BDBSA Project Metadata Detail

Survey/Project Number: 946          Total No. of Sites: 0
Survey/Project Name: DALHOUSIE FISH SURVEY -KODRIC-BROWN 2003
Abstract: The project will resurvey Dalhousie Springs at Witjira National Park, South Australia, to evaluate the general hypothesis that the system is highly dynamic, with the springs experiencing episodic periods of connection and isolation and the fish undergoing coincident colonization and extinction events. We will visit and re-sample all of the approximately 43 active springs at Dalhousie Springs in Witjira National Park to document the distribution and community structure of six fish species [goby (Chlamydogobius gloveri ), gudgeon (Mogurnda thermophila), catfish (Neosilurus gloveri), perch (Leiopotherapon unicolor) and two species of hardyheads (Craterocephalus gloveri and C. dalhousiensis)]. We will use the same procedures as described in Kodric-Brown and Brown (1993). Data will be analysed as in Kodric-Brown and Brown (1993) to determine correlations between presence-absence of individual species or overall fish species richness and environmental variables (area of source pool, discharge rate, temperature of source pool, length of outflow channel, and area of riparian vegetation). Logistic regression will be used to estimate the threshold values of environmental variables required for the occurrence of each species. Since many of the environmental variables are correlated, we will again use Principal Component Analysis or a similar multivariate technique to quantify the independent patterns of variation. Using the same methods of collecting data and similar analyses will allow us to compare current patterns with those we observed in 1991 (Kodric-Brown and Brown 1993), and to a certain extent to make comparisons with Glover's data collected in the 1970s. How many changes in the fish distributions and community structure, and also in the spring environments, will have occurred in the 12-year period is an open question. Regardless of the answer, the results of the resampling should be informative. If few or no changes are recorded, this will provide valuable information on the temporal stability of the springs and their fish faunas. If, however, as we expect, we document substantial changes, this will provide equally valuable information on the dynamism of the springs and their fish communities. Relating the colonization and extinctions of the fish populations to the changes in spring characteristics and the record of flood events will lead to much better understanding of the mechanistic processes. Several previous studies based on repeat sampling suggest that if the intervals between surveys are too long, colonization and extinction events can be missed, and the magnitude of species turnover can be substantially underestimated (e.g. Diamond and May 1977; Wright 1985).
 
Start Date: 01/08/2003      End Date: 01/10/2003
Survey Type: Fauna only
   
Study Area Description: *** No study area description recorded
Objectives
         Vegetation: *** No vegetation survey objectives recorded
         Fauna: Refer to Abstract.
Methodology
         Vegetation: *** No vegetation methodology recorded
         Fauna: We will use the same procedures as described in Kodric-Brown and Brown (1993). We will revisit each of the 43 springs that we sampled in 1991, and search the area for new springs that may have formed in the past 12 years. Microhabitats in the springheads and outflows will be intensively sampled using a variety of methods (minnow traps, seines, and dip nets). Each spring will be visited on at least two consecutive days. Minnow traps will be set overnight in the source pool and along the outflow (three traps in each location providing that the water is sufficiently deep) and vegetated areas will be sampled with dip nets (for very shallow water) and a small seine to sample adjacent vegetation in both source pools and outflow channels. All fish will be captured alive, identified to genus, measured with calipers, and returned to the spring. Our sampling methods are designed to determine accurately the presence or absence of each species in every spring, not to estimate absolute or relative abundance. We already know the microhabitats of each species, and we will use this information to direct our sampling, to check the previous records of occurrence (Glover 1989, Kodric-Brown and Brown 1993), and to search especially carefully for any species not obtained in the initial collecting. Using a combination of on the ground measurements and remote sensing, we will measure the characteristics of both source pool and outflow stream. Water chemistry (temperature, dissolved oxygen, conductivity, and salinity) of the source pool and the outflow will be measured with a water quality meter (hydrolab). Flow (discharge) rates of outflows will be measured with a flow meter (Flowprobe). Length of the outflow channel and the area of riparian vegetation will be measured by digitizing updated aerial photographs (South Australian Department of Lands). Current measurements of environmental variables and fish distributions will be compared to published records (e.g. Smith 1989, Glover 1989, Kodric-Brown and Brown 1993). This will give a 30-year record of changes, both in the physical characteristics of the springs and in their fish communities.

Data Distribution Rules: Sensitive Dataset: Written permission required from Information Authority
Project Basis: Fauna : Threatened community targeted survey.
Information Authority: University of New Mexico
CSIRO