Impact of channel habitat deterioration on distribution
The gradual encroachment of woody vegetation in conjunction with receding channel width has caused a marked shift in the distribution of sandhill cranes along the Platte River. The present distribution of sandhill cranes is shown in Fig. 7. Although there is a paucity of information on crane distribution before the era of water development the birds probably utilized most of the Platte in south central Nebraska. The river was formerly wide, shallow, and bordered by meadows throughout its length within the study area.
Observations published by Walkinshaw (1956) indicate that by 1954 sandhill crane distribution along the Platte River was largely restricted to the reach between Cozad and Grand Island, a distance of 150 km (93 miles). Large numbers of cranes remained at that time in the Cozad-Lexington Area. By 1979, cranes had abandoned an additional 37 km (23 miles) of the Platte and use of the river channel by cranes in the Overton-Elm Creek section was restricted to a few sites (Fig. 7). High densities of roosting cranes are now centered on staging area 1 eastward from Kearney to a few miles east of Grand Island, in the Elm Creek-Odessa reach of staging area 2, and on staging area 3 between Sutherland and North Platte (Fig. 7).
The distribution of roosting cranes along the river is influenced by both man-made and natural factors. Characteristics of river segments utilized by roosting cranes are listed in Appendix I. Bridges crossing the river and roads immediately adjacent to it have a significant (P<0.05) negative effect on crane use of the river. The mean number of cranes roosting on segments with bridges or adjacent roads, or both, is less than half the mean number on segments without these disturbances. Among 385 0.8 km (0.5 mile) segments of river considered in this part of the study, 61 (16%) had either a bridge or an adjacent road. Within the study area, 35 bridges cross the North Platte or Platte Rivers, and 18 of these occur within the staging areas.
The availability of wide, unobstructed sections of river channel has a highly significant correlation with the roosting distribution of cranes (P<0.001). Over 99% of all cranes observed roosting on the Platte and North Platte Rivers in 1979 were within 0.8 km (0.5 mile) segments with an unobstructed stretch of river at least 50 m wide; almost 70% of the cranes were in segments over 150 m wide. Only 25% of the segments studied had unobstructed channel widths of over 150 m, and 82% were over 50 m wide. The fact that crane use of segments of the river with wide stretches of open water was proportionately higher than the availability of such segments implies a preference for these areas.
The height of vegetation on river islands and along the banks of the main channel also seems to affect cranes use, with the birds apparently preferring to roost where the vegetation is short. However, preference for short vegetation does not appear to affect crane use as much as their preference for wide, open channels. When the two factors are considered together, the relationship is more easily understood. In Table 9, the 385 0.8 km (0.5 mile) segments considered are divided into three categories of unobstructed channel width and five categories of vegetation height. The first parameter indicates the width of the widest, unobstructed channel within the 0.8 km (0.5 mile) segment, and the latter reflects the mean of the predominant height classes of vegetation on the north, south, and island banks within that segment. Each cell of the table, therefore, represents a different combination of channel width and vegetation height. Each cell indicates the percent of all roosting cranes observed and percent of the 0.8 km (0.5 mile) river segments in that category. These values are also summed across the rows and down the columns. A percent use value greater than the percent availability value for that cell implies cranes prefer those conditions, whereas if use is less than availability, this would imply that cranes avoid such conditions. A chi-square analysis strongly rejected (P<0.001) the null hypothesis of a random distribution of cranes.
Cranes avoided all segments of river channel in which the widest available open channel was less than or equal to 50 m, regardless of vegetation height (Table 9, Fig. 19). In fact, although 18% of the 0.8 km (0.5 mile) segments were in this width class, less than 1% of the cranes roosted in these conditions. In the intermediate width category, cranes avoided all segments except where the vegetation was less than 0.5 m, in which situation a slight preference is indicated. Among segments with unobstructed channels wider than 150 m, a preference is seen in all instances, regardless of vegetation height.
The last row of Table 9, which presents the sums of the percents of all widths within each vegetation height class, shows that, when channel width is not considered, cranes tended to avoid segments where the vegetation is greater than or equal to 8.0 m in height, and they seemed to prefer areas where the vegetation is shorter, with the 4.0-7.9 m vegetation height class receiving the narrowest margin of preference (the percent difference between the values). The last column of the table sums the percents of all vegetation height classes for each width category. These data indicate that, when vegetation height is excluded, cranes strongly avoid roosting in segments of the river where the widest unobstructed channel width is less than or equal to 50 m; they also avoid, although not as strongly, segments where the open channel is 51-150 m wide; and prefer portions of the river with open channels over 150 m wide.
In summary, the combined factors of vegetation height and unobstructed channel width strongly influence the roosting distribution of sandhill cranes; they prefer river segments with channels over 150 m wide and exhibit a slight preference for channels 51-150 m wide if the predominant bank vegetation is less than 0.5 m high. Only 26% of the 0.8 km (0.5 mile) segments exhibit these characteristics, but they support 70% of the roosting population.
Land use and food requirements
Continued deterioration of the Platte River channel with further crowding of roosting areas and changes in agricultural practices could have serious impacts on crane food resources. The estimates of crane food requirements (Fig. 12) and data on food availability (Table 5) provide a basis for evaluating the impact of potential future changes in land use.
If woody vegetation encroachment eliminates crane use of staging area 2 and the cranes were to shift to staging area 1, the population of staging area 1 could increase to over 400,000 birds. The population would not be limited by available food resources, however, assuming that cranes are dependent on agricultural food resources (Table 10), that cranes can utilize 50% of the waste corn on each field, that fall tillage continues at the current level of 0-25%, and that native grassland habitat remains available.
Availability of food would change drastically, however, if fall tillage increases or corn production decreases (Table 10). If fall tillage approaches or exceeds 50%, crane food resources would be seriously affected (Table 10). If this occurs, changes in crane distribution, movements, and perhaps physical condition would follow. These data underscore the highly beneficial impact upon cranes of the current practice of grazing livestock on harvested cornfields during winter because it discourages fall tillage and maintains the supply of waste corn. A shift to alternative crops could adversely affect food supply. If corn production is replaced by soybeans, for example, crane food resources probably would decline drastically. The effect of certain other crops would be less deleterious, but none provide comparable quantities of energy-rich food.
Factors affecting trends in use of the Platte River
The number of confirmed sightings of whooping cranes on the Platte River has fallen markedly in recent decades. In the years 1890-1919 there were 17 confirmed sightings on or near the river (Allen 1952). From 1920-49 there were 65 sightings, but from 1950 through 1980 there have been only 5. Possible causes for the decrease in sightings include: (1) a decrease in the amount of effort to observe and report cranes on the river; (2) reduced visibility of cranes while on the river due to encroachment by woody vegetation (3) a change in the population of whooping cranes; and (4) an actual decrease in use of the river by whooping cranes.
It seems doubtful that an actual decrease has occurred in efforts to observe and report cranes along the Platte. Although biologists Swenk, Brooking, and Allen actively solicited firsthand observations from 1912 through 1948, there has been a marked rise in public interest in observing cranes in the Platte River Valley. Personnel of the Nebraska Game and Parks Commission and the U.S. Fish and Wildlife Service also have monitored crane activity in recent years. Although encroachment by woody vegetation in recent decades has markedly reduced visibility on the river channel, this has been offset, at least in part, by improved access to fields adjoining the river. Therefore, it seems probable that whooping cranes have as much chance of being detected now as in years past.
Changes in size of the whooping crane population do not appear to account for the decrease in sightings. The total population of migratory whooping cranes was declining from 1890 through 1941, over which time there was an increase in sightings along the Platte. From 1942 to spring 1980, the wild population which migrates through the Great Plains increased from an estimated 15 (Allen 1952) to 76 birds (K. Johnson, pers. comm.). Yet the number of sightings on or near the Platte River fell markedly during this period.
Available evidence suggests there has been an actual decrease in the use of the Platte River by whooping cranes during the past 30 years. This conclusion is reinforced by the increased number of sightings in areas near the Platte, especially in the Rainwater Basin Area to the south (Appendix J). Although the reasons for this decrease are not fully understood, the factor that appears most likely to have caused reduced use of the channel is the same as that affecting sandhill crane use of the river, i.e., changing habitat conditions brought on by reduced channel width, growth of woody vegetation on islands and channel banks, and increased human activity, particularly adjacent to roads and bridges. Other disturbance factors include the construction and use of Interstate 80 to the north, farming activities in areas which were formerly part of the river channel or adjacent wet meadows, gravel pit operations, and home construction. All of this activity along the Platte River probably has inhibited whooping crane use.
The relationship of the Platte River to whooping cranes
The importance of the Platte River to whooping cranes has been the subject of much discussion in recent years because of increased pressures for diverting much of the remaining flows from the river for various purposes. Additional diversion of flows is likely to cause further habitat degradation, thus concern has been voiced for the welfare of the whooping crane population. To address the issue adequately requires some understanding of how whooping cranes use the Platte River, including an examination of whether the Platte serves as a staging area for whooping cranes similar to that of the sandhill crane population. By staging area, reference is made to an area where migrating birds traditionally congregate and remain for a period of time in order to ready themselves physiologically for the next stage of their migration.
The whooping crane population apparently does not use the Platte River as a staging area because the cranes do not congregate there on a year-to-year basis; nor do they stay in that area for a sufficient period of time to replenish significantly or increase their lipid reserves. It is doubtful that whooping cranes ever used the Platte River as a staging area in the same manner as sandhill cranes do today. Whooping cranes migrate as individuals, pairs, families, and small flocks; they probably never congregate into large flocks in a restricted geographic area as do sandhill cranes. There is little evidence to suggest that individuals or pairs use the same area on the river from year to year. Whooping cranes rarely stay on or near the Platte for longer than 5-7 days, with overnight or two-day stops being the most common. The brief stopover interval suggests that the Platte accounts for a relatively small part of the lipid reserves deposited during spring migration. Whooping cranes probably accumulate substantial lipid reserves during spring migration but do not restrict lipid deposition to a few staging areas on the Great Plains migration route as is characteristic of sandhill cranes. The Platte River appears to best fit the description of a nontraditional stopover for whooping cranes. A nontraditional stopover area is a site which offers suitable habitat but is used on an irregular basis.
Information presented in the preceding discussion does not diminish the need to maintain habitat for whooping cranes along the Platte River. During migration, the species requires access to the types of habitat still available along some reaches of the Platte. Therefore, the river offers the potential for meeting needs of both the sandhill crane population and whooping cranes on a long-term basis. The habitat requirements of the whooping crane and sandhill crane populations for roosting and foraging appear to be similar. This is an important consideration given the rate of loss of other potential stopover sites in the region, especially in the Rainwater Basin Area, and the great importance of this site in the past.
The decreased use of the Platte River in recent times and associated rise in use of the Rainwater Basin Area by whooping cranes is a serious concern. With most of the rainwater basins destroyed, waterfowl and crane populations are confined into the relatively small habitat base that remains. As discussed in a later section, major avian disease outbreaks have occurred in recent years in south central Nebraska. Whooping cranes have been reported at several rainwater basins where large waterfowl die-offs from avian cholera have been recorded (Appendix J). Because sandhill cranes and presumably whooping cranes are susceptible to avian cholera, use of the Rainwater Basin Area as an alternative to the Platte Valley may represent a direct threat to whooping cranes. Twice in the last 6 years (spring 1975 and spring 1979) whooping cranes have been chased from wetlands experiencing avian cholera outbreaks and on one of these occasions the birds settled on the Platte. Limited use of the rainwater basins by sandhill cranes also has been recorded in recent years (Appendix K). In all likelihood, the disease problem will persist in the Rainwater Basin Area, underscoring the need to maintain adequate habitat along the Platte River which provides the species with sites that minimize exposure to avian cholera.
Impact of riverine conditions on distribution and abundance of eagles
Distribution of bald eagles on the study area changed only slightly during the 1978-79 and 1979-80 winters although the magnitude of ice-free channel habitat varied widely between years (Fig. 20). Eagles tended to concentrate in those reaches of river where availability of ice-free habitat was most stable from year to year due to discharges from reservoirs or power plants. The average percent open water ranged from 17%~ in January and February 1979 to 68% during this same period the following year. The 1978-79 winter was the coldest on record in Nebraska, and most of the river was frozen during that year. Ice-free areas were restricted primarily to 2.5 km (1.6 miles) of river channel downstream from Kingsley Dam, a few seepage areas, and downstream from the Tri-County Diversion Canal discharge site. Warm return flows from the canal maintained the south channel in an ice-free state for approximately 13 km (8 miles) in the Lexington-Overton Area. During the mild winter that followed (1979-80), ice-free habitat occurred in nearly every reach of the river (Fig. 20). Slush ice frequently formed in the channel but often melted by late afternoon. Ice break-up occurred by 15 March during both years.
Although the distribution of bald eagles and waterfowl was similar during both years (Fig. 20), numbers of both more than doubled in 1980. An improvement in foraging conditions appears to have been at least partially responsible for the higher use of the study area during the latter year.
Relationship of riverine conditions to food resources
The status of riverine conditions in winter had a marked influence on prey availability to eagles. Food habits data indicate that fish are a preferred item in the diet. Foraging success on fish, however, varies markedly with river conditions. Eagle success at capturing fish is enhanced by low flow conditions and by other effects of power plants, in particular, changes in currents and water temperature. Gizzard shad, a major food item of bald eagles along the Platte River, occur near the surface when water temperature exceeds 7 degrees C and, therefore, are more vulnerable to predation. When the temperature drops below this level, shad seek greater depths in the channel, and are less available to eagles. With cold atmospheric temperatures during the 1978-79 winter, water temperature remained below 7 degrees C and shad were not available. Substantial numbers of shad also became available to eagles after being stunned or killed by the power turbines. Greater access to fish in the south channel downstream from the Tri-County Canal discharge site during the peak water release period caused large numbers of eagles to gather in the area during the 1979-80 winter. For example, 106 eagles were counted along 8 km (5 miles) of river downstream from the canal discharge site on 21 January 1980.
When fish are not readily available, bald eagles shift to alternative prey. The shift in diet under certain conditions could increase the risk of mortality from lead poisoning in the wintering eagle population because of increased predation on mallards carrying lead shot. The extent to which mallards occur in the diet appears related to availability of certain other potential prey. During the 1979-80 winter, for example, with a general lack of snow cover, eagles preyed heavily on jackrabbits. During the previous winter heavy snow cover and limited access to fish caused increased predation on waterfowl. Although many of the ducks eaten by bald eagles were pirated from other raptors, overall mallard losses presumably are to a considerable extent attributable to eagles because other raptors continued to prey upon mallards until their needs were met.
A bald eagle living exclusively on mallards would require approximately 23 mallards per 100 days during winter to meet its dietary requirements. Projecting these data to an overwintering population of 200 eagles and assuming 37% of the diet to be mallards as suggested from egested pellet analyses, a total of 1679 mallards would be needed to meet eagle population food requirements during a 100-day winter period. These data probably are indicative of the most severe conditions to be experienced but do provide insight into the potential degree of impact on this wintering mallard population.
Perhaps most importantly, high predation rates on waterfowl increase eagle susceptibility to lead shot poisoning. The lead shot incidence rate is not known for the Platte River wintering mallard population. However, lead shot were found in nine egested eagle pellets. Most shot were associated with waterfowl remains although one shot each was obtained from remains of a pheasant and a mammal.
Specific recommendations concerning manipulation of riverine flow do not appear warranted at the present time although it is recommended that close monitoring of the wintering bald eagle population be continued.
Impact of changes in hydrology and land use on populations
Before settlement, the Platte and North Platte Valleys of Nebraska contained extensive lowland meadows interrupted only by the broad open channels of the rivers. Reduced flows, the concomitant encroachment of woody vegetation upon the channel, and the cultivation of much of the low grasslands have markedly changed the distribution and abundance of breeding bird populations.
Populations of those species associated with riparian woodland habitat in the Valleys have increased substantially during modern times as this habitat base has grown. In the mid-19th century, woodland was largely absent west of Grand Island, whereas an inventory of riparian woodland along the 203 miles of channel from Kingsley Dam to Chapman (based on August 1979 color photography at a scale of 1:8000) indicated that 21,300 ha (52,600 acres) were present. Fifty-seven species of birds were encountered on wooded islands in the channel including 35 considered to be breeding. Five species reached their maximum density on wooded channel islands (willow flycatcher, Bell's vireo, common yellowthroat, cardinal, and indigo bunting).
One species that has benefited from woody vegetation encroachment is Bell's vireo. This species is a common resident in many riparian systems of the southwestern United States and its range has been expanding slowly northward; breeding pairs are encountered regularly as far north as southwestern Wisconsin. During this study it was found only sparingly in the river systems of the sandhills, most notably in the Loup River System, where it occurs locally in limited numbers. In the Platte Valley, however, Bell's vireo is a common breeding species on river islands with woody vegetation. The breeding population was estimated at about 7000 pairs. A qualitative assessment of the impact of woody vegetation encroachment on the abundance and distribution of various other species occurring in the Platte Valley is shown in Appendix L.
Species that nest on the open sandbars of the Platte River have been affected adversely by the encroachment of woody vegetation. The most profound impact has been on the distribution and abundance of the least tern and piping plover. Both species require broad expanses of unvegetated river channel and sparsely vegetated sandbars. Research on the nesting ecology of the least tern and piping plover reflects their preference for open sites; the microhabitat within a l-m2 area with the nest at the center averaged 72% and 75% bare ground for least terns and piping plovers. Although both species nested exclusively on river islands, the least tern generally nested on the highest ridges. Plovers nested at lower sites (mean depth to moisture at 17 least tern nests was 26 mm and at 39 piping plover nests averaged 10 mm). Reduced flows in the Platte River have resulted in less sorting and moving of the alluvium thereby reducing the frequency at which dry unvegetated ridges occur on the islands.
Originally, least terns were widely distributed on the study area with confirmed nest records from near Lexington and at North Platte. The least tern is now restricted almost exclusively to a reach of river channel bordering Mormon and Shoemaker Islands, a distance of about 3 km (1 8 miles). In 1979, the breeding population was estimated at 30 pairs.
The largest group of species in the Valley to undergo a major population decline in recent times has been those species associated with the native meadows (low grassland). Thirty-five species were found in this habitat type during the study, including 27 that were considered breeders. The magnitude of grassland habitat loss is reflected in the present habitat composition of the three sandhill crane staging areas. Among the estimated original 790 km2 (305 mi ) of grassland habitat on the staging areas, approximately 240 km2 (92 mi2) or 30% remain.
The loss of approximately 70% of the native meadows presumably caused sharp declines in the abundance of the upland sandpiper, bobolink, eastern meadowlark, grasshopper sparrow, and other species breeding in this habitat type. Cropland, the habitat type that is rapidly replacing the native meadows, supports markedly fewer species. For example, corn, which accounts for much of the cropland acreage in the Valley, supported only three breeding species (horned lark, killdeer, and western meadowlark), an 89% decline in number of species present.