crease the numbers of many kinds of wild bees. Details for carrying them out would depend upon many local factors; local conditions would probably call for certain additional measures.

1. Apply insecticides to blossoming plants only when there is no other way to control the harmful insects. Such applications should be made between. 7 p. m. and 7 a. m. and should contain only toxaphene, methoxychlor, or other toxicants demonstrated to be relatively safe for bees when used at the = proper strength.

2. Provide a continuous supply of bloom throughout the season. Forage crops such as vetch, clover, and alfalfa make a good series lasting from late spring through summer. Fruit trees, maples, hawthorns, elderberries, and other hedgerow plants generally provide needed spring forage. Of course, each area would be best served by the plants suited to its own climate and agricultural needs.

3. Establish and maintain hedgerows around agricultural fields and along roadways, ditch banks, and canals. Pithy-stemmed plants such as elderberry, sumac, and tree-of-Heaven should be encouraged in such hedgerows. Light browsing would make them more suitable for nesting than if they were left undisturbed.

4. Hollow-stemmed plants such as milkthistle, wild parsnip, canebrake, and teasel should be broken over after the stalks are well developed. These will provide nesting places for leafcutting bees and harbor many hibernating species.

5. Establish and protect areas of bunch-type perennial grasses, especially along the tops of banks. They will provide nesting places for bumble bees and tend to stabilize and shelter the banks. Banks so protected, especially if nearly vertical, are ideal nesting places for many kinds of bees.

6. Preserve known nesting sites of gregarious bees from being cultivated, flooded, trampled, or encroached upon by dense vegetation. Expand the available nesting ground if necessary, and

establish new areas with the same conditions as populated sites. In the past few years many nesting sites of the alkali bee have been discovered by alfalfa-seed growers. Once apprised of their value, the growers have usually been willing and even anxious to keep them in an unaltered state. Several and perhaps most of the gregarious species of bees migrate in large groups to newly prepared areas. If other conditions for population increase are favorable, it should not take long for new areas to be populated.

Another approach to the problem is through better utilization of available populations of native pollinators. The following principles should apply to

many crops.

1. Grow the crop in areas where native pollinators are known to be abundant. In most cases such areas will be adjacent to or surrounded by untilled land.

2. Limit the acreage of the crop in bloom at one time to that which the native pollinators can handle.

3. Reduce competitive sources of pollen and nectar.

4. Time the blooming of the crop with the period of greatest natural abundance of the pollinator. (In general, only forage crops would be concerned here.)

GEORGE E. BOHART, a member of the division of bee culture of the Bureau of Entomology and Plant Quarantine, is in charge of the pollination studies in connection with the production of legume seed, conducted at the Legume Seed Research Laboratory in Logan, Utah.

The attention of the reader is directed to the section of color drawings in which appears a drawing of an alkali bee (Nomia sp.) tripping an alfalfa blossom and the nesting sites and life stages of the bees. Opposite the drawing is a description of the life history and pollination activities of alkali bees.

Breeding Bees

Otto Mackensen, William C. Roberts

Because honey bees produce honey and beeswax and help pollinate many plants, improving them through breeding benefits beekeepers and farmers.

Man has kept bees for ages, but selective breeding of bees has lagged far behind that of other domesticated animals and plants. The main reasons therefor stem from the social nature of honey bees, the mating of the queen and drone away from the hive, and a lethal mechanism that may kill a large percentage of eggs and brood.

Honey bees will not mate, reproduce, or survive in isolated pairs as nonsocial insects do. Each colony consists of a fertile queen and her many infertile daughters, the worker bees. All contribute to the performance of the colony. The colony rather than the individual is the unit upon which the selection of breeding individuals must be based. After a superior colony has been chosen, one can only use as breeding individuals the virgin queens that are sisters to the workers and the drones that are sons of the queen.

The breeding quality of a colony can be obscured by environmental factors. A colony's large honey crop might be the result of its robbing activities rather than its industry in bringing in nectar from the field. A queen might have a high egg-production potential, but the actual number of eggs she lays each day depends on the size of the population of the colony, its food, and space.

The drone, which develops from an unfertilized egg, is haploid-he carries only a single set of chromosomes and genes, the tiny elements of heredity. The sperms he produces are all genetically identical; they carry the same genes as the drone himself. The queens

eggs, are diploid; they carry a double set of chromosomes and genes. In the production of eggs, the genes segregate, so that each egg carries a sample half of the genes of the queen. A queen may mate with one or more drones, but after her mating period she does not mate again.

The sperms are stored in the queen in a spherical structure called the spermatheca, and released a few at a time as the eggs are laid. If a queen is mated to one drone, all the workers of the hive receive identical genes from the drone and all the genetic variability comes from the queen. The same is true of queens reared from the colony. If the queen mates with two or more drones, there will be greater variability in the workers.

The control of parentage obviously is essential to breed improvement. In this the honey bee presents a special problem because the queen leaves the hive to mate. She returns in about 15 or 20 minutes, bearing evidence of having copulated, but how and where mating takes place is still a debated point. Attempts to mate queens in confinement have failed.

Of the various methods employed to control mating, two have proved most practical: Isolation, by placing colonies containing the 'breeding individuals (virgin queens and drones) in a location far away from other bees; and artificial insemination, by taking semen from one or more drones and injecting it into the queen by means of special instruments.

Each method has advantages and disadvantages and a place in breeding programs.

virgin queens can be mated at once and At a mating station several types of in large numbers. Only one type of drone can be allowed to fly there at one time, however. Because stray swarms may drift into the area unnoticed, one can never be certain that isolation is entirely effective, even when requirements of distance have been met. Individual matings cannot be distin

guished because many queens mate more than once.

With instrumental insemination, on the other hand, control of parentage can be absolute. Many types of drones can be used simultaneously in the same queen yard and individual- or multiple-drone matings can be made at will. Because the operation is time-consuming, however, only a limited number of inseminations can be made.

THE DISTANCES required for complete isolation at a mating station under various conditions have not been determined fully. In some experiments, virgin queens did not mate when the nearest source of drones was 6 miles away. A shorter distance may be all right when enough drones of the desired type are provided or when the location is geographically isolated, as by mountains or a body of water.

For 2 years the division of bee culture of the Bureau of Entomology and Plant Quarantine maintained a mating station at Grand Isle on the marshy coast of Louisiana. It was 20 miles from other bees and was adequately isolated. Queens and drones of a highly selected yellow strain that were mated there showed no evidence of having mated with strange drones. Kelleys Island, Ohio, 4.5 miles from the mainland in Lake Erie, has been used since 1948 as a mating station for the mass production of hybrid queens. It also is adequately isolated. Virgin queens did not mate during a complete lack of drones on the island although mainland colonies 5 or 6 miles away contained drones. Islands, marshy coasts, and desert areas probably offer the best locations for isolated mating stations.

INSTRUMENTAL INSEMINATION can be learned by anyone, but because expensive equipment is needed and it takes so much time, few beekeepers have made use of it.

The principal instruments needed are a stereoscopic microscope, a device for administering carbon dioxide as an anesthetic, and the insemination ap

paratus itself (the manipulating stand, holding hooks, queen holder, and syringe). The procedure, as we practiced it in 1952 and the results to be expected are given in the paragraphs that follow.

The queen is allowed to back into the queen holder tube until the end of her abdomen projects. She is secured by means of a stopper, through which carbon dioxide is flowing, and placed in the manipulating stand. Then the chitinous plates at the tip of the abdomen are separated with the holding hooks to expose the genital opening. Semen is taken into the syringe from the drone and injected into the genital opening of the queen. The operation takes about 5 minutes.

The drone penis is relatively large and in copulation turns inside out, bringing the semen to the end of the everted penis. In artificial insemination, partial eversion is brought about by exposure to chloroform fumes and completed by pressure. The semen appears at the end of the everted penis as a cream-colored fluid accompanied by a white mucus. After a little practice, the operator can easily take the semen into the syringe.

A tonguelike structure, the valve fold, obstructs the opening to the oviduct. The fold must be pulled aside to permit the end of the syringe to pass. The semen must be deposited in the oviduct for successful insemination. The complete or partial failure of early investigators was probably due in large part to ignorance of this structure or disregard for it. From the oviducts, the sperms migrate into the spermatheca.

During the insemination operation, carbon dioxide is allowed to flow through the queen holder. It acts as an anesthetic and it also stimulates early oviposition. Properly used, it reduces the average age at initial oviposition from about 40 days after emergence in untreated queens to 11.5 days, which is about the age of initial oviposition after natural mating. Three exposures of 10 minutes' duration are given at 1-day intervals. With or without in

semination, this will stimulate egg production in queens. Other anesthetics and electric shock have the same effect.

By using a counting chamber slide to count samples of sperm with a microscope, we found that the spermathecae of naturally mated queens contained an average of 5.73 million sperms and those of queens inseminated with sperm from a single drone contained an average of 0.87 million. About 2.5 cubic millimeters can usually be taken from three drones. The average number of sperms reaching the spermatheca when this amount was given one, two, three, and four times were 2.97, 4.11, 4.85, and 5.52 million, respectively. The number varied greatly in individual matings, but the variation was not so great when the amount of semen and number of injections increased. When three and four inseminations were given, the variation was less than that of naturally mated queens. Two inseminations of about 3 to 4 cubic millimeters have given excellent results. Queens so inseminated have performed as well as naturally mated queens of the same parentage.

Queens that lay unfertilized eggs in worker cells, where they intend to lay fertilized eggs, are called drone layers. With single-drone inseminations, a few queens will be partial or complete drone layers at the start and others will later become drone layers. Some have performed satisfactorily in small hives for a year or more, however. When larger inseminations are given, drone layers are rarely found at the beginning of egg laying.

The percentage of good laying queens obtained varies with the stock and rearing conditions. The results we got in two lines during the 1950 season

give an idea of what might be expected. Of 38 queens inseminated in one line, 36 (95 percent) started laying; 33 (87 percent) were considered good queens. In another slightly inbred line, 32 queens were inseminated; 25 (78 percent) started laying and 22 (69 percent) were considered good queens. All in both groups produced good worker brood.

Several men have made noteworthy contributions. Lloyd R. Watson, of Alfred, N. Y., developed a workable syringe and was the first to demonstrate (in 1926) a successful technique that could be used by others. W. J. Nolan, of the division of bee culture, developed the basic manipulating apparatus and queen holder. Harry H. Laidlaw, of the University of California at Davis, made a detailed anatomical study of the reproductive or. gans of bees and pointed out the importance of the valve fold.

We have been striving to improve the instrumental insemination apparatus and technique since 1935. We found that semen could be collected more quickly from the completely everted penis than from the seminal vesicles or dissected parts of the partially everted penis, as earlier workers had done.

The greatest improvement in apparatus was that made in the syringe. A glass syringe tip was constructed that tapered at the end to an outside diameter under 0.3 millimeter (0.012 inch) so that it could be inserted into the oviduct, which usually has a diameter of 0.33 millimeter. The plunger barrel had an inside diameter of 0.41 millimeter; that was large enough to give the syringe a capacity of about 3 cubic millimeters, or the amount of semen

usually obtained from three to four drones. Later a less breakable tip of the same type was made of plastic.

A syringe was finally designed that does away with the troublesome tightfitting plunger. It employs a rubber diaphragm. The base of the tip fits against the diaphragm and has a coneshaped depression into which the diaphragm is pushed by a plunger activated by a screw. In use, the tip is first filled with water, then some of the water is pushed out, and semen taken up in its place. This tip has a main. barrel of 0.025 inch (inside diameter) and then tapers 0.156 inch to an inside diameter of 0.006 inch. The outside diameter at the end is 0.010 inch (the average oviduct diameter is 0.013 inch).

A LETHAL MECHANISM in bees complicates breeding. To clarify this mechanism some preliminary explanations

are necessary.

We have explained the location of the hereditary determiners, the genes, in linear order on chromosomes and how they are inherited in bees. When genes are paired as in the diploid queen, and one member has a different action than the other, each gene is called an allele of the other. Individuals in the population may carry still other alleles at the same locus on the chromosomes, and there may be a large series, each having a slightly different action. Fertilization may bring together various combinations of these alleles.

Genes have various effects. Some have detrimental effects. In some instances this effect is so great that the gene kills the individual inheriting it. Such a gene is called a lethal.

The lethals we are concerned with are members of a series of alleles which we have designated as a, b, c, d, et cetera. Females (queens and workers) are always heterozygous—that is, they contain two of these alleles that are different, as, for example, a and b. A queen of this composition produces eggs one-half of which are a and one

970134°-52-10

half are b. Since the drones develop from unfertilized eggs, one-half of the sons are a and produce only a sperm, and the other half are b and produce only b sperm.

When an egg such as a is fertilized by a sperm carrying a different lethal allele such as b, a queen or worker having the composition a/b results. If it is fertilized by a sperm carrying a similar allele (a), this homozygous combination (a/a) causes the individual to die before maturity, usually in the egg stage. When a queen (a/b) is mated to a single drone carrying a different lethal allele such as c, then all the progeny resulting from fertilized eggs will have lethal alleles dissimilar (a/c, b/c) and will be viable-able to live to maturity. Efficiency in the brood nest will be high, and a populous colony will result. If, on the other hand, she is mated with a single drone having a similar lethal allele such as b, then one-half of her progeny resulting from fertilized eggs will be a/b and viable, and one-half will be homozygous (a/a) and will die. Because most of the dying eggs are not removed until hatching time, 3 days after they are laid, efficiency is low in such a brood nest, and a weak colony will result.

Failure of selection to eliminate the lethals indicates that a nonlethal gene does not exist at this locus. Inbreeding reduces the number of lethals in the population and increases the chances of similar lethals meeting to produce low viability. Outbreeding brings new lethals into the population and this increases the frequency of high viability.

A similar series of lethals in a related insect Bracon hebetor, better known to geneticists as Habrobracon, has been studied by P. W. Whiting, of the University of Pennsylvania. In that insect a definite association with sex has been established. Such an association has not been proved in the honey bee, where promotion of outbreeding may be justification enough for the existence of such a wasteful lethal mechanism.

As long as individual matings are made, the percentage of viable fertil