1. Mechanics
Most stocks can be successfully cultured by periodic mass transfer of adults to fresh food. Bottles or vials are tapped on the pounding pad to shake flies away from the plug, the plug is rapidly removed and the old culture inverted over a fresh bottle or vial. Flies are tapped into the new vessel, or some shaken back into the old one, as necessary, and the two are rapidly separated and replugged. Good tossing technique combined with plugs that are easily removed and replaced result in very few escapees. You will learn from experience which stocks require a medium or large inoculation of adults and which do better with only a few.

The frequency with which new subcultures need to be established depends on the health and fecundity of the genotype, the temperature at which it is raised, and the density of the cultures. Temperature has a large effect on the rate of Drosophila development. Generation time (from egg to adult) is approximately: 7 days at 29oC, 9 days at 25oC, 11 days at 22oC, 19 days at 18oC. For most purposes stocks are maintained by live culture, transferring adults to fresh medium every few generations. Stocks kept at room temperature should be transferred to fresh food every 20 to 30 days. Mites and mold are more likely to be a problem in older stocks, so it is good policy to set 30 days as an absolute upper limit for room temperature stocks. This period can be extended by keeping stocks at lower temperature. 18oC buys more time than 22oC, for example, but a significant number of genotypes fail to thrive at 18oC, and mold can be a serious problem. It is wise to keep a room temperature backup of stocks to be maintained at low temperature for the first two or three transfers in case the stocks do poorly. If the quality of your fly food is unreliable it is wise to have at least two cultures for each stock, staggered to assure the use of different batches of medium (at least until you find a new cook).

Cryopreservation of ovaries (see Ashburner, 1989) or embryos (Cole et al., 1993; Steponkus and Caldwell, 1993) are viable alternatives to continuous culture for some purposes. Genotypes that are unstable due to reversion, breakdown, or accumulation of modifiers, especially those with non-visible phenotypes that are time consuming to select, are good candidates for freezing. Also, if you are generating hundreds of stocks that will not be in use but must be kept for many years it might be cost-effective to maintain these as frozen stocks. For most routine stockkeeping purposes, however, live culture remains the preferred route.

Identify stocks with tags showing the complete genotype of the stock, sans shorthand. Writing the genotype on the vial or bottle at each transfer invites transcription errors and takes longer than moving a tag. Don\'t use a stock center stock number or other potentially cryptic symbol as the only identifier of a stock. Stocks are often kept for many years and what is obvious to you now may be meaningless even to you in a few years, and is easily misinterpreted by someone inheriting your stocks. Unless you are careful to maintain complete stock data elsewhere, record all relevant information on the tag.

2. Balanced stocks and balancers
Mutations that are homozygous viable and fertile are most easily kept as homozygous stocks. Lethal or sterile mutations must be maintained in a heterozygous state. A balanced stock is one that regenerates the same set of heterozygotes each generation so the stock can be maintained by mass transfer of adults instead of by mating specific genotypes each generation. A simple balanced lethal stock carries different recessive lethals on each of the two homologues, allowing only heterozygotes to survive. Dominant male or female steriles (Ms or Fs) can be maintained in stock without selection by double balancing - one Ms, one Fs and one recessive lethal. Fs/lethal male and Ms/lethal females will be the only fertile genotypes present in the stock each generation.

In most cases, balanced lethal schemes work only if one of the lethal chromosomes is itself a balancer chromosome. Recombination between lethal (or sterile) mutations on different homologues can produce one homologue with both mutations and one wild-type homologue. The wild-type chromosome will rapidly predominate or become fixed in the stock. Balancers are structurally rearranged chromosomes that prevent recombination between homologues in females (meiotic recombination is absent in D. melanogaster males and in the tiny 4th chromosome in females). This is accomplished in part by reducing recombination directly and in part by preventing transmission of recombinant chromosomes. The most commonly used balancers carry overlapping sets of inversions and prevent recombination throughout most of the length of the chromosome. Some special purpose balancers work well only for specific regions of a chromosome. Suppression of recombination is less effective when balancers for two or more heterologues are present in a stock.

Drosophila is relatively pestilence-free, but mites, fungi and bacteria can be problems in laboratory cultures. It is good practice to clean your bench top and fly pushing equipment regularly. This is particularly important if a problem is evident. Clean the bench top and all equipment that comes into contact with potentially contaminated stocks with 10% bleach, 70% ethanol or soap and water after use. Sharing pounding pads, CO2 pads, fly pushers and sorting plates can aid the spread of contaminants. If sharing is unavoidable, the need for cleanliness should be understood by all and enforced.

1. Mites
\"I have mites\" is not an admission you want to have to make to your fly colleagues (but you must make it, if true). The most dangerous species are egg predators, but even those that simply feed on the medium can out-compete weak genotypes and compromise experimental observations. Frequent stock transfer, tight plugs, and zero mite-tolerance by all the fly workers in a building are the best defenses. In general, cultures grown at 24-25oC should never be kept for more than 30 days. If mites are known to be a problem in your lab or building, cultures should be checked and discarded after 18 to 20 days. Lining stock trays with benzyl benzoate-treated cheese cloth (soak cloth in 10% benzyl benzoate in 95% ethanol, air dry; replace the cloths every 6 months) may help prevent infestation. Some kinds of plastic are dissolved by benzyl benzoate, so test first if you use plastic vials or trays (the cloth will fuse with the plastic within 24 hours) and protect paper items such as stock tags from direct contact with the cloths.

To prevent the importation of mites from outside sources all stocks new to your lab should be quarantined for at least two generations. Never open a foreign bottle or vial at your fly bench (or your neighbor\'s) without first inspecting the culture for mites. Using a microscope, examine the surface of the medium and the walls of the container, especially around pupae or pupal cases. If no mites are evident, replace foam or paper stoppers with tight cotton plugs and isolate cultures in a quarantine tray. As an added precaution, cultures can be wrapped in the mite cloths described above. Keep the original bottle or vial for about 20 days, even though you have established fresh cultures, rechecking for mites every 5 to 10 days. We check the new cultures too, just to be safe, but we have never found mites in a subculture when the parental vial was mite free.

Any culture found to contain mites should be frozen or autoclaved immediately if it can be replaced from a mite-free source. If replacement is not possible, use one of the methods described in Ashburner (1989) to disinfect the culture, such as daily transfer of adults for about a week, using only the final transfer to establish a new and mite-free, it is hoped, culture. Keep infected cultures wrapped in mite cloths until they have been mite free for three generations.

2. Fungi and Bacteria
If mold is a problem in isolated cultures it can usually be eliminated by daily transfer of adults for 7-10 days. Visually inspect cultures from the later transfers for hyphae (look around the pupal cases) and use one that appears to be free of fungal growth for further subculture. In extreme cases this process may need to be repeated for an additional generation. If fungal contamination is a wide spread problem be sure that fungal inhibitor (p-hydroxy-benzoic acid methyl ester) is being added to the medium after it is cooked (boiling destroys the inhibitor), add a small amount of live baker\'s yeast to every culture (the yeast tends to inhibit the growth of unwanted fungi) and check for sources of infection in the lab, such as incubator fan housings. Clean any contaminated or suspicious areas with disinfectant.

A variety of bacterial contaminants can occur in fly cultures. The most common problems are caused by mucus producing bacteria. Although not directly toxic, larvae, and to some extent adults, become trapped in the heavy layer of mucus that coats the surface of the food. Large numbers of larvae overcome the effects of the bacteria in a healthy stock, but weak stocks or pair matings can be seriously compromised. A widespread bacterial problem may indicate that the pH of your medium is too high; try lowering the pH to about 5. Individual stocks can be treated with antibiotics for one generation. A quick approach that often works: add 100 祃 of penicillin-streptomycin solution (10,000 u/ml and 10,000 礸/ml, respectively) to the surface of the medium in a vial and allow it be absorbed. Add a small amount of yeast and transfer flies to the treated medium. Discard adults before progeny eclose; subculture progeny on untreated medium. Other antibiotics may be tried if the contaminant proves to be resistant to penicillin and streptomycin.

Alternatively, clean cultures can be established from embryos dechorionated with 5.25% sodium hypochlorite (liquid household bleach, full strength). A convenient method is to transfer eggs to a bleach soaked wedge of filter paper, wick away bleach after chorions have dissolved (3-5 minutes), wash eggs several times with water and transfer to a fresh piece of filter paper (small enough to sit on the surface of the medium in a vial) moistened with water. Place the filter paper with eggs into a fresh vial of food and place a larger strip of filter paper along the wall of the vial. Wet this strip of paper to maintain high humidity in the vial until the eggs hatch.

1. Matings
While mass transfer of adults works well for most stockkeeping purposes, it often results in overcrowded cultures. Overcrowding can effect the outcome of crosses and experimental procedures. Development time is slowed, different genotypes may be disproportionately affected by competition for food and pupation sites, and many pupae and adults will drown in the soup of larvae and liquified medium. The best yield of healthy adults is obtained from cultures established with an optimum number of animals. Expect 50-100 adults from a vigorous 8 dram vial culture, 300-600 from a comparable half-pint bottle culture. For most genotypes the optimum number of females will range from 1 to 3 per vial and 5 to 20 per bottle. Set up a few test bottles or vials to determine this number empirically for the genotypes involved; control the age of the food, the age of the females, and the number of days the females are left in the vials. One or two males are usually sufficient to rapidly inseminate several females, but some genotypes will require an equal or excess number of males to insure rapid mating. If necessary, the effects of overcrowding can be reduced mid-culture by adding baker\'s yeast and a tissue (such as a Kimwipe ) to provide additional nutrition and pupation sites, respectively, for the excess larvae. Extremely crowded cultures are best dealt with by distributing larvae (scoop them out with a spatula) to several fresh bottles or vials.

If you cannot distinguish parent from progeny by phenotype, parents should be discarded before the progeny begin to hatch. Experimental crosses maintained at 25oC should be discarded after 18 days to prevent recovery of second generation progeny. An effective schedule is to establish crosses on day 0 (start on a Friday if you want to begin virgin collection on a Monday), discard adults and add yeast and papers (optional) on day 7, collect virgins or score progeny on days 10 through 18.

Some mutant phenotypes are affected by temperature or genetic background. Before setting up a large scale effort such as a screen, make a test cross of the relevant genotypes under the conditions to be used and confirm that all phenotypes are scorable. It is also prudent to assure the absence of \'background\' lethals in a stock to be used for mutagenesis by isogenizing a chromosome for use in a screen. To isogenize a ri e chromosome, for example, cross to an appropriate balancer stock, recover 10-20 progeny heterozygous for ri e and the balancer, backcross them individually to the balancer strain, cross sibling ri e/balancer progeny, and then recover ri e homozygotes from one of the lines and establish a stock. Only lines carrying lethal-free chromosomes will produce homozygotes among the progeny of the sib matings.

2. Virgins
Most experimental schemes require virgin females. D. melanogaster adults do not mate for about 10 hours after eclosion, allowing virgins to be collected within 8-12 hours after the culture has been cleared of adults. The timing of this virgin window appears to be genotype dependent. Collect females within 8 hours to be safe, or determine empirically for a given strain how long you can wait and still recover virgins. Ashburner (1989) describes a variety of environmental and genetic tricks to facilitate the collection of large numbers of virgin females.

For many mating schemes virginity is desirable for efficiency\'s sake, but not essential because the progeny of non-virgin females can be distinguished phenotypically from the progeny of interest. If your scheme requires virginity (e.g., male fertility testing, or the progeny of non-virgin females are indistinguishable from those of the intended mating), hold females for 3-4 days and check for larvae in the holding vial before using the females in matings. Don\'t overcrowd females in holding vials - 50 or so in an 8 dram vial, fewer if you aren\'t sure of their virginity (you\'ll have to discard all of the females in the holding vial if any have mated). The peak of female fertility is genotype-dependent, but on average females are best used between 4 and 10 days old.