Search

Mechanically reshaping stream banks is a common practice to mitigate bank erosion in streams that have been extensively channelised and lowered for land drainage. A common perception regarding this activity is that fish populations will be largely unaffected, at least in the short term, because the low‐flow wetted channel remains undisturbed. However, the response of fish populations to this practice has rarely been quantitatively evaluated. Using a Before‐After‐Control‐Impact design, we assessed fish community responses to a catchment‐scale bank reshaping event in a fourth‐order low‐gradient stream that drains an intensive agricultural landscape. Quantitative electric fishing and fish habitat data were collected 2 months before and annually for 3 years after the reshaping event. After reshaping, deposited fine sediment levels increased in impact reaches, and there was a significant reduction in anguillid eel biomass (by 49%). In contrast, densities of obligate benthic gobiid bully species increased significantly in impact reaches—potentially due to reduced predation pressure from eels. Three years after bank reshaping, fish community structure had largely returned to its preimpact state in the reshaped areas. Our results suggest that, even in highly modified stream channels, further bank modification can reduce instream habitat quality and displace eels for at least 1 year. Managers should endeavour to use bank erosion control measures that conserve bank‐edge cover, especially in streams with populations of anguillid eels, because these fish are declining globally.

Using a space-for-time substitution design, we investigated the response of structural instream habitat and fish populations to different riparian management practices throughout a Dairy Best Practice Catchment. We found a significant negative correlation between the upstream area of stock exclusion fencing and deposited instream fine sediment cover. Furthermore, we determined that this relationship emerges when ≥300 m lengths of upstream riparian area were included in the analysis, indicating the scale at which stock exclusion fencing results in a positive instream habitat response. Specifically, for this historically degraded spring-fed stream, our findings indicate that riparian segments with 5 m wide stock exclusion fences (both banks) are required to achieve instream fine sediment cover below 20% in downstream reaches. Fish were sparse and evenly spread throughout the catchment. Fish distributions were not correlated with reach-scale riparian or instream habitat variables, possibly because th...

We investigated changes in seasonal biomass of benthic macroinvertebrates over whole floodplain transects upstream and downstream of existing flow diversions in the Wilberforce and South Ashburton Rivers, Canterbury, New Zealand. We tested the predictive performance of a 1-dimensional hydraulic-habitat model informed with existing general habitat suitability curves (HSC), as well as river- and channel-specific HSC developed from the studied rivers. Diversions resulted in significant declines in wetted width, habitat complexity, and resulting declines in benthic biomass, but community composition was unchanged. Weighted usable area (WUA) predictions based on general HSC were often poorly correlated with cross-sectional macroinvertebrate biomass. River- and channel-type-specific HSC greatly improved the accuracy of predictions; however, all HSC generally under-predicted the total proportional change in cross-sectional macroinvertebrate biomass associated with flow change. We conclude that WUA was an insensi...

We developed habitat suitability curves (HSC) using generalised additive models (GAMs) for nine benthic macroinvertebrate taxa from a small New Zealand river for hydraulic-habitat modelling assessments of instream flow requirements. We included interaction terms between the primary variables (water depth, velocity, substrate) when significant, to address a longstanding criticism of univariate HSC. To date, only large-river univariate HSC have been available and these have been used in hydraulic-habitat applications on small rivers, despite doubt over the transferability of HSC between rivers of different size and type. We tested the outcome on the predicted abundance–flow relationship of applying the small-river habitat suitability GAMs versus large-river GAMs for two taxa on the same small river. We found the effects of flow allocation were overestimated by the large-river GAMs relative to the small-river GAMs. Further research to develop general HSC for categories of river size and type is needed to bet...

We tested a model [Benthic Invertebrate Time series Habitat Simulation (BITHABSIM)] for simulating the effect of changes in flow on benthic macroinvertebrate habitat and relative abundance. The model calculates a habitat index (WUA2) based on weighted usable area (WUA) modified to account for reduction of invertebrate abundance by flood disturbance and drying followed by recolonization. The test involved a comparison of WUA and WUA2 with a 1-year abundance time series of the common New Zealand mayfly Deleatidium spp. in a small gravel-bed river with naturally variable flows. The fit of WUA and WUA2 to the Deleatidium spp. abundance time series was judged on correlation and regression analysis of the magnitudes and slopes of the mean-standardized indices and abundance versus time. WUA fit the low flow part of the abundance time series fairly well, but not the portion after flood disturbance. Over the entire time series, WUA fit Deleatidium spp. abundance and rate of change poorly. WUA2 fit Deleatidium abundance better, but the correlation was not quite statistically significant at the 95% level. However, it did explain the essential temporal pattern. The fit of the slopes of standardized WUA2 to the slopes of standardized Deleatidium spp. abundance was significant (R2 = 0.66), but with a systematic bias; high slopes were underestimated and low slopes overestimated. BITHABSIM adds biological realism to traditional hydraulic-habitat modelling based on WUA and so improves the reliability of assessments of effects of flow change on benthic macroinvertebrates over the entire hydrograph. Parameter uncertainty and research needs to improve BITHABSIM and future process-based models are discussed. Copyright © 2014 John Wiley & Sons, Ltd.

We describe and demonstrate a model (Benthic Invertebrate Time Series Habitat Simulation) for calculating the effect of changes to flow regimes on benthic invertebrate habitat and population dynamics. The following inputs are required: a hydrograph (discharge time series), habitat–discharge relationship, disturbance–discharge relationship, wetted width–discharge relationship and a recolonization time series. Habitat–discharge, disturbance–discharge and wetted width–discharge relationships are common outputs from instream hydraulic habitat models (e.g. Physical Habitat Simulation, River Hydraulic Habitat Simulation and River2D). Hydraulic habitat models calculate a combined habitat suitability index from physical habitat suitability curves for water depth, velocity and substrate composition and weight this by area to give a weighted usable area (WUA). Because conventional invertebrate habitat suitability curves are based on density estimates, the combined habitat suitability index can be treated as an index of density and WUA treated as an index of potential relative abundance (at the reach scale) in the absence of disturbance due to flow variation (flooding and drying) and biotic processes. Our approach begins with WUA and calculates realizable suitable habitat (i.e. relative abundance) by taking into account the resetting of benthic invertebrate densities by floods and drying and recovery (or accrual) rates and times. The approach is intended mainly to compare the relative amounts of productive invertebrate habitat sustained by natural and modified flow regimes, but it also has the potential for investigating the influence of flow variation on invertebrate population dynamics. We anticipate that the model will be particularly useful for assessing effects of changes in flow regimes caused by diversions, abstractions or water storage on annual benthic invertebrate productivity. Copyright © 2013 John Wiley & Sons, Ltd.

The effects of natural flow variation on juvenile brown trout population dynamics were investigated by biannual sampling over 5.5 years in the Rainy River, a tributary of the Motueka River. A large flood in late March (50-year return period) substantially reduced the density (by 66%) and biomass (by 73%) of 0+ trout over autumn–spring, but the cohort responded with compensatory survival to achieve similar density and biomass by spring as in other years. A low-flow event in February–April (return period >8.4 years), when 7-day low flows fell to 56% of the 7-day mean annual low flow (MALF) and were less than the MALF for 46 days, had no adverse affect on the population. We found no evidence for density-dependent growth. However, there was strong evidence for a two-phase self-thinning response in density, with no self-thinning occurring over summer (i.e. the 0+ population remained below carrying capacity) until a threshold mass of 22.08 g (length=123.7 mm) was attained in autumn after which severe s...

We assessed whether taxonomic structure and density of aquatic drift could be predicted from the benthos in three New Zealand rivers. The three main orders contributing to both the benthos and drift were Ephemeroptera, Diptera, and Trichoptera. Drift and benthic densities for all taxa and all rivers combined were not significantly correlated (adults inclusive and exclusive). There were significant positive correlations between benthic and drift densities for the three main drifting orders—Ephemeroptera, Diptera, and Trichoptera when data from all rivers were combined. However, these relationships were not always detected in individual rivers. The propensity for Deleatidium to drift was negatively related to chlorophyll a concentration; suggesting density‐dependent drift mediated by food limitation. Drift was reduced when periphyton chlorophyll a concentration was high in relation to benthic Deleatidium density. This highlights an unexpected effect of periphyton proliferation on invertebrate drift...

We investigated temporal (day‐to‐day and season) and spatial (reach) variability of drift with the aim of guiding sampling protocol for quantifying drift at the whole river or reach scale. Overall, we found aquatic drift density and biomass varied considerably seasonally (CV = 72.9, 88.1) and to a lesser extent spatially (CV = 31.3, 30.7) and from day‐to‐day (CV = 45.2, 39.4). Although spatial and day‐to‐day variation in drift density and biomass were similar, sampling logistics suggest spatial sampling would be more cost‐effective and less time consuming. Drift density and biomass estimated from top samplers was often higher than estimates from samplers near the streambed or mid‐water column. A reliable estimate of mean densities and biomass at a site may require only two samplers— a top sampler and either a middle or bottom sampler. In our study, we calculated that sampling at four sites over 1 or 4 days at one site would be required to obtain a 95% CI within 50% of the mean drift density. Eigh...

We compared raw drift‐dive counts of adult brown trout (Salmo trutta Linnaeus 1758) with population estimates calculated from mark‐resight techniques on two occasions in two back‐country rivers near Nelson, New Zealand. Trout were caught by angling and marked with colour‐coded dart tags between 1 and 7 days before drift dives were carried out. Drift‐dive counts in the Owen and Ugly Rivers were 57–66% and 21–43%, respectively, of the population estimates calculated using the mark‐resight techniques. Although underwater census is the cheapest and easiest way to census adult trout populations in clear‐water rivers, a large proportion of the trout population may not be seen in rivers with abundant in‐stream cover. Raw drift‐dive counts provide an index of relative abundance that is useful for monitoring changes in trout populations of specific river reaches over time. However, for comparisons between rivers more accurate population estimates are required. We recommend a combined approach of raw drift...

The post‐spawning movements and distribution of 15 brown trout (Salmo trutta Linnaeus 1758) and 21 rainbow trout (Oncorhynchus mykiss Richardson, 1836) were monitored weekly between August 1996 and March 1997 in the hydro‐electric reservoir, Lake Otamangakau, New Zealand, using radio‐telemetry. The most frequently used areas were the vicinity of the main inflow and the channel running through the main body of the lake. After spawning, brown trout regained rapidly their usual habitat. Rainbow trout were using the inlet intensively before recolonising the main channel of the lake. Brown trout made greater use of the weedy parts of the lake and littoral margins, had a more confined home range, and were less mobile than rainbow trout. Rainbow trout occupied the main body of the lake by day, and moved nearer the surface, and closer to the margins, at night. High lake levels allowed brown trout to exploit wetland margins of the lake.

New Zealand landscapes were extensively modified after European settlement in the late 1800s. Large areas of native forest were converted to pasture and few large areas of unmodified forest remain. Trout (Salmo trutta and Oncorhynchus mykiss) were introduced into the rivers at the same time, and these two changes are thought to have affected native fish communities. Kahurangi National Park is a large area of unmodified native forest with few introduced trout in smaller rivers draining to the west. A survey of these streams was made to determine whether their communities differed from those in other areas of New Zealand. No non‐diadromous native fish were found, suggesting a biogeographic separation from adjacent eastern and southern areas since the Pliocene or early Pleistocene. Contrary to popular expectation, fish diversity and abundance was often lower than at equivalent elevations in other areas of New Zealand. Lowland fish communities comprised a small stream community dominated by inanga (G...

Otolith daily growth increment patterns were used to discriminate between early life‐history patterns of 0+ rainbow trout and to determine the contribution of each to adult recruitment. Inlet stream residents had narrow increments (2.13–3.18 μm) and Outlet Creek residents generally had wide increments (3.63–4.26 μm) across the entire otolith radius. The dominant otolith growth pattern in lake juveniles comprised narrow increments before emergence (3.23–3.66 μm) with wide increments (4.34 μm) soon after emergence. This “early lake migrant” pattern was consistent with incubation in cool inlet streams followed by emigration to the warmer lake upon emergence. Discriminant function analysis (DFA) using mean increment widths at 50, 150, and 200 |im from the primordia discriminated between these patterns (78% overall classification success; 96.5% for inlet stream residents, 60.0% for Outlet Creek residents, 77.0% for early lake migrants). Estimated proportions (from DFA) of the early life‐history types ...

Spatial and temporal variation in relative density of 0+ and 1+ brown trout was examined over three summers in the Kakanui River, North Otago, New Zealand. Both 0+ and 1+ trout were distributed throughout the river. Spatial variation in relative density was much higher than annual variation. Within years, relative density of 0+ brown trout varied by 5 to 92 times between sites (mean coefficient of variation, CV = 1.08) and of 1+ trout by 18 to 84 times between sites (CV = 1.13). Mean relative density of 0+ brown trout varied by 1.5 to 2.0 times between years (CV = 0.35) and of 1+ trout by 3.6 to 23.2 times between years (CV = 0.96). The pattern of 0+ trout spatial distribution was not closely related to the distribution of redds (which also were distributed throughout the river) and not related to the distribution of benthic invertebrate food. By their second summer, most trout occupied the middle and lower reaches of the river. Mean length of both 0+ and 1+ trout increased with distance downstre...

Underwater census and single‐pass electrofishing were compared for estimating relative abundance of juvenile brown trout in the Kakanui River, South Island, NZ. Mean sampling efficiency was lower, and the variability of sampling efficiency was much greater, for underwater census (0+ trout: x = 0.38, s = 0.368; 1+ trout: x = 0.62, s = 0.822) than for single‐pass electrofishing (0+ trout: x = 0.61, s = 0.143; 1+ trout: x = 0.74, s = 0.171). Sampling efficiency of both methods was dependent on temperature. Electrofishing became less efficient at higher temperatures whereas underwater census became less efficient at colder temperatures. The low, and highly variable, sampling efficiency for underwater census of 0+ brown trout was related to substrate hiding behaviour which is dependent on temperature. A ratio method for comparing relative abundance estimates is presented. Minimum significance values for the ratio (R) were derived for 0+ trout using temperature adjusted sampling efficiencies. To be sta...

Stomach contents of Gobiomorphus cotidianus,Retropinna retropinna, Gambusia affinis, and Anguilla australis were compared between two shallow lakes in the lower Waikato River basin, to examine the relationship between turbidity and diet. Lake Waahi and the south arm of Lake Whangape had been turbid (20–40 g suspended solids (SS) m−3) and devoid of submerged macrophytes since the late 1970s and early 1980s, respectively. The main basin of Lake Whangape had been generally clearer (5 g SS m−3) with dense beds of submerged macrophytes, but at the time of sampling (1987) water clarity had deteriorated (> c. 10 g SS m3) and submerged macrophytes had declined. The mysid Tenagomysis chiltoni was an important prey for all species of fish from turbid water bodies but was less important in stomachs of fish in the main basin of Lake Whangape. Apparently, mysids were not an important prey in Lake Waahi before it became turbid. Chironomid larvae and pupae dominated the diets of small fish in the main basin of ...

Growth rates of shortfinned eels (Anguilla australis) from various locations in the Waikato basin, New Zealand, were determined by otolith examination. Two colour‐morphs from the Whangamarino Swamp exhibited different growth rates. Dark‐coloured eels (96% of the catch in low water levels) resided in the swamp; their growth was much slower than that of light‐coloured eels (40% of the catch in floods) which probably entered this habitat in flood waters from Lake Waikare, the Whangamarino River, or main wetland channels. Overall, growth rates were fastest in Lake Waikare followed by Lake Waahi, Lake Whangape, and Hakarimata pastoral streams. Dark‐coloured swamp eels were the slowest growing. Relative annual length increments ranged between 10 and 47 mm. Fast growth of eels in Lake Waikare may be the result of low densities owing to poor recruitment and heavy fishing pressure in recent years. Estimated age at threshold commercial size varied by 11 years throughout the Waikato.

Species composition, population changes, and emigration of 0+ trout in Scotts Creek, an inlet tributary of Lake Alexandrina, were studied from 1979 to 1982. Trout populations were dominated by late emerging 0+ rainbow trout. Rainbow trout juvenile migration consisted primarily of newly emerged nocturnally migrant fry, whereas that of brown trout was mostly of fish which had undergone a period of stream residence. Pattern of initial downstream migration of early migrants was consistent with a passive dispersal mechanism. Although rainbow trout dominated the 0+ salmonid populations, brown trout made greater proportional use of the stream for rearing. This situation arose from differences in patterns of emigration between the species, particularly that of newly emerged fry, with rainbow trout showing a much greater tendency to emigrate upon emergence. The significance of migratory strategies and interspecific competition in relation to lotic versus lentic rearing of brown and rainbow trout is discussed.

Social interactions between 0+ brown trout (Salmo trutta L.) and rainbow trout (S. gairdneri Richardson) were examined in a pair of stream observation troughs. The two species exhibited similar agonistic behaviour and interacted socially. As fry (< 35 mm), rainbow trout were dominant over brown trout in the shallow, faster water (< 15 cm deep, < 15 cm s‐1) nearest the incoming food supply. In fingerlings (c. 43–67 mm), social status was reversed throughout the troughs with brown trout dominant over rainbow trout. A size ratio of 1.7 between early‐emerged brown and rainbow fingerlings and late‐emerged rainbow fry reduced aggression between these fish. Prior residence had only a minorpositive influence on social dominance.

This paper describes the first tests of a fine mesh trap net suitable for sampling shallow‐lake fish communities in New Zealand. The catch from 2 shallow lower Waikato lakes of this and 5 other gear types frequently used in freshwater fisheries research in New Zealand are compared. The trap net caught the widest range of fish species and sizes, and was the least selective for species composition and relative abundance of any individual gear type. A combination of trap net and gill net caught the widest range of species and sizes for individual or pairwise combinations of gear types, and is recommended as being the best combination for sampling shallow‐lake fish communities in New Zealand.

Nine native diadrornous and 2 exotic fish species were recorded in an intensive survey of tributaries of the Mokau River. At the site level, species diversity was low and much of the fauna had a very restricted distribution. Sites were grouped on the basis of their species composition using the classification procedure 2‐way indicator species analysis. Four groups of sites were identified, characterised by: (1) a longfinned eel‐elver assemblage; (2) a longfinned eel‐adult redfinned bully assemblage; (3) an inanga‐adult redfinned bully assemblage; and (4) a torrentfish‐bluegilled bully‐redfinned bully‐elver assemblage. Relationships between fish assemblage distribution patterns and environmental factors were examined with multiple discriminant analysis. The overriding feature influencing patterns of fish distribution was the prevalence of diadromy in the fauna with species varying in their ability to penetrate upstream. Distance from the sea and gradient from the river mouth were the environmental...

Density and growth of mixed populations of 0+ brown and rainbow trout, and population changes of the latter, were studied in Scotts Creek, a spawning/nursery inlet stream of Lake Alexandrina. Stream rearing capacity and the potential for interspecific competition between resident 0+ brown and rainbow trout were assessed. The 0+ trout populations were dominated by rainbow trout with maximum density of rainbow fry being 7.7 fry m‐2; whereas thai: of brown fry usually was less than 0.16 frym‐2. Observed loss of 0+ rainbow trout over much of the summer was constant whereas mortality increased. Loss for the first six months of life was c. 96%. Mean length at the end of the summer growing season (April) was c. 114 mm and 69 mm for 0+ brown and rainbow trout, respectively. Size‐selective emigration for large fish affected observed growth of 0+ rainbow trout in the autumn and early winter. Maximum biomass of 0+ fish ranged between 19.2 + 3.5 gm‐2and 26.0 ± 4.5 giti‐2 over the three summers. Brown trout o...