The results of the study indicate relatively stable population growth through the 1990s and early 2000s with a significant decline in population density between 2005 and 2011 at Mahaxai. Similarly at Thathot, near to the regulating pond of the NT2 dam, N. aperta population densities in 2011 were observed to have fallen to around 20 % of their 2004 levels. Unfortunately, no data for the period from 2006 to 2010 are available so that it is not possible to determine when the decline began, and in particular if it began only after the NT2 dam was closed in 2010. Nevertheless, together with other factors the impact of the NT2 impoundment should be evaluated as a potential cause of the decline in snail populations.
The eradication of N. aperta populations from the Xe Bang Fai river may have occurred before 2010; however, the river at, and downstream of, Mahaxai is no longer a suitable habitat for this snail (Additional File 1: figure S1). The mean flow rate in the river during the dry season (March to May) has been calculated to show an increase from 13 m3/s to around 140 m3/s after operation of the dam. In addition, the concentration of suspended solids during the low water period is expected to have increased from 45 to 95 mg/l. The diversion of Nam Theun waters into the Xe Bang Fai (via the downstream discharge channel) has led to increases in dry season river depth of 3.5 to 5 m above pre-impoundment levels . The combination of increased current, turbidity and water depth would prevent N. aperta from becoming established in the Xe Bang Fai river downstream of the NT2 discharge channel. The impact of back-flow also appears to have eliminated habitats at least 5 km upstream, but areas further upstream have not been surveyed.
The decline in N. aperta population density at Ban Khi Lek, Northeast Thailand, 400 km downstream of the NT2 project may be part of a natural cycle or fluctuation because a drop in population density was seen in 2004, although this was not as marked as in 2011. The GSS model predicts a slight but steady decrease in population density from 1992 to 2001 with an upturn and population growth thereafter. In view of this, the population could be in a natural growth-decline cycle, but the density observed for 2011 is significantly less than that predicted by the model and much less than the observed 2001 value. The NT2 project could effect population reductions in the Mekong river, far downstream in Thailand in two main ways. First, the Xe Bang Fai river had much higher levels of calcium than the Nam Theun river (about one order of magnitude higher), with a total hardness of 131–149 mg/l (as CaCO3) recorded in February 1995 . Consequently, the influx of Nam Theun waters, further diluted by rain water collected in the reservoir of the dam, will have lowered considerably the calcium content of the lower and middle Xe Bang Fai. Three main rivers, which drain central Laos, enter the Mekong upstream of Ban Khi Lek. The northern most is Nam Kading, then the Xe Bang Fai 150 km further South, and finally the Xe Banghieng about 2 km upstream of Ban Khi Lek and 200 km South of the confluence of the Mekong and Xe Bang Fai (Figure 1). Of these three rivers only the Xe Bang Fai drains a limestone platform and so this river is the main source of hard water flowing down the Mekong river through Ban Khi Lek. The snail N. aperta requires hard water to provide the calcium needed for rapid shell growth as populations re-establish in the Mekong river each year following the annual flood . In view of this, it is possible that the dilution effect of the NT2 discharge into the Xe Bang Fai could be responsible for the decline in snail populations at Ban Khi Lek.
The NT2 project could also impact snail populations at Khemmarat through its effects on snail colonization. Past studies have indicated considerable gene-flow or genetic affinity between N. aperta populations of the lower Mekong river and the Xe Bang Fai river . In addition, it has been proposed that N. aperta populations in the lower Mekong river are replenished following the annual flood by colonists from source populations in the lesser rivers of southern Laos. The hypothesis supported by the fact that the annual spate is less severe in these rivers and they tend to show greater snail population densities earlier in the dry season than habitats in the Mekong river itself . The higher calcium levels, shallower waters and less severe spate may explain the much higher population densities at Mahaxai than at Ban Khi Lek. Consequently, the decline in population density at Ban Khi Lek could be due to a reduction in the upstream sources of snails from tributaries draining into the Mekong (i.e., the loss of the Xe Bang Fai populations).
Alternatively, the fall in population density could be attributed to operations of the sand dredging industry at 16°02′09′′; 105°16′55′′; this activity has led to increased silting at Ban Khi Lek and might lower snail population densities there. Sand dredging at Ban Khi Lek began around 2000 and it is unlikely that its activity would have a sudden and marked effect in 2011. Other factors that could explain the present observations are, an unreported snail control experiment at Ban Khi Lek in 2011, or some unknown polluting event such as a chemical spill from a factory in Laos or in Khemmarat. Snail population densities at Thathot also fell markedly between 2004 and 2011. Thathot is not directly affected by changes in flow caused by the NT2 project; this suggests that N. aperta populations in the region are declining for some reason unconnected with the NT2 project. Thathot is, however, close to the regulation pond of NT2 and the area may have been affected during construction by increased human activity in the area, such as increased numbers of people using the stream where the snails live for laundry, washing vehicles and other polluting activities. The NT2 project itself supports more people living in the area, through economic opportunities, and this alone could lead to increases in pollution of water bodies.
At both Mahaxai and Ban Khi Lek the GSS model appeared to fit the data best if non-stationarity was assumed. The finding implies that at both sites the snail populations were still growing towards the carrying capacity of their habitat. The explanation for this could be a recent de novo colonization of the river at Mahaxai and Ban Khi Lek or recovery from some past population crash; the latter is the most likely. It may be that N. aperta populations in these rivers are not stable in the long-term and experience regular local extinctions followed by recolonization events. If so, this would make the impacts of HEPPs more difficult to determine, in that they must be extracted from a background of natural population fluctuations.