Factors influencing the distribution of kokopu and koaro (Pisces: Galaxiidae)

• Main Author: Main, Malcolm Richard
• Language: en
• Published: University of Canterbury. Zoology 2012
• Online Access: http://hdl.handle.net/10092/6900

Banded kokopu (Galaxias fasciatus Gray), giant kokopu (G. Argenteus (Gmelin)), short jawed kokopu (G. postvectis Clarke) and koaro (G. Brevipinnis Guenther) (all "large galaxiids") are four species whose juveniles are members of the New Zealand "whitebait" catch. Adult fish occur most often in forested streams, and are rare in (or absent from) poorly-forested regions, such as Canterbury. Four possible reasons for their association with forests (availability of terrestrial invertebrate prey, low maximum temperatures, presence of acidic waters, and presence of favourable microhabitat in forest streams) were investigated experimentally, and a fifth (interactions with brown trout, Salmo trutta L.) is discussed briefly. The diets of large galaxiids were examined using gut content and stable carbon isotope analyses, and by simple laboratory food selection experiments. These studies showed that all three kokopu species relied heavily on terrestrial prey. Terrestrial prey were equally abundant in surface drift samples from streams in forested and pasture catchments in Westland and on Banks Peninsula however; so it is unlikely that kokopu depend on forests to supply that food. Over-representation of terrestrial prey in kokopu diets instead may be related to size-selective predation, or, more likely to their habit of feeding from a station in the water column. Koaro, in contrast, are benthic fish, and terrestrial prey were of lesser importance in their diet. Nevertheless, koaro in forest streams take terrestrial prey in moderate numbers, particularly when drift-feeding in riffles. The upper thermal tolerances of post-whitebait stage juveniles of large galaxiids were determined using a critical thermal maximum (CTM) test, and compared with the thermal regimes of a pair of open and forested Banks Peninsula streams. The mid-summer temperature maximum of the open stream exceeded that of a comparable forest stream by 11°C. However, it did not exceed the CTM for juveniles of any kokopu species (30°C), but came close to that for koaro (28°C). The relatively low CTM of koaro may explain why that species is most abundant in streams at moderately-high altitudes (> about 100m). Choice experiments were carried out in the field in small artificial channels in which the pH of stream water was manipulated, in an attempt to determine whether whitebait of large galaxiids have a preferred pH range. Whitebait of koaro entered circum-neutral pH channels preferentially, whereas kokopu species entered acidic (down to pH 4.5) and circum-neutral pH channels about equally often. Microhabitat use by banded kokopu and koaro were determined in the field by plotting distributions of fish at night, and by comparing use of habitat features with their availability in the environment. The swimming ability of banded kokopu was examined in the laboratory to help interpret results of the field study. Banded kokopu maintain station in the gentle water of pools and backwaters, usually where there is woody debris or large streambed materials (large cobbles and boulders), which can provide cover. Preference for pools may be related to the fishes' habit of feeding in the water column, but swimming performance tests also showed that they were unable to withstand rapid flows for long. Woody debris and large streambed materials (cobbles and boulders) probably provide protection from the effects of flooding, during which banded kokopu are susceptible to being washed dowmstream and injured by abrasion. The availability of appropriate microhabitat was the most important factor influencing the distribution of banded kokopu identified in this study. Similar microhabitat conditions may be required by giant and short jawed kokopu, but koaro inhabit riffles equally as often as pools. Interactions with introduced brown trout and their possible effects on the distributions of large galaxiids are discussed briefly in the light of observations made in the laboratory, and published distributional data from South Westland. Distributions of large galaxiids and brown trout are largely allopatric. Laboratory observations showed that trout are more aggressive than banded kokopu, and have the potential to influence the kokopu's distribution by competitive interference.