BOSTON PUBLIC LIBRARY 3 9999 06317 796 6 RESPONSE OF BREEDING BIRDS TO AERIAL SPRAYS OF TRICHLORFON (DYLOX) AND CARBARYL (SEVIN-4-OIL) IN MONTANA FORESTS JJf UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE Special Scientific Report — Wildlife No. 224 Library of Congress Cataloging in Publication Data Main entry under title: Response of breeding birds to aerial sprays of trichlorfon (dylox) and carbaryl (sevin-4-oil) in Montana forests. (Special scientific report— wildlife; no. 224) Supt. of Docs, no.: I 49.15/3:224 1. Trichlorfon— Environmental aspects— Montana— Beaverhead Na- tional Forest. 2. Carbaryl— Environmental aspects— Montana— Beaver- head National Forest. 3. Western spruce budworm— Control— Montana— Beaverhead National Forest. 4. Birds— Montana— Beaverhead National Forest. I. DeWeese, Lawrence R. II. Series. SK361.A256 no. 224 [QH545.P4] 639'.9'08s [632'.951] 79-607091 NOTE: Use of trade names does not imply U.S. Government endorsement of commercial products. RESPONSE OF BREEDING BIRDS TO AERIAL SPRAYS OF TRICHLORFON (DYLOX) AND CARBARYL (SEVIN-4-OIL) IN MONTANA FORESTS By Lawrence R. DeWeese Charles J. Henny Randy L. Floyd Kathie A. Bobal Albert W. Schultz UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE Special Scientific Report— Wildlife No. 224 Washington, D.C. • 1979 Contents Page Abstract 1 The Study Area 2 The Flora 3 The Fauna 4 Methods 5 Bird Census Techniques 6 Breeding-pair estimates 6 Fixed-station census 6 Statistical Techniques 7 Nesting Studies 7 Casualty Searches 7 Stomach Contents Analysis 7 Occurrence of Spray on Birds 8 Results and Discussion 8 Breeding-pair Estimates 8 Fixed-station Census 11 Nesting Studies 14 Casualty Searches 17 Food 18 Occurrence of Spray on Birds 20 Relation of Insecticides to Cholinesterase in Birds 21 Recommendations for Future Studies 22 Conclusion 22 Acknowledgments 22 References 23 Appendix 25 Response of Breeding Birds to Aerial Sprays of Trichlorfon (Dylox) and Carbaryl (Sevin-4-Oil) in Montana Forests1 by Lawrence R. DeWeese, Charles J. Henny, Randy L. Floyd, Kathie A. Bobal, and Albert W. Shultz U.S. Fish and Wildlife Service Patuxent Wildlife Research Center Laurel, Maryland 20811 Abstract Breeding density, food, nesting success, and mortality of 20 bird species were monitored at Beaverhead National Forest, Montana, in 1975 in conjunction with experimental applications of trichlorfon (Dylox) and carbaryl (Sevin-4-oil) to western bud worms (Choristoneura occi- dentalis). Bird species on nine 350- to 550-ha forested plots (three controls and three treated with each pesticide) were studied before and for 14 days after the spraying of trichlorfon at 1.1 kg in 9.4 L of Panasol AN3 per ha (1 pound active ingredient in 1.0 gallon/acre) and of car- baryl at 1.1 kg in 4.7 L of diesel oil per ha (1 pound active ingredient in 0.5 gallon/acre). No significant decrease in bird numbers was detected from breeding-pair estimates or live bird counts after the spraying. Of the breeding pairs present before spraying, 92% remained on control plots, 89% on trichlorfon plots, and 92% on carbaryl plots. Counts of live birds made before and after spraying in three types of habitat supported the results of the breeding-pair estimates. Nests with eggs or with young at the time of spraying were 74 and 97% successful, respectively, in control plots, 83 and 100% in plots sprayed with trichlorfon, and 86 and 100% in plots sprayed with carbaryl. No sick or dead birds were found after the spraying, although budworms were found in bird stomachs, and tracer-dye from the pesticide occurred on the feathers or feet of 74% of the 202 birds collected. Species dwelling in the tree canopy encountered the dye (and thus the pesticide) at a slightly higher rate (80%) than did species below the treetops (71%) or near the ground and in open areas (70%). Larvae of the western budworm (Choristoneura occi- dentalis) defoliate conifers— primarily Douglas-fir (Pseudotsuga menziesii)—'m the western United States where populations of the budworm are high (McKnight 1968). The U.S. Forest Service tradition- ally used DDT to control the budworm. For example, the Service applied DDT for western budworm control at the rate of 1.1 kg/ha (1 pound/acre) to 810,405 ha in Montana from 1952 to 1959, and at 0.6 kg/ha to 230,850 ha in 1960-1962, and 168,075 ha in 1963. In 1963, the last year in which DDT was routinely used, the Service tested two other insecticides, and since then has investigated various alternative chemicals in attempts to find an effective and environmentally acceptable method of controlling the western bud- worm. Control agents must be evaluated by reliable methods for appraising their environmental effects 'This study was funded by the U.S. Forest Service and the U.S. Fish and Wildlife Service. (Leedy 1959). Once a candidate chemical shows poten- tial to control a target pest, it is first tested in small- scale field experiments and eventually in a pilot control project of larger scale (to simulate operational conditions) before registration for operational control. The results of a pilot control project with trichlorfon (Dylox) and carbaryl (Sevin-4-oil), with particular reference to the effects of the chemicals on 20 bird species, are reported here. Among insecticides used on U.S. forests, trichlorfon ranked first and carbaryl third by volume in 1973; car- baryl ranked first and trichlorfon second in 1974, if one excludes the emergency use of DDT in the Northwest (Fowler and Mahan 1975:50). Trichlorfon (0,0- dimethyl-l-hydroxy-2,2,2-trichloroethylphosphonate) is an organophosphate, systemic insecticide that is effective in the control of endoparasitic and ectopara- sitic insects in livestock. It is toxic to many species of flies, wasps, bugs, and beetles (Matsumura 1975:74). Its lethal dietary concentration (LC50) for young bobwhite quail (Colinus virginianus) on a 5-day con- taminated diet is about 720 ppm (Hill et al. 1975). Car- baryl (1-naphthyl N-methylcarbamate) is one of the most widely used carbamate insecticides (Heath et al. 1972; Matsumura 1975:80). It has wide-spectrum effects on 100 to 150 species of insects, including the gypsy moth, Porthetria dispar (Matsumura 1975:80). The LC50 for carbaryl on young bobwhite quail exceeds 5,000 ppm (Hill et al. 1975). Both chemicals are known to inhibit brain cholinesterase (Zinkl et al. 1977). The present report has two purposes: (1) to discuss the methods and procedures we used to evaluate the impact of these two chemicals on nontarget wild bird populations (published methodology for such studies is limited), and (2) to present our findings on the effects of trichlorfon and carbaryl on bird populations in a pilot study conducted in Montana in 1975. The U.S. Fish and Wildlife Service previously assessed the effects of trichlorfon and other chemical control candi- dates on bird populations during similar field experi- ments with forest insecticides (Pillmore et al. 1971a; Pillmore 1973). The present study was carried out in cooperation with the U.S. Forest Service. Tests of field efficacy for control of western bud- worm require a relatively high population of the insect. Also, the test area, as recommended by the Forest Service, should be representative of the forest type and living communities that could be treated if such chemicals were approved for future use. The Beaver- head National Forest in southwestern Montana met all the requirements for the 1975 pilot project, including a western budworm outbreak in an advanced stage. The highest kill of western budworm is effected when it is treated during fourth, fifth, and sixth in- stars (Flavell et al. 1978); these instars normally occur in late June in the study area. Additionally, this period coincides with increased food demands of forest birds during their breeding season. The adults and young of many forest bird species eat insects, either principally or exclusively, during the breeding period. During this time, the growing western budworm larvae may become an increasingly important food (Gage et al. 1970). A potential route for the contamination of in- sectivorous forest birds arises from their feeding on budworm larvae and other insects that become af- fected by the broad-spectrum sprays. Furthermore, a reduction in food supply during peak demands could retard growth or induce mortalities of nestlings, thus lowering productivity. Consequently, the time when entomologists believe the budworm is most sus- ceptible to the insecticide is also a period of high vulnerability for birds. In addition to following the two approaches recom- mended by Stickel (1973) for studies of organophos- phates and carbamates— searches for sick or dead birds, and analyses of brain tissue or blood for cholin- esterase inhibition— we determined the nesting suc- cess (the death of breeding birds or their departure from the area would result in lowered nesting success), censused live birds, and determined the food habits of resident birds. These techniques enabled us to monitor an important physiological variable, the population numbers, and the reproductive performance of the 20 most common species. This multi-faceted approach was used to reduce the chances of overlooking impor- tant effects of the spray applications on birds. The common and scientific names of plants and ani- mals mentioned in this report are given in the Appen- dix (Table A- 1) The Study Area The test area lies in southwestern Montana within Madison County in the Tobacco Root Mountains and Gravelley Range. The nearly continuous mountain ranges extend from Butte south to West Yellowstone, Montana, and are bordered on the west by the Ruby and Jefferson rivers and on the east by the Madison River (Fig. 1). The central base for cooperative opera- tions was established at Ennis, Montana. Butte • / Two N^~>JL Forks 1 • ) Bozeman / • \ Ennis* / Dillon vr N. 40- 5 \ M=J N^^J 3 x§ nk\ *) 7 f ai\ ^ West -Yellowstone Montana j E * 0 20km N Fig. 1. Study areas showing drainages and placement of major study plots (1-9). Nine 350- to 550-ha major plots were established at 1,890- to 2,380-m elevation in drainages on the east and west slopes of the two mountain ranges (see Table A-2 for plot descriptions). Hereafter, the term major plot (MP) is applied to one of these large areas (sprayed or unsprayed), and the term subplot (SP) is applied to an 8.1-ha area established within each major plot that was used for studies of breeding pairs. Fixed stations (FS) refer to localities inside the MP's, but separate from the SP's, where birds were counted while the ob- server stood still. Spring weather in 1975 was unusually cool; rain fell in the lower valleys and snowfall continued in our study area through mid-June. Plant development was apparently 2 to 4 weeks later than in most other years. However, deciduous woody species were fully leafed by about 1 July, 10 days before treatment began. The early-season cool weather may have delayed entry of migrant birds into the study area. After the weather changed to a more typical spring and summer pattern, near mid-June, several species— including western tanagers which had still been congregated at lower ele- vations—moved into higher elevation breeding areas. By late June most species of breeding birds were nest- ing. Near-normal summer weather continued through June and July, except for frequent afternoon showers and cool nights that slowed the growth of buds of co- niferous trees. As a result, the western budworm, whose growth depends on bud developments in Doug- las-fir, did not reach "treatable" instar stages until 10 July. Spraying began on that date. The Flora Vegetation in the study area is a mosaic of forested and nonforested habitats (Fig. 2), dominated by tim- bered types with considerable nonforested clearings. Vegetation in the clearings included sagebrush, grasses, and forbs (collectively called grassland, Pfister et al. 1977). Timber on the five MP's on the east side of the Gravelley Range (1, 2, 5, 8, and 9) were logged sporadically 30 years or more ago (Virgil Lind- say, personal communication), but were well estab- lished forests during our study. Among ' the four remaining MP's (3, 4, 6, and 7) on the west side of the range, some timber in parts of MP 4 was cut during 1959-62; the other plots had no history of recent cut- ting (John R. Hook, personal communication). A program of controlling sagebrush with herbicides in MP's 1, 2, and 4 during 1968-70 killed some of the sagebrush, but irregular, apparently untreated patch- es were present during our study. The other six MP's (3 and 5-9) had no history of sagebrush control (John R. Hook and Virgil Lindsay, personal communi- cation). Some MP's supported nearly dead but regen- - » i*--\ >_, ]3cz jls^ A Fig. 2. A mountainous area where major plots were located, showing the pattern of vegetation distribution. The forest has a considerable number of openings, typical of plots 3, 4, 6, 7, and 9. Plot 7 is in the immediate foreground. erating stands of sagebrush, whereas sagebrush in other plots appeared healthy, or partly decadent and partly healthy. We established the FS's in dead, regen- erating, or vigorously growing sagebrush stands, with- out considering their condition as a criterion for selec- tion of the stations. Douglas-fir occurred in continuous stands on slopes facing north and east but was sparse on other slopes. It was the dominant tree species, although it inter- graded with lodgepole pine, which replaced the Douglas-fir at the upper elevations (Fig. 3). An occa- sional Umber pine was found on rocky, xeric slopes with southern exposures at lower elevations. No pon- derosa pine occurred in the study area. Species of true firs and spruce were generally less common, growing mostly in mesic sites restricted to stream bottomlands and spring-fed areas. Woody riparian habitat was limited to a narrow belt of willow and birch along the small permanent streams that ran through all MP's except No. 5. Quaking aspen (Fig. 4) was common in all MP's except No. 8. Aspen was usually in small homogeneous groves; extensive stands were found only in MP 7. Aspen was most prevalent at the interface of the upper elevation limits of grassland and sagebrush areas, where coniferous stands begin, as well as along streams and in wet areas (Fig. 2). Wet areas were characteristically in the stream bottoms and along the lowest elevation of the MP's. An extensive description of the dominant Douglas-fir and lodgepole pine climax series in this area, including the understory associa- tions with these forest types, was given by Pfister et al. (1977). Because sagebrush-covered openings occurred on all MP's, and were considered important, we located SP's Fig. 3. A slope of South Meadow Creek with a northern expo- sure, showing dense Douglas-fir forest (foreground) blend- ing into a lodgepole pine forest (background). This combi- nation is more typical of subplots (SP's) 1, 2, 5, and 8 than of the others. Subplot No. 2 for breeding-pair surveys included the sagebrush, aspen, and Douglas-fir habitats in the foreground. and FS*s in this habitat: it closely resembled the big sagebrush-Idaho fescue series described by Mueggler and Handl (1974:32). Major plot 5 was slightly atypical in that it had only small sagebrush areas available for bird counts. Habitat for bird studies was similar in all MP's, although the relative proportions of nonforest types varied among plots. We recognized three broad habi- tat types for stratification of the FS bird census, although we realized that our habitat characterization was somewhat artificial. The three habitat types were delineated by their characteristic overstories: (1) Douglas-fir— all conifer stands. (2) aspen— mostly pure aspen, and (3) sagebrush. These habitat types were common to all MP's, which had varying understories of grasses, forbs, and some shrubs. The three habitats of SP 2 shown in Fig. 4 are representative of the habi- ----•■.... Fig. 4. Interior of the subplot (SP) on major plot (MP) 2, showing typical habitats of Douglas-fir on the north slope (left and background) and aspen and sagebrush on the south slope (right and foreground). tats in the other SP's. A more extensive description of each treated MP was given by Flavell et al. (1978). The Fauna The Beaverhead National Forest is well known for its populations of big game and upland game (Mussehl and Howell 1971) and its fisheries (Brown 1971). And, as in most forest ecosystems, abundant insects support a rich diversity of birds. The abundance of some insectivorous bird species, and perhaps their predators as well, may have been increased as a result of the high population of western budworms. Elk, moose, and mule deer were abundant on their winter range in the lower valleys and foothills and dis- persed in early May through the MP's and upward into their summer range as the snow cover receded. Cattle that were herded into the MP's near the beginning oi the postspray counts rapidly reduced much of the dense grass and forb understory. These wild and domestic animals trampled some ground nests that we monitored. We primarily studied birds nesting during June and July, but considered other species as well. Most were migratory. Of the 31 most abundant species (2 Falconi- formes, 2 Piciformes, 27 Passeriformes) 4 normally arrived in March, 7 in April, and 17 in May (Skaar 1969). The other three species were residents, but their numbers increased during the breeding season. We were unable to assess the effects of the spray program on the only game birds present— blue grouse, ruffed grouse, and mourning dove— because they were either too scarce or nested too early. Peak egg laying and incubation of most species was in late June (Fig. 5). The fledging of young began in late June, peaked in July at about the time the MP's were treated, and ended by early August. Nesting of most passerine species begins in May and lasts about 21/2 months, being restricted to a short period when food is abundant at this latitude and altitude. How- ever, the dark-eyed junco, which winters in the area, and the American robin, which arrives early, both commonly nest twice per season. This was evidenced by robins on eggs in their first nests by late May (even during snowstorms), and juncos that built their second nests in middle to late July. May June July August Fig. 5. The phenology of nesting of 27 forest breeding species studied. The solid line represents frequency of nest ini- tiations (1 day before first egg laid) and the dotted line the frequency of hatching. All of the 247 nests that were started are included, regardless of whether they were successful. Vertical lines indicate the 10-17 July period when the six major plots of the study area were sprayed. Each tick on the X axis represents 5 days. Methods The experimental design was completely randomized with three treatments (trichlorfon, carbaryl, and control) and three replicates. Three of the nine MP's were sprayed aerially, through 18-8010 Tee Jet noz- zles, with carbaryl diluted in diesel oil 1:1 at the rate of 1.1 kg in 4.7 L/ha (1 pound active ingredient in 0.5 gallon/acre) and three were sprayed through 37-8010 Tee Jet nozzles, with trichlorfon diluted in Panasol AN3 oil 1:3 at the rate of 1.1 kg in 9.4 L/ha (1 pound active ingredient in 1.0 gallon/acre). No adhesive sticker was used in the formulations. An automate-red dye that persisted for several days was mixed into both formulations by the U.S. Forest Service for the assessment of spray coverages. Three MP's— the controls— were not treated. The spray was applied from a Bell 205 A-l helicopter fitted with a pump capable of maintaining 2.8 kg/cm2 pressure in the spray boom during spraying. The 15.2-m spray boom fitted on the helicopter yielded an effective swath width of 61 m. Spraying began at dawn and was completed by 1000 h; only one MP per day was sprayed. Spraying of plots extended from 10 to 17 July (Table A-3); rain pre- vented spraying on 2 of the 8 days. All plots were sprayed when winds above the forest canopy were less than 10 km/h (Scuderi et al. 1975), and the spray ap- peared to fall into the forest with little noticeable drift. A broken spray-boom hose caused a spill of trichlorfon over MP 5. However, the spill was corrected after one swath of undetermined length. Although some of the spilled chemical may have fallen on parts of the SP's and FS's on MP 5, most of the excess spray apparently fell 400 to 600 m away. We are uncertain whether the spill had any unusual effects on birds in our study areas in MP 5. During May and June we established a rectangular 8.1-ha (402 x 201 m) SP in each of the nine MP's, using a staff compass and surveyor's tape. Counts of birds on these SP's provided estimates of the density of breeding pairs (Fig. 6). Inasmuch as breeding bird density and species richness decrease progressively from heterogeneous mixtures of deciduous and conifer types along moist creek bottoms, up the slopes into dryer homogeneous conifer stands (Balda 1975), we established each SP along the principal stream course to insure the highest possible breeding bird density and number of species. The principal stream occurred near the edge of some MP's (4, 6, and 7) rather than through the middle as in MP's 1-3, 5, 8, and 9 (Fig. 6). Douglas-fir, aspen, and sagebrush habitats occurred in each SP. In each SP we established 66 points, each marked by an orange-painted wooden stake (122 cm high), along a grid at 40.2-m intervals (Pillmore 1973:145). This grid was duplicated on field forms for the recording of bird observations at precise locations within each SP. In addition to having one 8.1-ha SP within each of the 9 MP's, we established a separate series of 9 to 15 roughly circular or oval FS's in each MP (Fig. 6). The FS trend counts made at these plots were similar to those made by Fowler and McGinnes (1973). These FS's were stratified by dividing them among Douglas- fir, aspen, and sagebrush habitats (Table A-4) to reduce the variance contributed by unequal bird densi- ties in different habitats. We monitored all three habi- tat types because we expected different habitats to support unlike species and unequal densities of birds (Kendeigh 1944:100). One to five FS's were marked with surveyor's flagging at their central point and Fig. 6. Nine major study plots (MP; see Fig. 1 for locations), showing relative size, shape, relation to drainage, direction of drainage; and 8.1 -ha subplots (SP; solid rectangle); and fixed-station plots (FS; dots). Plots 1, 3, and 9 were con- trols; 4, 5, and 7 were treated with trichlorfon; and 2, 6, and 8 were treated with carbaryl. periphery in each of the three identified major habitats in each MP. Adjacent FS's were all more than 91 m apart, thus providing nearly independent counts at each FS. The numbers of FS's varied from 9 to 15 per MP and the average areas from 0.09 to 0.74 ha, depending on the visibility through, and availability of, habitats (Table A-4). The dimensions of each FS were paced, and the area was calculated. Because counts at FS's were scattered throughout the MP's (except for MP 8), birds counted at FS's generally were farther inside the MP's (i.e., nearer the center of the sprayed areas) than were those in the SP's (Fig. 6). Bird Census Techniques Breeding-pair Estimates Two types of systematic living-bird censuses were made in the MP's during the prespray and postspray periods. The first was a 2-h breeding-pair count on each 8.1-ha SP, which began at sunrise (0540-0610) on each count day (see Table A-3 for count days). An observer walked a standardized route, counting all birds either seen (stationary or flying) or heard (terri- torial or nonterritorial), and plotted them on the 66- point grid form for the SP. Weather data, bird behav- ior, nest locations, and the presence of sick or dead birds were also recorded. We used the breeding-pair mapping method (Williams 1936) for an estimate of breeding pair density (see also Hall 1964 and Svensson 1970). Subjective biases inherent in the breeding-pair map- ping technique (Best 1975) were considered: each ob- server plotted species maps for his or her three SP's, but one person interpreted all of the species maps and determined the numbers and locations of breeding pairs. We recognized that this technique, at best, may reveal only about one-fourth of an actual population reduction (Robbins 1963), and that rapid repopulation may mask any actual reduction (Stewart and Aldrich 1951). Because of this shortcoming, we also employed the FS census as a second method. Fixed-station Census The second census was a count of birds from the center of each of the total of 114 marked FS's (Fig. 1), which provided a separate estimate of bird abundance and species composition. Counts at the stations began 2 h after sunrise (0740-0810) and took 2 to 3 h to com- plete. Stations were occupied by the counters in the same order on each count day. Upon arrival at the center of each station, the observer stood motionless for 1 min to allow for the resumption of bird activity, and then recorded individuals of all species from this central point for 5 min. The locations where birds were seen were plotted on duplicated sketch maps of the FS plots. A bird that moved about during the count was tallied as one bird. Preliminary trials indicated that often no birds were tallied inside the flag-marked FS's. These zero counts were attributed to a number of factors: the shyness of birds; the inevitable smallness of FS's caused by the sparseness of some habitats (aspen and sagebrush) within the limits of the MP boundary; obstruction to bird observation imposed by large trees (Douglas-fir and aspen); and the necessity to have areas set aside within MP's for the collection of live birds by other researchers for cholinesterase studies. Consequently, to counter the low numbers of birds tallied inside, we combined the birds inside FS's with those seen or heard outside the FS's and categorized our obser- vations into two primary groups: (1) birds registered only inside the boundary of each FS, and (2) total birds— the sum of birds inside and outside the bound- ary. Both primary groups included birds either seen (sta- tionary) or heard (territorial or nonterritorial). Only birds perched either within the flagged FS boundary or in the outside area were counted. The maximum dis- tance from the center point to the point at which birds were recorded did not exceed about 75 m (station radii were 10 to 60 m). This distance was arbitrary, and varied because of differences in habitat and topog- raphy. However, two rules governed this type of count. First, the same bird detected at the same loca- tion from two adjacent stations was recorded at only one station. Independency of counts at all stations was required for the statistical analysis; counts at each station were to represent different birds. And, second, once the counts began, the observers made maps and plotted landmarks marking the outer boundaries of the "outside area" for each station, and restricted counts to this same outside area throughout the study. Data from this census were treated as three inter- related numerical values from each of the three desig- nated habitat types: (1) a crude estimate of number of birds per hectare from birds inside the station bound- ary, (2) an estimate of numbers of singing males from inside and outside the station boundary combined, and (3) an estimate of total bird numbers, in which all birds (males and females) inside and outside the station periphery were combined. These three estimates were used in prespray and postspray comparisons. Esti- mates (2) and (3) included birds in the designated habitat, inside the fixed stations, and also from adja- cent habitats, which sometimes differed from that of the stations; thus, unlike estimate (1), they included birds from more than one kind of habitat. plots when plots of like treatment were combined for comparison. We scheduled postspray counts so that all plots were censused nearly the same number of times on days representing about equal post- treatment intervals. All postspray counts began on the 1st day after treatment and were repeated at inter- vals 2-4, 6-7, 9-11, and 13-14 days after treatment. Cen- suses on control plots were divided into prespray and postspray periods similar to those on the sprayed plots. Five censuses were made on all SP's and FS's in the 2 Vi-week period before and after the insecticide application. Nesting Studies To supplement the census data, we recorded bird nests seen in each of the MP's. All nests located were monitored to determine fledging success. Nests in cav- ities were observed with cavity viewing devices de- scribed by De Weese et al. (1975) and Seidensticker and Kilham (1969). Nesting success information was gathered only from nests that were active (i.e., that contained either eggs or young) on the date of spraying. No attempt was made to evaluate the few nests that were initiated after the spraying dates. Statistical Techniques Casualty Searches The methods for analyzing many of the data pre- sented in this report are discussed in later subsections. Fixed-station bird counts presented a particular problem common to studies in which animal numbers are counted repeatedly in the same place over a period of time. Two sources of variation involved in com- paring prespray with postspray bird counts con- founded the detection of treatment effects: an elapse of time and pesticide treatment. We removed the time factor from our FS data by analyzing 27 values (3 treatments x 3 habitats x 3 replicates), derived by sub- tracting the postspray mean from all stations in a given habitat, and in a given plot, from the correspond- ing prespray mean. We then used a split-plot 3x3 fac- torial analysis of variance (Steel and Torrie 1960:236) to analyze the data. Three observers were each assigned three MP's, each with a different treatment (control, trichlorfon, and carbaryl); only one observer worked in any given MP during the study. Bird censuses were conducted during the morning and nest studies and carcass searches during the afternoon. One plot that received each treatment was surveyed on each count day from mid-June through July (Table A-3). This procedure tended to reduce the effects of weather, observer, and We searched for "sick" birds showing toxic signs, and dead birds, during the prespray and postspray periods. Hours expended on this effort were recorded for each plot. Birds found sick or dead were also noted during the early-morning censuses. Stomach Contents Analysis Stomachs of 183 birds of eight species that were col- lected for a cholinesterase study were retained for food content analysis. Birds were collected alive with shot- guns during the morning at points adjacent to the MP's before treatment, and in the MP's after spray- ing. Birds were not collected in treated MP's any closer than 90 m to an SP or FS and never on census days. Birds were kept on dry ice or in a freezer until the stomach contents were removed. Food items washed from the esophagus, proven- triculus, and ventriculus of each bird were analyzed collectively as stomach contents. Plant, animal, and grit materials were sorted and visual estimates of the percentages by volume were made as described by Crase (1977). The percentage of the total number of insect food items was measured by counting individual 8 insects (Martin et al. 1951). It was sometimes neces- sary to determine the number of insects by counting body parts (e.g., the total parts that come from dif- ferent insects equalled the minimum number of insects represented by those parts). Where actual counts were not possible because of advanced digestion, insect numbers were estimated. Most insects were identified to order, and the western budworm to species. Occurrence of Spray on Birds Birds collected for cholinesterase studies for the 5 days during and after the spraying were examined for automate-red dye. Comparisons among the bird species for exposure to the dye was made on the basis of the forest canopy level that they usually inhabit. Results and Discussion Breeding-pair Es tima tes Sixty-six bird species were observed on at least one of the 8.1-ha SP's and 34 species were found breeding on nearly all SP's. Of the breeding species, 20 (1 con- sisting of 3 species of flycatchers combined) exhibited territorial or breeding behavior and were numerous enough to support comparisons between treated and control plots (Table 1). Evening grosbeaks and pine siskins were abundant on all plots, but their erratic and wide-ranging movements precluded a usable esti- mate of breeding density. The postspray estimate of breeding pairs for the 20 selected species were 92% of the prespray estimate in the control plots, 89% in the trichlorfon plots, and 92% in the carbaryl plots. Post- spray estimates were consistently lower than the pre- spray estimates, but these decreases occurred on treated as well as untreated plots and were not sig- nificant (P = 0.50; Steel and Torrie 1960:101). The 20 species were also categorized by their spatial feeding habits (Thomas et al. 1975; Bent 1939, 1942, 1946, 1948, 1949; Martin et al. 1951; present study, Table 2 and Table A-l). Simple broad feeding cate- gories were used, so that we could easily classify a species and yet maintain enough ecological difference between it and others for comparison. In this approach we assumed that the farther below the treetop level a species foraged and the greater the distance and amount of foliage between the spraying device in the helicopter and a species' activity area, the less the species would be exposed to the spray. Changes in postspray breeding-pair estimates, when grouped by feeding habits, were observed in control and sprayed SP's (Table 2). In categories containing more than 20 pairs before the treatment, the total numbers of breeding pairs of aerial feeders decreased 27% in control plots and 24% in the trichlorfon plots (Table 2). The only other categories that indicated noteworthy postspray changes were tree-canopy feeders on all SP's (12 to 18% decrease) and understory feeders on control SP's (43% increase). Postspray changes in other categories were either minor or the samples were too small for meaningful analysis. Changes in total breeding-pair estimates after spray- ing were slight, but some species seemed to decrease more than others. A decrease in density of a particular species could be more severe than for other species. Prespray and postspray breeding-pair estimates for the 12 most abundant species are presented in Fig. 7. The eight other species included in the breeding-pair count were not abundant enough for comparison. The Empidonax group accounted for 100% of the post- spray drop of the two aerial feeders (Empidonax spp. and tree swallow) in control and trichlorfon plots; tree swallows decreased only in carbaryl plots. The yellow- rumped warbler, warbling vireo, and western tanager accounted for about 90% of the decrease of tree- canopy feeders in trichlorfon plots. The disappearance of mountain bluebirds after carbaryl treatment and of white-crowned sparrows and Cassin's finches on control plots after the spraying was noticeable, but densities of these birds were too low to suggest mean- ingful comparisons. The increase of lazuli buntings after treatment probably reflected their late arrival on the study area. Postspray censuses indicated that no species that was abundant before spraying was deci- mated by either of the treatments. For three species (chipping sparrow, yellow-rumped warbler, and house wren) the estimated number of breeding pairs after treatment was reduced in treated plots, but not in control plots. When averaged over replication, estimated numbers of postspray breeding pairs of these three species were lowered by 14, 40, and 14% in trichlorfon plots and 6, 8, and 28% in carbaryl plots, compared with no decrease in control plots (Table 1). Pooled postspray estimates of the numbers of yellow-rumped warblers decreased more in trichlorfon plots than did any of the other 12 selected species in all other plots. Further, SP 5 accounted for 50% of the yellow-rumped warbler pairs that were missing after spraying in the trichlorfon treated plots. The treat- ment of MP 5 possibly had a greater impact and also a greater general effect on birds than that of any other treated plot. Not only did the observer note a general decrease of bird activity there after treatment, but breeding-pair numbers dropped 29%, compared with 0- 12% in all other plots (Table 1). However, the spray treatment of MP 5 differed from all other plots: treat- ment was made on 2 consecutive days, and, as men- tioned earlier, a spillage from the aircraft occurred Table 1. Estimated pairs of breeding birds of the 20 most common species on the nine 8.1-ha treated and un- treated subplots, based on five prespray and five postspray counts." Prespray rank of Control Trichlorfon Carbaryl Prespray Postspray Prespray Postspray Prespray Postspray Species abundance 1 3 9 1 3 9 4 5 7 4 5 7 2 6 8 2 6 8 Common flicker 19 0 1 1 0 1 1 1 1+1 1 0 2 P 1+P P 2 0 Yellow-bellied sapsucker 16 2 P 1 2 1 1 1 1 1 0 0 1 1 0 1 1 1+0 Empidonax spp. 2 7 5 6 4 4 4 7 4 3 4 3 2 7 6 5 9 4 6 Tree swallow 12 2 1 1 2 1 1 2 1+4 2 0 5 1 3 P 0 2 0 Mountain chickadee 7 2 4 4 2 4 3 4 3 3 3 2 4 3 4 5 2 4 2 House wren 10 2 2 2 3 2 3 3 2 2 3 1+ 2 3 4 0 2 3 0 American robin 4 8 3 3 5 3 6 6 2 3 8 1+ 4 6 8 3 4 6 5 Hermit thrush 16 0 2 1 0 1 0 P 1 0 0 1 P 1 0 3 1 1 3 Swainson's thrush 13 4 1 1 3 0 1 1 0 1 1 0 1 2 0 3 2 0 3 Mountain bluebird 20 1 0 0 1 P 0 1 P 1 0 0 1 0 1+ 0 0 0 0 Ruby-crowned kinglet 8 1 4 4 0 2 3 3 2 P 2 3 P 3 4 3 3 2 2 Warbling vireo 1 7 3 7 4 3 6 8 5 6 9 4 3 9 6 4 10 6 3 Yellow-rumped warbler 6 4 4 5 4 4 5 6 3 1 5 1+P 5 3 4 5 3 3 MacGillivray's warbler 13 0 0 2 1 1 2 2 2 2 3 2 0 2 0 3 1 0 4 Western tanager 9 1 3 3 2 3 3 2 2 0 0 2 0 5 2 3 5 3 3 Lazuli bunting 11 3 0 1 5 0 3 1 2 0 3 3 0 4 2 2 4 2 3 Cassin's finch 18 P P 1 0 0 0 2 P P 1 0 2 2 2 0 1 2 0 Dark-eyed j unco 5 4 4 4 4 6 2 6 2 5 6 2 5 3 6 3 3 6 2 Chipping sparrow 3 3 6 5 4 6 4 6 5 3 4 4 4 5 6 5 5 6 4 White-crowned sparrow 15 1 1 0 0 P 0 2 0 3 3 0 3 2 0 0 2 P 0 Plot total 52 44 52 46 42 48. 64 38 39 58 29 39 64 58 47 60 53 43 Grand total 148 136 141 126 : 169 156 aP = species that occurred on a plot in insufficient numbers for pair determination; 0 = species that did not occur on a plot; + = at least one observation suggested an additional pair, but the evidence was insufficient to allow inclusion among full breed- ing pairs; of 180 possible instances of occurrence (9 plots x 20 species), 10 species (none of which ranked higher than 8th) were absent from one to three plots in 16 instances. bProbably includes Empidonax traillii (willow flycatcher), E. hammondii (Hammond's flycatcher), and E. oberholseri (dusky fly- catcher); the dusky flycatcher was probably the most common. Table 2. Estimated numbers of breeding pairs of 20 selected species, grouped by spatial feeding habits, in sub- plots (SP's) with differing treatments, and (in parentheses) percentage change after treatment. Spatial feeding habit8 Treatment Number of Control Trichlorfon Carbaryl species Prespray Postspray Prespray Postspray Prespray Postspray 2 22 16 (-27) 21 161-24) 22 21 (-4) 6 57 481-16) 50 411-18) 67 591-12) 1 3 4 ( + 33) 3 1 (-67) 2 2(0) 4 14 20 (+43) 21 23 ( + 10) 22 21 (-4) 6 51 47 (-8) 44 44(0) 55 53 (-4) 1 1 1(0) 2 1 (-50) 1 01-100) 20 148 1361-8). 141 126(-11) 169 156(-8) Aerial Tree canopy Tree trunk Understory Ground Air-Ground Total "Species in relation to the spatial feeding habits designated here are given in Table A-l. 10 Warbling vireo Chipping sparrow Dark-eyed junco Mountain chickadee Western tanager Tree swallow Empidonax spp. American robin Yellow-rumped warbler Ruby-crowned kinglet House wren Lazuli bunting 20 10 20 10 20 10 20 10 20 10 20 10 20 10 20 10 20 10 20 10 20 10 20 10 1 ill ill Ji Jl ill i i ■ i ■ 1 i ■ I 1 ■ ■ ■■J: c o O c o **— k_ o o CO n k- O Fig. 7. Estimated numbers of breeding pairs of the most abundant species found in the trichlorfon and carbaryl treated plots and controls (three plots for each treatment) before spraying (solid bar) and after spraying (stippled bar). during treatment. Other investigators of the effects of trichlorfon and carbaryl on wild bird populations have used various techniques, sizes of treated areas, chemical formu- lations, rates of application, and a variety of vege- tative types (Table A-5). Only one of the studies have shown measureable effects of treatment: Moulding (1976) reported that mean numbers of forest birds, based on transect counts, declined significantly by 2 weeks, and were de- pressed for as long as 4 to 6 weeks after the forest was sprayed with carbaryl (1.1 kg active ingredient per ha in water with pinolene sticker). Canopy-feeding birds tended to be more affected than ground-feeding species. The bird populations had not recovered 8 weeks after spraying, presumably because of a reduc- tion in the food supply in the forest canopy. Our breeding-bird data suggest that canopy-feeding birds were not significantly sensitive to either insec- ticide, but that they were more sensitive to trichlorfon than to carbaryl. Other studies have not shown significant effects on forest birds: although data shown by Pillmore et al. (19716) suggested a decline in breeding pairs of the yellow-rumped warbler and ruby-crowned kinglet (both canopy dwellers) in a forest area sprayed with trichlorfon, his results in general were inconclusive because the area had been treated with synthetic pyre- thrum (2.2 kg active ingredient per ha) in the preceding year. Also, breeding pairs of both species decreased after the spraying by 67 and 28% in two plots treated with trichlorfon but likewise decreased 28% in one control plot. Doane and Schaefer (1971) reported a reduction in the numbers of 13 bird species heard singing on 20-ha plots in an urban forest in Connecticut treated with tri- chlorfon or carbaryl (1.1 kg active ingredient per ha). They concluded that a 99% kill of gypsy moth larvae forced birds to forage outside the sprayed plots, thereby affecting the singing rate or the numbers of singing males inside the plots, or both. One consid- eration that is overlooked is the possible response of territorial males when food in existing territories is drastically reduced. We believe that lower food sup- plies within a bird's normal foraging range may have far-reaching effects on singing activity, intraspecific strife, and overall productivity. There is evidence, at least for mammals (Boyd and Taylor 1969), that a lowered dietary intake of protein increases sus- ceptibility to some insecticides. Conner (1960) found no effects of carbaryl (1.4 kg active ingredient per ha) on numbers of common song- birds and singing locations of male songbirds in a New York mixed forest after treatment. No adverse effects on birds were observed by Richmond et al. (1979) when carbaryl (2.2 kg active ingredient per ha) was applied in forests of northeastern Oregon— although 2 of 55 11 Table 3. Means (±95% confidence intervals in parentheses) of 5-min bird counts made at fixed stations (FS's) before and after the treatment. Prespray Postspray Within Within Habitat Number fixed fixed type and of Areab station Total Territorial station Total Territorial treatment counts3 (ha) boundary0 birdsd malese boundary0 birdsd malese Douglas-fir Control 70 5.5 5.6(1.3) 9.8(0.7) 5.1 (0.7) 7.0(1.7) 11.1(0.8) 6.4(0.5) Trichlorfon 75 6.3 4.2(1.0) 8.0(0.7) 3.6 (0.5) 2.9(0.8) 8.0(0.8) 4.2(0.6) Carbaryl 65 5.5 4.1(1.0) 10.2(0.8) 5.4(0.8) 4.0(0.8) 10.9(1.0) 6.0 (0.6) Aspen Control 70 3.8 15.3(4.2) 8.6(0.8) 4.0(0.5) 10.2(3.1) 8.7(1.0) 4.7(0.5) Trichlorfon 65 3.6 21.9(5.8) 10.3(1.0) 4.5 (0.6) 15.1(4.5) 10.2(1.0) 5.2(0.6) Carbaryl 45 3.2 11.4(2.3) 10.2(1.2) 4.9(0.7) 7.8(1.5) 10.2(1.0) 5.4(0.7) Sagebrush Control 65 6.7 3.6(1.0) 10.4(1.0) 5.5(0.6) 2.1(0.7) 12.1(0.8) 7.310.6) Trichlorfon 55 7.0 2.1(0.7) 10.9(1.0) 6.1(0.8) 1.6(0.8) 10.3(0.9) 6.0(0.7) Carbaryl 60 8.3 2.6(0.8) 11.1(1.0) 5.4(0.7) 2.0 (0.6) 11.2(0.9) 6.2(0.7) aTotal number of counts made in all of three plots per treatment before and after the treatments. bArea of all fixed stations totaled for three plots by habitat. cBirds counted per hectare within fixed-station areas. dIncludes complete fixed-station bird counts, within and outside the fixed-station area. eIncludes birds counted within and outside the fixed-station area. birds showed reductions in brain cholinesterase. Car- baryl was observed to have little effect on a variety of bird species in wetland (McEwen et al. 1963, 1964) and grassland (Finley et al. 1963; McEwen et al. 1972) eco- systems. Fixed-station Census The total-birds variable was used to analyze counts from FS's because it offered larger numbers (Table 3), fewer zeros, and a lower coefficient of variation. Data were analyzed in two ways. First, all recorded species were tested, which resulted in a nearly significant F value (P = 0.10; Table 4). Also there was no significant interaction between treatment and habitat (P = 0.48), and the response to treatment was not different among habitats (P = 0.49). The mean differences expressed as the grand mean numbers of birds per habitat prespray minus the grand mean number of birds per habitat postspray, over replication, were not significantly different between control and treated plots. Negative numbers in the following comparison indicate a postspray in- crease and positive numbers denote a postspray decrease. Control Trichlorfon Carbaryl Douglas-fir Aspen -1.39 0.34 0.08 0.00 -0.78 -0.17 Sagebrush -1.70 0.41 -0.27 Counts in control and carbaryl plots in all habitats (except for aspen habitat in the control FS's) indicated an average postspray increase whereas counts in tri- chlorfon plots either decreased or were unchanged. Although only suggestive, there appeared to be some effect from the sprays on the total population, and the effect of trichlorfon seemed to be greater than that of carbaryl. In the second test, species that were not common to all plots or that had large or undefinable home ranges were excluded from the total birds data (leaving mostly the species listed in Table 1). Species excluded were raptors, grouse, corvids, evening grosbeaks, and pine siskins. The probability of a larger F value from the modified data was not significant (P = 0.43). Table 4. Split-plot, 3x3 analysis of variance of the difference between the mean of prespray and post- spray counts of all species in three habitats, with three treatments and three replicates. Source of Sums of Mean variation df squares square F Treatment (A) 2 5.276 2.638 3.430" Error (a) 6 4.616 0.769 Habitat (B) 2 2.786 1.393 0.864b Interaction (AXB) 4 5.402 1.350 0.837b Error (b) 12 19.338 1.612 Total 26 37.418 aP = 0.10. bP = 0.50. 12 The average number of birds per hectare at the FS's (Table 3) was intended to indicate abundance over time. However, the presence of the observer affected the counts, thus diminishing their value. For example, some birds, such as Swainson's thrushes, tended to avoid observers, whereas others, such as the ruby- crowned kinglet and warbling vireo, seemed to be attracted to observers or were at least less wary. The relative smallness of some FS's precluded a sound esti- mate of the actual population because birds main- tained a certain "safety" distance from the observer. This behavior was apparently related to habitat, since some species seemed to maintain a greater safety distance in the open sagebrush than they did in the Douglas-fir or aspen habitats. Total birds per count in all habitats except those in the FS's in aspen consistently represented the highest mean values of the three variables (Table 3) and also had the lowest coefficient of variation. The total bird variable appears to be the best for statistical purposes; however, the technique is difficult to describe and is not precise in practice. Furthermore, the plotted line of this variable (Fig. 8) leads to the same conclusion as the other variables (i.e., that there was no appreciable population change). The general trend of mean total birds per count remained about even before and after spraying in all plots, whereas males per count increased slightly but insignificantly (Fig. 8). Conversely, birds inside the boundaries of the FS's tended to decrease in all habi- tats after the treatment. The increase of singing males was probably related to the arrival of two inter- mediately abundant species— the western tanager and lazuli bunting. An evaluation of changes in species composition for the 14 most abundant species after the treatments showed that none of the species disappeared from the counts, or decreased on only treated plots, in any of the three habitats (Fig. 9). Only the ruby-crowned kinglet, western tanager, lazuli bunting, dark-eyed junco, and chipping sparrow differed in appreciable numbers before and after the treatment. Ruby-crowned kinglets were observed less fre- quently after treatment in all plots, but the decrease was most noticeable in trichlorfon plots. This change may have resulted from a shift in male activity from that of singing to that of feeding of young, or from insecticide effects, thus decreasing the chance of their being heard or seen. Emigration of males (which com- prised a majority of the observations) is unlikely because breeding pairs of kinglets did not drop greatly in any SP (Fig. 7). When the frequencies of occurrence in all habitats are compared (Fig. 9), ruby-crowned kinglets were observed less frequently at FS's during postspray in plots treated with trichlorfon (89%) than in plots treated with carbaryl (68%) or control plots t- > 13 o a Cl en CL LU CD « 40 a> 30 2 20 to c 10 Aspen 5 4 3 2 1 1 2 3 4 Prespray Postspray CENSUS PERIOD Fig. 8. Periodic means of bird counts inside the fixed stations (FS) only, the total birds, and territorial males-only ob- served during five counts before and after treatment from FS's positioned in Douglas-fir, quaking aspen, and sage- brush habitats. Lines are pooled data from counts at FS's in three major plots: solid lines control, dotted lines tri- chlorfon plots, and dashed lines from carbaryl plots. 13 Warbling vireo Pine siskin Chipping sparrow SC 0 50 0 Dark- eyed junco Mountain chickadee Western tanager Tree swallow Empidonax spp American robin Yellow- rumped warbler Ruby-crowned kinglet Evening grosbeak I I I I I I I I I I I I I House wren Lazuli bunting ill I ■ I ■ ■ Wm I I ■ I ■ I Sagebrush All Habitats Fig. 9. Frequency of occur- rence of 14 major species observed at fixed sta- tions (FS) during five prespray (solid bars) and postspray (stippled bars) counts in three habitats. 14 Table 5. Number of prespray encounters (birds sighted or heard) before treatment and percentage change after treatment for 14 abundant species recorded at fixed stations (FS's) in three habitats in treated and untreated plots." Treatment Habitat Control Trichlorfon Carbaryl Number of encounters Number of encounters Number of encounters Possible Actual no. Postspray Possible Actual no. Postspray no. prespray change (%) no. prespray change (%) Possible no. Actual no. prespray Postspray change (%) Douglas-fir 980 382 + 12 1,050 322 -3 910 369 + 2 Aspen 980 324 + 10 896 323 -4 616 228 + 8 Sagebrush 910 299 + 4 840 277 -2 840 307 -9 aOne encounter = one species tallied (seen or heard singing) during a 5-min count; number of possible encounters = number of counts per period per plot x 14 species; see Fig. 9 for the 14 species. (66% decrease). Although the difference was not sig- nificant (P = 0.12), we believe that the decrease of ruby-crowned kinglets in trichlorfon-treated plots, as compared with control plots, was possibly related to the trichlorfon treatment. Observations of western tanagers and lazuli bunt- ings increased substantially after the spraying, prob- ably because both species were among the last to establish breeding territories, and males sang more frequently during the postspray period. Chipping sparrow and dark-eyed junco encounters likewise increased after treatment— mostly in the Douglas-fir habitat. Both nested late in the season; male dark-eyed j uncos often exhibit a surge of singing during their second nesting. After treatment, total encounters (birds detected by sight and sound) increased about 9% in control plots, decreased (insignificantly; P = 0.32) about 3% in tri- chlorfon-treated plots, and remained essentially un- changed in carbaryl plots (Table 5). Birds counted after the spraying inside the bound- ary of FS's tended to increase in Douglas-fir habitats and to decrease in aspen and sagebrush habitats (Table 3). The relation persisted for control plots but not for treated plots, wherein a consistent decrease of birds occurred for the inside station variable in all habitats. It is conceivable that birds in untreated plots used the Douglas-fir to a greater extent, perhaps for foraging, than the other habitats after the treatment date. The suppression of the western budworm by the insecticides apparently reduced some of the attrac- tiveness of the Douglas-fir areas for birds. Although the breeding-pair estimates and FS counts generally agreed, some results were inconsistent. Our most conflicting results involved the western tanager on all plots. Breeding pairs of this species decreased in trichlorfon SP's (Fig. 7), whereas the FS census indi- cated an increase on all plots (Fig. 9). One reason for this disagreement might be that factors causing changes in tanager numbers or conspicuousness varied within the sprayed MP's. Also, since tanager activity, but not numbers, increased after treatment, the breeding-pair census on SP's would not reflect such an increase in numbers as the FS count method would. We suspect that species' activity changes probably affected the FS counts more than the breeding-pair census, because the FS's were spread out in the major plots. For tanagers, where territorial density probably remained constant and the breeding activity increased after treatment, this increased postspray conspicuous- ness resulted in higher FS counts. Despite the incon- sistencies, the general agreement of the two types of bird counts added confidence to our overall assess- ment of population responses to the spray treatments. The FS count gave us a check on population changes that the breeding-pair estimate might not detect, because the FS's were scattered throughout the MP's and thus increased our chances of detecting effects on birds on the whole plot. Also, the breeding-pair esti- mate yielded only one value for each treatment period, whereas the FS count provided trends in species abundance. Nesting Studies A total of 227 nests that were active at spray time were located, of which 122 were cavity nests (Table 6). About 52% of the active nests held young at spray time; the rest had eggs. By taxonomic grouping, the relative abundance of nests was as follows: thrushes 22%; woodpeckers 17%; flycatchers and swallows 16%; house wren 14%; sparrows and juncos 13%; and warbling vireo 6%. Sixty-seven additional active nests observed before treatment were not included in the evaluation because they became inactive before treat- ment. Because it was impossible for us to obtain accurate 15 Table 6. Number and type of nests found on the study area that were active immediately before the spray applications. Species and group Nest type3 Treatment Control Trichlorfon Carbaryl Goshawk Sharp-shinned hawk Red-tailed hawk Golden eagle Common flicker Yellow-bellied sapsucker Williamson's sapsucker Hairy woodpecker Downy woodpecker Northern three-toed woodpecker Empidonax spp. (flycatchers) Tree swallow Black-capped chickadee Mountain chickadee House wren American robin Swainson's thrush Mountain bluebird Warbling vireo Yellow warbler Yellow-rumped warbler Western tanager Pine siskin Green-tailed towhee Dark-eyed junco Chipping sparrow White-crowned sparrow Total N N N N C C c c c c N c c c c N N c N N N N N N N N N 0 1 1 0 3 6 0 2 1 0 3 8 1 2 8 14 0 6 8 0 0 0 0 2 5 5 1 77 1 0 0 0 1 1 1 0 4 6 6 3 1 0 2 2 1 1 0 1 7 3 13 3 0 0 6 3 14 10 8 10 0 2 6 3 1 5 1 0 1 0 0 2 2 0 0 0 7 3 5 3 1 0 89 61 aC = cavity, N = noncavity (based on nest sites used in the study area). counts of eggs laid or young fledged in many nests (e.g., in cavity nests), we used a simplified definition of nest failure or success (Table 7). A successful nest was defined as one with advanced age nestlings in or near the nest, and with adults carrying food to the nest, or one from which the young had probably fledged after our preceding visit. Another possible method for eval- uating incomplete nesting data in which nest exposure days are used was described by Mayfield (1961, 1975); however, we believe our data were sufficient for direct comparisons. The success of nests in control plots did not differ significantly from the success in trichlorfon- and car- baryl-treated plots (Table 7) when either nests with eggs, or with young, or all nests combined were tested (P > 0.50; Sokal and Rohlf 1969:599). Therefore, the insecticides did not affect nest success, as we defined it. In the control plots, 74% of the nests with eggs and 97% of the nests with young at spray time were suc- cessful; these respective percentages were 83 and 100% in trichlorfon plots and 86 and 100% in carbaryl plots (Table 7). A weakness in this method of eval- uating nesting success is the lack of information about the numbers of young produced per nest. The fates of eight unsuccessful nests containing eggs in control plots were as follows: two were de- stroyed by predators, three were abandoned, and two fell to the ground; one nest with young was destroyed for reasons unknown. In trichlorfon plots, two nests with eggs were destroyed by predators, one each was abandoned with eggs and with young, and one was de- stroyed by livestock. In carbaryl plots two nests with eggs were abandoned and one was accidentally de- stroyed by our activities following treatment. It appeared that birds— even those living well inside the treated plots— did not abandon the plot or their nests in abnormal numbers. Our breeding-bird and FS counts indicated that birds were still using the plots regularly after treatment with insecticides; further- more, no impact of the treatments on nesting success could be detected. Even though budworm populations were reduced in treated plots by about 70% at 1 week 16 Table 7. Percentages of successful nests among nests of known outcome, as determined at the last postspray visit to nests that were active at the time of treatment. Nests with young at time Nests with eggs at time of treatment of treatment Known outcome Outcome Known outcome Outcome Treatment, nest type. No. Nests Fledged unknown Nests Fledged unknown and feeding category species (no.) (%) (no.) (no.) (%) (no.) Control Nest type Cavity 9 16 73 1 19 100 1 Noncavity8 9 13 75 1 14 92 2 Total or weighted average 18 29 74 2 33 97 3 Feeding category Raptorial 2 1 100 0 1 0 1 Aerial 2 7 29 0 3 100 0 Tree canopy 2 0 0 0 3 100 0 Tree trunk 3 2 100 0 7 100 0 Understory 3 8 100 0 3 67 0 Ground 5 8 75 1 14 100 2 Air-Ground 1 3 50 1 2 100 0 Total or weighted average 18 29 74 2 33 97 3 Trichlorfon Nest type Cavity 9 19 93 4 33 100 6 Noncavity" 12 19 72 5 10 100 1 Total or weighted average 21 38 83 9 43 100 7 Feeding category Raptorial 3 0 0 0 3 100 0 Aerial 2 18 100 4 1 100 0 Tree canopy 3 0 0 0 9 100 2 Tree trunk 4 0 0 0 10 100 2 Understory 3 7 67 1 9 100 2 Ground 5 11 67 2 7 100 0 Air-Ground 1 2 0 2 4 100 1 Total or weighted average 21 38 83 9 43 100 7 Carbaryl Nest type Cavity 9 13 91 2 18 100 3 Noncavity8 8 12 80 2 7 100 1 Total or weighted average 17 25 86 4 25 100 4 Feeding category Raptorial 1 0 0 0 1 100 0 Aerial 2 4 100 0 1 100 0 Tree canopy 3 2 100 0 4 100 0 Tree trunk 4 0 0 0 7 100 1 Understory 1 7 100 2 2 100 0 Ground 5 10 75 2 9 100 2 Air-Ground 1 2 50 0 1 0 1 Total or weighted average 17 25 86 4 25 100 4 "Most warbling vireo nests that were active at spray time are not included because we could not see into me nests, oi eigia iuui- tional nests on control plots, seven were active and the status of one was unknown, one nest on trichlorfon-treated plots was active, and of three nests in carbaryl-treated plots, two were active and the status of one was unknown at the last postspray check. 17 Table 8. Results of systematic searching for dead birds in three untreated and six treated major plots (no sick birds were found). Prespray Postspray Search time8 No. dead birds found Search time3 Treatment Days Hours Days Hours birds found Control Trichlorfon Carbaryl 10 10 12 11.0 13.0 13.5 0 lb 4d 8 8 7 7.5 8.5 9.0 0 2C 0 aOne search day = one person searching in one plot for 1 day, for 0.5 to 2.5 hours; one search hour = number of people search- ing times number of hours searched. bBlue grouse adult. cPartly feathered nestling dark-eyed j uncos found partly ingested by garter snakes on two occasions. dOne adult each, yellow-rumped warbler, blue grouse, common flicker, and western tanager. after the treatments (see following subsection on food), breeding pairs were inhabiting sprayed plots up to 2 weeks after treatment. Apparently, for the 2 weeks after treatment, there were sufficient arthropods to sustain adults and nestlings. Numbers of insects in sweep-net samples from flowering plants increased in treated and untreated plots after the treatments (Schmidt et al. 1978), suggesting that insects were available as food, at least in the lower vegetative cover. We believe nesting success may be an important consideration in field testing of a pesticide. If adults die, lose their ability to care for young, or emigrate from the sprayed areas, increased nest failures would probably occur on treated plots. If so, other infor- mation could be used to help explain nest failures. For example, census results could confirm die-offs or emi- gration, carcass counts could indicate die-offs, and brain cholinesterase measurements could indicate physiological impairments in the birds. In other studies of trichlorfon and carbaryl, as in ours, no measureable changes in nesting success have generally occurred. Pillmore et al. (19716; design of study same as Pillmore 1973, Table A-4), found that 84% of 25 nests fledged or were active in two plots treated with trichlorfon (in oil at 1.1 kg active ingre- dient per ha) during 2 weeks postspray; all nests in a control plot held or had fledged young during the same period. Bednarek and Davidson (1967) applied car- baryl at 1.1 kg active ingredient per ha over a mixed forest in Massachusetts during the fifth breeding season of a 5-year study of nest success by six bird species using artificial nest boxes. Percentage of eggs that hatched, clutch size, and nestling mortality after treatment during peak laying was essentially the same in 34 nests during years before spraying and in 8 nests during the one season postspray. The only evidence of a possible pesticide effect was in one clutch that hatched near the time of spraying; all five young tree swallows had been dead for some time when found and contained 0.4- to 2.0-ppm concentrations of "apparent Sevin." In another study (Conner 1960), the young of three bird species were successfully reared in several nests after a New York mixed forest was sprayed with carbaryl (in oil) at 1.4 kg active ingredient per ha. Rich- mond et al. (1979) found no major differences in the probability of success of nest contents in control plots and plots treated with 2.2 kg active ingredient per ha carbaryl. The same was true 1 year later. Four of five nests representing three species that were active in a New Jersey area sprayed with carbaryl fledged young after treatment (Moulding 1976). No nestling mortal- ity was observed, and the only control nests also fledged young. In a similar study involving trichlorfon and carbaryl (Doane and Schaefer 1971), 15 nests representing nine species were all successful after treatment, except for one which was lost to predation. Casualty Searches Searches for dead or sick birds (Table 8) indicated that mortality did not increase after the treatments. A few dead birds, but no sick birds were found, sug- gesting that the dead birds did not die from direct insecticide poisoning or related complications. It is extremely difficult under most circumstances to locate small, dead birds in forest vegetation. However, since no sick birds were noted, it is unlikely that we overlooked significant mortalities. Furthermore, these insecticides have relatively low acute and dietary toxicities to laboratory-treated birds (Heath et al. 1972; Hill et al. 1975; Tucker and Crab- tree 1970; and Zinkl et al. 1977), and other workers have not observed mortality of birds under field condi- tions in other regions (Moulding 1976; Pillmore 1973; and Richmond et al. 1979). 18 Table 9. Average percentage (range in parentheses) of animal matter, plant matter, and grit, estimated by volume, in 183 stomachs from birds of selected species collected in treated plots and untreated areas away from the study area. Species and category of area Treatedb Mountain chickadee American robin Warbling vireo Western tanager Evening grosbeak Pine siskin Dark-eyed j unco Chipping sparrow Totals Untreated0 Mountain chickadee American robin Warbling vireo Western tanager Evening grosbeak Pine siskin Dark-eyed j unco Chipping sparrow Totals aAll insects and spiders except for an occasional snail. bSamples collected 10-19 July. cSamples collected 2-8 July. Material Number of birds Animal" Plant Grit 21 99(90-100) < 1 (0-10) <1 (0-5) 24 93 (0-100) 2(0-15) 5 (0-95) 10 99(95-100) 0 - 1 (0-5) 17 99(95-100) 0 - 1 (0-5) 17 80 (5-100) 10 (0-75) 9 (0-50) 10 50 (0-90) 6 (0-60) 42 (5-90) 31 95(80-100) <1 (0-7) 5 (0-20) 23 95(50-100) 2 (0-40) 3 (0-50) 153 91 (0-100) 2(0-75) 6 (0-95) 3 88 (85-95) 10(0-15) 2 (0-5) 7 59 (0-100) 25 (0-90) 16 (0-95) 1 100 — 0 - 0 - 2 100 — 0 - 0 - 6 63 (5-95) 34 (0-95) 3(0-10) 3 75 (60-85) 0 - 25(15-40) 5 72 (25-95) 11(0-45) 17 (0-75) 3 65 (5-100) 25 (0-75) 10 (0-25) 30 71 (0-100) 18(0-95) 11 (0-95) Food Animal matter (mostly insects) was the dominant food in the stomachs of eight bird species in the treated areas, although some plant material and grit were present (Table 9). It is likely that the dying or dead insects that were made more readily available by the insecticides' effects were consumed at a high rate. Although differential rates of digestion of stomach contents can bias the interpretation of food habits toward insects that digest more slowly than others (Coleman 1974), we found that numbers of individual soft-bodied insects in the stomachs (lepidopteran larvae) still exceeded those insects with more resistant parts, such as adult Coleoptera and Hymenoptera. Larvae of Lepidoptera contributed 68 to 95% of the individual identifiable insects found in all bird species. Larvae of the black-headed budworm, cone moth, and perhaps loopers (Geometridae) were possibly sometimes confused with the larval western budworm when the stomach contents were partly decomposed. However, most of the western budworm larvae in the stomachs in our samples were large and well devel- oped, and we were able to use the key characteristics of the western budworm in its late instar stages (Patrick J. Shea, personal communication) for identification. Four species, the American robin, evening grosbeak, warbling vireo, and western tanager, had the highest occurrence of western budworms per bird in their stomachs (Table 10). All the species listed in Table 10 except the American robin, dark-eyed junco, and to some extent the pine siskin and chipping sparrow, are canopy-feeding species. The robin is an opportunistic ground-feeding omnivore (Martin et al. 1951), and probably ate fallen budworms after the spraying. Robins {n = 5) had more budworms in their stomach contents on the day of spray (40 per bird) than any of the other seven species (one to seven per bird; n = 49). Budworms that had fallen to the ground and others affected by the spray that were falling from the trees on their silk threads were readily visible within a few hours after the treatments. Western budworms were found in stomachs of the eight species in 66% of the individuals collected in tri- chlorfon-treated MP 4, where we collected one-third of the samples of the eight species (Table 10). Budworms were found in only 30% of the birds of the same eight species collected in the other five treated plots. West- ern budworms were found in 17% of the birds collected in untreated areas, but budworm populations were not 19 Table 10. Occurrence of western budworm larvae in stomachs of eight bird species collected in treated and un- treated areas. Average number of western budworms per bird Species Mountain chickadee American robin Warbbng vireo Western tanager Evening grosbeak Treated areas Untreated areas Pine siskin Treated areas Untreated areas Dark-eyed junco Chipping sparrow Totals or means Number Percentage Birds of birds occurrence8 All birds with larvae 24b 21 0.4 2.0 31b 48 13.6 28.2 lib 18 1.7 8.5 19b 53 2.9 5.6 17 65 12.8 19.7 6 67 14.0 21.0 10 50 1.5 3.0 3 33 0.7 2.0 36b 42 1.4 3.5 26b 23 0.5 2.3 183 40 4.9 12.1 "Percentage of all bird stomachs containing one or more western budworms. bAdditional numbers of each species— three chickadees, seven robins, one vireo, two tanagers, five juncos and three sparrows— that were collected in untreated areas adjacent to the study plots did not contain western budworms; all other birds listed except as noted were collected in sprayed plots. determined in these untreated areas. The high inci- dence of western budworms in the stomachs from MP 4 could be a function of availability due to the insecticide kill, density of the budworm population, or both— although an apparent higher population of budworms occurred in plot 4 (Table 11). Budworm populations also were high in MP 8 (Table 11); how- ever, we collected few birds there. Table 11. Numbers of western budworms counted per 100 buds in the nine major plots, before and after the treatments." Days after treatment Plot and treatment Control 1 3 9 Trichlorfon 4 5 7 Carbaryl 2 6 _8 "Reprinted with permission of Flavell et al. (1978). espray 7 14 27.4 27.7 22.8 17.4 18.0 18.4 13.3 12.4 8.4 25.5 8.0 5.1 19.3 6.2 4.4 11.6 3.8 1.6 18.5 6.0 4.0 13.7 2.5 1.9 25.5 6.5 3.1 Western budworm larvae represented nearly 50% of the total insects in stomachs of the evening grosbeak, American robin, pine siskin, and western tanager, in decreasing order (Table 12). Assessment of food intake in larger birds, such as the American robin and evening grosbeak, was more complete because most of the food they had eaten could be readily identified. Because prespray birds were collected in untreated areas adjacent to the control plots, where population estimates of western budworms are lacking, prespray versus postspray comparisons are not clear-cut; how- ever, we believe that budworm availability was prob- ably lower in the areas of prespray collection. We developed an avian feeding potential for the species we examined (Table 9) by multiplying the number of birds per hectare (from Table 1) by the average number of budworms per bird (a procedure modified from Mitchell 1952). The American robin showed the highest feeding potential (12.5; Table 13). The densities of evening grosbeaks and pine siskins could not be estimated with the same accuracy as those of the other six species from our own counts. Although not directly quantifiable from our data, grosbeaks were probably about as abundant as the western tanagers, giving them a feeding potential of 5.9. The warbling vireo was the most abundant species examined; however, it had a low western budworm feeding potential, 1.9. Feeding potentials of all species except the robin and evening grosbeak were below 2.0. 20 Table 12. Percentages of total insects composed of western budworms, and of other insects or spiders in birds taken from plots treated with carbaryl or trichlorfon. Total Organism stomachs identifiable Western Total Other Unidentifiable Species examined insects (no.) budworm Lepidopti 3raa arthropod sb arthropods Mountain chickadee 21 97 10 68 22 10 American robin 24 778 54 95 4 1 Warbling vireo 10 73 23 79 14 7 Western tanager 17 149 40 70 17 13 Evening grosbeak 17 360 60 98 1 1 Pine siskin 10 31 48 97 0 3 Dark-eyed junco 31 201 27 70 16 14 Chipping sparrow 23 138 11 71 19 10 aEssentially all larval forms; includes western budworms. bNearly all were adult Coleoptera or Hymenoptera, or Araneida. Table 13. Feeding potential of selected bird species on western budworm larvae during the postspray period in treated plots. Species No. of pairs per 100 haa Number of stomachs sampled Number of larva budworms per birdb Feeding potential0 American robin Evening grosbeakd Warbling vireo Western tanager Dark-eyed junco Chipping sparrow Mountain chickadee Pine siskine 34.3-57.6 17.2-28.8 37.6-75.4 17.2-28.8 30.2-50.8 40.3-60.3 39.8-43.9 30.0 31 17 11 19 36 26 24 10 13.6 12.5 12.8 5.9 1.7 1.9 2.9 1.3 1.4 1.1 0.5 0.5 0.4 0.3 1.5 0.9 aRange of breeding-pair density in the nine 8.1-ha plots; the high value = the total number of pairs estimated (border pairs called a full pair) for the 8.1-ha plot and the low value = the density for 13.6 ha to account for additional peripheral area occu- pied by pairs with their territories dissected by the plot border. bMean number of western budworm larvae per stomach for all birds sampled. cAverage number of pairs per hectare x 2 x average number of western budworms per bird = feeding potential. dAn estimate based on relative abundance of other species determined in this study. eEstimates from a large area adjacent to our study plots (Skaar 1969). Our food investigation revealed some changes in predation on the western budworm by birds as the budworm population decreased in treated plots; how- ever, our samples were too small to make statistical comparisons. The numbers of western budworm larvae in the pooled stomachs of tree-canopy feeding species in Table 9 (see Table A-l for feeding categories), ex- pressed as an average precent of total insects per bird over several plots, was 30% on spray day in MP's 2, 4, 6, and 8 (n = 31 birds); 52% on 2 days postspray in MP's 5 and 6 (n = 16); and 61% on 5 days postspray in MP's 2 and 4 (n = 19). Percentages for ground feeders in the same MP's and at the same times were 60 (n = 23); 24 (n = 16); and 77 {n = 16). It appears that ground-feeding species, led by the robin, keyed on the budworm beginning with spray day, whereas the tree- canopy feeders used the budworm more heavily 5 days postspray. Both foraging groups increased the con- sumption of budworm at 5 days postspray, at a time when larvae populations were becoming reduced by 70% in the coniferous foliage (Table 11; 7 days post- spray). Occurrence of Spray on Birds All birds collected on days 2 and 3 postspray showed evidence of the dye (Table 14). The dye was found most frequently on the feet of the 10 species examined (66% of 202 samples); in only 8% was it found only on the 21 plumage. Dye marks were found on the plumage or feet 5 days after treatment, but at a lower frequency than on birds collected on the day of spraying. Spray on the bodies occurred in decreasing frequencies among birds found in the treetops (80%), within the canopy (71%), and near the ground (70%). A test of the equality of two percentages (Sokal and Rohlf 1969:607), however, showed that the probability of not encountering the dye was not significantly different among groups (P — 0.08). The forest canopy appar- ently provided only limited protection to birds by intercepting the spray and lowering the amount that reached the understory vegetation. The Lincoln's sparrow, found exclusively in open willows along streams, was exposed to the spray at a frequency similar (82%) to that found on the species dwelling in treetops (80%) (Table 14). This seems logi- cal, since the willows were in open areas, and thus were not protected by an overstory of trees. All of the treetop dwelling western tanagers and warbling vireos that we collected had dye on them, suggesting that the birds living in the treetops came in greater contact with the spray than did other birds, either directly or from contaminated foliage. Relation of Insecticides to Choline sterase in Birds Cholinesterase (ChE) determinations were made on brain tissue of 13 bird species collected in treated areas. Ninety were collected in carbaryl plots and 103 in trichlorfon plots through 5 days postspray. Ludke et al. (1975) considered that at least a 20% reduction in normal brain ChE activity was necessary to indicate exposure to a ChE inhibiting compound. ChE activ- ities that were depressed by two standard deviations or 20% below prespray levels were found in nine birds of five species (Zinkl et al. 1977): one dark-eyed junco, two evening grosbeaks, three mountain chickadees, two western tanagers, and one Lincoln's sparrow. Six were from trichlorfon plots and three were from car- baryl plots. The greatest depression (> 20% below pre- spray levels of ChE activities) was found in two evening grosbeaks and two western tanagers; two tan- agers and one grosbeak were from trichlorfon plots, and one grosbeak came from a carbaryl plot. These data, along with the results of our bird counts, and occurrence of dye, suggest that effects of the sprays on Table 14. Occurrence of automate-red dye after aerial applications of insecticides containing the dye, on the plumage or feet of birds that generally live in different parts of the forest. Days after treatment 0 1 2 3 5 Overall With With With With With With Total dye Total dye Total dye Total dye Total dye Total dye Species and habitat no. (%) no. (%) no. (%) no. (%) no. (%) no. (%) Treetop Yellow-rumped warbler 0 — 3 66 2 100 1 100 0 — 6 83 Western tanager 11 100 3 100 4 100 0 — 1 100 19 100 Warbling vireo 9 100 0 — 1 100 0 — 1 100 11 100 Pine siskin 11 82 1 100 0 — 0 — 6 33 18 67 Evening grosbeak 6 66 5 60 1 100 0 — 8 50 20 60 Total 37 89 12 75 8 100 1 100 16 50 74 80 Below treetop Mountain chickadee 10 80 6 66 0 — 0 — 5 20 21 62 Chipping sparrow 13 92 2 100 4 100 0 — 8 38 27 78 Total 23 87 8 75 4 100 0 — 13 31 48 71 Near ground, open area American robin 11 36 10 90 1 100 1 100 8 62 31 64 Dark-eyed junco 19 84 4 100 4 100 2 100 9 22 38 74 Total 30 67 14 93 5 100 3 100 17 41 69 70 Near ground, bushes Lincoln's sparrow 4 100 1 100 0 — 2 100 4 50 11 82 Grand total 94 82 35 83 17 100 6 100 50 42 202 74 22 populations (although statistically insignificant) were small, but that birds living in the treetops probably received the greatest exposure to the chemicals. Recommendations for Future Studies Improvements in study design that would increase the probability of detecting effects of organophos- phorus or carbamate insecticides on forest birds in similar studies are as follows: (1) spray plots of no less than 750 ha, (2) monitor the effects on canopy-dwelling species more intensively than the effects on other species, and (3) estimate the abundance of insects that insectivorous birds eat before and after treatment. A minimum of three replicates of treated and untreated plots (as used in the present study) is also needed for statistical purposes. A minimum plot size of 750 ha is necessary for these reasons: (1) treated areas that are well away from bird- count sites are needed to collect birds for studies of brain cholinesterase activities, (2) plots need to be large enough to adequately encompass the singing and foraging territories of birds, and (3) a large buffer of sprayed area is needed between bird-count sites and the plot edge so that birds cannot easily move between sprayed and unsprayed areas. Since canopy-dwelling bird species seem to be most exposed to aerially applied forest insecticides, more effort should be extended to those species. This could be done by (1) collecting adequate numbers (8 to 10) per collecting time of each species for cholinesterase study during prespray and at specific intervals post- spray as an indicator of exposure to insecticides, and (2) making an intensive effort to locate nests of canopy-dwelling species. Locating nests in tree canopies is difficult, but some information about canopy nests could be collected. Candidate insecticides that have reached the pilot- testing stage by passing laboratory and small-scale field tests of efficacy and hazard are seldom Likely to cause obvious bird mortalities. However, a subtle but potentially severe effect on insectivorous birds can result from a persistent reduction of insect abundance after treatment. This may occur a few days after the immediate kill of insects and last for an unknown length of time. Changes in consumption of insects by birds and in insect abundance after treatment should be measured by (1) identifying and quantifying insects in stomachs of birds collected concurrently before and after the treatment in sprayed and unsprayed areas (these birds serve also as samples for cholinesterase studies) and (2) using entomological techniques to quantify populations of insects that are important bird foods. Incorporation of the above methodologies, along with searching for dead and affected birds, behavior studies, monitoring nests of the general population, and conducting population estimates are the best approaches for detecting short-term effects of aerially applied organophosphorus and carbamate insecticides on forest birds. Richmond et al. (1979) recommended the following research methods that should receive the highest priority (in order of decreasing importance): (1) studies of brain cholinesterase activity, (2) nesting studies, (3) behavior studies, and (4) population studies. The main value here is that any one of the four listed methodologies, all of which measure specific aspects about the same population, can be helpful in interpreting the results from any of the other three approaches. Conclusion We detected no significant effects of trichlorfon or carbaryl sprays on the numbers of breeding pairs, bird abundance, nest success, mortality rates, or the activ- ities of brain cholinesterase in the breeding population. However, two types of live-bird censuses showed that statistically nonsignificant postspray declines of bird numbers occurred more on trichlorfon plots than on carbaryl plots, and that the species involved lived in the treetops. Three factors indicated that at least some birds were exposed to the insecticide sprays. First, some species ate large numbers of western budworms after the spraying (presumably because they became more readily available); second, a tracer dye in the sprays was found on about three-fourths of the birds exam- ined, indicating direct contact with the sprays; and third, a few individuals (9 of 193) sampled had brain cholinesterase inhibitions great enough to indicate physiological exposure to the insecticides. Trichlorfon showed more potential effects on birds than carbaryl, but neither compound, at the rate applied and under the test conditions, posed a threat to birds that could be measured by the techniques that we used. Acknowledgments Laboratory facilities were provided by the Ennis school district. We thank many people of the Forest Service— especially the project coordinators from Mis- soula, including W. M. Ciesla, H. Flake, T. H. Flavell, and H. Meyer, for their helpful assistance in com- pleting this study; J. R. Hook, District Ranger at Sheridan, Montana, for providing quarters for our crew; and V. Lindsay, District Ranger, Ennis, Mon- tana, for smoothing the way for access to the study plots through private lands. We thank D. W. Ander- 23 son, R. D. Deblinger, P. H. Eschmeyer, S. Federighi, E. R. Levy, L. C. McEwen, R. E. Pillmore, and M. L. Richmond for their constructive comments during preparation of the manuscript; K. P. Burnham for valuable guidance in statistical treatment of the data; C. J. Nasman for assembling the cavity devices; C. and B. Yancy for storing frozen specimens; and K. Miller for data compilation. References American Ornithologists' Union. 1957. Check-list of North American birds. Fifth ed. Port City Press, Baltimore. 691 pp. American Ornithologists' Union. 1973. Thirty-second supple- ment to the American Ornithologists' Union check-list of North American birds. Auk 90(2):411-419. American Ornithologists' Union. 1976. Thirty-third supple- ment to the American Ornithologists' Union check-list of North American birds. Auk 93(41:875-879. Balda, R. P. 1975. Vegetation structure and breeding bird diversity. Pages 59-80 in Dixie R. Smith, technical coordi- nator. Proceedings of the symposium on management of forest and range habitats for nongame birds. U.S. For. Serv., Tech. Rep. WO-1. 343 pp. Bednarek, R., and C. S. Davidson. 1967. Influence of spray- ing with carbaryl on nesting success in a sample of bird- boxes on Cape Cod in 1965. Bird-Banding 38(l):66-72. Bent, A. C. 1939. Life histories of North American wood- peckers. U.S. Natl. Mus. Bull. 174 (Reprinted by Dover Publishing Co., N.Y., 1964). 334 pp. Bent, A. C. 1942. Life histories of North American fly- catchers, larks, swallows and their allies. U.S. Natl. Mus. Bull. 179 (reprinted by Dover Publishing Co., N.Y., 1964). 555 pp. Bent, A. C. 1946. Life histories of North American jays, crows, and titmice. Part 2. U.S. Natl. Mus. Bull. 191. (Re- printed by Dover Publishing Co., N.Y., 1964). 495 pp. Bent, A. C. 1948. Life histories of North American nut- hatches, wrens, thrashers and their allies. U.S. Natl. Mus. Bull. 195 (Reprinted by Dover Publishing Co., N.Y., 1964). 475 pp. Bent, A. C. 1949. Life histories of North American thrushes, kinglets and their allies. U.S. Natl. Mus. Bull. 196. (Re- printed by Dover Publishing Co., N.Y., 1964). 454 pp. Best, L. B. 1975. Interpretational errors in the "mapping method" as a census technique. Auk 92(3):452-460. Booth, W. E. 1972. Grasses of Montana. Montana State Uni- versity, Bozeman. 64 pp. Borrer, D. J., and D. M. DeLong. 1970. An introduction to the study of insects. Third ed. Holt, Rinehart and Winston, New York. 812 pp. Boyd, E. M., and F. I. Taylor. 1969. Toxaphene toxicity in protein-deficient rats. Toxicol. Appl. Pharmacol. 1 8(1): 158- 167. Brockman, C. F. 1968. Trees of North America. Golden Press, New York. 280 pp. Brown, C. J. D. 1971. Fishes of Montana. Endowment and Research Foundation, Bozeman. 207 pp. Burt, W. H., and R. P. Grossenheider. 1952. A field guide to the mammals. Third ed. Houghton Mifflin Company, Boston. 284 pp. Coleman, J. D. 1974. Breakdown rates of food ingested by starlings. J. Wildl. Manage. 38(4):910-912. Connor, P. F. 1960. A study of small mammals, birds and other wildlife in an area sprayed with Sevin. N.Y. Fish Game J. 7(l):26-32. Crase, F. T. 1977. Food of nestling tricolored blackbirds. Condor 79(2):265-269. DeWeese, L. R., R. E. Pillmore, and M. L. Richmond. 1975. A device for inspecting nest cavities. Bird-Banding 46(2): 162- 165. Doane, C. C, and P. W. Schaefer. 1971. Aerial application of insecticides for control of the gypsy moth: with studies of effects on non-target insects and birds. Conn. Agric. Exp. Stn. Bull. 724. 23 pp. Finley, R. B., Jr., R. E. Pillmore, J. E. Peterson, M. H. Mohn, R. A. Wilson, and G. H. Ise. 1963. Grasshopper control: aldrin, dieldrin, and Sevin. Pages 58-59 in Pesticide- Wildlife studies: a review of Fish and Wildlife investi- gations during 1961 and 1962. U.S. Fish Wildl. Serv. Circ. 167. 109 pp. Flavell, T. H., S. Tunnock, J. W. Barry, R. B. Eckblad, and W. M. Ciesla. 1978. Pilot project of carbaryl and trichlorfon against the western spruce budworm, Beaverhead National Forest, Montana, 1975. Part I in Western spruce budworm, a pilot control project with carbaryl and trichlorfon 1975. U.S. For. Serv. Rep. 78-5. 255 pp. Fowler, D. K., and B. S. McGinnes. 1973. A circular plot method of censusing post-breeding bird populations. Proc. Southeast. Assoc. Game FishComm. 27(1973):237-243. Fowler, D. L., and J. N. Mahan. 1975. The pesticide review 1974. U.S. Department of Agriculture, Washington, D.C. 58 pp. Gage, S. H., C. A. Miller, and L. J. Mook. 1970. The feeding response of some forest birds to the black-headed bud- worm. Can. J. Zool. 48(2):359-366. Hall, G. A. 1964. Breeding bird census— why and how. Audubon Field Notes 18(3):413-416. Heath, R. G., J. W. Spann, E. F. Hill, and J. F. Kreitzer. 1972. Comparative dietary toxicities of pesticides to birds. U.S. Fish Wildl. Serv., Spec. Sci. Rep.-Wildl. 152. 57 pp. Hill, E. F., R. G. Heath, J. W. Spann, and J. D. Williams. 1975. Lethal dietary toxicities of environmental pollutants to birds. U.S. Fish Wildl. Serv., Spec. Sci. Rep.-Wildl. 191. 61pp. Jacques, H. E. 1959. How to know the weeds. Wm. C. Brown, Dubuque, Iowa. 230 pp. Kendeigh, S. C. 1944. Measurement of bird populations. Ecol. Monogr. 14(1):68-106. Leedy, D. 1959. Pesticide-Wildlife problems and research needs. Trans. N. Am. Wildl. Conf. 24:150-165. Ludke, L. J. 1975. Cholinesterase (ChE) response and related mortality among birds fed ChE inhibitors. Arch. Environ. Contamin. Toxicol. 3(1): 1-21. Martin, A. C, H. S. Zim, and A. L. Nelson. 1951 American wildlife and plants. McGraw-Hill Book Co., New York (Reprinted by Dover Publishing Co., N.Y., 1961). 500 pp. Matsumura, F. 1975. Toxicology of insecticides. Plenum Press, New York. 503 pp. Mayfield, H. F. 1961. Nesting success calculated from expo- sure. Wilson Bull. 73(31:255-261. Mayfield, H. F. 1975. Suggestions for calculating nest sucess. Wilson Bull. 87(4):456-466. McEwen, L. C, J. Macdonald, R. B. Finley, Jr., J. E. Peter- son, M. H. Mohn, and G. Ise. 1963. Grasshopper control: Sevin. Pages 57-58 in Pesticide- Wildlife Studies: a review of Fish and Wildlife Service investigations during 1961 and 1962. U.S. Fish Wildl. Serv. Circ. 167. 109 pp. McEwen, L. C, J. O. Ells, J. E. Peterson, M. H. Mohn, and G. H. Ise. 1964. Grasshopper control: Sevin. Pages 57-58 in Pesticide-Wildlife Studies 1963: a review of Fish and Wild- 24 life Service investigations during the calendar year. U.S. Fish Wildl. Serv. Circ. 199. 130 pp. McEwen, L. C, C. E. Knittle, and M. L. Richmond. 1972. Wildlife effects from grasshopper insecticides sprayed on short-grass range. J. Range Manage. 25(31:188-194. McKnight, M. E. 1968. A literature review of the spruce, western and 2-year cycle budworms (Choristoneura fumi- fernana, C. occidentalis. and C. biennis (Lepidoptera: Tor- tricidae)). U.S. For. Serv., Res. Pap. RM-44. 35 pp. Mitchell, R. T. 1952. Consumption of spruce budworms by birds in a Maine spruce-fir forest. J. For. 50(5):387-389. Moulding, J. D. 1976. Effects of a low-persistence insecticide on forest bird populations. Auk 93(4):692-708. Mueggler, W. F., and W. P. Handl. 1974. Mountain grassland and shrubland habitat types of western Montana (interim report— 1974). U.S. Forest Service, Ogden, Utah. 89 pp. (Unpublished report) Mussehl, T. W., and F. W. Howell, editors. 1971. Game management in Montana. Montana Fish and Game Depart- ment, Helena. 238 pp. Pfister, R. D., B. L. Kovalchik, S. F. Arno, and R. C. Presby. 1977. Forest habitat types of Montana. U.S. For. Serv. Gen. Tech. Rep. INT-34. 174 pp. Pillmore, R. E. 1973. Toxicity of pyrethrum to fish and wild- life. Pages 143-165 in John E. Casida, ed. Pyrethrum: the natural insecticide. Academic Press, New York. 329 pp. Pillmore, R. E., E. L. Flickinger, and M. L. Richmond. 1971a. Forest spraying of Zectran and its safety to wildlife. J. For. 69(10):721-727. Pillmore, R. E., M. L. Richmond, and L. R. DeWeese. 19716. Effects of selected pesticides on wildlife in forest environ- ment. U.S. Fish and Wildlife Service, Denver Wildlife Re- search Center. 8 pp. (Unpublished report) Richmond, M. L., C. J. Henny, R. L. Floyd, R. W. Mannan, D. M. Finch, and L. R. DeWeese. 1979. Effects of Sevin-4- Oil, Dimilin, and Orthene on forest birds in northeastern Oregon. U.S. For. Serv., Pac. Southwest For. Range Exp. Stn. Pap. In press. Robbins, C. S. 1963. Bird census by two methods. Page 85 in Pesticide-Wildlife studies: a review of Fish and Wildlife Service investigations during 1961 and 1962. U.S. Fish Wildl. Serv., Circ. 167. 109 pp. Robbins, C. S., B. Brunn, and H. S. Zim. 1966. Birds of North America. Golden Press, New York. 340 pp. Schmidt, D., T. Olson, G. M. Moore, J. C. Elliott, and S. Gilbert. 1978. Effects of spruce budworm control on polli- nating insects. Part III in Western spruce budworm, a pilot control project with carbaryl and trichlorfon 1975. U.S. For. Serv. Rep. 78-5. 255 pp. Scuderi, J. A., J. A. Boegler, B. M. Earlewine, F. L. Moon, J. W. Barry, R. B. Ekblad, and W. M. Ciesla. 1975. Meteorological data supplement, 1975 spruce budworm pilot test, Gallatin and Beaverhead National Forests. U.S. Army Proving Ground, Dugway, Utah. 69 pp. Seidensticker, J. C, and L. Kilham. 1969. A device for exam- ining nesting cavities. Bird-Banding 39(3):228-230. Skaar, P. D. 1969. Birds of the Bozeman latilong. P. D. Skaar, Bozeman, Montana. 130 pp. Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Company, San Francisco. 776 pp. Steel, R. G. D., and J. H. Torrie. 1960. Principles and proce- dures of statistics. McGraw-Hill, New York. 481 pp. Stewart, R. E., and J. W. Aldrich. 1951. Removal and repopu- lation of breeding birds in a spruce-fir forest community. Auk68(4):471-482. Stickel, W. H. 1973. Effects on wildlife of newer pesticides and other pollutants. Proc. West. Assoc. State Game Fish Comm. 53:484-491. Svensson, S., chairman. 1970. An international standard for a mapping method in bird census work recommended by the international bird census committee. Audubon Field Notes 24(6):722-726. Thomas, J. W., G. C. Crouch, R. S. Bumstead, and L. D. Bryant. 1975. Silvicultural options and habitat values in coniferous forests. Pages 272-287 in Dixie R. Smith, techni- cal coordinator. Proceedings of the Symposium on Manage- ment of Forest and Range Habitats for Nongame Birds. U.S. For. Serv. Tech. Rep. WO-1. 343 pp. Tucker, R. K„ and D. G. Crabtree. 1970. Handbook of tox- icity of pesticides to wildlife. U.S. Fish Wildl. Serv., Resour. Publ. 84. 131 pp. Williams, A. B. 1936. The composition and dynamics of a beech-maple climax community. Ecol. Monogr. 6(3):317- 408. Zinkl, J. G., C. J. Henny, and L. R. DeWeese. 1977. Brain cholinesterase activities of birds from forests sprayed with trichlorfon (Dylox) and carbaryl (Sevin-4-oil). Bull. Environ. Contamin. Toxicol. 17(4):379-386. Appendix 26 Table A-l. List of common and scientific names of birds, mammals, insects, and plants used in the text and tables.3 Group and common nameb Scientific name Group and common nameb Scientific name Birds Goshawk— R Sharp-shinned hawk— R Red-tailed hawk— R Golden eagle— R Blue grouse— G Ruffed grouse— G Mourning dove Common flicker— G Yellow-bellied sapsucker— TT Williamson's sapsucker— TT Hairy woodpecker— TT Downy woodpecker— TT Northern three-toed woodpecker— TT Empidonax (flycatchers)— A Tree swallow— A Black-capped chickadee— TC Mountain chickadee— TC House wren— U American robin— G Hermit thrush Swainson's thrush— G Mountain bluebird— AG Ruby-crowned kinglet— TC Warbling vireo— TC Yellow warbler Yellow-rumped warbler— TC MacGillivray's warbler— U Western tanager— TC Evening grosbeak Lazuli bunting— U Accipiter gentilis A. striatus Buteo jamaicensis Aquila chrysaetos Dendragapus obscurus Bonasa umbellus Zenaida macroura Colaptes auratus Sphyrapicus varius S. thyroideus Picoides villosus P. pubescens P. tridactylus Empidonax spp. Iridoprocne bicolor Parus atricapillus P. gambeli Troglodytes aedon Turdus migratorius Catharus guttatus C. ustulatus Sialia currucoides Regulus calendula Vireo gilvus Dendroica petechia D. coronata Oporornis tolmiei Piranga ludoviciana Hesperiphona vespertina Passerina amoena Cassin's finch Pine siskin -TC Green-tailed towhee— U Dark-eyed junco (Oregon race)— G Chipping sparrow— G White-crowned sparrow— G Lincoln's sparrow Mammals Elk Mule deer Moose Insects Western budworm Spruce budworm (Eastern) Cone moth Black-headed budworm Gypsy moth Plants Douglas-fir Lodgepole pine Limber pine Ponderosa pine True firs Spruce Willow Birch Quaking aspen Sagebrush Idaho fescue Carpodacus cassinii Carduelis pinus Pipilo chlorurus Junco hyemalis Spizella passerina Zonotrichia leucophrys Melospiza lincolnii Cervus canadensis Odocoileus hemionus Alces alces Choristoneura occidentalis C. fumiferana Dioryctria reniculella Acleris uariana Porthetria dispar Pseudotsuga menziesii Pinus contorta P. flexilis P. ponderosa Abies spp. Picea spp. Salix spp. Be tula spp. Populus tremuloides Artemisia tridentata Festuca idahoensis aAuthorities for the names follow: birds— Robbins et al. (1966), American Ornithologists' Union (1957. 1973, 1976); mam- mals—Burt and Grossenheider (1952); insects— McNight (1968) and Borrer and DeLong (1970); and plants— Jacques (1959). Brockman (1968), and Booth (1972). bSymbols following common names indicate feeding categories used in text tables 2 and 7 (symbols are not shown for species not represented in these tables); A— aerial; AG— aerial and ground; G— ground; R— raptorial; TC— tree canopy; TT— tree trunk; and U— understory. 27 Table A-2. Location, direction of exposure, approximate size, and treatment assignment of the nine major study plots. Direction of Approximate Date of Major Elevation - tributary Major area treatment plot (m) T R S drainage tributary (ha) Treatment in July 1 1890-2380 3S2W 29,31,32,33 E North Meadow Creek 350 None — 2 1890-2240 4S 2W 28,29,33 E South Meadow Creek 440 Carbaryl 10 3 1950-2250 9S3W 29,32 NE Clear Creek 350 None — 4 2270-2290 8S3W 26,27,34,35 S North Fork Warm Springs Creek 490 Trichlorfon 14 5 1890-2190 8S2W 35,36 9S2W1.2 E Cherry Gulchb 500 Trichlorfon 16 6 1890-2380 9S 3W 24,25 9S2W 17,18,19,20 W Middle Fork Warm Springs Creek 500 Carbaryl 11 7 2010-2380 9S3W 21,25 26,35,36 NW South Fork Warm Springs Creek 550 Trichlorfon 17 8 1900-2380 9S1W 7,8,17,18,19,20 NE Ruby Creek 460 Carbaryl 12 9 1860-2250 9S1W 34,35 10S1W2.3 NE Nickerson Creek 350 None ~~ aThe distance between the two plots farthest apart (1 and 9) was 50 km. bAH drainages except Cherry Gulch had permanent running water. Table A-3. Schedule of bird counts (C) and insecticide treatments (T) for nine major study plots.a Treatments were made on 10-17 July (except 13 and 15 July, when rain intensified). No bird counts were made on 30 June or 7-10July. Plots prespray Date 1 8 9 Plots postspray Date 8 June 20 C C C 21 C C C 22 C C C 23 C C C 24 C 25 C C 26 C C C 27 C C C 28 C C C C 29 C C C 30 July ICC C 2 C C C 3 C C C 4 C C C 5 C C C 6 C C C 10 T 11 C C T C July 12 C C C T 13 C C C 14 T C C 15 C C 16 C C T 17 C C T C 18 C C C 19 C C C 20 C C C 21 C C C 22 C C C 23 C C C 24 C C 25 C C C 26 C 27 C C C 28 C C 29 C C 30 C C aPlots 1, 2, and 4 were assigned to one observer; plots 3, 5, and 6 to a second; and 7, 8, and 9 to a third. Bird counts included breeding-pair and fixed-station (FS) counts. 28 -o a -*: v. O 'S1 5 co *- k, o •2 o a 2 5 e 53 B Cu C3 C O k 9) -O s < cu CO Eh > < O m c a 60 2 CO ■p. h CO O ^ CO o co o •S3 cu S c - i s I ■ lOTrTrifiN^^^T t— CO O CM CO ^ CO ^P ododddddo ifiwtmiCininHifl B« CO rS 00 CO CM CO CM CM CM CM -T IN CM T d d d d d d d © d HI > < uoin^iflio^ico^io oo to cm" CO V ■V c CO I G o *- u o 3 o o ^H CM CO ^ lO tC t^ o E- 29 6c e c a 5? "a 3 -~ <^ •5 o oo ~-< ' CO "— » ft) e a -o *- t co 8 if ° p^ oi Q e = o c 'G 1 C O •2. o "a 3 3 9) C a S g 3 CO < CD 3 cd Jb o S ° Z 3 5 "2 z 3 co -a a c CO 5. ■*-3 ^- B o S3 cd .c B -a P3 ^ Q< ^ c o oj to ^3 ft cd co *j co 5 M r1 E en o CO ^ C7> CO CO i— 1 +j iO h » n o >> CO c CD CJ> c a g c! CO b Q Z Q Z o Z CO CJ> ■° s •s c — CO -t-j — ~ cc O *J CD r- „ O CD § - « " 8 CD « > S t- c co m a> ■9 Q u = Q Z a z a z m a z a z 60 3 ■a 3 CO r-H Q) T3 CD T-H CO CD £ cd 55 o 5. "5 s o m I Q Oh -a a CD CO CD CM ~H ^ H> CO z Q Z o O a ■v -o c 60 c o Z a z CD CD Q Z P Z -a en c CD as Ih O "co M-i CO CO CO c O s CO -u u QJ 0) p G •^ & CO ■a c _co ^ o o & CO o 3 o 3 •O 'u CD Q o 3 O •O c o o o c o o 4-1 — co — —. Q Z J- — — +j . P o o CN S cdg0gcMCco •"^hcoX-^^-",xcoX CD .^nCD^CD'-'.^H O cc A O O co A ™ io 2 °~ o c « c i C X W) Oi O >— • <— i CJS •— ' 60 J! 3°~C CD 60 J! CO CO ^ '^ 61 o m co 3S « 'S 'S cs c » c i, c - o — ;s — cd •- 60— 60 , 60 P 60 rt n ph % ^ CS ,J) ,H f; rn §- rHCN^nX S (N T3 H ^ H CO^H hJ CO ^ Z •S-3 ■* S — > co CD CD z s I CO jo _a; o -5 > <- a ■" a -a co e *■ 3 co ^ .3 zl cl ■ CO -t^ .. CO CO T3 s § 3 C ° II CN Z ■a 5 co 2 . o CN CJ * s tj) co Z 3 I O h 60 £ CO CD ]= 3 CD *J CO at co _ £ c -5 .5 "75 « +j o W _ C MT! o C O co o c-3 £ | b S ^ ° o b -r S ° u ° 3 co « BJ S2S a> co 3 CD 2 -S & 3 rH 3 ■« E CD b J3 3 CD nfl s « CO CD *9 c 1= S 3 CS E CO CD 3 E IS X X « CO CD 0) u ~ CD '-3 -O CD ° I c i? 8g CD • — " -s CO CD S CD > 3 to C CD O CD ^ | CO 1/5 S oj co 3 ^ « CO 53 o 0 » tt£ _CD -3 *J 3 .3 3 'co * O •" c S g — CO x -2 | III e» ti o CO i C^ cc Tf gif-i -3 ft IB 1 Z * g fe OJ •^3 t*-i "O ■ a S t~i CO ™ 2" « °° ^ CO n co co £ CD CD a 3 t3 co -B 3 CO 'to *J CD CD CO >-. t- 3 CO — H e >> 3 w o a CX cj ■a 2 CD I « 2 ^ a. E s 3 c a CO o Si - CO ^ CD ill 3 c ~o -a C 3 CO CO CD CO o a a ft -X' CD V Cu -3 -o ■: co -3 CM CO 6 J52 c ft >> S ■a A .St--3 J3 CO —• ^ ••= -° CD > CO « *• Ji °g'2 rt^ t- 3 £ co JD CD T3 CD S b -a _ CO CO bo. E CO 3 CD ca to ? *^ -a 3^ .s -° 4J CD & 2 2 ° • ° "O co CD . u » "2 <2 s 3 co d-E-a c -a 2 3 co c 3 CD • - co "a 2 ■s "a ■ O j CO ft 3 £ >,£ ° T3 CD ' CO > .5 CD J s « 5 ft r- O ftu 3 O CD (- S £ 3 ■a « 8 3 « -a o o -^ ft S oo CD £ CN w CO m o 3 O 6 CO c a CO >, >> S -a -c •>5 3 ■a co CO C 3 3 CO HH ^^ As the Nation's principal conservation agency, the Department of the Interi >r has responsibility for most of our nationally owned public lands and natural res nirees. This includes fostering the wisest use of our land and water resources, protecting our fish and wildlife, preserv- ing the environmental and cultural values of our national parks and historical places, and providing for the enjoyment of life through out- door recreation. The Department assesses our energy and mineral resources and works to assure that their development is in the best interests of all our people. The Department also has a major responsi- bility for American Indian reservation communities and for pe >ple who live in island territories under U.S. administration. UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE EDITORIAL OFFICE AYLESWORTH HALL, QSU FORT COLLINS, COLORADO 80523 POSTAGE AND FEES PAID U.S. DEPARTMENT OF THE INTERIOR INT 423 NOTE: Mailing lists are computerized. Please return address label with change of address.