troglobitic condition (restriction to subterranean habitats).In addition, several putative troglophilic fish populations have also been recorded in Brazilian caves (see Bichuette & Trajano, 2003 for definitions).These species differ in their degree of troglomorphism from those with slightly, but statistically significant, reduced eyes and pigmentation, such as Pimelodella spelaea (Heptapteridae; Trajano et al., 2004), to those completely anophthalmic and depigmented.
In 1975, during the exploration of a newly discovered network of galleries inside the Santana Cave by a speleological team of the Centro Excursionista Universitário (CEU), the presence of small, pale armored catfish was noticed in a small upper tributary of difficult access.A few cavers reaching those conduits on the subsequent years confirmed the report.More than 30 years after the first notice, within the scope of a large project aiming to survey the diversity, ecology, behavior, and evolution of subterranean fish throughout the country, our research team visited that small tributary and collected some specimens fitting the description of the armored catfish formerly observed there, presently identified as Isbrueckerichthys alipionis (Fig. 1).The long-term permanence of this population in a subterranean vadose stream, a quite specialized habitat, makes it worthy of interest both from the scientific and conservation points of view.
Herein we present a detailed description of the habitat, notes on behavior, and comparison to an epigean population.

Morphometry
The hypogean population of Isbrueckerichthys alipionis is morphometrically indistinct from the epigean population living in the rio Betari, the watershed where the Santana Cave is included.That population was studied by Pereira & Reis (2002) and the comparison with the cave population is in Table 1.The examination of the orbital diameter, snout length, and interorbital width reveals that differences in eye size are not noticeable when comparing both populations.
We used Principal Components Analysis (PCA) to investigate morphometric variation among epigean and hypogean populations of Isbrueckerichthys alipionis.The analyses were performed on the co-variance matrix of six (head only) and 24 (entire body, Table 1) log-transformed measurements taken from 24 specimens representing comparable size of both populations.The resulting first principal component included a large proportion of the total variance (77.5% in the head measurements and 79% with all measurements included) and all variable loadings were positive and varied little in magnitude, suggesting that it represents a general size factor.Factor scores were plotted for the PC II and III, interpreted to represent size-free shape (Bookstein, 1989).
Plots of factor scores of principal component II vs. III grouped specimens into two largely overlapping clusters (Fig. 2).For the head measurements PC II and III included 10.8 and 4.9% of the total variance, respectively.Measurements  with heavier loadings on PC II are length of mandibular ramus (0.866) and interorbital width (-0.329); on PC III are snout length (-0.531) and orbital diameter (-0.593).For the 24 body measurements PC II and III included 7.9 and 5.7% of the total variance, respectively.Measurements with heavier loadings on PC II are body width at anal fin (-0.682) and body width at dorsal fin (-0.403); on PC III are width of caudal peduncle (-0.757) and orbital diameter (0.461).

Coloration
An individual variation in coloration was observed in the cave population, with some specimens conspicuously paler than others from the same habitat (but still showing the typical coloration pattern with irregular spots) and also in comparison to the epigean conspecifics.A few specimens were brought alive to the laboratory and the differences in coloration were still observed, but they died after some days.Therefore, it is not possible to say how permanent is the paler condition.

Habitat and distribution
The Santana Cave, currently with estimated 7,000+ m of passageways (E.C. Igual, pers.comm.), is one of the largest caves in the State of São Paulo.The loricariids were found in the riacho Taqueupa, a vadose stream (see Trajano, 2001, for habitat classification) crossing the distal end of the Rede Tatus ("Armadillo Net"), an upper net with dry, narrow conduits and large rooms.The riacho Taqueupa appears inside the Rede Tatus, crosses a distance of 15-20 m and disappears again.In one of the occasions when the lowest water levels were observed (April 2007), the maximum depth was around 20 cm in the small pools inhabited by the catfishes; a film of water connected these pools (A.Camargo, pers.comm.).Eight or nine specimens have been seen along 10 m of water course, varying from approximately 3 to 12 cm of total length.These fish displayed a strong avoidance behavior towards the approaching observers, and some went out of the water, moving through humid gravel towards more distant pools (A.Camargo, pers.comm.).The Rede Tatus is accessed through climbing from the base-level stream (the lowest regional erosional level, corresponding generally to a major river that represents the upper limit of the phreatic zone), the rio Roncador, approximately 1.5 km upstream of the cave main entrance, which is the resurgence of the Roncador.The catfishes were noticed for the first time during the discovering and first exploration and mapping of the Rede Tatus, in January 1975.Due to the great beauty and fragility of the speleothems, visitation has been restricted and controlled by the PETAR authorities.Such necessary and welcome restriction, associated to the difficulties of access and progression and the paucity of biologists with sufficient interest and skills, explain why it took decades until the first specimens were collected.

Discussion
The principal component analyses of both head and entire body measurements failed to discriminate clusters among the two populations of Isbrueckerichthys alipionis being investigated, indicating that morphology is highly homogeneous and suggesting no species-level separation.Despite being isolated from one another, the two populations are not morphologically distinct.
The riacho Taqueupa is located 15 m above the Santana base level.Its discharge point into the rio Roncador is not known.Likewise, no permanent epigean creeks uphill could be associated to the riacho Taqueupa.Apparently, this drainage is fed by infiltration water from several small autochthonous depressions, running through small rock fractures.The riacho Taqueupa must be permanent otherwise it would not support a differentiating fish population.There is no confirmed record of Isbrueckerichthys catfish in the base level rio Roncador.Therefore, the Taqueupa population seems to be currently isolated, but not long enough as for morphological differentiation.
Isbrueckerichthys alipionis has been found in the rio Betari and epigean tributaries.Isbrueckerichthys catfish are quite common in these streams, representing 16% of the total number of specimens belonging to 26 fish species reported Table 1.Comparative morphometric and meristic data of cave and epigean populations of Isbrueckerichthys alipionis.Values are given as percent of standard length or head length.Data of epigean population includes holotype and paratype.
for the basin (Buck, 2000).However, none was recorded in caves, except for the population from Taqueupa, supporting the hypothesis of isolation in the latter.
The long-term permanence in the cave habitat and its topographic isolation points to the troglophilic status for the Isbrueckerichthys alipionis population in the riacho Taqueupa.In spite of the diversity of non-troglomorphic fishes in Brazilian caves, in few cases there is strong evidence of the occurrence of troglophilic, i.e., self-sustained populations, in which each individual is able to complete its life cycle in the subterranean environment (see Bichuette & Trajano, 2003, for examples).Among loricariids, there is another case of a probable troglophilic armored catfish, which is a large population of Hypostomus sp.found in an upper tributary inside the São Bernardo Cave, São Domingos karst area.Dozens of these catfish, apparently well nourished and in good physical conditions, have been found in syntopy with the troglobitic heptapterid Pimelodella spelaea (E.Trajano, unpubl. data).Interestingly, their subterranean habitat, a vadose upper tributary, is similar to that of Isbrueckerichthys alipionis.
Conservation remarks.Conservation policies for karst systems in Brazil, and in most other countries as well, have used, as the sole biological criteria for protection, the presence of troglobites.Such an oversimplification is clearly an error that jeopardizes the efficacy of these policies.Subterranean ecosystems are the result of complex interactions between unique phenomena, of which the origin of troglobites is one, an extremely interesting indeed, but still one of many aspects.The presently studied Isbrueckerichthys alipionis catfish is a rather rare case of non-troglomorphic isolated population adapted to a specialized habitat and a candidate for differentiation leading to speciation in the subterranean environment.It would illustrate a first step in the evolution of troglobitic species, and deserves protection measures more effective than the current restriction of visitation to the Rede Tatus.So, the access to the conduit crossed by the riacho Taquepa should be prohibited, except for strictly scientific purposes, and the quality of water and other environmental variables uphill the Santana Cave should be monitored.