A genetic bottleneck is a crisis in a species, subspecies, breed, or other population that permanently reduces the available gene pool of that population.
Pretty much all domestic breeds were affected by the two World Wars, when resources were taken over by the war effort and people were conscripted, and breeders were simply not there or could not afford to continue their bloodlines. The psychological scars of the wars also led to some breeds of Germanic origins being rebranded to obfuscate their associations, such as German Shepherd Dog to 'Alsatian', German Boarhound to 'Great Dane', and poodles to 'French poodles'.
Bottlenecks due to wars and other outside influences are beyond the control of breeders, but some bottlenecks occur because of poor decisions by breeders themselves. The most obvious and notorious bottleneck event is the popular sire. This is a male who produces an inordinate number of offspring with a large proportion of the available females in the breed, to the point that it becomes difficult to find males who are not related to him a few generations later. Although popular sires and other bottleneck events are often not inbreeding events, they inevitably result in inbreeding and loss of overall heterosis downstream as it becomes extremely difficult to avoid having the popular sire on both sides of the pedigree, often multiple times. A poodle dog born in 1981 was successful at conformation shows, and as a result this dog produced in excess of 100 litters, and it is now impossible to find a black or brown poodle descended from British bloodlines that passed through that era that does not have this dog in the pedigree at least once. Fortunately people seem to have been more sensible since then, as no dog who has enjoyed success at conformation since then has been so overbred since. However, sires can sometimes be overused more insidiously, or for well-meaning reasons, such as many people using a dog because he has good health test results, or because an imported dog offers (or is thought to offer) the chance of an unrelated mate and low COI offspring, and ironically the problem is only realised two generations later.
Bottlenecks have always occurred in the wild as well as in domestic breeds, so they are to some extent unavoidable. What breeders need to strive to do is avoid what is avoidable, and mitigate the damage caused by what isn't avoidable.
Registration trends in Poodle (Standard)
Poodles as a breed and poodle breeders are living through what can probably be politely termed 'interesting times' in recent decades. Most recently, we have the pandemic, which has increased demand for puppies and it is thought may be attracting people into breeding dogs for the wrong reasons, while simultaneously perhaps putting a damper on more established breeders by making it difficult to travel to use a male, to go abroad, or to go to a vet for AI or accurate progesterone tests. Concern has also risen amongst careful breeders about the suitability of the sorts of people applying to buy a puppy. Similar concerns hit breeders in the early 2000s due to the emergence of the novelty mongrel fad which saw poodles attract puppy farmers and deadbeat breeders as components to produce mix-breed puppies to market as 'hypoallergenic'. Many older more established breeders decided at this point to cease breeding because of concerns about their puppies and bloodlines ending up in bad breeding establishments. Older breeders are always going to leave the breed at some point, but situations like this perhaps precipitate their departure and make it more likely they will not have a suitable protégé to leave their line with.
'Boom and bust' cycles of demand in any breed are harmful to the gene pool. At the start of the boom, the breed may experience an influx of 'bandwagon jumpers' with little serious interest in it, while at the same time suffering an exodus of jaded older breeders pushed into retirement after becoming frustrated with having to field enquiries from unsuitable buyers. When the cycle busts and the popularity of the breed declines, all the bandwagon jumpers jump off the bandwagon and the bloodlines they got hold of disappear, for better or worse. Bloodlines that have ended up in puppy farms tend to be viewed as contaminated despite how noble their origins might have been, and end up dying out. The dedicated breeders who stay with the breed are unlikely to have a proportional representation of all the bloodlines that were present before the boom started, so when the gene pool contracts, it does not do so evenly.
In addition to trends in popularity of whole breeds, breeds can sometimes experience internal trends, particularly if certain colours suddenly become either unpopular or popular. In poodles, there are two such trends that have occurred in recent years. Firstly, the colour red either didn't originally occur in standards as it did in the other sizes of poodle, or it was lost during one of the World Wars. Red was recovered or introduced to standards from the other varieties, in countries outside the UK because the KC considers the varieties of poodle to be separate breeds and won't register their offspring. Imported red poodles started to appear in the UK around the start of the 2010s and the colour wasn't available here prior to this. Since then, the number of puppies registered as red has quickly increased.
Secondly, a similar trend has occurred with the registration of puppies recorded as 'colour not recognised by KC', with few, sporadic registrations occurring prior to the early 2010s and increasing numbers of litters that are largely or sometimes entirely registered with this description. These animals are most likely mainly S-locus and agouti mismarks (there is a contradiction in the breed standard in that it states all solid colours are acceptable, but doesn't allow silver beige and café au lait to be recorded when registering pups, but looking at the pedigrees I suspect they aren't SB or CAL). S-locus and agouti mismarks are produced by combinations of recessive genes that have always been present in the breed and have probably always occurred at a low level.
Some recent phenomena of note:
An imported dog registered as being an unrecognised colour with a hip score almost three times higher than the median for the breed produced a number of litters, not an extremely high number of litters, but many of the dog's offspring seem to be subsequently being used for breeding. This is one of the more insidious ways a dog can end up being a popular sire, even if the owner of the dog is cautious not to let him be overused. Several offspring of this dog have been hip scored, and not one of them has a result better than the median for the breed. This isn't a criticism of the merits of the careful use of dogs with not quite perfect attributes or a discussion about hip scoring, but it demonstrates how popular sires can have an effect on the distribution of genetics in the breed, in particular genetics that might not be beneficial.
Of the dogs who have sired the most litters in standard poodles in the past decade, the three highest-siring dogs produce large numbers of puppies registered as either red or CNR, and have produced 24, 17, and 14 litters respectively. All three dogs are relatively young, having only produced litters in the past 2-4 years. There are a number of dogs producing traditional-coloured puppies who have sired 10-13 litters and most of these are much older and have produced a number of litters more commensurate with their age.
Bottleneck Events in Alpacas
Surprisingly, alpacas seem to have not experienced any significant bottleneck events since they started being bred in the UK. This is quite unusual for a breed, particularly one that wasn't native and was thus reliant entirely on imports to start the population here. Alpaca pedigrees at most trace back up to around 8 generations before reaching animals imported from Peru and Chile on which no pedigree data is available. It's possible that bottleneck events occurred prior to this, and alpacas have a woeful genetic history since their heyday during the Inca empire, but I think it's likely that breeding practices were chaotic and serious bottleneck events probably didn't occur.
There are a couple of recurring ancestors that seem to be present in quite a lot of pedigrees, but none of them are so prevalent that they're likely to become a popular sire issue in future. Interestingly my best alpaca is the most inbred, with a negligible COI of 3% (which most breeders of other domestic species would be thrilled to achieve).
How to Avoid Bottleneck Events
In this post I've only talked about two species/breeds I'm familiar with, and I can't speak for those I'm not. Some bottleneck events occur because of circumstances outside of anyone's control (such as pandemics and wars) but there are general rules that should be capable of being applied to a person working with any population that will help.
Firstly, try to avoid using males who have already sired a large number of offspring. This unfortunately doesn't help if you were the first person to use a male who was subsequently overbred, or if a male was used once to produce a litter and all the offspring in the litter produced large numbers of grandchildren, but it does go a long way to impeding the emergence of a popular sire. An underutilised brother of a popular sire isn't as good from a genetic standpoint as an unrelated, unused male, but is still the better choice if he's the only realistic alternative.
Do not get involved in a breed and do not choose a male to use for any reason to do with 'popular' colours or trends. To get involved with breeding anything for the purpose of an easy way to make money is a bad idea. When you take on a breed, you are becoming a custodian of that breed and you are responsible for the bloodline you create and preserving it, and eventually for passing it on to someone you trust. All breeds need new breeders, but they need to be in it for the right reasons. Don't do it if you're not prepared to stick at it and carry that responsibility.
If you are coming into a breed, try to find a bloodline that is not being widely bred. Some of this is difficult because there's only so much legwork you can do and a knowledge ceiling you're going to hit before you have to take the plunge and obtain something and actually, well, breed, and continue your learning and your development as a breeder. Most people make mistakes with their first acquisitions, but hopefully with enough research it's something you can still improve on later. Most of this is deep pedigree analysis. There are some DNA analyses around but the development of this sort of technology is still at a level where it needs to be seen as a complement to accurate pedigree research and not a replacement.
If reproductive technology is available for your animal, THEN USE IT. Frozen semen can be a lifeline for a breed if mistakes get made or uncontrollable disasters occur, and in theory it can last indefinitely, meaning semen you store might benefit your breed centuries after your death. Hopefully specialist vets should be more accessible soon, if they're not sick of deadbeats asking them to reproduce dogs that can't mate naturally and pandemic puppy farmers requesting they inseminate Chihuahua semen into Bullmastiffs for the next novelty mutt fad.