Update AI Experiment

[OogieM]

Our USDA ARS experiment with a non-surgical VAI procedure has concluded for this year.

58 ewes in the experiment
29 AI 48 hours after sponge pull
29 live cover rams in 48 hours after sponge pull

8 AI ewes lambed
10 live cover ewes lambed to the first cycle

The synch with sponges worked much better than the sync with the CIDRs. Looks like it also did not mess up the rest of the ewe cycles because based on being in heat when we did the AI and the actual dates of lambing all who did not settle to the first cycle came into another normal heat about 14 days later.

The live cover rams were in with batches of 10 ewes in sync and it looks like they were a bit tired as they had also been collected for the AI but all rams still got ewes pregnant to the first cycle. A lot settled to the second cycle though. Further review indicates we inseminated the AI girls a bit too early but put the rams in a bit too late based on results from Norway and Ireland. I then pulled rams and has a set of backup rams.We have one backup ram who failed to work properly so I had a bunch of open ewes. Most of them were ones that did not settle to AI and also were in a group with a BU ram that did not work well.

Several notes: Do not cut the strings from the sponges no matter how long they look at insertion. The ewes draw them up inside and it makes sponge retrieval very difficult. Running a gloved finger with some disinfectant around the sponge before attempting removal is likely to be helpful to remove any adhesions. We had several cases of sponges tearing upon removal requiring more invasive retrieval methods. If you are sponging to get ewes in sync for live cover rams then put them in 24 hours after sponge removal or even as you remove the sponges. Next year AI will be attempted 52-56 hours after sponge removal in next year's experiment. We did 48 but it was too early. Some research indicated earlier was better but I don't think so based on ewe behavior and results. We amy also go with 2 inseminations timed a certain number of hours apart to cover more of the individual ewe variation that researchers are reporting.

Semen quality is a huge issue. We had some problems and had to substitute rams and I still don't think we got as good a semen as we needed for the AI. Even pre breeding exams and semen evaluation on rams is not quite enough. We really need semen frozen and then tested with in-vitro fertilization. Latest current data indicates that even with semen looking good post thaw that it may fail to get ewes pregnant but you can tell easily in-vitro. That's on the list to test with next year. That is how Ireland determines which rams to use for AI. Problem will be getting sheep oocytes to test with. Given the long term goal of being able to import semen and get new genetics in that way it is reasonable to expect that rams frozen semen be tested by in vitro methods so while difficult it's not an impossible requirement.

All in all I was very encouraged as the actual AI procedure was simple and easily done and not at all stressful to the ewes. We are making progress if a lot slower than we all wish.

[Darroll Grant]

Did any one consider checking the ewes with a teaser to determine when they came into heat after sponge removal or even if they came into heat? Were natural services determined on that treatment?

[OogieM]

We don't have a teaser but we did have intact rams near the ewes and they were observed. All ewes but 2 showed in heat to the rams (or us, I have a bunch of ewes who will display in heat to us really obviously, even more so if we've handled the rams before walking into the pen) after sponge removal but the timing was variable. All services were timed based on hours not heat detection per the experiment protocol. We did have a number of maiden ewes in the experiment, not something that is recommended by anyone.

[Tom Nichols]

Oogie,
Interesting. a few questions.
Was this a winter mating to lamb in June and July? What is or are the breeds?
Tom

[OogieM]

December mating to lamb in May with backup rams to produce lambs in June.
All purebred Black Welsh Mountain sheep

[Phil Crome]

No offense, but you have got to be kidding. I respect the effort you're making, but there is absolutely no way the USDA should be paying for any of this. Rams that settle only 30% either by AI or natural means, then backup rams that are also sterile, then ewes that windup open anyway? Using rams for live cover immediately after collection doubtless affects conception, but is that accounted for in your experimental design?

I get that you need outside bloodlines, preferably ones with strong fertility traits. There has to be a better way of getting them than this. Aren't there any real sheep problems that need research.

Phil Crome

[lambchop]

Phil,
In addition to the fine points you made, there is extensive research in vaginal AI that has already been done. Can we figure out more ways to waste research money?

[OogieM]

No offense, but you have got to be kidding. I respect the effort you're making, but there is absolutely no way the USDA should be paying for any of this. Rams that settle only 30% either by AI or natural means, then backup rams that are also sterile, then ewes that windup open anyway? Using rams for live cover immediately after collection doubtless affects conception, but is that accounted for in your experimental design?

I get that you need outside bloodlines, preferably ones with strong fertility traits. There has to be a better way of getting them than this. Aren't there any real sheep problems that need research.

Phil Crome

Well first off the USDA isn't paying for the research except for the lab time and researchers time and they consider it important to work on. They no longer have any access to research flocks that are of more typical commercial sheep that are willing to work with them so we're it. We are not an ideal flock by any means but we offer the flock and all of our time and our facilities for free. Other research flocks that are around have been unwilling to do the record keeping and individual handling necessary for this type of research. If you don't like them spending time on a rare breed then step up and offer a more typical flock, with all the required record keeping with all the required labor at no charge that is within driving distance of the lab to reduce any travel or other expenses and I am sure they will take you up on it.

You are not understanding the protocol, probably because I didn't explain it well. We separated the ewes into 2 groups. half were bred via AI and half via live cover. In normal years, we get 95%+ pregnancy rates with over 75% of them on the first cycle with live cover rams with no sync or other interventions. This is only the second time I've ever had a ram that did not work well so I am not skilled in dealing with that. He was very young, he passed his pre-breeding semen checks so who knows what the issue was. It is very unusual for that to happen in Welsh Mountain sheep at all. Both the live cover and AI groups had backup rams put in but quite a bit later so we could be sure who the sire was. We also restricted the time the backups are in with the ewes to limit the lambing season. Keep in mind a more typical way to deal with breeding is a bank of rams not individual rams in with individual groups of ewes, especially as the backups. I need to know full pedigrees so I can't do it that way. Most folks also leave rams in a lot longer than we do. I only have backups in for 2 weeks, so at most one cycle. In our system a ewe only gets 2 chances to get pregnant, once with the primary ram and once with the backup ram. Total time exposed to rams in our flock is usually 4 or 5 weeks at most split up by a waiting period between primary and backup. I typically have primary rams in for 2-3 weeks, then 2 weeks with no ram and then a backup ram for 2 weeks. I also butcher out ewes who do not get pregnant by that system unless there is a really good reason why they did not get pregnant. So for this year of the open ewes, any who were in with the ram that failed to work will get looked at and may stick around if I find no other issues with them. All the rest will be slaughtered.

The AI group was AI'd per a very specific hour protocol based on the researchers trying to get publishable results that take current research being done in Norway and Ireland a bit further. Is it an ideal protocol, heck no, but it was done to answer some specific questions left open by previous research. The live cover ewes had the same sync protocol and the same rams as we used to collect semen and do AI to help eliminate ram issues.

None of our rams are well trained to collection at the AI. The rams who were used the year previously were much better, both in terms of collecting semen and in terms of overall performance both via AI and live cover. Without any intervention the older rams have a history of getting 100% of their ewes pregnant within 21 days. Other rams used in the experiment have been good breeders before as well although not 100% but we also tried a number of new or maiden rams because part of the question is what do we need to do to the rams to get them to work in this sort of program.

Other researchers spend a long time training and selecting rams. Private correspondence indicates that they eliminate up to 90% of the rams they try to use for AI for one reason or another. It's clear that training and selecting the rams is a big part of success that other places are seeing. Of course, they also have huge government sponsored flocks (as in thousands of sheep available with rams of all breeds and ages) and can spend months teaching the rams the collection procedure without stress before they ever try to get semen for their AI experiments. We did know that and had worked with the rams prior to the experiment but not as much as other countries do. That's something we are going to change for next year.

Hope that explains things a bit more.

[OogieM]

Phil,
In addition to the fine points you made, there is extensive research in vaginal AI that has already been done. Can we figure out more ways to waste research money?

Unfortunately the "extensive research in vaginal AI" doesn't work consistently in the US at all no matter the breed used.

That is the focus of the research here. There are 2 basic questions that we're trying to answer:

Why don't the same procedures used on the same breeds work here in the same way with the same results?

What can we do to modify those procedures for better success under typical US conditions?

Some flocks in the US using vaginal AI get 50-75% pregnancies, some flocks are lucky to get 1% and that is within the same breed and using the same semen from the same rams. Throw in breed differences and you get results ranging from 0% to 75% pregnancies from vaginal AI.

Even using surgical AI there is a vast difference in results, some flocks and breeds get 75-80% or even higher and some are averaging 20% or less. No one seems to know exactly why either.

[woolpuller]

Must have a prodigal to become better or talk to the experts in Australia to make it better.

[Saffronsheepranch]

I know Finns get better conception rates. They probably leave one a broader margin of error by releasing so many eggs besides all the other reasons to choose them for such an experiment.

[OogieM]

Must have a prodigal to become better or talk to the experts in Australia to make it better.

It would be nice if the Australian procedure (which BTW is nearly always surgical AI not vaginal) actually worked with the same level of results on US sheep. Unfortunately, even when bringing in the people and semen from there to the US to do the procedures US producers do not get the same results with the same protocols. Hence the research. Try to find out what's so different here that is causing such widely differing rates and successes across all different breeds and styles of sync and timing.

[CDKfarm]

I was told by the guy that did AI for us one year that the reason AI results were so much lower in the US were the type of sheep we are breeding and using. Commercial people are not using AI. For the most part the people using it are show guys. These sheep have been bred for so many extremes that the fertility is not really there. He told me that black faced breeds were 10 percent or more lower than white faced show breeds which were lower than small breeds such as jakobs or a similar breed.

[Saffronsheepranch]

Maybe it is also because actual rams get ewes more than once too quite often, right? And different rams can get the same ewe. I have seen that.

But yeah, probably choosing sheep just based on conformation and appearance as opposed to the commercial guys who concentrate on productivity as the number one priority, I guess fertility could go by the way side. Because I do think that if you AI-ed sheep that tend to have trips and quads, that you would have a better chance of making lambs more easily.