I echo Becky's excellent advice, especially to map your seedlings, as markers can fade, get broken or lost, or be carried off by critters. Also to keep notes both paper (as computer hard drives can crash, and you may not have backed up your files
), and on the computer (as paper can get lost or waterlogged or otherwise damaged beyond readability).
Between lost or waterlogged paper notebooks, and any number of missing, broken, or faded labels, I now have several NOID seedlings.
As for numbering or labeling schemes...
I don't know what the people who crank out thousands of seedlings do, but I can't imagine that they number and label every single seedling prior to bloom.
In my case, I am just a backyard pollen dabbler, so the labeling is not too complicated, and I actually put some thought into my labeling scheme when I started converting to it a couple/few years ago. The scheme is as follows (and yes, you MUST record cross information down on the computer, and preferably on paper copy (you can print out your computer file)):
C#-YY-nn where C indicates Cross, # is a # assigned to a particular cross, YY indicates the year that the cross was made, and nn indicates the number seedling within the cross. (I may or may not assign nn before bloom; it can save on labeling material to only assign nn after I have seen bloom, as some seedlings are instant discards.)
For example, C1-11-1 indicates a seedling from Cross # 1 (whatever parentage that may be, say 'Arctic Lace' x 'Ballerina on Ice'), made in 2011 (seeds are harvested in 2011), and "1" indicates a particular seedling.
I could, of course, just label the seedling as "AL x BoI" - YY- nn (following that same example), but apart from taking more label space, I think that knowing the parentage (it's on the label, practically in your face) when looking at the plant is at odds with being able to dispassionately evaluate the seedlings on their own individual merits. "Hiding" the actual parentage behind a generic C# helps, I think, to be a bit less emotional in evaluating the resulting seedlings.
That said, it is sometimes advantageous or desirable to modify this approach a bit... thus, I might designate a group of seedlings by P#-YY-nn, if the intent of the cross was to (hopefully) produce polymerous daylilies. That exception to the rule tells me to look for any sign of polymerous blooms, and track the % of each on these particular seedlings. This "special cross" approach can be used to accomodate whatever traits your particular hybridizing focus is on (D for double, E for early, or EMO for early morning opening and so on) - all while still hiding the particular cross. When the seedling blooms, you can look at the tag and then ask yourself if the seedling exhibits whatever trait(s) you were aiming for. (The thing to be aware of here is
to not be blind to any good traits the seedling has, which may fall outside the scope of whatever the particular cross was aimed at. In other words, and as an example, don't throw out a good EMO-CMO or rust resistant seedling just because the plant produces only single blooms, and that cross was aimed at producing doubles or polys or crested or whatever daylilies.)
Of course, I think that whatever scheme you end up using, you must be able to accomodate exceptions. I have experienced the following situations where the usual labeling approach was useless (and there may be more):
M#-YY-nn I use this nomenclature when somehow seeds (or seedlings) are known to have been mixed. Maybe I absent-mindedly threw pods from two different crosses into the same baggie, and then realized my error. Maybe, when trying to pry seedlings out of an unmapped multi-cell tray, all of the labels fell out. In such cases, I resign myself to
"M" = Mix, and record what the possible/probable crosses are under that umbrella. (One year I had 4 such "mixes"...)
SV-YY-nn This nomenclature is used for
Spring Volunteers. I usually have at least a few of those, due to seeds flying as I collect pods (I tend to collect seeds after the pods have opened). In this case, the YY designates the year that I discovered the volunteer, and the presumption is that the cross was made the year prior to YY. Earlier this week I had FFO on SV-14-13, a seedling that I found in spring of last year, presumably from a 2013 cross.
While you may never know with certainty what the Spring Volunteer parentage is, you can sometimes figure out what at least the pod parent may have been. Seeds tend to fall near their parents, and if there was only one parent you set pods on in the neighborhood of the seedling, that is probably your pod parent. If you keep
really detailed notes, you will know what every single pollen parent used on that pod parent was (and if the cross produced pods or not), and that will give you some idea of the possible pollen parent of your volunteer. (I don't go to such detail.)
In the above example, SV-14-13 was found several feet away from 'Belle of Ashwood' but in the same raised bed - information that I had dutifully recorded. (Yes, this locality information for the volunteers must be recorded somewhere (again, computer and paper files).) As BoA was the only pod parent within 12 feet or so of the seedling, and as the seedling bears some resemblance to BoA, it is a pretty good guess that BoA was the pod parent.
Similarly, SoH-12-1 was a volunteer found underneath 'Sacrament of Healing' in the spring of 2012, and the next closest pod parent was 12 or more feet away, so it is almost certainly the case that SoH was the pod parent for that seedling (and the seedling looks it). (Yes, the labeling here is another exception... one might argue that this should have been a SV-YY-nn label, but I was about 100% certain that SoH was the pod parent, and so I labeled the seedling as such.)
(Regarding the pollen parentage, alas, I do not obsessively record all of the attempted crosses on any one plant, though I do know some of them, and try to record the crosses that produced seeds. Sometimes, though, the clip or label on one or more pods gets lost, and depending on what crosses you made, it can be problematic concluding what the pollen parent was, even if you know for certainty what pods were actually produced. In the case of SoH-12-1, every pollen parent used on the few blooms was a near white... since the seedling is also a near-white (actually light cream), I can't with certainty say who the pollen parent was (though I have a suspicion).)
Sometimes you get bee pods which you don't bother picking off, or your pod clips fall off, or your pod label gets unreadable. If you want to keep these individual pods separate, you can use record them as a C#-YY-nn group for that pod, noting the situation (bee or lost clip). If you want to group the pods (from the same plant) together, you can use a C#-YY-nn, noting a mixed group of pods from the same pod parent - OR you could use M#-YY-nn. (I'm assuming nobody would want to go to the extreme of clumping
all bee or unlabeled pods from
all parents into the same M#... though that is a possibility, too.)
Finally, there is the good old
X#-YY-nn, for
"X", the Unknown. Two possible sources are as follows: You collected the pod, but you forgot to mark the baggie (and you collected so many pods from so many plants that you haven't a clue which plant this came from... because you don't obsessively record Every Single Pod on your computer). Or the baggie got labeled, and the seeds got thrown into the fridge with a bit of water to stratify them, but the baggie leaked and wiped out the label... and again, you can't reconstruct which cross this was.
Whatever seedling numbering scheme you come up with, I think that you need to be able to accomodate these odd cases.
Good luck to you!
) 
