How Can Water Be Scarce in a Wet Region?

The Great Lakes region is known the world over for its abundance of high-quality freshwater in lakes and rivers. This perceived abundance belies a scarcity of water that manifests locally in numerous watersheds throughout the Great Lakes basin. Even though there is no danger of people running out of water in the Great Lakes region, heavy and concentrated uses of water can create a scarcity of water for fish, insects, and aquatic plants, and other wildlife in specific wetlands, streams, and rivers, especially during late summer. Because all of the life in an area is connected to the streams and wetlands, this water scarcity indicates degradation of the surrounding ecosystem.

The ways in which people withdraw, consume, discharge, store, and release water for social and economic purposes have a profound impact on the vitality of freshwater ecosystems. These impacts interact with the natural hydrology and climate to determine a watershed’s ecological vitality. Researchers are working to understand how many different social and economic activities combine to affect Great Lakes streams and watersheds, creating a ‘footprint’ of society’s total effects on a watershed. How can the needs of the ecosystem be balanced with important human uses of water? How can we better manage our water resources so that both fish and aquatic wildlife thrive along with increasing human water use needs? To better answer these questions, we need to understand how the cumulative water use impacts freshwater stream ecology at scales from small streams to large rivers.

The Concept of Ecosystem Water Scarcity

Consumptive water uses occur when water is withdrawn and not returned to the source. When too much water is consumed, this can reduce streamflows to the point where fish and other parts of the ecosystem are harmed. It is common sense that water withdrawn from a stream or river can deplete the flowing water. More recently water managers and regulators have understood that people can also affect freshwater ecosystems when water is withdrawn from groundwater aquifers using pumps and wells. However, detailed information for the timing and location of human water uses has only recently become available. This information, in combination with recent scientific advances relating streamflow to the health of aquatic ecosystems, has now made it possible to map the combined effects of all socio-economic water uses on the freshwater ecosystems in a watershed.

The Study

The Kalamazoo River watershed, located in southwest Michigan, is studied as an example of a typical Great Lakes watershed. Two mid-size cities are located in this watershed, along with major industries and a large electrical power plant, and six percent of the watershed’s land area is irrigated for agriculture to produce high-value crops including fruit and vegetables, seed crops, and turf. For each of 133 stream reaches within the Kalamazoo River watershed, from the smallest headwater stream to the Kalamazoo River itself, a computer hydrology model is used to estimate the combined effects of all upstream water uses and compare the combined ecological impacts within a given the stream reach.

The model incorporates the concept of an Adverse Resource Impact (ARI) threshold, which is the allowable reduction in ‘streamflow’- or water flowing in a stream- before the ecosystem is degraded. ARI’s were recently established by the State of Michigan through the Michigan Water Withdrawal Assessment Process (MIWWAP). The MIWWAP employs a publicly available online Water Withdrawal Assessment Tool (WWAT http://www.deq.state.mi.us/wwat/) which allows anyone proposing to make a new or increased large water withdrawal of over 70 gallons per minute to see how the Michigan Department of Environmental Quality would assess the likely impact of the proposed withdrawal on nearby streams. The WWAT applies ARI thresholds to determine whether a proposed use will create an adverse impact on the ecosystem.

This study is the first to use these ARI thresholds in conjunction with large public water use data to map ecosystem water scarcity and to identify places and times where this scarcity may exist due to large concentrations of water use. The study explores the following questions related to water scarcity within the watersheds of the Great Lakes, using the Kalamazoo River as an example:

  1. In what places and at what times does ecosystem water scarcity manifest, according to the MIWWAP?
  2. To what extent do different types of water uses contribute to freshwater ecosystem impacts?
  3. At what spatial and temporal scales are freshwater ecosystem impacts likely to occur in the future?

The Results

An ARI exists when combined streamflow depletions D are larger than the ARI threshold T. (This is where the water scarcity ratio D/T is greater than one.) ARI’s are observed to exist in late summer months when streamflows are lowest, due to a concentration of upstream water uses in seven of the 133 stream reaches in the Kalamazoo River watershed (see Figure 1). However, people can also add water to augment streamflow, for example when deep aquifers are tapped for use and subsequent wastewater is discharged into a surface stream, or when people store springtime floodwater in a reservoir and release it during summer months. This augmentation was found to offset ARI’s in some streams of the Kalamazoo Watershed, and even to cause a net increase in streamflows as compared with normal streamflows (Figure 1). If the effect of human flow augmentation is not included in the ‘footprint’ calculations, then water managers would mistakenly over-estimate the harmful impacts of human water use on the freshwater ecosystem, especially for watersheds with many reservoirs or where deep aquifers are a major source of water.

Figure 1. Kalamazoo Watershed in Southwestern Michigan. Darker shading indicate a higher degree of the water scarcity ratio, which is the ratio D/T between net flow depletion ‘D’ and a stream’s Adverse Resource Impact Threshold ‘T’. The type of stream is also indicated; ‘T’ tends to be lower, and the stream more sensitive to Depletion, in small, cool streams.
Figure 1. Kalamazoo Watershed in Southwestern Michigan. Darker shading indicate a higher degree of the water scarcity ratio, which is the ratio D/T between net flow depletion ‘D’ and a stream’s Adverse Resource Impact Threshold ‘T’. The type of stream is also indicated; ‘T’ tends to be lower, and the stream more sensitive to Depletion, in small, cool streams.

On the main reaches of the Kalamazoo River watershed near the Lake Michigan outlet, ecosystem water scarcity does not currently occur and is unlikely to occur in the future, because cumulative flow depletions are outweighed dramatically by total streamflow. However, water scarcity is observed to occur in some small streams draining watersheds smaller than 300 square kilometers (see Figure 2). This scarcity tends to be caused by large summertime water uses in cities and by irrigated agriculture located immediately upstream, and by highly sensitive ‘cool’ and ‘small’ streams where trout and similar fish characterize the ecosystem.

fig2-booklet-1024x748
Figure 2. Water scarcity ratio (D/T) of upstream streamflow depletion’D’ to scarcity threshold ‘T’ for the watersheds of the Kalamazoo River, plotted against the area of the drained watershed. Markers above D/T = 1 indicate the presence of ecosystem water scarcity in that stream and watershed.

As streams flow from their headwaters and toward the Great Lakes, small-scale streams draining small areas of land combine to form larger-scale streams and rivers draining larger areas of land. It is important to understand how large water uses can combine to create downstream impacts on a larger portion of the watershed. This research determined that the largest ARI’s do currently create ecosystem water scarcity and ecological degradation during summertime in certain smaller streams and watersheds. However, these water uses do not combine to create a scarcity of water in the greater Kalamazoo watershed.

This resilience exists because in the Kalamazoo, human water uses are spaced randomly throughout the watershed, but with a tendency for larger uses to be located near larger streams. As a result of this randomness, streams with a large ARI usually drain into other streams without any ARI, and impacts are therefore rapidly dissipated as water moves downstream. This is a fortunate characteristic of watersheds like the Kalamazoo because it means that the greater ecosystem of the entire river watershed is resistant to degradation from small and localized streamflow depletions. However, if there were more intense human water uses located throughout the entire watershed, or if those water uses were all concentrated in one part of the watershed, large scale ecological degradation could be caused by an accumulation of the individual uses.

For the first time it is now possible to develop a statistical model for the ecological impacts of human water uses in the Great Lakes region. Figure 3 shows the statistical probability model based on the Kalamazoo findings; this is a statistical ‘fingerprint’ for how human socio-economic water uses relate to the natural streamflows in a watershed. In the Kalamazoo River watershed, the average value of the water scarcity ratio D/T tends to be much smaller than the critical value of D/T = 1. The distribution’s with- its ‘variance’- is much higher in smaller watersheds than for the watershed as a whole, meaning that it is much harder to model and predict the ecological impacts of human water uses for smaller streams and watersheds. This statistical model can now be used to design monitoring and management strategies for parts of the Great Lakes region where detailed streamflow, water use, and ecological data may not be available.

Figure 3. This is the probability distribution function for the water scarcity ratio ‘D/T’ which is the ratio of the Net Flow Depletion ‘D’ to Adverse Resource Impact threshold ‘T’ for all stream segments ‘g’ in the Kalamazoo River watershed. This is a statistical probability distribution which gives the likelihood that a randomly chosen stream segment will have a certain ratio D/T. The form of the probability function is Logistic, with distributions separately estimated for different watershed areas from 10 to 10,000 square kilometers (km^2).
Figure 3. This is the probability distribution function for the water scarcity ratio ‘D/T’ which is the ratio of the Net Flow Depletion ‘D’ to Adverse Resource Impact threshold ‘T’ for all stream segments ‘g’ in the Kalamazoo River watershed. This is a statistical probability distribution which gives the likelihood that a randomly chosen stream segment will have a certain ratio D/T. The form of the probability function is Logistic, with distributions separately estimated for different watershed areas from 10 to 10,000 square kilometers (km^2).

Implications

Location of Water Uses in the River Network

The computer hydrology model demonstrates that even the largest water uses located along the main reaches of the Kalamazoo or on the Great Lakes themselves are unlikely to contribute to ecosystem water scarcity. Conversely, the model shows that water scarcity is already present in some smaller streams draining small watersheds, particularly those that are headwaters, are in urban areas, or are in areas with a concentration of irrigated agriculture. Large future water uses should be strategically located on a major, lower river segment or the Great Lakes themselves to avoid contributing to water scarcity and its attendant negative impacts on the surrounding ecosystem. Currently, most large consumptive water uses, such as public water supply and industrial users, are concentrated in or near the main reach of the Kalamazoo River where base flows are larger and ARI thresholds are higher. However, some large summertime users, especially irrigated agriculture, tend to be located in smaller upland watersheds that are more sensitive to streamflow reductions. Water users in these areas might reduce summertime impacts on ecosystems by using stored water or groundwater from deep aquifers that are not directly connected to streams in order to avoid or minimize negative ecosystem impacts. Overall, this study’s results imply that the current and future water use profile of the Great Lakes region is likely to remain sustainable and resilient during growth, as long as the location and timing of larger water uses is carefully chosen in keeping with a management system such as the MI-WWAP.

Climate

Because rainfall and air temperatures affect streamflow, ecosystems are more vulnerable to water scarcity during years that are warmer and have less rainfall, especially during summer. Drought conditions will increase the region’s vulnerability to ecosystem water scarcity and can create situations where normal water uses suddenly cause ARI’s during drought. Climate adaptation protocols should consider the possibility of adjustments to adapt ARI thresholds if future climate regimes result in significantly higher or lower streamflows. In addition, it would be wise to avoid establishing water use patterns that push current ARI limits, because those limits may need to be lowered in the future.

Potential Future Studies

The ‘fingerprints’ of human impacts on freshwater streams can be used by scientists to identify and manage the pattern of human water use and ecological impacts, including both water-rich watersheds like the Kalamazoo and the water-scarce watersheds in heavily populated and arid regions. Within the Great Lakes, the models established by this study may be used to estimate these patterns in watersheds where all the necessary data is not available. As human pressure on water resources increases, this basic science is important to inform the sustainability of the watersheds that we all share.

Additional questions that the researchers would like to explore in future studies include:

  1. What are the roles of seasonal and annual water storage and climate variability in determining how water uses and the seasonality of water uses affect ecosystem water scarcity?
  2. Is it feasible to manage local water scarcity by outsourcing water use through the import of traded goods and services produced using water in another location?
  3. What are appropriate thresholds for different types of freshwater ecosystems? At present, Michigan is the only state in the Great Lakes basin that has legally established ARI thresholds.