90 Monitoring Threatened Species and Ecological Communities
The majority of the threatened freshwater fishes considered in this chapter
occur in southern Australia, which is where the majority of threats to freshwater
fish are most active (Lintermans 2013a). Many taxa occur across jurisdictional
boundaries (particularly in the Murray–Darling Basin), but notable exceptions
include the Tasmanian galaxiid species, and species from northern or Western
Australia where biographical boundaries are more likely to be contained within a
single jurisdiction.
The median ranks of EPBC-listed species were compared with non-EPBC-
listed species (i.e. but recognised by the ASFB) for nine evaluation metrics
(fit-for-purpose, coverage, periodicity, longevity, design quality, coordination,
data availability and reporting, management linkage, and demographic
parameters) using Wilcoxon rank sum test with a normal approximation. Using
just EPBC-listed taxa, the median ranks of state endemics with multi-
jurisdictional species, and taxa without a recovery plan were compared with taxa
with a recovery plan (even if lapsed or not current, or if the plan is for an
ecological community rather than a single species), again with Wilcoxon rank
sum test. Species considered threatened by ASFB but not EPBC were not included
in this analysis because several state endemics have only been identified as
threatened by ASFB in 2015/16 (e.g. Malanda rainbowfish Melanotaenia sp.,
Running River rainbowfish Melanotaenia sp., black-stripe minnow Galaxiella
nigrostriata and salamanderfish Lepidogalaxias salamandroides in Western
Australia) and would not be expected to have monitoring programs established.
Median ranks between the states and between the three EPBC classifications
(Critically Endangered, Endangered, Vulnerable) were compared using Kruskal–
Wallace non parametric tests. Extinct in the Wild classification was excluded
because only a single species is so-classed. Significant differences between
individual states (for endemics) or EPBC classifications from the Kruskal–
Wallace analyses were compared using multiple comparisons on the least-squares
differences of the mean ranks (Neter et al. 1996). Multiple comparisons were
adjusted for type 1 error using Scheffe’s correction method. Only states with
more than one endemic species listed (i.e. Queensland, Western Australia and
Tasmania) were included in the state comparisons.
Results
Distinguishing between monitoring and data collected as a result of other research,
management or survey programs was sometimes problematic. Of the 57 species
examined, monitoring programs were located for 31 (54%) (including 22 EPBC-
listed), with 26 (including 15 EPBC-listed species) having no current national
monitoring. Seven of the 26 with no current monitoring (including five EPBC-
listed) had monitoring programs that had been discontinued for >12 months. Of