Synchronization of larval development is important for staging and making sure we are comparing "apples" to "apples" across strains (i.e. making sure that differences in gene expression reflect strain differences at the same stage, not just relative developmental rates of each strain). However, even when larvae are developmentally synchronized by collecting freshly hatched 1st instar larvae, there is often as much as 12 hour difference in how long it takes such strains to reach the beginning of pupation. In addition to variation among different wild type strains, many developmental mutants also show developmental delay. So absolute age is not a good indicator of developmental stage (and we generally want the tissues at the same developmental stage not the same temporal age).
When developing at 25C wild type Drosophila melanogaster stages take approximately (with lots of variation among stages)
- fertilization of egg to hatching of 1st instar larvae: ~22 hours
- 1st instar - molting of 2nd instar: ~ 24 hours
- 2nd instar - molting of 3rd instar: ~ 24 hours
- 3rd instar to onset of pupariation: ~ 48 hours
Within each of these larval instars there is also considerable growth of tissues we care about (such as the imaginal discs). Indeed even during the ~48 hours (at 25C under good growth conditions) gene expression (both amounts and spatial distribution) of many key developmental genes is very dynamic.
In general we need to find the average time of developmental events (hatching, larval moultings and onset of pupariation) for each strain we use (it also can vary a few hours by sex as well). There is a brief discussion of this in Ashburner et al, Chapter 6. (pages 151-154). These include using hatching of larvae from the eggshell to get a "0" point for larval growth, and then the size and number of teeth of the larval mouthparts to distinguish larval instars. Finally, the end of larval development can be assessed with some clear developmental stages including the onset of larval wandering (although this is influenced by crowding and humidity), darkening of the spiracles, and purging of the remaining food from the gut (done by adding 0.05% bromophenol blue to the food). From these data you can generate the mean times of developmental events that you need to use. Keep in mind that growth conditions such as temperature, food quality and quantity and larval density can all influence these. Even if you use the same incubator, temperature, humidity, food recipe and larval density, you will still see some changes in mean developmental time among batches. That is just the nature of biological variation!
Generally then we want to get a "0" hour by generating freshly hatched larvae (see below).
We (in our lab) don't need to worry about transition from 1st to 2nd instar We need to estimate the mean time of molting (based on mouthparts) from 2nd to 3rd instar. Then using a number of characteristics (such as eversion of anterior spiracles, body shortening and being glued to the vial, and finally seeing the posterior spiracles turning orange which is the beginning of pupal development (prepupa, stage P1).
Once you have this information you can calculate (with associated variation) the times you will need for stages for your dissections (i.e. for late third instar wandering larvae (the final 24 hours of larval development at 25C), you could work back from the onset of pupal development or eversion of the anterior spiracles to the time you need (probably 6 hours before eversion of the spiracles for late 3rd instar).
For mid third instar you could do the half way point between the mean time to moulting of third instar, and the onset of pupariation.
In Ashburner et al. Drosophila A laboratory Handbook (big book, light blue cover) see chapter 6, in particular figures 6.15 (mouthparts for different larval instars), 6.16 (anterior spiracles), table 6.7 (number of teeth in mouthhooks), Fig 6.18 and 6.22. Essentially pages 150-157.
In Demeric (ed.) Biology of Drosophila (smaller forest green book, usually in the fly room), see chapter 4, but in particular pages 283-287, 294-298. Figure 4 in this chapter has a useful picture of both moutparts and anterior spiracles.
Oliviera et al. (2014). Coordination of Wing and Whole-Body Development at Developmental Milestones Ensures Robustness against Environmental and Physiological Perturbations. PLoS Genetics Link
Cohen S (1993) Imaginal disc development. In: Bate M, Martinez-Arias A (eds) The development of drosophila melanogaster. Cold Spring Harbor Laboratory Press. Available at McMaster Library. Some very useful information.