Where are the ducks?

The question haunts us. For the past four seasons waterfowlers in all four flyways have stared forlornly at mostly empty skies. When we find an occasional concentration of ducks --as always happens, even in the poorest of seasons – the bonanza generally is short-lived. The ducks quickly move on or are dispersed by heavy gunning.

At the same time, authorities turn a blind eye to our observations, assuring us that our flocks are in good health. They note the 2003 mallard breeding-population estimate of 7.95 million is virtually equal to management’s population goal of 8.1 million. This prompts organizations such as Ducks Unlimited to declare, “With a few important exceptions, waterfowl populations are doing extremely well.”

Yet, suspicion remains. It forces us to ask: Are recent mallard breeding-population estimates accurate? Did the 1994-2001 surveys artificially boost mallard numbers because the U.S. Conservation Reserve Program made the Dakotas and Eastern Montana more attractive to breeding ducks? Is our basic biology flawed?

Waterfowl biologist Cliff Feldheim examined these issues for Madduck.org. As you can see, his analysis strongly suggests large numbers of previously uncounted mallards moved from poorly surveyed, outlying regions into the CRP-altered Dakotas to artificially boost the mallard breeding-population estimates.

More importantly, we can create a simplified, back-of-the-envelope conversion equation to compare pre-CRP surveys with post-CRP surveys. Pre-CRP data for the years 1970-86 reveals a 1973 peak mallard breeding population of 8.1 million. Post-CRP data reveals the highest mallard breeding population totaled 10.8 million in 1999, falling to 7.95 million last spring, a decline of 26 percent. If we assume the pre-CRP total equals post-CRP’s 1999 peak of 10.8 million mallard, a 26 percent decline suggests the 2003 mallard breeding population is equivalent to 5.99 million pre-CRP mallards, a total significantly below management’s oft-stated population goal.

This tells us – contrary to what DU, the U.S. Fish and Wildlife Service and others say –the high counts of recent years are not indicative of more ducks. It further suggests that we cannot compare post-CRP breeding populations to pre-CRP populations for reassurance that all is well with our breeding mallard flocks.

This is alarming. The finding explains why we are seeing far fewer ducks than expected with current breeding populations. More importantly, the results strongly suggest Adaptive Harvest’s biological analyses that base today’s season lengths and bag limits on current and past populations are fundamentally flawed.

We commend Feldheim’s analysis to your attention. . -- The Editor.

The Conservation Reserve Program (CRP), signed into law in 1985, provided a financial incentive to farmers to convert highly erosive cropland into planted grassland, usually for a period of 10 years. By 1992 over four million acres had been enrolled in CRP in North Dakota, South Dakota, and eastern Montana, the region that constitutes the U.S. prairies in the traditional waterfowl breeding-ground survey. The CRP program marks a landscape-scale perennial grassland program that provides a unique opportunity to evaluate how mallard breeding populations changed in response to large-scale habitat changes.

Mallard Populations in the U.S. Prairies: Classical waterfowl management theory predicts that more nesting habitat should result in more nesting ducks. In fact, the number of mallards nesting in the U.S. prairies, as a percentage of the total mallards counted within the U.S. Fish and Wildlife Service’s traditional survey area, has more than doubled since 1992 (Figure 1).

Figure 1. The percentage of breeding mallards within the traditional North American survey area that were counted in the U.S. increased significantly after water returned to the northern prairies in 1994. Source: USFWS.

However, population totals can be a misleading indicator of reproductive success.

Mid-continent Mallard Production: The number of May potholes has long been considered an indicator of juvenile productivity (recruitment), as measured by the immature-to-adult age-ratio in the harvest. As a general rule, the greater number of potholes, the more young in the fall flight and the higher the age-ratio. We can examine juvenile productivity and habitat changes by comparing the relationship between May potholes and age-ratios in the harvest from 1974-91 to 1992-2001. These time periods reflect periods before and after the establishment of CRP and the return of water to the northern prairies.

The data reveals significant changes in the relationship between May ponds and mallard age-ratios in both the Mississippi and Central Flyway. Not only are the relationships weaker after 1991 (suggesting May potholes are no longer as good as a predictor of juvenile production), but the age-ratio is significantly lower than what would be expected based on the 1974-1991 data.

In the Mississippi Flyway, age-ratios averaged 41 percent lower than expected, while in the Central Flyway age-ratios averaged 11 percent lower than expected. Because the Mississippi Flyway gets more birds in its harvest from the U.S. prairies than the Central Flyway, we would expect the relationship to be more pronounced in the Mississippi Flyway if low production was largely a result of the U.S. prairies.

Reasons for Poor Production: The lower than expected age-ratios should not surprise those who have followed the most current research on nest success and factors associated with mallard productivity in the U.S. prairies.

In mid-continent mallards, Hoekman et. al. (2002) found that nest success and hen survival were the two most important variables in the rate of mallard population growth. Reynolds et. al. (2001) examined mallard nest success in the U.S. prairies between 1992-95. He found nest success generally high enough to maintain a stable population, but highly variable (11-34 percent nest success rates) across the landscape. Similarly, Garrettson and Rohwer (2001) examined nest success of upland nesting ducks between 1994 and 1997 in North Dakota and found success rates varied from 4-42 percent. Reynolds also found that one of the most important variables in determining mallard nest success was the amount of surrounding land planted in grassy cover. For nest success to reach a level necessary to the population (estimated to be from 15-20 percent), at least 30-40 percent of the landscape needed to be planted in perennial cover.

Although there is little doubt that the planted perennial cover from CRP has been a benefit to nesting waterfowl when compared to cropland, existing data and completed research suggests that recruitment and nest success in the U.S. prairies may still be too low to maintain mallard populations across the entire landscape. In some areas, the CRP perennial cover may not be sufficiently extensive (30-40 percent of the landscape) to protect nests and hens from excessive predation by red foxes, skunks and raccoons.

An increasing breeding population in the U.S. prairies coupled with decreasing juveniles points to only one thing: immigration. Between 1992-2001 mallard breeding populations in the U.S. prairies were, on average, 121 percent higher than expected based on the relationship with habitat between 1974-91. This tells us mallards from other breeding areas must be coming in to the U.S. prairies to be counted in the breeding-ground surveys.

Interestingly, mallard populations within the traditional survey area, but outside the U.S. prairies, ranged from stable to increasing during this period (from 5.0 million in 1992 to 5.1 million in 2001, with an average of 5.9 million form 1992-01). However, since 1992, breeding populations in northern Manitoba are negatively correlated with populations in the U.S. prairies and this relationship had changed when compared to 1974-1991.

Northern Manitoba and the surrounding northern areas are surveyed relatively poorly. Is it possible that mallards from areas in Northern Canada that are poorly surveyed, or not surveyed at all, are showing up in the U.S. prairies? The data does not exist to answer that question, but mallards are immigrating into the Dakotas from somewhere.

Mid-Continent Mallard Harvest: Since 1995, high mallard breeding populations in the mid-continent have led to liberal hunting regulations. Between 1970 and 1979 populations were also high and regulations were liberal. Interestingly, if we compare the total mallard harvest and the estimate for the number of active adult hunters in the Central and Mississippi Flyways between 1970-79 and 1995-01, the total mallard harvest is up an average of 13 percent despite the average number of active hunters declining by 19 percent.

Conclusions: Since 1992 the number of mallards nesting in the U.S. prairies has more than doubled to record highs. During that same period, mallard production from the mid-continent, as measured by the immature-to-adult ratio in the Mississippi and Central Flyway harvest, is about 25 percent lower than expected based on habitat conditions. Mallard populations in the U.S. prairies appear to be increasing as a result of mallards coming from other areas, perhaps areas that are poorly surveyed or not surveyed at all. When compared to the last string of liberal hunting seasons and high mallard populations (1970-79), the average 1995-2001 mallard harvest is significantly higher despite significantly fewer hunters.

If the relationship between potholes and breeding population size in the U.S. prairies has changed, this presents significant a challenge to Adaptive Harvest Management (AHM). AHM uses the historic relationship of habitat conditions and mallard breeding population size to predict juvenile production and population change. If juvenile production is less than expected, then hunting regulations may have been more liberal than appropriate for the actual recruitment rate. If Hoekman’s finding that nest success and hen survival are the two most important variables affecting population growth, then over time poor nest success and increased hen mortality as a result of liberal hunting regulations (the most significant source of mortality outside of the breeding season) could work in concert to limit or reduce the breeding population.