By Mike Anderson and Scott Stephens

This article, the third in the series "Prairies Under Siege," highlights how science underpins DU's choices about investing habitat dollars in a region crucial for waterfowl conservation and beset by new and continued threats to nesting habitat

Imagine that you are DU's director of conservation programs. Board a Northwest Airlines A320 bound for Minneapolis from Edmonton, Alberta, on a sunny day in May. In just under three hours you'll fly a narrow transect across the Prairie Pothole Region-North America's duck factory. These 300,000 square miles of farmlands and wetlands support more than half the continent's breeding ducks.

As the jetliner climbs and banks east from Edmonton, you cast your eyes north to the horizon and glimpse the patchwork of dark and light green marking the southern fringe of Canada's vast boreal forest. This has been a land of ducks and trees for thousands of years but is now a place of rapid resource development. Directly below the plane lie the aspen parklands-a matrix of aspen and willow clumps, grazing land, and fields dotted with thousands of small wetlands that reflect the rising sun. You can't see the ducks from up here, but this is mallard country. Where the wetlands are deeper and fringed with cattail, it is also canvasback country.

As the jetliner levels off near 37,000 feet you can see below the climate-driven transition from forest to prairie. This high up, approaching the Alberta-Saskatchewan border, you can take in all at once the gradual change from mostly trees to mostly grass. Everywhere there are still potholes-remnants of the last glacial scouring. Most of the ponds are small; some lie in dense concentrations, more are sparsely scattered; many have been drained or filled.

You notice that the landform is anything but uniform. Patches of higher, hilly ground-often with less tillage and more wetlands-stand out from the flatter, intensively farmed plains. As you approach southern Saskatchewan, the largest of these clumps of hills takes the form of a long, mostly continuous range stretching to the southern horizon (to South Dakota, in fact, if you could see that far). This is the famed Missouri Coteau, an awesome landscape that comprises the single greatest sweep of native grassland and wetlands in the world. In wet years this is truly the best of the duck factory, where wigeon, gadwalls, pintails, and others teem, but in many years much of the coteau is dry.

Floating on south along the eastern edge of the hills, you cross the 49th parallel, separating Saskatchewan from North Dakota. The landforms look much the same; here the national boundary is clearly an arbitrary human imprint. Soon you notice that more frequent blocks of grassland appear in the farmland matrix-the product of USDA's Conservation Reserve Program (CRP)-totaling several million acres in the Dakotas alone. At the same time, the drift plain stretching eastward shows the marks of long-standing agricultural use. Except for CRP fields, and large federally protected wetlands, this is intensively farmed country and vast areas have been ditched, drained, and plowed for human purposes.

CONSERVATION CHOICES IN A COMPLEX WORLD

Now suppose that you had $1 million to invest in conservation actions to help secure the productive capacity of this duck factory for all time. Where within this diverse region would you invest those precious dollars? Should you focus on protecting intact habitats or restoring lost habitat? Should restorations be focused in badly degraded landscapes or mostly intact landscapes? And in each place, is restoring wetlands or upland nesting cover more important? Should you invest in areas best suited to northern pintails, mallards, lesser scaup, or canvasbacks? The questions quickly get complicated.

Then there's the matter of how to accomplish the work. Should you invest in engineering works to restore drained wetlands? . . . work with the forest industry to sustain wetlands where agriculture is gnawing away at the forest fringe? . . . purchase intact parkland habitats from willing sellers? . . . improve grazed landscapes in Saskatchewan and the Dakotas? . . . fund grassland easements in the Missouri Coteau? . . . work with farmers to promote cultivation of duck-friendly winter wheat? . . . or, work with legislators in Ottawa, Washington, and the provincial/state capitals to promote changes in government policies that will benefit ducks and people?

H.L. Mencken is credited with the advice that "There is a simple answer to everything and it's wrong." He wasn't talking about designing habitat conservation programs, but he could have been. While few fixes are "wrong," certainly the best approach differs from place to place.

Within the Prairie Pothole Region, places differ in the abundance of wetlands, human land-use, agricultural policies, predator communities, soil, climate, and the frequency and duration of wet and dry periods. Not only are there manifold possible combinations and tradeoffs among conservation actions, but each management decision also involves uncertainty about the benefits of alternative decisions. And of course, the world is not DU's to do with as it pleases. Much of the best waterfowl habitat is privately owned by people with their own dreams and aspirations for the land.

So how does DU make choices about conservation actions? The history of DU's work under the North American Waterfowl Management Plan (NAWMP) illustrates the challenges and how science helps guide managers in making choices.

NAWMP ON THE PRAIRIES-18 YEARS OF LEARNING AND ADAPTING

With the advent of the NAWMP in 1986, biologists were challenged to develop new programs on an unprecedented scale, with the goal of restoring waterfowl populations to the high levels seen in the 1970s. No conservation initiative of this magnitude had ever been contemplated before anywhere in the world.

In the early 1980s there was still much debate about what factors determine prairie duck numbers. New evidence was emerging, however, from both the U.S. Fish and Wildlife Service (USFWS) and the Canadian Wildlife Service (CWS), that pointed toward poor nesting success associated with intensification of agriculture. Wetland loss, while still serious, was no longer judged to be as limiting for most duck species as the loss of nesting habitat. Thus, the waterfowl conservation community began to address the conservation of whole landscapes-wetlands, grasslands, woods, and fields.

Planners were still faced with a wide array of possible conservation actions and no obvious way to choose among them. The approach taken by DU in Canada and its NAWMP partners was first to select target areas for upland cover enhancement based on wetland abundance, which largely determines local duck abundance. Planners then chose among candidate conservation measures by using a modified form of a computer model of mallard production (the "mallard model") developed by USFWS researchers. Canadian planners focused most of their attention on the aspen parklands rather than the more arid grasslands, believing that in most years the parklands would be wetter and thus have greater potential to produce ducks. In the United States, where previous conservation efforts consisted mostly of restoring larger wetlands and purchasing federal waterfowl production areas, planners followed a similar path, but based upon USFWS wetland management districts.

The mallard model was not designed from the outset to plan habitat programs, but rather as an aid to understanding mallard biology. And in the mid-1980s, such models were unable to incorporate information about the spatial arrangements of habitats: For example, how close planted nesting cover is to wetlands or what geometric shape a conservation project might assume. Despite these limitations, the mallard model was then state-of-the-art technology. By using the model to generate estimates of duck production before and after simulated applications of conservation programs, planners chose an array of actions expected to achieve NAWMP duck population goals. This work culminated in the first-ever comprehensive conservation plans for the Prairie Pothole Region.

So, conservation delivery changed and accelerated beginning about 1990 as new U.S. and Canadian federal dollars, matched by DU funds, flowed toward the prairies. Recognizing the huge conservation challenges ahead, the uncertainties about how ducks and people would respond, and the imperative of using funds wisely, DU committed to learning and improving its performance while it delivered NAWMP conservation projects.

With the support of the North American Wetlands Conservation Council, DU launched the ambitious Prairie Habitat Joint Venture Assessment Study in 1993. Scientists from DU's Institute for Wetland and Waterfowl Research followed the fates of more than 3,600 radio-marked mallard hens and more than 16,000 duck nests on 27 25-square-mile study sites in Alberta, Saskatchewan, and Manitoba between 1993 and 2000. The study was designed to determine the effectiveness of the new habitat programs and to test the major assumptions and data used in the planning model. Feedback from this study substantially modified NAWMP conservation programs and spawned development of a second-generation computerized planning tool to help guide decisions about conservation investments.

It turned out that as a planning tool for the parkland region, the original mallard model needed adjustments. In the end, planners decided that a simpler, multispecies model would provide a better solution. Fortunately, much of the research needed to build the new planning tool had been accomplished during the assessment study.

MODELERS AND FIELD BIOLOGISTS-AN ESSENTIAL TEAM

Computer simulation models of how we think the world works are imperfect; however, they can be useful as aides for conservation planning. Biologists still have to do the hard thinking-concocting arrays of program options that make sense. For example, will a proposed management action address those factors limiting local duck production? Which fixes are practical given regional soil and climatic conditions? What options are likely to be of economic interest to private landowners or a good policy match for government priorities?

What computers do is remove the burden of endless calculations and thereby allow managers to consider and compare complicated combinations of actions across landscapes and over time, and they greatly accelerate the speed with which planners can play "what if?" games.
DU's new decision support system (DSS) represents another leap forward. Dr. Karla Guyn, head of the DU DSS development team, notes that, "By synthesizing our current knowledge about landscape factors affecting duck production, DSS helps ensure that conservation programs are targeted based on the best scientific information available."

DSS is also proving useful for the design of public policy. The model's results already have helped shape DU's approaches to Canadian government agencies and public forums where DU advocates its vision for a sustainable prairie landscape.

There is no better basis for targeting DU's work today than the knowledge and ideas incorporated in these production models. The value of any planning tool, however, depends crucially on the soundness of the assumptions that underpin it. Managing adaptively means, among other things, testing key assumptions that, if wrong, could drastically change conservation outcomes. For instance, while the density of wetlands in spring largely determines the number of birds settling to breed (and also affects nesting effort, duckling survival, etc.), the amount of permanent vegetation in a landscape seems to impact the likelihood that nests hatch. Such a relationship may exist because more grass means predators are less likely to detect duck nests that are widely dispersed, or because the amount of vegetation affects the mix and density of predators present on a landscape. Regardless, for upland-nesting ducks like mallards, data on wetland abundance and land cover together provide DU planners with advice on where to deliver which kinds of conservation actions.


Our understanding of how landscape features affect ducks has improved recently, but the quest for better knowledge continues. For example, the rate of gain in nesting success with increases in permanent cover may be greater in grassland regions than parkland regions. If verified, this would have important ramifications. Consider also how dynamic wetland conditions are in the grassland-pothole region, cycling from years of extreme wetness to severe drought. Layer on top of that changing land-use resulting from changes in farm programs and commodity prices, fluctuating populations of small mammals such as voles and mice (alternative prey for duck predators), and varying levels of diseases such as rabies and mange that affect predator numbers. With this variability, conservation actions that are effective in one area may not be in another, and activities that produce the desired results one year may not be successful the next. According to Dr. Jim Ringleman, head of conservation programs in DU's Great Plains Regional Office, "The million dollar question, really, is how to achieve desired results of conservation programs a high proportion of the time amidst all this variation and complexity."

To address these important questions, DU scientists have launched new studies to help refine their planning tools even further. In Canada, researchers are testing the relationships assumed by the DSS between wetland conditions, duck abundance, nesting effort, land cover, and nesting success. This is being done on landscapes that vary from dry to wet and with nesting cover that varies from poor to excellent. Monitoring sites are scattered from the southern grasslands north to the edge of the boreal forest. In the U.S., nesting success is likewise being monitored on three clusters of sites in North and South Dakota, using a sampling scheme designed to reveal how nesting success is affected by both the surrounding landscape and factors that vary from year to year.

"Coping with this variability may seem challenging," observes Ringelman, "but this is why our new studies are designed this way. Past research has not been able to span either the geographic space nor the year-to-year differences that allow clear separation of the effects on duck production of landscape features and temporal changes." By accomplishing this, Dr. Guyn adds, "DU planners will be able to create even more powerful planning tools in the near future."

And so the cycle of improvement in management performance continues. Plan-implement-evaluate-refine. A simple formula that allows DU's conservation programs to get better year after year.

EYES TOWARD THE FUTURE

DU remains keenly focused on its waterfowl mission and conservation vision. We are in business to sustain duck populations for the long term, and we are committed to maximizing the long-term impact of conservation investments. That has led us to an unswerving commitment to manage adaptively. We embrace science, landscape-scale solutions, and broad partnerships as our approach to the business of conservation. The challenges before us are daunting, but never in its 67-year history has DU stood as ready and well prepared for the work ahead.

The choices we make today are wiser than the ones we made yesterday, and the ones we make tomorrow will be better still. That's what the union of research and wise conservation program delivery is about.