Ecological restoration and large-scale ecological disturbance: the effects of drought on the response by fish to a habitat restoration experiment

Summary

In some areas, human-induced erosion in uplands has resulted in deposition of large volumes of sediment in lowland streams forming ‘sand slugs’ i.e. relatively static body of sediment hundreds to thousands of metres in length, and up to several metres depth. Such sedimentation inundates channel features such as pools and fallen timber, resulting in a flat uniform riverbed. Loss of refugia such as deep pools and backwaters that allow aquatic fauna and flora ito survive  through floods and droughts can be particularly detrimental. In an attempt to restore some habitat features to benefit fish, the response of fish populations to wood addition in two streams in southeastern Australia impacted by sand slugs was investigated.

Study area: The experiment was conducted in the Granite Creeks in the Strathbogie Ranges, central Victoria. Human-induced erosion has resulted in deposition of vast quantities of sand into creeks of the Goulburn River floodplain. This has changed the creeks from a series of deep, seasonally connected but permanent pools, to a mostly shallow and uniform channel. The fish assemblages in the sand slug–affected reaches are now much altered with several species being lost and some only present in low numbers relative to undegraded streams.

Timber addition: The habitat manipulations were carried out along a roughly 20 km length of two streams, Castle Creek and Creightons Creek. Castle Creek ceases to flow in most years and has few refugia compared to undegraded stream sections which have ponds chains. Creightons Creek generally flows all year round, which is presumed to be why river blackfish Gadopsis marmoratus persist in Creightons Creek but not Castle Creek. Other more robust species manage to persist in Castle Creek due to the presence of some permanent pools.

Individual sites conmprised a 100 m length of stream devoid of large timber (500 to 2,000m  apart).

Experimental manipulations: A total of 18 sites, 9 on each creek, were selected. Sites were randomly assigned to one of three treatments: control (no wood structures added), one wood structure added, or four wood structures added. The wood structures consisted of red gum railway sleepers bolted together in an overlapping fashion, spanning the channel, perpendicular to the flow. The idea was that during high flows scour downstream would create localized areas of deep water behing the sleepers. Sleepers were installed using a large excavator, extended approximately 0.5 m into each bank to ensure stability.

At one-structure sites, the structure was located approximately 50 m from the downstream end of the site (100 m total length), at the four-structure sites they were located at 40, 48, 56 and 64 m from the downstream end. This distance between the structures was adopted based on the estimated maximum size of pools likely to form around them. In addition, keeping the manipulations away from the upstream and downstream ends of the 100 m reach provided some independence from conditions (particularly wood density) immediately adjacent to sites, which were not otherwise controlled for.

Fish surveys: The fish populations were assessed by electrofishing each site three times before and four times after timber placement on 11–15 May 2001. Fish length was measured before being returned to the stream.

Effect of timber structures on habitat: Due to drought, there were no significant high-flow events during the monitoring period. Thus although there was some scour immediately around the structures with localized increases in depth, these were not apparent at the site scale. However, the wood structures tended to accumulate large quantities of coarse organic matter such as leaves and small branches, which greatly increased habitat structure immediately adjacent to them.

Changes in fish abundance: A total of 3,579 fish of six species were caught over the seven survey periods. Mountain galaxias Galaxias olidus, accounted for 75% of all fish. Mountain galaxias, river blackfish Gadopsis marmoratus and southern pygmy perch Nannoperca australis, together accounted for 97% of all fish. The other three species, Hypseleotris klunzingeri, European carp Cyprinus carpio and goldfish Carassius auratus (the latter two introductions), were comparatively rare and were not considered in further analyses.

Excluding the final survey prior to the stream drying, there was clear evidence that, in Creightons Creek, the three common species were positively affected by the habitat manipulations. Despite a drought during the study, there was a short-term increases in the abundance of mountain galaxias at the four-structure sites. River blackfish and southern pygmy perch generally declined in abundance over the course of the experiment due to drought, but for both, these declines were buffered by the habitat manipulations. The drought eventually caused the complete loss of surface water, and thus fish, from the study sections. So although the study supports the use of timber structures as a means of increasing local fish abundances, these beneficial effects were negated by drought.

Conclusions: The results demonstrate that the introduction of large timber may be a successful in restoring fish habitat in streams degraded by sand slugs, providing there is sufficient flow. Rather than focusing on so-called residential habitat, as this study did, the authors suggest that future work may be better directed toward restoring specific refuge habitats that enhance the ability of the fauna to resist and recover from disturbances such as droughts.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at: http://www.blackwell-synergy.com/journal.asp?ref=1061-2971