One of the major effects of dam construction on river fish commúnities is the decline and possible disappearance of diadromous species. This is because the adults are prevented from migrating upstream to spawn and the downstream movement of juveniles is delayed (Welcomme, 1985). Negative effects may be tempered if there are breeding areas upstream, in the upper reaches of the river, in the tributaries or at the entrance to the reservoir.
The reasons why fish accumulate downstream of the dams are probably associated with feeding, but fish also accumulate below dams when they are interrupted in their migrations (Welcomme, 1985). The major concentrations below the Salto Grande dam in the Uruguay river are due to the latter cause. Large concentrations of Serrasalmus spp., Astyanax spp. and small pimelodids have also been observed (Delfino, Baigún and Quirós, 1986). Experimental catches of Prochilodus platensis and Salminus maxillosus (size classes ranging from juveniles to adults) have declined in five years since the filling of the reservoir. The trend is the same with respect to Leporinus obtusidens. Catches of Luciopimelodus pati, however, have clearly tended to increase (Delfino, unpublished information). Large concentrations of Prochilodus platensis and Salminus maxillosus occur in the endorheic Dulce river basin, below the Los Quiroga dam (Quirós, personal observation). During low water, fish mortality downstream of the dam is high (Castello, 1982). There are large stocks of both species in the reach separating that dam from the Hondo river dam as well as in the reservoir upstream of the Hondo river dam (Quirós, unpublished information).
There has been a decline in abundance of the highly appreciated migrant species, Salminus sp. and Pseudoplatystoma sp. upstream of the Tres Marías dam on the San Francisco river in Brazil (Milward de Andrade, 1976). The fish community of the Rio Grande, which divides the states of Sao Paolo and Minas Gerais, in which 13 dams have been built, is now very different from what it was originally. The situation in the Tiete river is similar. The construction of 37 dams virtually converted the river into a chain of reservoirs. Pseudoplatystoma coruscans was one of the species which rapidly disappeared (Welcomme, 1985). None of the dams in the Grande and Tiete rivers have been provided with fish pass facilities (Machado, 1976).
Large shoals of Semaprochilodus theraponura accumulate below the Curuá-Una dam in the Amazon basin. This migrant species usually forms large shoals which ascend black water rivers to feed and, later, spawn. The concentrations occur at the turbine outflow and below the weir and, at the present time, Semaprochilodus theraponura is not caught in the reservoir upstream (Ferreira, 1984). Valderrama Barco (1986) reports a similar situation with respect to Prochilodus reticulatus magdalenae at the El Prado dam on the El Prado river in the Magdalena river basin. Catches of this species upstream have declined significantly since the closure of the river.
The Guri dam in Venezuela's Caroni river impedes the upstream migration of the prochilodontids, Prochilodus mariae and Semaprochilodus spp. and many species of pimelodid catfishes (Welcomme, 1985).
The only generalization one may be tempted to make is that where there are no spawning grounds upstream of a dam, migrant species tend to disappear gradually. In cases where these grounds do exist, it may happen that the river may be completely closed off at a particular stage in its hydrological cycle when the migrant stocks are downstream of the dam. This would prevent the development of these stocks upstream of the dam.
