Background

In the July 2007 issue of The Scottish Beekeeper (Vol. 84, No. 7, pp. 188-192) there appeared an article with a similar title to the above, this being the translation by A.E. McArthur of a paper by Professor Kaspar Bienefeld, at that time working at Das Institut für Bienekunde, J. W. Goethe-Universität, Oberursel, Germany and published in German in Die Neue Bienen Zeitung (1991), Issue 8, pp. 33-36. It drew highly critical comments from two Scottish beekeepers, Jane Slight and Dave Stokes, which were published in the SBK September issue (Vol. 84, No.9) along with a point-by point response by Eric McArthur, which however leaves some issues unresolved. The present article is intended as a clarification, critique and extension of those ideas.

Eric McArthur did us a considerable service in translating Professor Bienefeld's paper, as it describes what I believe to be one of the most important contributions to honeybee genetics of recent years, which, so far as I know, is otherwise inaccessible in the English language.

Introduction

Those new to genetics need to know that all members of any species carry a characteristic number of genes for that species, that are distributed in (virtually) the same order along the chromosomes they inherited from their parents. Bees are unusual in that, whereas workers and queens each carry two sets of 16 chromosomes (i.e. 32 in toto), normal males have only one 16-piece set. Female bees therefore have two copies of each gene, normal male bees only one. Each gene controls a specific feature and in the population as a whole most genes have several versions, called"alleles". A female having two identical copies of an allele is "homozygous" for that allele, or "a homozygote". One with different alleles for a specific gene is a "heterozygote", while normal male bees, with only one copy of any gene, are described as "hemizygous".

It appears that Bienefeld's original article contained several major errors, including an inappropriate diagram and the idea that chromosomes in outbred individuals are "stimulated" by chromosomes from the other source. An even more fundamental error was to overlook chromosomal crossing-over during meiosis, and the consequent re assortment of genes that occurs routinely during the production of every ovum and drone and so, indirectly his sperm. This could mean that Bienefeld's concept of degrees of inbreeding are serious over-estimates. Nevertheless, despite these errors and deficiencies, some important findings come across. The study has considerable power, as the volume of data was large and the degree of inbreeding recorded, being based on 5,500 colonies led by instrumentally inseminated, or island-mated queens. What Bienefeld and his team set out to do was to compare the effects of inbreeding in queens with inbreeding by queens. Put another way, their analysis compares the effects of inbreeding on a colony when the queen has closely related parents, with when the workforce is produced by queens mated with closely related drones.

Degree of inbreeding is defined (in this paper) as the percentage of chromosome pairs in an individual that are identical because both are derived from an ancestor common to both parents. The study covers inbreeding of queens over the theoretical range of 0-45%, i.e. the least inbred have none and the most inbred theoretically 45% of their chromosome pairs identical by descent (although this apparently makes no allowance for crossing-over). If the latter estimate is correct these queens should be homozygous for at least 45% of their genes.

McArthur's translation refers to the (biologically impossible) crossing of "half-sisters" (in German, Halb-Geschwisteren). I suggest a better term is "half-siblings" and I assume what was meant is crossing between colonies led by queens with the same mother but different fathers.

Effects of inbreeding

Inbreeding is usually deleterious for three rather different reasons.

1. Recessive alleles

Alleles can be dominant or recessive to one another, but recessive alleles remain unexpressed in heterozygotes that also carry a dominant, i.e. recessives are expressed in females only when homozygous. They can be expressed in every normal hemizygous drone, in whom the concepts of recessivity and dominance are inapplicable, although for reasons of their function, expression of some genes is restricted to one or other sex.

Evolutionary considerations ensure that most beneficial alleles are dominant. Harmful recessives can be carried in outbred populations with little adverse effect, by virtue of heterozygosity. But they can be responsible for problems following inbreeding, as this favours their being brought into the homozygous state, so allowing their expression, with harmful consequences.

Inbreeding thus favours expression of maladaptive or harmful recessive traits, in workers and queens, including poor performance and susceptibility to disease.

2. Sex determination

In bees, development of female anatomy and physiology is triggered in the larva by heterozygosity at a specific locus carrying the sex determining gene, i.e. by the presence of two different sex determining alleles. By contrast, an embryo develops as male when only one kind of sex determining allele is present. This occurs normally in unfertilised eggs, described as "haploid", meaning having only one set of chromosomes. It can however also arise in fertilised eggs (described as "diploid", meaning that they carry two sets of chromosomes) if they are homozygous at the sex gene locus. There are 15-20 different alleles of the sex determining gene and in a truly random mating situation most eggs should be fertilised by sperm carrying alleles different from those in the eggs they fertilise. However, if a virgin queen mates with her brothers or cousins there is an increased probability that some of the resulting embryos will be homozygous at the sex gene, the lack of variation in allele type defining maleness. In this way inbreeding favours creation of "diploid drones" that develop in worker cells.

Diploid drone embryos are usually recognised and destroyed by nurse bees, which leaves empty cells in the spread of sealed worker brood, known as "pepperpot brood". This loss creates a steady toll on worker bee production and a relative weakening of the forager force, but it can provide an indicator that inbreeding took place at the mating of the queen and this may explain other deficiencies. It is, however, an unreliable indicator, as a similar effect is caused by European Foul Brood, while some queens will fill up the gaps with a second egg in each empty cell. Inbreeding by queens therefore leads to retardation of colony development through a constant loss of diploid drones.

3. Polygenic conditions

Quantitative characters, such as honey yield, tend to be polygenic, i.e. controlled by the combined effects of genes at several loci. Outbreeding gives rise to new gene combinations, simultaneously creating heterozygosity for many genes and offering scope for improvement with respect to polygenic, complex characters such as foraging capacity and disease resistance. This is the basis of "hybrid vigour", or "heterosis".

Inbreeding favours the opposite effect, called "inbreeding depression." Genotype is not the only important influence: genes define potential, but realisation of genetic potential requires environmental factors. These include availability of suitable forage at times that match colony development, appropriate weather, and the management skills of beekeepers.

Heterosis and inbreeding depression are probably most relevant to the typical lowland situation where forage may change year by year, exotic queens may sometimes be imported and present populations are not at the optimum degree of adaptation. On the other hand, if a population is already well adapted to an environment that is stable, such as one of native bees that has been resident beside the same heather moor for a thousand years, outbreeding may undermine that adaptation and lead to reduction in health and productivity. In such circumstances lack of completely unrelated mating partners may ensure retention, or retain "fixation", of adaptive alleles that have accumulated by long term elimination of less suitable alternatives. In these circumstance, and in this respect, inbreeding may be deemed beneficial.

According to this paper, Moritz showed as long ago as 1981 that only slightly inbred queens produce drones with a 75% reduced sperm count, so we should expect serious fertility impairment in the sons of inbred queens.

Honey and wax yield

Inbred colonies produce less brood and honey than outbred ones. The question Bienefeld and colleagues addressed is whether this is due to deficiencies in queens, caused by inbreeding by the workers' grandparents, or to poor performance by workers due to inbreeding by the queen herself. Somewhat surprisingly they found that when queens are themselves inbred, there is no detectable influence on honey or wax production. By contrast, every 1% degree of inbreeding in the workers reduced the yield of honey by 140g per colony and that of wax by 7%.

The most productive colonies are those led by queens that have mated with unrelated drones, creating hybrid vigour in the foraging force.

Colony behaviour

Two behavioural characters were also examined by Bienefeld's group: aggression toward the handler and calmness on the comb. They found that when the queen is inbred the workers are restless and aggressive. By contrast if the workers themselves are inbred they are generally calmer and more docile, this latter feature probably being an aspect of inbreeding depression. Bienefeld's explanation of the former effect, that inbreeding in the queen creates restless and aggressive workers, is that there are probably deficiencies in queen substance, making the colonies short tempered, as if partially queenless. If this explanation is correct, inbreeding in the grandparental generation may foreshadow supersedure of the laying queen.

In summary, the gentlest colonies are led by outbred queens mated to related drones, but the most productive by queens mated to unrelated drones.

Bee breeding strategy

The level of inbreeding in a population is raised if you breed from just a few colonies to the exclusion of the majority and in the long run this can be detrimental in many ways. Bienefeld suggests a workable compromise is to maintain a minimum of 50 colonies (of presumably initially unrelated bees) and to raise similar numbers of daughter queens from as many as 30% (i.e.15 colonies) per generation. If it is impractical to maintain such large colony numbers, no one queen line should be especially favoured as a producer of queens or drones. This he suggests should ensure retention of as many sex alleles as possible and could halve the rate of increase of inbreeding that frequently occurs with artificial selection.

Inbred stocks need not be rejected entirely, however, as they can be used to create first-generation inter-line crosses, which can be highly productive.

Conclusion

In summary, queens that are themselves inbred, but who mate with unrelated drones can give rise to strong, but restless and aggressive colonies with honey and wax yields that are high, but that produce drones with seriously reduced fertility. Queens that have mated with related drones produce inbred colonies, indicated by "pepperpot brood". These inbred workers tend to be calm and docile, have reduced workforce strength and produce less honey and wax than outbred stocks. Colonies derived in the first generation from inbred lines crossed with unrelated inbred lines can however be highly productive as they profit from heterosis.

It should not be taken from the above that the inbreeding/outbreeding parameter is necessarily the most important genetic consideration. Honey yield, aggression and restlessness, etc. are also under the control of individually selectable alleles and their effects may mask those of inbreeding. Degree of inbreeding is however very important in broadening or narrowing the range of performance characteristics of a honeybee colony.