Doing experiments in rivers: it’s hard but not impossible


Broadly speaking, there are two ways to do science: make observations or do experiments. Of course, there is also just plain thinking about how things work, although even Einstein conducted thought experiments. Modelling is slightly different, but can be thought of as a type of experiment without actually doing anything physical. While observational studies in rivers are commonplace and relatively straightforward, experiments are generally much more challenging. Hey, but that’s half the fun.

Observational studies doesn’t mean just standing on the bank of a river looking at the water recording what you see.  Observations can mean actually getting in the water with an electrofisher or nets or light traps or whatever and catching things.  The difference, of course, is that with observations, you do not change or manipulate the conditions in the environment (besides putting your sampling gear in there), but with experiments, you do change or manipulate the environment.

An example of an observational study on fish movement might be to trap fish moving into and out of a wetland when it is connected to the main channel of the river (Lyon et al. 2010).  Even though Jarod Lyon and colleagues used fish traps, they did not deliberately manipulate river heights or other factors that might encourage the fish to move. Another example of observational study may be to attach transmitters to fish and see where they go (Crook et al. 2001), as long as their behaviour is not affected by the attachment process or having the transmitter on their backs (Crook 2004a).

An example of an experimental study of fish movement might be to move fish from one location to another and then see if they return from whence they came or establish new territories (Crook 2004b). Another might be to deliberately place fish larvae in a channel and determine how many die following passage through different types of weirs, attempting to simulate ‘undershot’ and ‘overshot’ weirs (Baumgartner et al. 2006).

From Baumgartner et al. (2006)

Experiments are often thought of as being superior to observational studies, because you can work out cause and effect: I change X and get Y result; so Y is caused by X. Often, however, observational studies are the only ones possible, since the systems are just too large or complex to manipulate. Also, with experiments (and observations for that matter), it is important to replicate or have several experimental units and several control ones, or ones you don’t manipulate. Imagine trying to change the course of a river or several rivers, for example. 

But people have done large-scale experiments. Often these are unreplicated, because it is just impossible to do things to more than one system, or there are no other systems around which are comparable.  A good example is the whole lake experiments in Wisconsin that were done in the 1990s (Carpenter et al. 2001).  In that situation, Peter Lake was manipulated and the nearby Paul Lake was not. Piscivorous fish (those that eat fish) were removed, and planktivorous fish (those that eat plankton) were stocked in Peter Lake. Paul Lake remained piscivore-dominated. The phytoplankton (or plant plankton and primary producers that photosynthesize) ended up being in much greater concentrations in Peter Lake, because planktivorous fish kept zooplankton (animal plankton) numbers down, which in turn allowed the phytoplankton to grow well. These results had important implications for management of nutrient-enriched lakes.

Other unreplicated experiments include when scientists and managers released flood flows into the highly regulated Colorado River from Glen Canyon Dam on several occasions from the mid 1990s (Patten et al. 2001; Cross et al. 2011). These are challenging experiments for all sorts of reasons, not the least because of how the community responds.  They are such large events that experimenters have limited control over their effects. But they can tell scientists a lot about how large ecosystems work.

A type of experiment which can get out of hand, through no fault of the scientist, is the natural experiment.  In this case, something major happens and a scientist happens to be there at the time or shortly after to measure the effect of some natural, but usually substantial change.  An example of this was when Mount St. Helens in Washington State, US, erupted in 1980. A large area was affected by ash and lava, and amount of sediment and ash entering rivers increased dramatically as a result. Many scientists took advantage of the event to investigate recovery of both terrestrial and aquatic environments (Wood and Del Moral 1987; Lisle 1995) and the effects on animals in the short term (Brzezinski and Holton 1983).


Mount St Helens erupting in 1980 (Photo: Austin Post and Lyn Topinka, USGS)

Of course, not all experiments have to be big and can often, therefore, be replicated.  We did some experiments in 2002/03 in the Broken River, north-eastern Victoria, in which we aimed to see how young fish and shrimp responded to changing the flow in slackwaters (spots in the river with little or no current), since these are important nursery grounds and are often adversely affected by river regulation (Humphries et al. 2006). We did the experiment this way, because our main project, which was an experimental flow release from a dam, could not take place because of an ongoing drought. We spent a lot of time and effort to create and destroy slackwaters using sand bags, by diverting the path of the current in one reach of the Broken River.  It was back-breaking work, but paid off. We found simply by experimentally decreasing the current, we could create conditions that young fish and shrimp liked and vice versa. 


Diagram showing the experimental set-up we used for creating flow patches and creating slackwaters. From Humphries et al. 2006.

Often, experiments in rivers are difficult, but they are not impossible. It takes imagination, determination and patience….and sometimes, in the case of the Mount St Helens eruption, contacts in high places.

References: Baumgartner LJ, Reynoldson N and Gilligan DM (2006) Mortality of larval Murray cod (Maccullochella peelii peelii) and golden perch (Macquaria ambigua) associated with passage through two types of low-head weirs. Marine and Freshwater Research 57, 187-191. Brzezinski MA and Holton RL (1983) A report on the macroinvertebrates of the Columbia River estuary found in deposits of volcanic ash from the May 18, 1980 eruption of Mount St. Helens. Estuaries and Coasts 6, 172-175. Carpenter SR, Cole JJ, Hodgson JR, Kitchell JF, Pace ML, Bade D, Cottingham KL, Essington TE, Houser JN and Schindler DE (2001) Trophic cascades, nutrients, and lake productivity: whole-lake experiments. Ecological Monographs 71, 163-186. Crook DA (2004a) A method for externally attaching radio transmitters to minimize dermal irritation. Journal of Fish Biology 64, 258-261. Crook DA (2004b) Movements associated with home-range establishment by two species of lowland river fish. Canadian Journal of Fisheries and Aquatic Sciences 61, 2183-2193. Crook DA, Robertson AI, King AJ and Humphries P (2001) The influence of spatial scale and habitat arrangement on diel patterns of habitat use by two lowland river fishes. Oecologia 129, 525-533. Cross WF, Baxter CV, Donner KC, Rosi-Marshall EJ, Kennedy TA, Hall Jr RO, Kelly HAW and Rogers RS (2011) Ecosystem ecology meets adaptive management: food web response to a controlled flood on the Colorado River, Glen Canyon. Ecological Applications 21, 2016-2033. Humphries P, Cook RA, Richardson AJ and Serafini LG (2006) Creating a disturbance: Manipulating slackwaters in a lowland river. River Research and Applications 22, 525-542. Lisle TE (1995) Effects of coarse woody debris and its removal on a channel affected by the 1980 eruption of Mount St. Helens, Washington. Water Resources Research 31, 1797-1808. Lyon J, Stuart I, Ramsey D and O’Mahony J (2010) The effect of water level on lateral movements of fish between river and off-channel habitats and implications for management. Marine and Freshwater Research 61, 271-278. Patten DT, Harpman DA, Voita MI and Randle TJ (2001) A managed flood on the Colorado River: background, objectives, design, and implementation. Ecological Applications 11, 635-643. Wood DM and Del Moral R (1987) Mechanisms of early primary succession in subalpine habitats on Mount St. Helens. Ecology, 780-790.


This entry was posted in River ecology, River ecosystems, River management, River research, Rivers and tagged , , , , , . Bookmark the permalink.

One Response to Doing experiments in rivers: it’s hard but not impossible

  1. Kim Barrett says:

    The site is great and I look forward to the updates. Very informative – I wish I’d seen it before I’d handed in my assignment, lots of observational and experimental study design examples. However, I’m sure it will help with my future studies and also just because I’m interested in what you do and the results that you get.
    Cant wait for the next update…..

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s