Joe and Kent,
I disagree with some of your statements, not because I don't understand what you are saying or agree with you in the strictest sense but because I believe your taking things to an extreme that is not realistic. At the strictest level, I think this extreme is dangerous to breeding efforts because its one step removed from "F- it, lets breed everything to everything without any attempts what-so-ever to maintain a natural state given the best information and knowledge we have available to us." The reality is that any coordinated effort to preserve any state (natural or unnatural) requires a certain degree of compromise to accommodate realistic constraints that we cannot escape.
I believe the biggest problem that lies between your standpoint and my own is with how we each are defining locale and the line between adjacent locales. We name pardalis locales based on geographic points in the proximity that help us communicate to others where something can be found. Some locales are heavily isolated (islands) while others are less so. Additionally, some locales that have been coined are closer together then others. The problem becomes establishing a reasonable standard for differentiation between locale assignments. If you say that "Locale AGH" is found along a 300m stretch between 1st and 3rd St and that "Locale BFE" is found along the same 300m stretch between 4th and 6th St, obviously your geographic resolution is so high that natural movements of individuals interferes with locale differentiation and the establishment of a differentiable locale standard is impossible. On the other side of the coin, if your resolution is too large, natural movements are no longer an issue but we have no ability to account for variations seen within the range that do occur in a discernible pattern.
Realistically, there needs to be recognition that just because a locale is named after a town does not mean that anything immediately outside the town's city limits are suddenly not that same locale. Lines are not that clean cut and defining them completely based on our own geopolitical lines is useless. The complication is that at the same time, if we are going to use that town as a guide-point to define a locale, a certain distance away needs to be recognized as different enough to warrant exclusion from that locale-type. This is a very fuzzy line and designation when you are talking about locales on the mainland without clear boundaries. Islands are easy but the issue comes with animals collected a certain distance away from a given geographic guide-point. I don't know that we can create a clear boundary standard and while this is problematic for absolute definition purposes of a locale, it should not be reason to outright ignore our best ability to group breeding efforts into a framework of locales based on our best knowledge and information. This is not an unreasonable task by any stretch of the imagination when you are comparing locales such as Ankaramy, Ambanja, Ambilobe, Diego Suarez and Tamatave, to name a few mainland locales with fairly distinct differences between them. When you start differentiating Ambanja, Ankify and Ambato (Nosy Faly could also be thrown into this group), it becomes more difficult but I would argue that with care, not impossible or unreasonable either.
One of the first steps in ensuring locale type purity is establishing a semi-loose definition of a locale based on appearance. As I said, this definition needs to be semi-loose because there are variations within populations and not every animal will match a definition perfectly. This same issue is seen in the definition of different species yet scientists are able to get around it and still differentiate in most cases (this is why holotypes and paratypes are used). By examining specimens at given locales in the wild, by looking at location specific photographs, by examining countless animals with reasonably reliable background information, etc., experienced individuals become much more accustom to differentiating locales. There are subtle indicators that end up being quite reliable in most circumstances, the same way certain features are good indicators of a species even when some might indicate something different. Developing an ability to do this takes time, practice and learning.
In a perfect world, when we bought something labeled as a particular locale, we could have reasonable blind faith that we are receiving what we pay for. Unfortunately we know that we can't simply rely on this all the time. Collectors can collect animals in places that are more accessible to them rather then in the area they tell their buyer. Exporters can fill orders with individuals from another locale because they have them available (particularly with females). Importers can pass off animals as a given locale when they don't have locale designations or they otherwise don't have that locale to provide for an order. A breeder could sell an animal as a pure locale without having raised up the offspring to check for outright signs of it not being pure. There are a lot of potential issues. The key is being an informed buyer and looking for signs of what locale they are, even in the face of possibly losing part of your investment. If a group of animals does not appear to be identified correctly, assume that none of the females ID can be relied on and that with the males you will need to compare general locale attributes to narrow down what they are. With offspring, the first time a female is bred, male progeny should be raised up and examined for signs of hybridization and locale in the case of CH clutches. There are ways to improve your ability to obtain accurately identified locale types.
Obviously there are instances when an accurate locale on an individual cannot be reasonably reached. This is particularly true of females especially when their CH progeny do not clearly indicate a particular locale. Even males from a well known locale can show unusual coloration that could bring into question their true identity in the absence of good information and knowledge to help define it. These animals should be regarded as unknown and not put into a locale type if the aforementioned techniques of IDing imports cannot come to a reasonably confident ID.
In my opinion, the goal of breeders should be to establish pardalis bloodlines that represent variations occurring in a discernible geographic pattern. These variations should be designated at a geographic resolution that is realistically conducive to dividing these geographic variations. Further, there should be recognition that depending on the variability at the locale type in question and adjacent types, the appropriate geographic resolution will vary. Expectations for different locale types should be broad enough to encompass typical individual variation within the geographic type but differentiation between similar types should be based on an effort to combine the best set of information available to us as a community and our own experience. Ultimately I believe breeders should employ their best ability to group breeding efforts into a framework of locales based on our best knowledge and information. Experience and care in selecting breeders and care in verifying the progeny a female throws are vital to establishing bloodlines with a reasonable expectation of purity. This falls on both the original breeder and the buyer and was one of the reasons I tried to setup the CCBTD. As a buyer, you should request photos of the bloodlines (sire, dam and male progeny from the dam). There are many bloodlines that well known breeders have going that I would not be comfortable adding to lines I had put together in an attempt to build a "pure" locale type bloodline. At the end of the day, I think we all need to attempt to do everything in each of our power to maintain purity in our own lines because who knows when importation will stop and we will be stuck with what we have.
Chris
Note: I had to walk away from writing this so many times that hopefully my train of thought and argument is logical. We'll see I guess.