PLANTEAMIENTO INICIAL

Environmental predictors of snail distribution and abundance were discernable at multiple spatial scales, from catchment to local levels. Previous ecological research on environmental requirements of aquatic spring snails in the U.S.A., suggest water temperature, current velocity, substrate type and channel structure (bank heights, angles and overhangs) are major factors contributing to the distribution of these snails (Sada 2001, 2008). The results obtained in this study, however, indicate that for these stream-inhabiting hydrobiid species, catchment size, flow, geology, altitude, forest community, aspect and disturbance are the major factors contributing to their distribution and abundance, although channel structure was not recorded in this study and therefore cannot be discounted.

The relationship between snail abundance and specific habitat variables in the current study was unique to each catchment. Altitude, geology, catchment size and flow were identified as influencing Beddomeia abundance in the Castra Rivulet catchment, although catchment size and geology contributed most to the model explaining the data. Within the Groom River catchment, catchment size, disturbance, forest community and aspect were the significant explanatory factors of snail distribution and abundance; however, disturbance was identified as a key explanatory variable in this catchment. Factors including flow and catchment size were negatively correlated with snail abundance, supporting findings of previous gastropod studies (e.g. Sada 2001, 2008), while the negative relationship between snail abundance and disturbance suggests that snails in some catchments may be sensitive to some types of anthropogenic disturbance.

Previous studies that report on the relationship between gastropod occurrence and stream flow have found that freshwater hydrobiids, and closely related families, are generally more abundant in low velocity flows (Ponder et al. 1989, Crowl and Schnell 1990, Collier and Winterbourn 2000, Richards et al. 2001, Spiers 2003, Sada 2008), while spring snails exhibit intolerance to fluctuations in temperature, salinity and dissolved oxygen and to desiccation (Ponder et al. 1989, O'Brien and Blinn 1999, Sada 2008). Davies & Cook (2002) and Spiers (2003) reporting on the distribution of Beddomeia launcestonensis, found this relationship

77 with flow to be equally true for this large-stream inhabiting species, but also noted that the species‟ habitat also experienced periods of energetic flows at times of extremely high river flows. Beddomeia launcestonensis occurs under large, stable rocks and boulders and in elevated scour pools. These apparent preferred habitat locations may reduce the impacts of extreme flows on the population, and the observation of minimal affects of flows up to 33 cm.s-1_{, may account for the species continued presence in such environments (Ponder et al.}

1993, Davies and Cook 2002, Spiers 2003). Several other Beddomeia spp. are also known to inhabit large streams, at least two species of which may be found under large boulders; however, their upstream range extent and habitat preferences remains unstudied (Ponder et al. 1993).

As catchment size increases so do water velocity, stream volume and depth downstream, and these variables also influence a stream‟s ability to retain CPOM and its composition,

possibly the preferred habitat for Beddomeia. At the same time, a reduction in stream slope in larger streams alters substrate conditions and stable rocks become more prevalent (Gomi

et al. 2002). Retention of such suitable in-stream micro-habitat also depends on the amount of overhanging riparian vegetation and hence the potential to create numerous small-scale debris dams and leaf packs (Gomi et al. 2002, Watson 2004). However, flow was only recognised as a contributing explanatory factor for the Castra data. A possible explanation for this is that the headwater streams in this catchment were, in general, steeper, thus the flows were greater and CPOM was more readily transported away during high rainfall events.

Geology was also identified as a major factor explaining snail distribution in the Castra Rivulet catchment. It is unfortunate that higher altitudes within the catchment correspond to a geological shift from basalt to siltstone, confounding the influences of the two variables, since this makes it difficult to determine the exact effects of each variable. Beddomeia was rarely found in streams on pure siltstone geology (coinciding with elevated parts of the catchment), and only one observation of reproduction was recorded on this substrate (and this was from a stream at the transition of the geological change, where rocks of both geologies occur). This result is contrary to the finding of limited difference in Beddomeia

abundance on basalt and siltstone rocks recorded in Chapter 2, although those findings also came from a stream exhibiting transitional geology. However, the relationship is supported by observational evidence collected from adjacent catchments indicating that Beddomeia and

78 unpublished data). The geology underlying the Groom River catchment is homogeneous and so contributes no further data to the relationship between occurrence and geology.

At least two explanations for the observed relationship between hydrobiids at Castra and geology (similar results observed in the Austropyrgus spp. data) are possible. Firstly, the different chemical composition of the geological substrates may limit the availability of calcium or other minerals directly to the snails. Secondly, differences in surface roughness may affect the snail‟_{s ability to adhere to rocks of the different geologies. Also, the}

mechanism making minerals available in the two substrates may change, altering the growth rate of the periphyton and bacteria on the different rock types. While these ideas need further investigation, the geological substrates at Castra do appear to influence the riparian

vegetation community structure, and hence the allochthonous input. Acacia mearnsii (Black wattle) and Nematolepis squamea (Lancewood) grow more readily in the upper catchment but are rarely present along streams occurring in the basalt soils. The surface structure and texture of siltstone rocks is faceted and smooth, but also more easily inclined to fracture, compared with basalt rocks which slowly and unevenly weather, creating a porous appearance, thus basalt may be more stable and more readily provide sheltering sites for snails. Field observations of rock surfaces also suggests that the rate of periphyton growth is slower on siltstone (K. Richards personal observation). Mineral accessibility, surface roughness, periphyton growth and riparian contributions combined are likely to explain the distribution of Beddomeia in relation to geology at Castra.

Catchment size and disturbance were identified as the most influential explanatory factors for the Groom River dataset. Despite high levels of anthropogenic disturbance in both river catchments, with more recent and widespread agricultural and forestry disturbances occurring within the Castra Rivulet catchment, disturbance was only identified as a significant factor in the models explaining snail abundance at Groom River. Evidence of historic mining disturbance was present in, or adjacent to, most of the study streams in the Groom River catchment, showing signs of significant channel modifications. Although natural regeneration has been allowed to occur throughout much of the catchment for approximately 90 years, there can be no certainty about the severity of the environmental impacts, caused by tin mining, on individual streams. Interestingly, despite this intensive stream channel disturbance, Beddomeia spp. still occur in most streams, indicating that populations can recover after time following high level disturbances.

79 While natural dispersal mechanisms such as downstream transport on CPOM may be

responsible for the widespread occurrence of Beddomeia spp. throughout Groom River, historic records show that alluvial tin mining practices did greatly impact upon the catchment causing severe degradation of many stream channels. Mining was also

responsible for removal of forest cover over wide sections of the catchment, to supply timber for the water-races and settlement buildings (Lesser 1987, Jackman 1998). Such

disturbances are likely to have resulted in extinction of some populations within more impacted streams, for example, those associated with the Anchor Creek mining operation. Conversely, mining operations may also have been the mechanism for translocation of

Beddomeia spp. to certain sites. It is probable that reintroduction of snails to highly modified or eroded streams could have resulted from modifications to the catchment hydrology, through the formation and operation of several large-scale systems of water races across the catchment to support the mining (Department of Mining data presented in Jackman 1998). These races operated for approximately 50 – 70 years, but have subsequently since been decommissioned, lapsing into a state of disrepair and no longer carry water; just how recently individual sections ceased functioning is unknown.