LIBRARY STATE PLANT BOARD E-545 August 1941 A REPORT OF INVESTIGATIONS OF THE EXTENT AND CAUSES OF HEAVY LOSSES OF ADULT HONEYBEES IN UTAH 1/ By A. P. Sturtevant, G. F. Knowlton, J. D. Hitchcock, G. H. Vansell, E. C. Hoist, and W. P. Nye. Division of Bee Culture, Bureau of Entomology and Plant Quarantine . United States Department of Agriculture, and Utah Agricultural Experiment Station INTRODUCTION For many years serious adult losses have occurred periodically in the State of Utah. While losses of varying degrees have occurred al- most every season, the most extensive death of bees occurred during the relatively dry season of 1939. Estimates of adult bee losses for the season of 1939 have ranged from $100,000 to $200,000. Extensive colony deaths and depletion of colony populations occurred during the latter half of June and up to the middle of July of that year, principally in Utah, Davis, Salt Lake, Box Elder, Juab, Wasatch, and Weber Counties. The suspected causes of these adult bee losses have varied from one locality to another and from year to year. These include (1) poisonous plants, (2) use of poisoned grasshopper bait, (3) orchard spraying, (4) spraying for the sugar- beet webworm, (5) dusting for the tomato fruitv/orm, (6) measures for alfalfa- insect control, (7) garden spraying or dusting, (8) mosquito larvicides, (9) smelter-smoke fumes, (10) infectious disease, and (11) starvation, un- favorable weather conditions, and faulty managem'ent of bees. Because of the apparent increasing seriousness of adult bee losses and the growing opinion among Utah beekeepers that at least certain of these losses were associated with grasshopper-control activities, the 1/ A contribution from the Bureau of Entomology and Plant Quarantine, United States Department of Agriculture, in cooperation with the Utah and Wyoming Agricultural Experiment Stations. A. P. Sturtevant, J. D. Hitchcock, G. H. Vansell, and E. C. Hoist are members of the Bureau of Entomology and Plant Quarantine. G. F. Knowlton and W. P. Nye are members of the Utah Agricultural Experiment Station. 2 - Utah Agricultural Experiment Station began investigational work early in August of 1939, and in May 1940 the investigation was continued v/ith the assistance and cooperation of the Bureau of Entomology and Plant Quarantine, United States Department of Agriculture. The object of the work was to study the extent and possible cause or causes of the losses of adult honeybees in certain beekeeping areas in the State. Sixty-three beekeepers in 18 counties of Utah were interviewed to obtain information concerning losses during 1940 and the history of previous losses. Commercial apiaries were visited periodically, in which records v/ere kept of colony conditions and of losses of bees. Information was likewise obtained as to orchard spraying, grasshopper and other insect control measures, the distribution of poisonous plants, and any other information pertinent to the problem. Field investigators collected at intervals samples of normal living bees and of dead or dying bees as well as samples of pollen, honey, leaves from sprayed plants and trees, soil, water, and other miscellaneous materials, such as dead grasshoppers. These samples were sent to the Intermountain States Bee Culture Laboratory, Laramie, Wyo., where examinations for internal parasites and chemical analyses for arsenic were made. Figure 1 shows the location of the chief areas of deciduous-fruit and alfalfa-seed production, poisonous plants in areas surveyed, and smelter areas of Utah. FIELD INVESTIGATIONS Poisonous Plants Death camass, whorled milkweed, larkspur, horsechestnut, and sea- side arrowgrass are poisonous plants of occasional occurrence in Utah, but they appear to have little connection with bee losses. On the other hand, locoweed (A stragalus sp.), occurs widely over the State (see map) and appears to have been responsible for serious losses of bees, especially in Beaver and Iron Counties in 1932. No losses from Astrag;alus were sus- pected in 1940, with the possible exception of one apiary in a mountain canyon east of Salt Lake City which also was within range of orchard spray. The heavy loss of bees in this yard occurred late in May, but the arsenical content of bees, while fairly high, was not convincingly so. Except in mountain areas, locoweed ceased blooming long before the periods of bee losses, consequently plant poisoning must have played a minor part in the extensive losses in 1939. Use of Grasshopper Bait The application of grasshopper bait in Utah has been done by farmers, the cooperating Federal-State control agencies making the bait available to them at mixing stations as needed. Persons in charge of grasshopper-bait mixing stations cautioned recipients as to the danger of poisoning livestock and explained the proper method of handling and spreading the bait. Occasional cases were ob- served, nevertheless, of carelessness in handling and spreading the bait. - 3 Heavy bee losses occurred in two apiaries in Utah County, namely, at Lehi and Pleasant Grove, late in July 1940, but the heaviest losses did not co- incide with the dates of nearby grasshopper-control operations, contrary to statements of beekeepers owning affected apiaries. The arsenic content of bees collected between July 15 and 25 at Lehi and Pleasant Grove was higher than that of samples collected before or after those dates. Both apiaries were within range of orchard-spraying activities. Irrigation water possibly may have become contaminated by grasshopper bait, as the bait had been spread heavily in a cornfield July 13 near one of the apiaries, and a surface-soil sample obtained 10 days later showed rather high arsenic content. Bees were observed coming to the field for corn pollen only, paying no attention to the irrigation water applied July 25. On the other hand, on July 9 con- siderable arsenic was found in algae growing in a nearby reservoir from which th9 bee? drank regularly, The source of the poison was not determined. To ascertain whether honeybees are attracted to grasshopper bait, large quantities of various bait mixtures were spread on sacks in an ex- perimental apiary at Kaysville, Davis County, June 25 and July 26, and also in a commercial apiary at Lehi, July 24 and 25. In spite of continu- ous observations for periods of 1/2 to 2 hours, including early and late morning, afternoon, and evening periods, the bees paid no attention to the baits even when these baits were placed at the bees' customary drinking places. Repeated experimental tests were made of the attractiveness of various mixtures of bran and of poisoned bait with sugar sirup, molasses, honey, and other materials, spread freshly each day on pieces of canvas 50 square feet in size near an experimental apiary in Weber County during August and to mid-Septem.ber 1939, and similarly throughout June, July, August, and September 1940, near an experimental apiary at Logan in Cache County. Even when the experimental baits were spread in quantities 10 or 15 times as great as recommended, the few bees that paused to inspect the bait seldom remained to feed. In actual field-control operations and in spite of cases of improperly spread bait, absolutely no evidence was obtained that during the summer months honeybees were attracted to the bait with or without molasses or amyl acetate. Tables 1 and 2 give summaries of the grasshopper-bait experiments for 1939 and 1940. It was fully realized that bees might be attracted to grasshopper bait during an extreme dearth of nectar and pollen even though they paid no attention to it during a honey flow. The colonies at Logan, where the bait tests of 1940 were made, failed to make appreciable gains, indicating a negligible nectar flow during at least part of the tests. The quantity of bait distributed in the State in 1938 was far in excess of that distributed in 1939. Greater bee losses occurred in the latter year, however, although this might have been influenced by weather conditions, the unusually dry season probably causing a dearth of both nectar and pollen. Amyl acetate was incorporated in every sack of bait distributed in Weber County in 1940; in other counties it was added only upon request. Large quantities of bait were obtained by residents of Orem, - 4 Table 1. — Summary of two grasshopper-bait experiments carried on during July and August 1939 in connection with bee losses in Utah Total number Total number Average number Material 1/ of of bees of bees per appl icat ions observed application Expe riment 1. Unpoisoned baits. Unpoisoned bran plus — Bran only (wet) 48 9 .19 Bran only (dry) - - 10 1 .10 Corn sugar sirup - 3 1 .33 Beet molasses - 16 5 .31 Amyl acetate - - - 16 14 .88 Total 93 30 .32 Experiment 2. Poisoned baits (sodium arsenite] Poisoned bran plus — Bran only (poisoned) Corn sugar sirup - Beet molasses - Amyl acetate - - - Beet molasses and amyl acetate 26 4 16 '17 8 2 3 11 .31 .50 .19 .65 .25 Total Grand total 67 160 25 55 .37 ,34 1/ 1 lb. of various bait materials was used per each 50 sq. ft. area under observation throughout the season. of - 5 - 'able 2. — Summary of three grasshopper-bait experiments carried on during June to September, inclusive, 1940, in connection with bee losses in Utah Material 1/ Total number of applications Total number of bees observed Average number of bees per application Experiment 1. Unpoisonea baits. Unpoisoned bran plus- Bran only (wet) - • Corn sugar sirup Beet molasses - - Beet sugar Amyl acetate - - ■ 58 2 4 1 9 0.05 3.00 .00 .00 .00 Total 74 .12 Experiment 2. Poisoned baits (sodium arsenite Poisoned bran plus — Bran only (poisoned) Corn sugar sirup - Molasses (beet) - - Molasses (cane) - - Honey Sugar Amyl acetate - - - Lemon ------- 75 64 62 49 58 6 65 49 .07 .06 .00 .02 .09 .33 .09 ,06 Total 428 26 ,06 Experiment 3. Poisoned baits (sodium arsenite] Poisoned bran and amyl acetate plus — Corn sugar sirup - Molasses (beet) - - Molasses (cane) Honey Sugar Lemon - 13 7 9 9 2 2 ,23 ,00 ,00 ,33 ,50 00 Total 42 17 Grand total 544 42 08 1/ Quantities of various bait materials used per each 50 sq. ft. of area under observation: June, 3/4 lb., or 10 gm.; July, 1/2 lb., or 10 gm. ; August and September, 1/2 lb. ,^ STATE Pl-ANT BOAW*' - 6 - Utah County, there being 232 calls for bait between May 28 and July 17, 1940. With one exception, calls were made every day on which the mixing station was open. Ten calls were for more than 5 sacks, 217 calls for 1 to 5 sacks, and 5 calls for less than a sack. In spite of this extensive use of bait, an apiary under observation at Orem suffered no poisoning. A few samples of bait were collected and analyzed and found to contain from 1.50 to 3.62 percent of arsenic. Dead grasshoppers, apparently poisoned, collected from the area contained 0.012 to 0.078 mg. of arsenic trioxide per grasshopper. The State has few mechanical bait spreaders, and the actual spreading is done largely by hand by the farmers, never by specially trained Orchard Spraying Orchard spraying was much in evidence during 1940 in the fruit belt (see.-fig. .1) and was reported as somewhat heavier and of more fre- quent application than in the last few years. Most orchards in Utah County near where bee losses occurred during July had been sprayed five times by July 25. Sweetclover, alfalfa, raspberries, thistle, and catnip were in bloom in various orchards, though many orchardists considerately cut such nectar-producing plants before spraying. Other orchardists grow only or- chard grass for a cover crop. Bees were observed gathering nectar from catnip flowers still wet from lead arsenate fruit spray. Bees were not seen to visit wet spray on apple leaves. Sweetclover blossoms under orchard trees and bees caught visiting them shortly after spraying operations showed a high arsenic content. A beekeeper at Pleasant Grove, Utah County, reported a heavy death rate of bees the morning after bees were observed working thistle in an orchard during spraying operations. A sample of these bees contained more arsenic than any other sample analyzed during the season from this apiary. It also is possible that orchard sprays or grasshopper bait may contaminate water in adjacent irrigation ditches. The beekeeper referred to reported that in 1939 he found "empty" lead arsenate sacks dumped into an irrigation ditch where his bees were accustomed to drink. Since water may run in a particular ditch only every ninth or tenth day, this might explain the irregularity of bee losses in a few cases. Fruit growers have complained of short crops in recent years, par- ticularly of apples and cherries. Many believe this is due to a decrease in pollinators, especially honeybees. Several commercial beekeepers stated that they had threatened to move their apiaries from orchards unless the fruit growers -were more careful to avoid spraying the bloom. Others follow the practice of moving to other locations as soon as fruit bloom is over. A beekeeper in Davis County who rented his bees for cherry pollination in- sisted that all spraying be done after 4 p. m. unless on cool, cloudy days, and reported his bees to have gone unharmed through four such sprays, al- though some of the colonies were directly under the trees. A custom sprayer for one community in Utah County reported that spraying had to be carried on - 7 almost continuously to make the rounds of successive orchards on his sched- ule. One sample of bees which was analyzed from this area -showed 0.00038 mg. of arsenic trioxide per bee and another 0.00022 mg. of arsenic trioxide per bee, and both showed only traces of lead. Spraying for the Sugar-beet Webwonn Paris green sprays were used extensively to control the sugar-beet webworm in 1938 and 1939. particularly in Weber and Davis Counties. Com- mercial beekeepers in this area suffered so heavily in those years that in 1C40 they did not attempt to keep bees in the old locations. Only a few infestations of webworms occurred in 1940, so few beets were sprayed. A commercial apiary near Kaysville, Davis County, was nearly wiped out in 1939, but an experimental apiary in the same location in 1940 showed no evidence of poisoning. Similarly, small yards in other locations suffered no losses. The evidence is indirect, therefore, that beet sprays may have caused serious bee losses in the past. At least some of these same apiaries, however, were also within range of orchard sprays. The only case of webworm spraying observed in 1940 was on July 5 at Smithfield, Cache County, and this caused no bee losses. Early in July in Weber County and later in Cache County a sugary exudate was observed on both upper and lower surfaces of the leaves of oc- casional beet plants infected with curly-top disease. The exudate occurred in distinct droplets 1/2 to 2 ram. in diameter. It was very sweet, gave a ref Tactometer sugar reading, and was attracting wasps, flies, and ants. No honeybees were observed. Under dearth conditions it is quite possible that this might attract honeybees. Sugar beets raised for seed in fields near Eden in Weber County omitted a distinct odor when in full bloom. Honeybees, bumblebees, and other insects visited the blossoms for both nectar and pollen on July 6. In connection with the spraying of beet fields, in addition to the nectar and pollen from the beets and the abnormal leaf exudate, it is pos- sible that moisture, or honey plants growing as weeds in or near the sprayed fields, might also have attracted honeybees. Several beekeepers watched beet fields in 1939 to see if honeybees were attracted by the paris green SDra- , but bees were not observed, even in early morning hours. Dusting for the Tomato Fruitworm uusting tomato plants with calcium arsenate to control tne tomaxo fruitworm has been suspected of poisoning honeybees, particularly in Weber and Davis Counties. In 1940 most dusting was done late in July and in August, and no bee losses in the neighborhood were reported. Observations for one whole morning and irregularly at other times at all hours of the day failed to disclose honeybees visiting tomato blossoms, and no beekeepers reported such activity. On one occasion, however, a few bumblebees and megachilid bees were observed gathering tomato pollen. - 8 Measures for Alfalfa-Insect Control One beekeeper reported serious bee losses in Beaver County in 1939 which he believed were due to arsenical dusting of alfalfa for the alfalfa weevil. This chemical control is practiced only occasionally and is con- fined largely to fields where alfalfa is raised for seed from the first crop. This type of dusting was not observed in 1940. Cultural control for the alfalfa weevil is practiced extensively wherever alfalfa hay is grown and in seed areas where the first crop is not saved for seed. This consists in cutting the crop early to expose the larvae and adults to the intense heat from the sun. Frequently the cutting is made just when the plants are commencing to bloom. Since al- falfa is ordinarily one of the most important nectar plants in the irrigated sections, this cultural practice probably has greatly reduced the honey crop and increased the dearth period between spring blooms and second- crop alfalfa, which generally constitutes the main flow for many beekeepers. It might be an indirect cause of certain bee losses by making it necessary for bees to search for moisture or sweet substances from abnormal sources, such as fields sprayed with insecticides. Alfalfa blossoms in all fields inspected were infested with varying populations of thrips, and a few beekeepers suggested that these insects might reduce the amount of nectar and be a cause of short honey crops. "A Lygus bug seriously affects the blossoming of alfalfa, the infested plants being short between the nodes and the flowers failing to develop, supposedly because of some secretion injected by the insect. Chemical control is not practiced against the thrips or L ygus bug, but these insects may be one of the factors responsible for the shorter honey crops obtained in recent years as compared with those of a decade ago. Garden Spraying or Dusting Beekeepers in Plain City, Weber County, reported bees actively work- ing asparagus plants for nectar and were afraid that arsenical insecticides might be applied during 1940 to control the asparagus beetle. This, however, was not done, and by July 27 the plants were through blooming. Considerable spraying of cabbage was reported in Utah County and a little dusting of cucumbers, although these operations apparently had no connection with bee losses in 1940. Potato fields in which fully half the plants were in bloom were systematically observed in Weber County, but no honeybees were seen visiting the blossoms. No beekeepers suspected potato sprays of giving them trouble. Use of Mosquito Larvicides One beekeeper in Salt Lake City believed that his bees were killed in 1939 by oiling operations for the control of mosquito larvae. However, mosquito larvicides have been applied in this district annually since - 9 1927, and crenaph has been incorporated in the spray since 1929. In 1940 oil was applied in a thin film on swampy ground and standing irrigation water, at weekly intervals beginning April 9, only where recent surveys had indicated that the water was actually infested with wrigglers. Sweet- clover adjacent to such places might accidentally receive a small amount of the spray, but this would occur too infrequently to be a serious danger to bees. Ducks suspected of having been killed by the oil in 1939 were reported autopsied and found to have died of botulism. Smelter Smoke Beekeepers in Salt Lake County suspect that arsenic in the smoke from copper smelters may be causing bee losses, and report that crawling bees are especially noticeable after a light rain following a drought period. Formerly about 20 smelters were operated in this valley, with no attempt to trap the arsenic in the smoke, but today only 5 are in oper- ation and practically all the arsenic is trapped from the smoke as a by- product. Some believe that past accumulations of arsenic in the soil may occur in dust that settles over the vegetation and thus might be eaten by bees working on plants covered with such dust. Soil samples collected near smelters at Murray and Midvale, Salt Lake County, were excessively high in arsenic, and one water sample collected near the Midvale smelter was also relatively high. In regard to the effect of prevailing winds on the distri- bution of smoke from the smelters, beekeepers claim that the smoke skirts the eastern edge of the valley, following the Wasatch Range, until winds from the various canyons blow it out over the valley. In Utah County the smoke from an iron blast furnace southeast of Provo has been suspected of injuring bees in the region. However, bees in a yard one-half mile south of this plant were not injured during 1939. The Tooele County bee inspector has reported that whenever colonies of bees have been moved to the neighborhood of the Tooele City smelter such colonies have always died out. In an experimental apiary in Salt Lake City eight 4-pound packages of bees established early in May had all dwindled to about three frames of bees by July 6, whereas four overwintered colonies brought to the same apiary had built up to strong 3-story colonies, during the same period. No satisfactory explanation for the difference is evident, unless a serious poisoning of adults occurred shortly after the packages were established. The emerging brood in the overwintered colonies, on the other hand, may have enabled these to build up, while the package colonies were so weakened that they never regained strength. Infectious Disease Practically all bee samples sent in for arsenic analyses were examined microscopically for parasites. No cases of Nosema or any other diseases were found. - 10 starvation. Unfavorable Weather Conditions, and Faulty Management of Bees June is reported to be a critical month for beekeeping in Utah, extensive dearth of nectar and pollen occurring then. Early cutting of first-crop alfalfa frequently has prolonged this period. Progressive beekeepers at this time feed their colonies sugar sirup or boiled honey, or move them to mountain areas. Most colonies in eastern Juab County had to be fed during the latter half of June in 1940 to prevent them from starving, although they had appeared to have abundant stores earlier m the month. The scale colony in an experimental yard in 1940 near Kaysville in Davis County lost weight from June 3 to 24 but gained after that until September 13, its gains during most of July being excellent. From June 1 to September 28 its total net gain was 338 pounds and 1 ounce,, and over five full-depth supers of extracted honey were produced. All 12 colonies in the yard averaged more than 3-g- supers of honey during the same period. Considering that these colonies had no attention throughout the spring, had old or poor queens, and contained much drone comb in the brood nests, this represents an exceptionally good honey flow. The average amount of sealed brood was only 550 square inches the first of June and 316 square inches the middle of July. Adult populations were not determined accurately, but it was quite evident that all colonies were weaker than normal colonies at Laramie, Wyo,, under honey-flow conditions. The drought of 1939 in Utah was reported as exceptionally severe. Data for 12 weather stations in the area of bee losses (Brigham City, Farmington, Heber, Levan, Logan, Morgan, Ogden, Prove, Salt Lake City, Santaquin, Spanish Fork, and Utah Lake) showed an average annual precipi- tation 9.89 inches below normal. Precipitation in June was above normal, but in most of the spring and summer months it was below normal. The area averaged 35 consecutive days, mostly from late in June to late in July, without measurable rainfall, while one of the stations had 56 consecutive days, beginning July 6, 1938, without a trace of rain. Such conditions might cause serious dwindling in colony strength, even if the bees were not attracted to poisons. Many Utah beekeepers, in trying to attain a large number of colonies, make increase by natural swarming and by artificial division, while compara- tively few of them purchase package bees. Several beekeepers even made more than 50 percent increase from overwintered colonies. Increase is often undertaken late in June, and thus the colonies are not given sufficient time to build up for the honey flow. A few beekeepers hesitated to add supers during a good honey flow, even when they had drawn comb on hand, and this resulted in badly honey-bound brood nests. A scale colony belong- ing to a beekeeper in Utah County showed its greatest increase July 22, and the following morning another beekeeper in the same territory reported the greatest death rate among his bees. 11 Many beekeepers carelessly allow drone comb to remain in the center of the brood nests. One beekeeper had used half sheets of foundation in the brood nests, with the result that over half the brood was drone instead of worker brood. The majority use burlap over the top bars both summer and winter. Many have poor equipment. A few give their colonies very little care. Better beekeeping practices in many apiaries should be of advantage in producing a honey crop and tend to counteract the losses of adult bees. LABORATORY INVESTIGATIONS During the summer of 1940 some 240 samples of bees and other materials from various parts of Utah were received and analyzed for arsenic at the Interraountain States Bee Culture Laboratory. A majority of the samples came from the fruit-growing areas (see fig. 1), although a few, mostly designated as normal bees, came from areas where arsenical poisoning was not a factor. The method of Cassil and Wichmann 2/ was followed in making the arsenical analyses. Samples of 100 bees were used where possible, the entire bees being used, and no differentiation being made as to arsenical material that possibly might have been on the body surfaces of the bees. The results of these analyses are given in tables 3 to 6. The first of these (table 3) summarizes the results for 153 samples of bees analyzed, on the basis of milligrams of arsenic trioxide (AS2O3) per bee. With the exception of a few samples from Logan, it was only in samples from Lehi and Pleasant Grove that exceptionally high amounts of arsenic were fcund, these being the two locations in which, as mentioned previously, serious adult bee losses occurred. Two samples collected at Lehi July 19 and 23 showed the presence of 0.023 mg. and 0.00449 mg. of arsenic trioxide per bee, respectively — ap- parently abnormally high amounts. (Samples omitted, table 4.) In these cases grasshopper bait had been spread July 13 and 14 on two small areas about one-eighth mile from the apiary. On one area bait was spread somewhat more heavily than recommended and just after fields had been irrigated. Nearby orchards had also been sprayed heavily July 18 and 20. Another sample collected July 24 to 25 shortly after a nearby -orchard had been heavily sprayed contained 0.00094 mg. of arsenic per bee. A sample collected at Pleasant Grove July 15, when bees were observed working on thistles in an orchard that was being sprayed, contained 0.00053 mg. of arsenic per bee. The three samples from Pleasant Grove which ranked next in amount of arsenic content were collected July 9, July 25, and August 6. During this period a slight amount of grasshopper baiting was done or had been 2/ Cassil, C. C, and Wichmann, H. J. A rapid volumetric micro method for determining arsenic. Jour. Assoc. Off. Agr. Chem. 22(2): 436- 445. May 1939. - 12 carried out previously in the vicinity, mostly in a strawberry patch close at hand, although nearby orchards were also heavily sprayed during this period. The sample of bees from Logan showing the highest arsenic content, 0.006 mg. per bee (sample omitted from table 4), consisted of bees captured while visiting sweetclover blossoms in a heavily sprayed orchard. In the case of the next highest sample from Logan, 0.00073 mg. per bee. apple orchards 2 miles away had been heavily sprayed shortly before the sample was collected. Table 4 gives the average arsenic trioxide content per bee for the samples referred to in table 3 after dividing them into two classes, (a) samples of bees including some live samples that had been designated as ap- parently normal when collected and (b) samples of bees suspected of having been poisoned. As will be seen in table 4, there is no marked difference between the average results for the various groups of these two classes, although the maximum results without doubt were significantly greater in most cases to account for the death of the bees. For purposes of comparison analyses were made of nine samples of normal bees obtained from laboratory colonies in and near Laramie, Wyo., which had not come in contact with arsenical poisons so far as known. These samples, ranging between 0.00004 and 0.00009 mg. AS2O3 per bee, aver- aged 0.00005 mg. per bee. This is considerably less than the averages for apparently normal bees in Utah, the results for which ranged between 0.00004 and 0.00036 mg. AS2O3 per bee, with an average of 0.00014 mg. per bee. Ten of the 153 samples of bees analyzed showed 0.00005 mg. AS2O3 per bee, or less. Some of these low samples came from areas in which no arsenical-con- trol operations were being carried on. It appears possible that bees in Utah have been subjected to contact with arsenic from various sources and that they contain somewhat higher amounts of arsenic per bee than do bees from other areas where insect-control measures have not been so prevalent or are lacking entirely. It is possible, furthermore, that the arsenic in samples of bees containing even only slightly more than the apparently normal amount may have been a contributing cause of death, since it has been demon- strated in experimental work carried on at the bee culture laboratory at Beltsville, Md. . that continued ingestion of small amounts of arsenic ma- terially shortens the length of life of honeybees. Table 5 summarizes the analyses of pollen, soil, and water from various sources. The pollen samples failed to show anything significant. None of the soil samples, except a few from heavily sprayed orchards and from the neighborhood of smelters, showed excessive amounts of arsenic. A few samples of soil that had had contact with grasshopper bait showed somewhat more than the average amount of AS2O3 per gram of soil. None of the water samples showed significant amounts of arsenic. Table 6 gives results of analyses of several miscellaneous materials. One sample of apple leaves and two samples of sweetclover blossoms from heavily sprayed orchards showed excessive amounts of AS2O3 per gram of material. In addition, two samples of Astragalus utahensis and two samples of Astragalus cibarius (Sheldon) , identified by Dr. Aven Nelson, Curator of the Rocky Mountain Herbarium, were found free of selenium by the Chemistry Department of the Wyoming Agricultural Experiment Station. 8, S 5 Qi 09 '^ 00X00*0 •^ OOIOO'O - 16000*0 '»n 06000*0 - 18000*0 '9m 08000*0 I^JOOO'O '3m 0^000*0 - T9000*0 'SB 09000*0 - 1^000*0 *9a 0^000*0 - 17000*0 •»» 0^000*0 - ICOOO'O '3m OCOOO*0 - 93000*0 *3m $2000*0 - T230OO*O •^ 02000*0 - 91000*0 ^ ■^ -^ r\ (*\ NO lO .—1 ■H txO C\2 i-H cv to C\2 O O o o o x: Si O O o o o O O o o o >3 >j CVI to -^ o c o c bC C c Xf CO : 'O c Q. o -Q QJ ^ a Ol 5< o w ^ 1° a I ■ S CD O) o o o o o o o o o o o o lO CD . — . fH o to o to to o o CD Q) L o o .—1 X X2 CD -P +-> •H to 00 c\; G (d f-. cv --I bD o o o o o o o o _) -u o o o t-, tn CD > o o f-, 4J CO w CO to CO CO >, 1 1 1 1 I-l CO (ti 1 1 >i ' > 1 a> ■H t- 1 o r-i a 1 1 nJ ' fn 1 o. ccJ • r-l {J5 s 1 Ch ' (0 o G 03 ^ . -"^ (d +-) CO cd 1 -H C -fJ -H , cd tM G !^ > (d _. CO o Q) to to H "^ > o •? r-H •H O rt u 0^ r-i o i:^ J 5 o J Q. O CO J-4 a 3 a a o X o 2 w u o Si bO ■H Xi bfl •H (d G bO x: Xi to c •H c © 1— 1 CO rH a a cd CO to © c^-4 Ot i-i r-t i-t HnOI ^ C«t H H «A » d H H H >|S ■i '•a •I u i g oa O CO cS » 13 ^ « 3 A •OH H CV CO i «> • +3 ^;^^§, .§1 ^ ° § ^ fe