Impacts of Managed Flows on Fish Spawning and Recruitment

"August 2008".

Project Number: Impacts of managed flows on fish spawning and recruitment - M/BUS/84.

MDFRC item.

This study tested aspects of the 'window-of-opportunity' hypothesis (a variation on the 'match/mismatch' hypothesis), which attempts to relate the timing of production of prey to the timing of production of fish larvae and subsequent recruitment of fish. The 'window-of-opportunity' hypothesis states that fish which spawn over a 'protracted' period have a recruitment advantage, in any one year, over fish which spawn over a 'brief' period. This is because the probability that a proportion of the larvae of protracted spawning species will encounter a period when conditions are optimal for recruitment is greater than for larvae of brief spawning species. This study tested the 'window-of-opportunity' hypothesis in the context of fishes in lowland rivers of the Murray–Darling Basin, with the primary aim of comparing the estimated hatch date distribution of early stage larvae with that of juveniles at the end of the spawning period (recruits) for a protracted spawning species: Australian smelt. This species was chosen because previous work has indicated that it can spawn for many months each year, and, given that it is widespread in the Murray–Darling Basin, it is likely to occur in a range of river types. We also sought to compare and contrast the relative timing of spawning and recruitment of other native and alien fishes in rivers which have substantially regulated flows (flow-managed rivers) and in rivers which have mostly unregulated flows (non flow-managed rivers), with the premise being that river regulation alters the rearing environment for young fishes and thus may result in different patterns of spawning and recruitment between these two types of rivers. We chose three regions of the Murray–Darling Basin - the 'Southern Region' in the south-eastern part of the Basin (north-eastern Victoria); the 'Lower Region' in the lower, south-western part of the Basin (western Victoria); and the 'Northern Region' in the north-eastern part of the Basin (south-eastern Queensland) - to look for broad-scale patterns throughout this large area, and established paired rivers of similar geomorphologies, sizes, discharge volumes, and ones which would have likely had similar fish faunas prior to regulation. We also recorded physicochemical variables - especially flow and temperature - and sampled food availability, to determine if there were any relationships between these variables and peak recruitment in the rivers sampled. The specific aims of this study were: -To determine whether recruitment is proportional or disproportional to the abundance of newly hatched larvae for a protracted spawner: Australian smelt. -To determine whether spawning and recruitment are synchronous between flow-managed and non flow-managed rivers for fish in general. -To determine whether spawning and recruitment are synchronous between years for fish in general. -To determine whether timing of peak recruitment is synchronous for protracted and short spawners. -To determine if environmental variables or food availability are correlated with peak recruitment of Australian smelt in a subset of rivers sampled. Sampling of larval and juvenile fish was carried out at three reaches of the Ovens (non flow-managed) and Goulburn (flow-managed) rivers in the Southern region, at three reaches in the Lindsay River (non flow-managed) and Mullaroo Creek (flow-managed) in the Lower region, and also in the Mole (non flow-managed) and Dumaresq (flow-managed) rivers in the Northern region of the Murray–Darling Basin, monthly during the breeding seasons of 2005/06 (between September/October and February) and 2006/07 (between August and February). Fish sampling was conducted using seines and hand trawls, mostly, although other (largely ineffective) methods were used in the first season. Seining was by far the most effective method and was used for the majority of analyses, except for descriptions of presence/absence and the occurrence of larvae and juveniles of species in rivers. Zooplankton samples were collected monthly from each reach. Temperature was logged hourly at all reaches, and duplicate physico-chemical variables were recorded each month. Hatching dates of early stage larvae (proto- and flexion larvae) were compared with those of juvenile recruits that survived until the end of the breeding season in 2006/07 using back-calculated ages estimated from otolith increments. A total of 19 species - 14 native and 5 alien - was collected from the six rivers, and most showed evidence of spawning and recruitment in both years of the study. Because of small and patchy abundances, however, only Australian smelt could effectively be compared among all rivers, whilst a small suite of species could be compared for a subset of rivers. Our results showed that overall faunal differences were regionally based. However, substantial differences in species composition were also found between pairs of rivers (flow-managed versus non flow-managed) within regions. These differences were consistent with previous work comparing unregulated and regulated systems. In addition to this, however, we found that in the Southern and Northern regions, common species tended to begin spawning earlier, and spawn for longer, in flow managed rives than in their non flow-managed counterparts. In the Lower Region the opposite was the case. The results, nevertheless, point to the fact that timing and duration of breeding of fish in rivers are likely influenced more by local differences in discharge regimes than broader climatic and geographic influences. The implication of these results is that fish are flexible and will probably respond to alterations to flow regimes if managers impose them. We concluded that classifying fish as either brief or protracted spawners does not include the element of flexibility that was apparent in many of the species studied. Instead we suggest that fish be grouped into 'brief' spawners (ones that only spawn for one or two months), 'flexible' spawners (ones whose spawning period varies by river and by year), and 'protracted' spawners (ones that spawn consistently for more than two months). Only Gambusia holbrooki fitted the last criteria. Our results do not support the 'window-of-opportunity' hypothesis, because the premise upon which this is built was not met - we now classify Australian smelt as a 'flexible' spawner, and its success cannot be purely because of its supposed protracted spawning behaviour. However, it is important to recognize that our results were limited to the sampling methods we used, and not entirely consistent with other studies in similar areas, and so we encourage others to modify this classification as more data are collected. Despite the lack of utility of the 'window-of-opportunity' hypothesis in explaining recruitment success, most species spawned over a long enough period to indicate that recruitment was not proportional to larval production. In other words, the hatching date distributions of early-stage larvae did not match that of juveniles at the end of the spawning season. This was shown best for Australian smelt - in most cases the early hatching events resulted in relatively poor recruitment, and later hatchings resulted in relatively good recruitment. The only exception to this pattern was for the Ovens River, where early larvae and juvenile hatching distributions coincided. Whereas timing and duration of spawning seemed to be influenced by local factors, timing of recruitment tended to be similar for rivers within regions, except for the Ovens/King and the Goulburn systems. Our preliminary comparisons of environmental variables with food availability suggested that temperature (positively) and discharge (negatively) influence production of the smallest size classes of zooplankton and that the first significant rise in zooplankton coincided with peak Australian smelt recruitment in the Ovens River. Management recommendations from this project are: -To classify native fish species in the Murray–Darling Basin based on their life history strategies, and to use these to select representative species for targeted management. -To recognize that the small species in the Murray-Darling Basin are critical to this system's function, and a greater emphasis on their study and management be made. -To recognize that river regulation-effects of fish are overlayed - and sometimes subservient to - regional factors and to, therefore, manage Murray–Darling Basin rivers and fish faunas not as a single unit, but as a number of climatic and zoogeographic zones. -To give high priority to amelioration of cold-water pollution in flow-managed rivers, as temperature depressions are likely to delay or prevent the increase in zooplankton biomass that may be critical to the recruitment of many species of fish. Scientific recommendations from this project are: -To develop and test other fish recruitment hypotheses, besides the 'window of- opportunity' hypothesis in lowland rivers. -To further investigate the timing of spawning and recruitment of Australian smelt in more flow-managed and more intact rivers in each region to increase our capacity to draw meaningful conclusions from patterns observed in the present study. -To initiate a series of adaptive management experiments to investigate the feasibility to influence spawning time and duration of native fish species in flow-managed rivers through manipulation of temperature and flow in flow managed rivers. -To process the rest of the zooplankton samples for non flow-managed rivers that were investigated as part of this study. -To initiate adaptive management experiments in which discharge and temperature are manipulated in flow-managed rivers, to determine if it is possible to influence zooplankton production to levels that are critical for fish recruitment.