ratio in 1958 is a shade low, while the ratio for 1960 is probably too high. These are our best estimates at the present time, however, and many of the figures in the table can be checked for reasonableness by two or more independent methods.

For ease in interpretation, pertinent information in this table are presented in figures 1 and 2. Important items revealed in the table and the figures are as follows:

1. Breeding population has varied considerably during the period but the estimated fall flight has varied even more. The primary reason for this is that production ratios were high during the mid-1950 period when habitat was good and breeding populations were high. Conversely, during the early 1960's when habitat conditions were poor and the breeding population was down, production ratios were low.

2. The gun is by far the most important cause of mortality once the birds are fledged. For the period 1955–65, bag and crippling loss accounted for an average of 37 percent of the fall flight, whereas losses due to causes other than the gun averaged only 16 percent.

3. The crux of the situation is shown in Column 5 (percent of fall flight bagged and crippled) and the corresponding line on the graph. The percent of fall flight lost due to the gun has ranged from a low of 26 percent in 1962 to a high of 47 percent in 1958. The relationship between percent of fall flight bagged and crippled and total annual mortality is shown in figure 2. We conclude that gun mortality is largely additive to natural mortality. In other words, if the birds are not shot, a large portion will survive.

The next problem to examine is: "What caused the proportion of fall flight bagged and crippled to change?" Those in the room who remember the years during the last 11 which had the most restrictive regulations will realize at once that lower proportions of fall flights, bagged and crippled, are associated with shorter seasons and smaller bag limits. For example, 1959 was the first year of restrictive shooting regulations associated with the recent drought and the percent of populations present that was harvested dropped from 47 percent in 1958 to 32 percent in 1959. The important point is that shooting regulations are effective, not only in changing the total number of birds killed, but also in changing the proportion of birds that are killed. During a year when the production ratio is low, or when the objective is to increase the breeding population the following spring, it is of great importance to realize that, when properly used, regulations can change the proportion of birds that are harvested. Emphasis should be given to "proper use" because mistakes can have quite serious consequences. This is demonstrated by what happened in 1964. In 1964, with a breeding population about equal to 1963, we experineced a lower production ratio than in 1963 and, consequently, a somewhat smaller fall flight. However, habitat conditions appeared to be somewhat better in 1964 and, in spite of a predicted decrease in fall flight, shooting regulations were relaxed a small amount. The number of birds killed increased moderately, but since this kill was taken from a smaller fall flight, the proportion killed increased to about 42 percent. As a direct result, the breeding population in the spring of 1965 dropped to the lowest on record. We must conclude that shooting regulations are powerful tools for managing the waterfowl resource and must be used with due respect.

These data demonstrate what has happened with mallards during the past 11 years, and the breeding population estimate for 1966 (7,560,00) provides us with the first clue as to where we are at present. As you can see, our mallard breeding population is approximately equal to 1963 and 1964 and is still considerably below the good water years of the mid-1950's when mallard kills in the 10 to 11 million category were recorded.

These data are presented with the thought in mind that they would prove of value in answering the question: "Where do we go from here?" It has been suggested that we might as well establish our objectives at the current population level and relax regulations to the point where all of the anticipated surplus this fall would be harvested. We cannot agree with this line of reasoning. In the first place, weather cycles are far from being predictable. It is entirely possible that reasonably good or even excellent water conditions will persist for several more years. Potholes without breeding birds, particularly at a time when the

population is relatively low, is a waste any way you look at it, and this was certainly the situation during the spring of 1966.

We have heard arguments to the effect that we should never have allowed the breeding population to decrease as much as it did. In line with this argument, our data suggest that it would have been quite possible to hold the mallard breeding population index at a 10 million level during the drought period. In our opinion, however, this would not have been the best management.

There are three aspects of the problem which must be considered. The first is production ratio. An analysis of existing data suggests that during periods when there are more breeders than can find suitable nesting sites (which is usually the case during drought periods) the number of young produced will be more a function of the amount of breeding habitat than it will of the size of the breeding population. To illustrate, in 1961 with a breeding population of 8.1 million, approximately 5.7 million young were produced. The data suggest that there would have been about this number of young produced in 1961 even if the breeding population had been somewhat smaller or considerably larger. In other words, if the breeding population in 1961 had been 10 million, the number of young would not have been significantly above 5.7 million, and the production ratio would have been closer to 0.6 immatures per adult than to the observed ratio of 0.8 immatures per adult. We are fully aware that there is not complete agreement with this line of thought. A reason for disagreement is that during even the driest year at the peak of the drought potholes were observed in the prairies with no ducks. This suggests there was unused breeding habitat even in years like 1959 and 1961. On the other hand, the evidence is clear that large numbers of pothole nesting species overflew the prairies during the dry years and spent the summer largely in northern Alberta and the western portion of the Northwest Territories. The fact that they were not productive in these northern areas is revealed by the low ratio of immatures to adults in the flight the following fall. The second aspect is mortality due to causes other than shooting. During the drought period (1959 through 1963) the average annual loss due to natural causes was about 20 percent of the fall flight. Had we been maintaining a 10 million breeding population during this period, rather than the smaller breeding populations which were present, the numerical loss to natural mortality factors would have been larger. Therefore, a larger portion of the young produced would have been required to offset this loss, and the already small harvestable surplus during those years would have been smaller still.

The third aspect is that shooting regulations required to maintain a 10 million breeding population of mallards during drought periods when production ratios are low, would need to be more restrictive than those in effect during most years following 1959. It seems likely that regulations similar to those in 1962, when the Mississippi and Central Flyways had 25-day seasons with a limit of 1 mallard and 2 birds in the daily bag, would have been required more frequently. In short, it appears feasible to manage so as to maintain a stable breeding population at a medium to high level but the restrictive_regulations required to do this during drought periods would be drastic. Further, offsetting the numerically larger loss due to natural causes associated with maintaining a larger breeding population would result in smaller harvestable surpluses. We have concluded, therefore, that the waterfowl resource can supply a greater harvest during a period including both wet and dry conditions if the breeding population is allowed to fluctuate.

If you agree with this interpretation of the records of the past 11 years, then the regulations established during most of those years constituted good waterfowl management. If not, then mistakes have been made. Regardless of whether you agree or not, the hard fact remains that duck breeding grounds are in condition to provide nesting sites for many more breeders than were available in the populations returning north this past spring. Had the 1966 mallard breeding population been 10 to 12 million instead of 7.6 million, both Canadian and U.S. hunters would have been able to enjoy longer seasons and larger bag limits than will likely occur. From the standpoint of weather cycles, we can see no reason to suspect that 1967 will be other than a good year insofar as breeding habitat conditions are concerned. Under this circumstance, there would be a big pay-off in terms of harvestable surplus in the fall of 1967 and beyond if the breeding population level is allowed to increase considerably. This can be accomplished with appropriate shooting regulations this fall. The choice is ours to make.

WATERFOWL BREEDING SURVEY SUMMARY, 1966

Weather and habitat conditions

Of

In the most important production areas both habitat and weather were favorable in 1966. The May pond index dropped slightly from 1965 but there was still more water than necessary to accommodate the breeding population. greater significance was a continued increase in the July water index. In other words, the habitat remains generally good going into late summer and fall (figures 3 and 4). There are some dry fringe areas, however, that bear watching such as parts of Alberta, South Dakota, Nebraska, Wyoming, Montana and Washington.

In contrast to the seemingly adverse weather in 1965, conditions in 1966 were considerably better over most of the area. Alaska was somewhat retarded as was much of Ontario and Quebec. Elsewhere an open winter was followed by an early breakup and rapidly advancing spring. There were a few pockets of adverse weather in Alberta and Manitoba, but nothing extensive nor prolonged. Breeding population indexes

Ducks.-Aerial breeding population indexes by survey area are shown in table 2. These indexes have not been adjusted for birds present that were not recorded by the aerial survey crews. Increases in breeding populations of game ducks were reported in most survey areas as compared to 1965. Exceptions were in Wyoming, Colorado and the tri-state area of the Dakotas and western Minnesota where breeding populations of game ducks decreased. (Using the results of air: ground comparison studies to adjust the aerial index figures for birds not recorded provides a different estimate of the change in the Dakotas and Minnesota ; the adjusted data suggest that here there was an increase between 1965 and 1966.) The overall increase between 1965 and 1966 in the game duck breeding population index in table 2 was 23 percent. When these data are adjusted with air:ground comparison data (to adjust the aerial index for birds present but not recorded) the overall increase suggested in the breeding game duck population was 26 percent. The trend of the index values (adjusted with air:ground comparison data) for total game ducks, excluding scoters, eiders, mergansers and oldsquaw, for the period 1955 through 1966 is shown in figure 5. These data show that the 1966 breeding population was 26 percent higher than the all time low observed last year (1965); 28 percent below the peak year in 1956; and 8 percent below the average for the last 11 years.

Several individual species deserve attention. These are the mallard, pintail, redhead, canvasback and blue-winged teal. Trends in the breeding population of mallards are shown in figure 6. Mallards made a good comeback from 1965 with an increase in 1966 of 29 percent. However, this species is still 16 percent below an 11-year average which now includes 6 years (from 1959 through 1965) of very low populations. Pintails, like mallards, showed a good increase in 1966 (about 23 percent) from 1965. However, pintails also are well below their average (29 percent) and have been at disturbingly low levels since 1961. The breeding population of blue-winged teal in 1966 increased from a year ago and appears to be at a level very near its average during the past 11 years. Considering the continental population of blue-wings, it appears that the experimental September season last fall did not have an adverse effect and that the population this spring, given reasonable nesting success, is sufficiently good to warrant the second experimental September season this fall. Populations of redheads and canvasbacks (figure 7) are of interest since we opened the season in 1964 after a closure of 4 years with a limited bag and then in 1965 further liberalized limits on redheads. The data show that both redheads and canvasbacks increased from 1965 and that both are near average levels but still below their peak population years. We conclude that the status of these two species is as good as most other ducks and better than that of mallards and pintails. Geese. Table 3 records data on goose and brant populations from the winter surveys in 1965 and 1966. These data suggest that in 1966 there was a marked decrease in cackling, white-fronted and snow geese in the Pacific Flyway and in snow and white-fronted geese along the Gulf Coast (Central Flyway). Field reports attribute the apparent decreases in these populations of geese to failure of the winter survey crews to find the birds. This explanation seems plausible because when the 1965 winter survey data are examined in light of John Lynch's evaluation of 1965 productivity, as measured by goose family count data and the 1965 harvest data, it is apparent that the 1966 winter survey figures must be

too low. We conclude, therefore, that the decreases referred to above are questionable and there is a chance that the breeding populations of these geese may not be appreciably different than in 1965.

Coots. Table 4 contains data on coot breeding populations.

The overall

trends suggested by these data is one of no change from 1965 but down slightly from the average.

Production indexes and fall flight forecast

Combining breeding population indexes with the results of the July production surveys provides an estimate of change in expected fall flight of ducks from each of the major breeding areas. These forecasts have been expressed in terms of increase, no change or decrease as shown in figure 8.

Starting with Alaska we began the season with an increase of 15 percent in the breeding population. Conditions were not as good as in 1965 but in spite of this there was a small increase in the number of broods and the fall flight is expected to increase in about the same magnitude as the breeding population. In the Northwest Territories and northern Alberta there was an 11 percent increase in breeding population. Although the production survey has not yet been completed, indications are that the fall flight from this area will be considerably larger than last year. In the northern portions of Sakatchewan and Manitoba, both breeding populations and production will be improved somewhat over last year, but to a lesser extent than in the Northwest Territories. There was a 6 percent increase in the breeding population with a commensurate increase in production. Data from Quebec and Labrador are more difficult to intrepret than those from the prairies but it appears that the flight of black ducks should remain about the same as last year.

In southern Alberta the breeding population increased over 1965 by 52 percent. Although the number of water areas decreased, weather conditions were favorable otherwise and a major increase in production is expected from this area. In southern Saskatchewan there was an increase of 52 percent in the breeding population with good weather and an excellent carryover of water through the rearing season. We should expect a major increase in the fall flight from southern Saskatchewan. In southern Manitoba there was only a 6 percent increase in breeding populations followed by a small increase in production over 1965. Water conditions remained good, but the weather was less favorable for optimum production.

In the tri-state area of South Dakota, North Dakota and western Minnesota, habitat conditions continued to improve throughout the season except for a part of southwestern South Dakota. The breeding population was the same as in 1965 but favorable conditions should allow for a small increase in fall flight from this area. For the second consecutive year aerial surveys, comparable with the remainder of the major breeding areas, were conducted in Montana. Although the habitat had deteriorated somewhat in the southeast part of the State, the overall breeding population increased by 27 percent. With adequate water remaining through the brooding season throughout the remainder of the State and favorable weather, a major increase in production is expected to come out of eastern Montana. Among western States contributing lesser numbers, production is expected to be poorer everywhere except in California. Conditions are dry particularly in Nebraska, Wyoming, eastern Oregon and Washington and will result in a smaller fall flight from these areas. In California a small increase in both breeding population and fall flight is expected.

Production data for coots suggest that, because of a very good reason in the Dakotas and western Minnesota, the fall flight of coots in 1966 will probably exceed that of a year ago.

Pacific flyway

FALL FLIGHT FORECAST BY FLYWAYS

Ducks-Because of a general increase in breeding populations in areas supplying the Pacific Flyway and conditions which appear to be right for good production in these areas, there should be an increase from 1965 in the fall flight of ducks to the Pacific Flyway (figure 9). Flights of both mallards and pintails into the Pacific Flyway should be greater this year than in 1965 and these two species will make up more than half of the birds bagged in this flyway.

Although it is judged that the 1966 fall flight into the Pacific Flyway will appreciably exceed that of 1965, the 1966 flight will be less than the average experienced in this Flyway during the last 11 years. This is because major production areas contributing to the flyway, although experiencing generally

good production this year, started out the season with breeding populations of ducks, especially mallards and pintails, that were markedly below the average. (It should be remembered that the last 11 years now include 5 or 6 years with very low duck populations.)

Geese and Brants.-Winter survey figures suggested a very sharp decline in 1966 wintering populations of Canada geese, white-fronted geese, and snow geese compared to 1965. However, as mentioned earlier, the winter survey in 1966 was unsuccessful in locating geese and the 1966 survey figures are probably too low. It is likely that, given a good production year, the fall flight of northern nesting geese to the Pacific Flyway will equal that of 1965. Concerning the Great Basin Canada geese, reports from Idaho suggested that the breeding population was down but production was up somewhat. Reports from Wyoming and Utah suggested record fall flights of Canada geese from these States. In Montana, where a few of the Canadas move to the Pacific Flyway, production was reported to be similar to that of 1965. California indicated that the fall population of Great Basin Canadas would show a slight increase from 1965.

Coots. Breeding populations of coots in areas supplying the Pacific Flyway were similar to those of 1965 but there was an indication that production among coots was better this year. The fall flight of coots to the Pacific Flyway, therefore, may be higher than 1965.

Central flyway

Ducks. Increases in breeding populations, along with good production estimates, in areas supplying ducks to the Central Flyway will result in a moderate increase in fall flight to the Central Flyway (figure 10). The predicted increase should include almost all species.

Here again, the 1966 fall flight, although exceeding that in 1965, will be below average. This is mostly due to lower than average flights of mallards and pintails originating in the southern prairie provinces of Canada.

Geese. Wintering populations of Canada geese were down somewhat in the Central Flyway from 1965. Wintering populations of white-fronts, blue and snow geese showed marked decreases from 1965. These declines, as in the Pacific Flyway, have been attributed to conditions that caused geese to be widely distributed and were, therefore, difficult to count by aerial survey. Consequently, we feel that populations of these birds were not down and with reasonable breeding success in 1966, fall and winter goose populations in the Central Flyway will probably be much the same as in 1965. However, the Tall Grass Prairie Population may be a matter of concern. Coordinated surveys in October in recent years suggest a decrease in this population.

Mississippi flyway

Ducks. It is expected that the fall flight of ducks in the Mississippi Flyway will increase moderately as compared to 1965 (figure 11). Increased populations and productivity, at least as good as last year's, contribute to this conclusion. Also wood duck breeding populations appear to be at least as high as in 1965.

Geese. The winter survey data for Canada geese showed a slight decrease from 1965 but some have attributed this to survey conditions and not to an actual change in the population. Winter survey data discussed earlier suggested that the blue geese in the Mississippi portion of the Gulf Coast held their own from 1965 to 1966 and snow geese showed an apparent increase, but overall Gulf Coast populations of both color phases of this geese decreased. However, because of the questionable validity of the winter survey of 1966 this decrease is probably not real and the 1966 fall flights of these birds will probably be as good as in 1965. White-fronts in the Flyway increased again in 1966 (there was a small increase between 1964 and 1965) and, if this flock of birds is truly distinct from the Texas white-fronts, these data may represent an increase in population.

Coots. The fall flight of coots in the Mississippi Flyway will probably increase in 1966 from 1965. This judgment is based on the fact that the estimated fall flight from Saskatchewan will be above that of 1965; that from Manitoba will be below the flight of 1965; and the fall flight from the tri-state area will be markedly increased over 1965.

Atlantic flyway

Ducks. Data from the experimental breeding ground surveys in Quebec and Labrador suggest that breeding populations may have been down but that production may well have been better than a year ago and, consequently, the fall flight