Our new paper shows a divergent response between native and non-native floodplain wetland fish of the Murray River to different temporal scales of wetland connectivity to the river. This finding has implications for the management of native fish communities with environmental flows. For a copy, click this link or send me an email (firstname.lastname@example.org).
Flow alteration is one of the most common threats to rivers around the world. Common causes of river flow alteration are flow obstructions such as weirs and dams and water withdrawal for human and agricultural uses. These changes to rivers tend to suppress the natural flood regime and create a disconnect between the river and its flood plain. This disconnection can have detrimental effects for fish communities because many species of fish evolved to rely on periodic access to floodplain wetlands for food, refuge and nursery habitat.
One tool available to managers of flow-regulated systems is environmental flows. Environmental flows are periodic water allocations used by managers to obtain environmental benefits. For example, in the Colorado River, U.S.A., environmental flows in the form of water releases from Glen Canyon dam are used to transport sediment from the bed of the river to the banks, creating backwater areas that are thought to serve as refuges for juvenile native fish. Similarly, managers periodically deliver water to the floodplain wetlands of the Murray River, Australia to maintain riparian forests and wetland faunal communities.
The use of environmental flows to manage river systems has been heavily studied, but very little attention has been paid to the conditions prior to water allocation when delivering water. For example, would you expect the same response from a fish community if floodplain wetlands were inundated after three months of disconnection from the river versus three years of disconnection from the river? Our study demonstrated that flow patterns on different timescales can be important for determining the fish community structure of the floodplain wetlands of the Murray River. We show that frequent wetland inundation by the river over a long time scale (e.g. 60 days of inundation every 3-6 months over 1-5 years) will favor a native fish community. Alternatively, long periods of wetland disconnection (e.g. 60 days of inundation once every 1-5 years) will favor a non-native fish community.
What does this mean for using environmental flows to promote native fish communities?
Environmental flows tend to be used sparingly because there are many competing demands for fresh water resources. Furthermore, in times of drought, delivering water to the floodplain would be of most benefit to the environment; however, this is also the time when demand for water is the highest for other purposes such as irrigation of agricultural land. Thus, delivering water to wetlands at regular intervals to support a native fish community is unlikely to be a politically popular or economically feasible water use policy in practice. This reality suggests that environmental flow policies, alone, may not be adequate to recover or promote native fish communities in flow-degraded systems such as the Murray River. However, a combination of management actions may prove more successful. For example, our analysis indicates that the common practice of allocating water only after a wetland has been disconnected from the mainstem river for multiple years is associated with high abundances of juvenile common carp, likely due to large spawning and recruitment events. Armed with this knowledge, a manager may be able to employ other strategies for carp control such as installment of carp exclusion devises or the use of daughterless carp technologies in combination with the periodic watering of wetlands. In this way the benefit of maintaining a wetland to the riparian flora and fauna may be realized with out creating a fish community dominated by common carp. The use of compound management strategies has not been evaluated in the scientific literature, to our knowledge, but would make a useful contribution to the field.
Another option is to implement environmental flows only at times, targeting or avoiding the spawning window of particular fish species. For example, common carp are known to spawn between August and April with peek spawning during September in the Murray-Darling Basin likely due to optimal water temperatures. Thus, implementing management experiments designed to evaluate the allocation of water to wetlands during various time periods or water temperature ranges may prove useful for moderating impacts of long-term disconnection of wetlands from the river on the fish community.
The difficulty of creating environmental flow policies that differentially promote native fish also highlights the importance of water conservation in general. The trade-offs among water uses is often not clear to the public, which can lead to inefficient water use practices. However, the reality remains that when water is wasted because of inefficient practices such as open earthen irrigation canals or non-essential purposes such as irrigating lawns or washing cars, it is not available for native fish conservation (i.e. flooding wetlands). Promoting a culture of conservative water use through education programs and water use regulations could relieve pressure on water resources. This would allow for a more liberal use of environmental flows for native fish conservation such as the flow scenarios our study highlights.
There is no doubt that additional research on environmental flows and the management of native fish is essential. The greatest benefit to the refinement of management strategies will likely be realized from targeted studies of flow policies from long-term monitoring programs and/or adaptive management experiments. However, our paper highlights one area where research of environmental flow and fish commonly fails. The issue is that studies of fish responses to flows most often ignore detectability and implicitly make the assumption that detection of fish is both perfect and constant. Unfortunately, this assumption is likely to never be met. In our study we explicitly accounted for incomplete detection of fish with our sampling methods and statistical analysis, and evaluated how our sampling efficiency varied. We showed that detection probability varied significantly among species, wetlands, and sampling gear. Most importantly we showed that detection probability declined as the size of a wetland increased for most fish species when sampling with a fyke net.
When this type of variation in sampling efficiency goes unaccounted for, the apparent variation in the catch data will be attributed to variation in true abundance, leading to incorrect conclusions about the effects of environmental flows.
We are only aware of a few papers, including ours, that account for detectability when evaluating environmental flows for managing fish. Although the ecological literature is dense with articles that warn against ignoring the detection process, some researchers of environmental flows may be unaware of the implications of unaccounted variation in detection. Additionally, accounting for detection is often viewed as too difficult or expensive and is, thus, ignored. We think this is a dangerous oversight that, in the long run, may incur costs in excess of those needed to account for detection because of the increased likelihood of spurious conclusions and mismanagement. Thus, an important message of our paper is that accounting for variable detection is not only possible but necessary to reliably inform environmental flow policies for native fish.
I am interested in any discussion, as there is much to learn about environmental flows and native fish. Please feel free to leave a comment.