Aquifer recharge and wild trout: can we have both?

In last week’s blog, I described a particularly busy, though not atypical, couple of weeks of my work here at the Henry’s Fork Foundation. The subject of aquifer recharge came up numerous times in that blog, so I thought I’d follow up this week with an overview of aquifer recharge, how it could affect wild trout on the Henry’s Fork, and what HFF is doing to ensure that recharge activities do not harm fisheries and perhaps could even help them.

Most of you are familiar with the Eastern Snake Plain Aquifer (ESPA), which underlies 10,000 square miles of the Plain, roughly from Ashton to Hagerman. It stores about 300 hundred million acre-feet of water (compare that to the 4 million acre-feet capacity of the all of the upper Snake River reservoirs, combined), is the sole source of drinking water for about 300,000 people, and is the sole or partial source of irrigation water for about 1.2 million acres of cropland. The Eastern Snake Plain region accounts for 21% of the economic activity in the State of Idaho, about $10 billion annually.

Prior to development of irrigation on the Plain in late 19th century, the aquifer was recharged primarily by seepage from streams and rivers and from groundwater flowing out of tributary basins directly into the aquifer (called “tributary underflow”). Because annual precipitation on the Plain averages only around 10 inches, direct precipitation was a minor source of recharge. Much of the seepage and tributary underflow came from the “sinks” drainages on the north side of the Plain, including the Big and Little Lost rivers and Birch, Medicine Lodge, Beaver, and Camas creeks. Most of the rest was provided by water that seeped out of the lower Henry’s Fork and South Fork and into the aquifer. Under pre-irrigation conditions, the Henry’s Fork lost an average of around 1,000 cfs to the aquifer between Ashton and its confluence with the South Fork. Estimated average recharge to the aquifer from all sources was about 6,000 cfs, and this was balanced by outflow (“discharge”) from the aquifer. About 4,000 cfs flowed out of the aquifer at Thousand Springs, in the Hagerman area, and the rest flowed into the Snake River in the American Falls area.

By the 1950s, the water budget of the aquifer had changed substantially. The amount of water recharged to the aquifer from stream seepage and tributary underflow was greatly reduced, primarily because most of the water in streams and rivers was being diverted for irrigation. Instead of flowing through stream channels and in the gravel underneath those streams, most of the surface water in the system was flowing through irrigation canals instead. These canals were unlined, earthen ditches that leaked water at higher rates than the natural stream channels. Furthermore, the total area of water in contact with the aquifer surface was much greater in the canal system than in natural stream channels. Thus, seepage from these canals recharged more water to the aquifer than stream channels had previously. In addition, irrigation water was applied directly to the soil, via furrows, ditches and flooding. Irrigation water applied in this manner exceeded the crop requirements, and the excess recharged the aquifer. The net result was an increase in annual recharge from about 6,000 cfs to about 10,000 cfs, producing increases in the amount of water stored in the aquifer, the height of the water table, and aquifer discharge. Thousand Springs discharge increased from 4,000 cfs to around 7,000 cfs. On the Henry’s Fork, lower Teton River, and other river reaches in the upper valley, this increase in aquifer recharge actually reversed the direction of groundwater flow between the aquifer and the river. Reaches of river that had lost water to the aquifer prior to irrigation now gained water from the aquifer. The Henry’s Fork between St. Anthony and Rexburg now gained about 250 cfs instead of losing 1,000 cfs. These groundwater returns created new wetland and spring habitats, provided a source of cool water in the summer, and maintained winter flows when water was being stored in upstream reservoirs. None of this recharge was intentional—it is what we refer to as recharge “incidental to irrigation.”

Starting in the 1950s, advent of deep-lift pumps and availability of inexpensive electricity made it possible to efficiently pump groundwater out of the aquifer to irrigate crops. The amount of cropland increased because of this, and soon, about 1.2 million acre-feet per year (about 1600 cfs) were being pumped out of the aquifer. The immediate result was decline in aquifer storage and discharge. Flow at Thousand Springs peaked in the early 1950s, around the time that commercial aquaculture facilities claimed water rights to that spring discharge and developed what became the largest culinary trout production area in the country. Then in the 1970s, farmers who irrigated with surface water began to convert their irrigation systems from flooding and other direct application methods to sprinklers. Much of this conversion was promoted by federal agriculture agencies and programs, and the goal was an increase in irrigation “efficiency.” Unbeknownst to most people at the time, this increase in efficiency reduced the amount of incidental recharge to the aquifer. Although diversion of surface water has declined somewhat since the 1970s, much of the water that previously had recharged the aquifer contributed to increases in crop production and yields, thereby increasing the actual amount of water consumed. As it turned out in the Henry’s Fork, the modest decrease in diversion, which could have been a benefit to fish, was offset by decrease in groundwater returns. In the Henry’s Fork watershed, diversion of water from the river and its tributaries decreased by about 200,000 acre-feet per year from 1979 to 2008, but groundwater returns to the river also decreased by that same amount, with no net benefit to fisheries from the increased efficiency in irrigation practices. In fact, there is some evidence that this tradeoff of lower diversion for lower groundwater return flows may have a negative effect on fisheries. The highest rate of diversion occurs early in the summer, when streamflows are high, but the majority of the groundwater returns in late summer, fall, and early winter, when streamflows are low and fish need it the most.

Hydrologists often refer to “consumptive use”, which is the net amount of liquid water removed from the watershed through evaporation, plant transpiration, and storage of water in the tissues of plants and animals, among a few other consumptive mechanisms. Consumptive use is important to distinguish from withdrawal, which is the amount of water pumped out of an aquifer or diverted from a river or other surface water body. Water that is diverted from a river and ends up seeping into the ground and then returning to the surface system has not been consumptively used; it has just taken a different path to get from point A to point B in the watershed. However, water that is diverted and then ends up being used by a potato plant, for example, is consumptively used. In our situation, consumptive use increases when water that previously had seeped back to the river now contributes to crop production. In fact, scientists from a number of disciplines have shown that increases in irrigation efficiency almost always increase consumptive use and almost never increase the amount of water left in a stream or aquifer. But, that’s a subject for a future blog.

To summarize where we are now, the net hydrologic effects of increased pumping and increased irrigation efficiency were: 1) substantial decrease in aquifer storage and discharge, 2) increase in crop production and yields, 3) increase in consumptive use, and 4) moderate decline in the amount of water withdrawn from streams and rivers. The legal, administrative, and economic effects of these hydrologic changes have been enormous. Water users spend millions of dollars fighting legal battles against each other and against the State regarding who owes whom how much water, as spring discharge and groundwater levels drop. The State spends millions of dollars of taxpayer money mediating these disputes, defending itself, and funding the technical and scientific work needed to manage water in what has become one of the most complex hydrologic and administrative settings anywhere. The bottom line here is that the State and its economy cannot afford to let the ESPA continue to decline; there is simply too much at stake.

Fortunately, the Idaho Water Resource Board (IWRB), the Idaho Department of Water Resources, and agency and university scientists have stepped up to the challenge. The State established a moratorium on new groundwater pumping on the ESPA in the 1970s and committed itself to managing groundwater and surface water as a single entity, something we call “conjunctive management.” The hydrologic aspects of this are challenging enough, but the administrative aspects are even more daunting. Because groundwater is impossible to see and difficult to measure, and because groundwater responds much more slowly than surface water to changes in inflows and outflows, mathematical modeling is the primary tool that scientists, managers, and watermasters use to manage the aquifer and administer water rights. Idaho has the most progressive conjunctive management program in the western U.S. and is far ahead of most of its neighbors in tackling the problem of declining groundwater levels head-on.

In 2009, the IWRB completed a Comprehensive Aquifer Management Plan (CAMP) for the ESPA. The goal of the CAMP is long-term sustainability of the ESPA and its outflows. The CAMP sets a long-term objective of a net change of 600,000 acre-feet per year (about 830 cfs) in the aquifer’s water budget through a combination of reducing pumping and increasing recharge. This time, however, the recharge will be managed. Physically, aquifer recharge is pretty straight-forward: divert water from the river and deliver it to a recharge basin such as a gravel pit or natural depression in porous lava rocks, where the water can readily sink into the aquifer. There are several recharge sites already in use on the Plain, including Egin Lakes, near the sand dunes southwest of St. Anthony.

However, the planning and administrative aspects of recharge are much more complicated. First, the benefits of recharge depend on when and where it is done. Recharge upstream of American Falls Reservoir will have short-term benefits on return flows to the river upstream of American Falls and medium-term benefits to increasing storage in the aquifer, but it would have very minimal benefit to discharge at Thousand Springs. On the other hand, recharge conducted downstream of American Falls will increase Thousand Springs Discharge in the short term but have little effect on increasing aquifer storage. Planning when and where to do recharge to achieve CAMP objectives on multiple time scales is challenging.

More importantly, especially for us, removing water from the river for recharge affects fish and wildlife and reduces surface water supply for downstream water users. These issues are being addressed through the water-rights permitting process. Any diversion of water for aquifer recharge can occur only with a valid water right, the same as diversion for any other use. A water right for groundwater recharge can be claimed and held by an individual water user, a canal company or irrigation district, or the IWRB. Although there are already some water rights in existence for aquifer recharge, accomplishing the CAMP objectives will require many more water rights for much larger quantities of water than currently exist. To meet this need, a number of applications for aquifer recharge rights have been submitted by the IWRB, by individual irrigation districts and canal companies, and by private companies in the business of facilitating mitigation arrangements by which groundwater users can avoid curtailment if their pumping is offset by aquifer recharge. These applications must go through a legal review process, in which anyone with reason to oppose the application can file a formal protest and participate in the process. If the applicant can provide evidence that exercise of the requested water right will harm another water user or a public-interest resource such as fish and wildlife, then the Director of IDWR has authority to deny or modify the application. This is where HFF comes in.

Currently, there are two sets of water-rights applications pending on the Henry’s Fork and its tributaries. One was filed by the IWRB in 1998, for a total combined diversion rate of 2,191 cfs from the lower reaches of the Henry’s Fork, Fall River, and the Teton River. A second application was filed in 2012 by Fremont-Madison Irrigation for diversion of an additional 1,200 cfs from the same river reaches. Given that flow in the Henry’s Fork at St. Anthony as I write this is around 1,400 cfs, it is obvious that these applications for new diversions have the potential to have a large effect on fisheries. However, if granted, these applications will have water-rights priority dates of 1998, and 2012, respectively, making them the most junior rights on the river. To put those priority dates in perspective, even Fremont-Madison’s 1935 storage right in Island Park Reservoir does not fill every year; it was not filled in either 2013 or 2014. These new recharge water rights cannot be used until and unless all of the storage reservoirs in the upper Snake system are full and all senior irrigation rights are filled. Detailed analysis shows that these new recharge rights could be used in only about half of all water years and that substantial amounts of water could be withdrawn from the Henry’s Fork only in the very wettest years, and then primarily during May and June. In no cases we have analyzed would water be available for recharge on the Henry’s Fork during the fall and winter. So, the prior appropriation system (“first in time is first in right”) itself greatly limits the negative effects that aquifer recharge could have on fisheries in the Henry’s Fork.

But what about the years when water is available? Could diversion under these new water rights have a negative effect on fisheries? Of course the answer is yes, potentially. However, the State has repeatedly expressed its commitment to conducting aquifer recharge only when and where it can be done without harm to fish and wildlife, particularly on the high-value resources of the Henry’s Fork and South Fork. IDWR is in the process of formalizing an environmental review committee that will provide input on State-sponsored recharge operations downstream of American Falls, and HFF has been an active participant in that process. Once formalized, the so-called “lower-valley recharge committee” will serve as a model for other such committees in the upper valley, which will review and provide guidance for State-sponsored recharge activities on the Henry’s Fork and South Fork.

Even more promising is that private applicants for aquifer recharge rights are willing to negotiate conditions on the new water rights that provide protection for fisheries and instream flows. Through such negotiations, one applicant agreed to include a minimum streamflow condition on a set of water rights for diversion from the Snake River near Blackfoot. Diversion under those water rights cannot result in flow dropping below a minimum value determined by IDFG scientists. These negotiations resulted from careful analysis of the hydrology and water-rights associated with those specific applications. I personally collaborated with IDFG on that analysis, and our technical report and recommendations were very well received by both IDWR staff and the applicant, resulting in a favorable outcome for everyone. More importantly, our analysis, report, and approach are being view by the State and the applicants as a model for how to proceed through negotiations on the rest of the outstanding water-rights applications, including those on the Big Wood River, Henry’s Fork, and South Fork. Our success in obtaining quantitative restrictions on the amount of water that can be diverted under these rights sets a precedent that we will be able to use all the way up and down the valley.

In light of this, one might ask the reasonable question of whether the recharge objectives can be met if all of the recharge rights are going to have restrictions on the amount that can be diverted. In general, the answer is yes, due again to the way water rights are administered in the upper Snake system. For example, in almost all cases, water available for recharge on the Henry’s Fork that is not withdrawn there will be available for diversion farther downstream, where its withdrawal will have little or even no effect on fisheries. One of the current operational objectives for the upper Snake system is to prevent spill of water past Milner Dam, near Burley, the so-called “0 flow at Milner” objective. Water not diverted from the Henry’s Fork can almost always be diverted at Milner and recharged there, without further impacts on a reach of river that is dry most of the time anyway.

In essence, all water available for recharge under these junior applications is water that would otherwise spill past Milner, as occurs primarily in wet years when the reservoir system is full or is expected to fill. The trick in planning and managing recharge is to distribute withdrawal and recharge of this water among the different locations up and down the Plain to maximize recharge benefits while minimizing negative effects on fisheries and other flow-dependent resources. In fact, there are some scenarios in which carefully timed and placed recharge will improve fisheries, via increased groundwater returns during the fall and winter. Some of us have even been thinking way outside the box and envisioning scenarios in which storage water could be potentially be used for recharge during times of year when release of that storage would benefit fisheries. The administrative and technical details of these scenarios are extremely complex, but the economic stakes involved in stabilizing the ESPA are great enough that no idea is too crazy to eliminate right now. Is it possible that the need for aquifer recharge could allow release of more water from Island Park Dam during the winter? It just might be. 

Idaho is at a critical point in management of the ESPA. Billions of dollars in the economy are at stake, as well as the fisheries of the Henry’s Fork, South Fork and other streams. Fortunately, we have great leadership from the IWRB and sound science and technical guidance from IDWR. On our end, HFF is fully engaged in all aspects of recharge—from review of water rights applications, to development of environmental review processes, to modeling of ground and surface water hydrology. As always, we give wild trout a seat at the negotiating table, bring the best science to that table, and work with agencies and water users rather than against them.