The Original
Online Alpaca Marketplace
Alpaca Forum at AlpacaNation
Alpaca Forum at AlpacaNation
Home | Profile | Register | Active Topics | Members | Search | FAQ
Username:
Password:
Save Password
Forgot your Password?

 All Forums
 Alpaca Discussion Forums
 3. Birthing & Neonatal Care
 Gestation length study
 New Topic  Reply to Topic
Author Previous Topic Topic Next Topic  

mythic

888 Posts

Posted - 09/28/2007 :  4:35:53 PM  Show Profile  Visit mythic's Homepage  Reply with Quote
I was curious, so I did a little study of registry reported gestation lengths, using the reported dates of conception and birth for ~1000 random alpacas. This ought to also remove any biases for area of the country, suri/huacaya, herd management practices, weather from year-to-year, etc.

I'll cover some of the highlights here, but I've got graphs to help visualize at: http://numeric.mythicalpacas.com/gestation_study.pdf

First, I excluded any conception dates that were 11 months prior to the date of birth. For example, for an alpaca born today (9/28/2007), I would have excluded any that reported that conception was on 10/28/2007. It turns out that more than 10% are reported this way.

The first result, though it's not very interesting to me as a statistic, the longest gestation among my random sample was 472 days, followed by 442 and then it drops off to several at ~400. So reasonably, the longest one can expect a gestation to go is ~400 days, though of course there will always be outliers.

Before I get to averages, let me tell you how I got there, because it's not as simple as just giving an average. There does appear to be a seasonal bias toward gestation lengths.

First I started looking at the data by birth month. Not surprisingly, it's heavily spring-summer-fall and very little winter. I calculated average gestation length and plotted them by birth month, noticing that earlier birth months (Mar-Jul) favored longer gestation times than later birth months (Aug-Nov), but not too significantly.

Then I started thinking that gestation length would be more dependent on the start (conception) rather than the end (birth), so I looked at average gestation by conception month and the difference is striking. When conceived in the earlier months, especially April-June, gestation is on average a week or more longer than in the later months. May is quite an outlier with an average gestation of 359 days! It does represent 8% of the data, but I think it would be tempered some with more data, as plotting gestation length by date of conception (day of year), there's no spurious spike in the May timeframe.

So with the clear and consistent trend of earlier month breedings having longer gestations, I broke the data into two similar sized groups: March-August and September-February. Not only is the average lower for the Sept-Feb breedings, but the distribution is tighter (lower standard deviation).

One thing about the distributions, they don't behave perfectly as normal distributions. Most of the data behaves in a classic normal distribution (bell curve), but then there are the high side outliers. So I calculated the average and standard deviation as representative of the normal distribution portion of the data, excluding the outliers. The average for early month breedings was 345 days, while the average for late month breedings was 338 days.

Another thing favoring the later month breedings is that there are fewer long gestation outliers. For the early month breedings, 15% of the data fell higher than 2 SDs (standard deviations) from the average and 7% were higher than 3 SDs. For the later month breedings, only 9% fell higher than 2 SDs and 5% higher than 3 SDs. When plotted together, the difference is more obvious.

For comparison sake, I took a look at the data Rick has been gathering here. At first glance it didn't match up with the random data I had (generally higher and more scatter), but in looking at birth months, Rick's data heavily favors early month breedings. So comparing Rick's data to the March-August data, there is much more similarity.

What does this mean? The spring birthing that a lot of people seem to favor has greater variability and generally longer gestation times than later month breedings.

This all makes sense in my head and I hope my fingers have translated it well enough, but in the likely event they haven't, I'm happy to explain and discuss more if anyone's interested (for the 3 of you that actually read this far! ).

Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559

alpacastarr

686 Posts

Posted - 09/29/2007 :  10:07:29 AM  Show Profile  Visit alpacastarr's Homepage  Reply with Quote
Surely I'm not the first one to respond?!?

I found this absolutely fascinating - admittedly in part because it supports my personal choice to do my breeding in the fall months (like November). I didn't know any of your research of course, it's just seemed to me that the fall born babies spend more time outdoors cavorting in the sunshine than babies born right when it starts to get all hot and sweaty. Spring born babies end up spending a lot of sunlight hours resting in the shed... on this farm.

Statistically it seems to me the only data you are missing -- and it's not in the registry so no way to get it -- is in which season is a breeding more apt to result in a live birth. Does it matter seasonally to have a higher breeding success rate? If you have a higher breeding success rate in the spring (before males are exposed to heat for example) you may have to trade that vs a longer or more variable gestation length but that might be worth it from a productivity perspective - that kind of question would be good to know the answer to but I don't think the data is available.

Every time I learn something it spawns 10 more questions!!
Thanks for sharing your analysis, Ryan. Excellent!


Starr
Venezia Dream Farm
Asheville, NC
http://alpacanation.com/farmsandbreeders/03_viewfarm.asp?name=11404
http://www.veneziadream.com/
Go to Top of Page

mythic

888 Posts

Posted - 09/29/2007 :  10:52:20 AM  Show Profile  Visit mythic's Homepage  Reply with Quote
First to respond, probably the first to read all the way through! Anyway, those are very good questions, and like you say, questions with no way to answer right now!

I have some ideas on how to proceed, just got to get my ideas written down, but I've got some mowing to do for now... If I can provoke curiosity and interest with the data available, hopefully I can motivate to record and collect more data. Time will tell.


Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559
Go to Top of Page

llamapelli

1401 Posts

Posted - 09/29/2007 :  11:10:25 AM  Show Profile  Visit llamapelli's Homepage  Reply with Quote
This is very interesting. Have you analyzed the numbers to see if the study is adequately powered, especially in the months with fewer births? I am interested in the p values of the data. I guess I'm being pretty geeky, but I am so used to going straight to the statistical analysis when reading studies before the study itself. I, too, tend to breed for spring babies. I've always wanted the babies to have some size on them before winter sets in, but I can definitely see the advantage of fall breeding further south. I do know one thing, I will enter actual conception dates rather than calculated when registering babies.

Melanie McMurry
Anasazi Alpacas & Kokopelli Llamas
Gladesville, WV
304-864-5210
Go to Top of Page

dshultz

96 Posts

Posted - 09/29/2007 :  3:43:02 PM  Show Profile  Visit dshultz's Homepage  Reply with Quote
This is a great study. Now I'd like to know how the births breakdown by sex based on month and time of day of conception. I heard a theory (probably wishful thinking) that more males are born from morning conceptions than afternoon. I know in our case that has been true. Coincidence? Probably, but it would be nice to know. Once again the data would be hard, if not impossible to acquire.
Diana

Diana & Dick Shultz
River Mist Ranch
Grants Pass, OR 97526
541-472-4060
www.rmralpacas.com
Go to Top of Page

richbye

750 Posts

Posted - 09/29/2007 :  7:35:15 PM  Show Profile  Visit richbye's Homepage  Send richbye a Yahoo! Message  Reply with Quote
Here in WNY, many (myself included) breed for early fall cria. Why? One reason is that the spring can be WAY too unpredictable.....one day snow, 2 days later 60 degrees and sun. In the fall, we KNOW that it will be generally cold and rainy. Weather we can actually predict! Summers also can be very hot (like 90 degrees and no rain for the past 3 years). I also like the idea of those late gestation females getting all that good grass for the final months of their fetal growth, rather than hay. Because we do fall birthing, all the cria get A&D. And, all of our females go 337-345 days gestation length. The one year we had a female due in the beginning of April, and she didn't give birth until mid-May! When we switched her to fall birthing, no more late babies.

Very interesting data!

Jeanne

Gemstone Alpacas, Inc.
11300 Savage Rd.
Chaffee, NY 14030
(716) 868-0883
http://www.alpacanation.com/gemstonealpacas.asp


Go to Top of Page

Alpacalady

685 Posts

Posted - 09/30/2007 :  7:36:24 PM  Show Profile  Visit Alpacalady's Homepage  Reply with Quote


Great study, Ryan, I find it all fascinating.

Right now I have one girl at 336 days and another at 330, one is a first time mom, the latter is having her third cria ( her other two were born at 342 days in June, her next one at 340 days in October).
In May of this year, I had one born at 346 days and another at 355 days.
Last year I had one give birth at 365 days! (October 25th).

I can see where some of my experiences fit into your study, even the outlier, very interesting!

Thanks a Bunch, I learn something every day on this forum.

Laila






Graceland Alpaca Farm
Lisbon Falls, Maine
gracelandfarms@verizon.net
207-353-2171
Go to Top of Page

bobvicki

2960 Posts

Posted - 10/01/2007 :  09:51:04 AM  Show Profile  Reply with Quote
Hi Ryan,

Thanks for confirming what I already believed. It has been by experience with suris that fall deliveries had shorter gestation times.

Just like Jeanne the weather here in the Midwest can be crazy in the spring so I prefer fall births, I also don't shear until late in May because of this.

Of interest would be if the suri numbers matched the data or if by chance they may lean toward the shorter gestation's.

Bob

Bob & Vicki Blodgett
Suri Land Alpaca Ranch
3288 Halter Avenue
Newton, Iowa 50208
641-831-3576
alpaca@iowatelecom.net
Go to Top of Page

mythic

888 Posts

Posted - 10/01/2007 :  10:25:21 AM  Show Profile  Visit mythic's Homepage  Reply with Quote
Hi Melanie,
I'm an engineer, not a statistician! That said, I know enough about statistics to be dangerous. A Student's T-test on the early vs. late data yields T=11.2 using my estimated mean & SD (neglecting the several percent of data that are outliers) and T=9.7 using a calculated mean and SD representing all data.

Considering the relatively high T and the sample size, the p-value is essentially zero. For a sufficiently large sample size, T only needs to be >3.291 to be statistically significant at the 99.95% one-sided confidence interval. That's how far my table goes at least.

Regarding sample size, I believe it's more than adequately powered with 573 data points for March-August and 491 for September-February! There's a table in the PDF I linked that has the number of observations broken out by month, so you can see that conceptions in May-November really drive the data. I could neglect December-April conceptions (80 total observations) alltogether and still arrive at the same conclusion.

Heather,
I could imagine there would be regional differences, and any one farm may even have all of their personal observations contrary to the trend, and who knows, maybe even certain lines have longer/shorter gestations, but none of that really matters when looking at the population as a whole. One can always have an example that doesn't fit the trend, but the trend can still be there. So, independent of region, herd management, weather, etc, the trend is that late-year conceptions gestate longer than early-year conceptions.

Diana,
More good questions without data to answer them with! I can imagine many questions based on time of day of birth, time of day of breeding, temperature during breeding (both ambient and individual), degree-days for some amount of time before breeding, number of breedings required for conception, sex of cria, cria weight, placenta weight, IgG...

The problem isn't just a matter of record keeping but record sharing. I've found that folks in alpacas are generally friendly, but when you ask for data they clam up. Secrecy to me is an indicator of fear, and in an industry where prices are all too often set subjectively rather than objectively, I can imagine there's a lot of fear to share data that can alter the price structure that person has invested what was previously their 401k into.

Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559
Go to Top of Page

mythic

888 Posts

Posted - 10/01/2007 :  10:33:00 AM  Show Profile  Visit mythic's Homepage  Reply with Quote
Hi Bob,
I could look into that, just not today. I have suri in my data, but not enough to really observe a trend using just them. From what I do have, huacaya are slightly shorter gestation, but not by enough to be statistically significant, at least not with the sample sizes I have.

Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559
Go to Top of Page

mythic

888 Posts

Posted - 10/02/2007 :  6:57:06 PM  Show Profile  Visit mythic's Homepage  Reply with Quote
I checked other sorts of things for correlation/deviation. I still need to add more suri data, but I find no difference between colors or genders, but curiously...

Full Bolivians tend to longer gestation by several days. With a 99% confidence interval, full Bolivians tend to gestate almost 5 days longer than others. I'd still like to add more data, but I went back to add another 50 alpacas (random again, other than selecting specifically for full Bolivians) and it still holds. Everything from half-Bolivian on up appears to gestate longer, with full being the longest. Again, no explanation, just an observation. No other heritage behaved this way (Peruvian, Accoyo, Chilean).

I also want to say, I have no personal bias in this. I'm not out to prove what I thought previously, just to read the data and see what it says. Originally, I only intended to calculate statistics on average gestation lengths and then it became clear that there were modifiers involved. So just to be clear, I'm not a pro/anti-fall-birthing advocate or a pro/anti-Bolivian advocate.

Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559
Go to Top of Page

Heidi Christensen

4211 Posts

Posted - 10/02/2007 :  7:44:39 PM  Show Profile  Visit Heidi Christensen's Homepage  Reply with Quote
Hey Ryan,

This may be more of a question for folks that have registered lots of cria, but any idea what kind of error you might expect for people who guestimate the conception date? I will be more careful from now on, but the couple I have registered I just count back 11.5 months. I have breeding dates on them, but didn't think about looking up on my herd software while I was registering them.

Heidi Christensen
WingNut Farm
Graham, Wa
(253) 846-2168
http://alpacanation.com/wingnutfarm.asp
http://wingnut-alpacas.com
Go to Top of Page

mythic

888 Posts

Posted - 10/02/2007 :  10:06:39 PM  Show Profile  Visit mythic's Homepage  Reply with Quote
Hi Heidi,
There are a huge number (~12%) of alpacas reported with 11 month gestation, to the day. Without those, there aren't any other dates that jump out as heavily used estimated dates. I'm not saying there aren't any other estimated dates, but they're fewer, well hidden, and are likely affecting all groups relatively equally.

Ryan

Ryan & Joanna Maas
Mythic Alpacas
Goode, VA
434.509.2559
Go to Top of Page
  Previous Topic Topic Next Topic  
 New Topic  Reply to Topic
Jump To:
Alpaca Forum at AlpacaNation © 2000 - 2009 AlpacaNation LLC Go To Top Of Page
Snitz Forums 2000

WELCOME TO ALPACANATION

Our family has 8,596 breeders, with more joining every day.

Join Now!