Direct Comparison of Natural Service vs. Timed Artificial Insemination

Despite the considerable advantages of artificial insemination (AI), a significant number of dairy producers use natural service (NS) for their breeding program. This article was created by Fabio Lima, Marie–Joelle Thatcher and Carlos A. Risco1and of the College of Veterinary Medicine, in conjunction with Albert De Vries and William W. Thatcher of the Department of Animal Sciences, University of Florida.
calendar icon 12 December 2008
clock icon 8 minute read


In a survey on bull management practices of California dairymen, 84 per cent reported use of NS as a component of their breeding program management (Champagne et al., 2002). The most common use of NS was after unsuccessful AI attempts. In dairy herds located in the northeast region of the US, reported use of NS, as a component of the breeding system, varied from 55 to 74 per cent (NAHMS, 2002; Smith et al., 2004).

In a study that compared pregnancy rates (PR) between AI and NS in Georgia and Florida dairy herds, the use of NS alone or in combination with AI was reported to be around 70 per cent (De Vries et al., 2005). A survey that examined current management practices in 103 herds participating in the Alta Genetics (Watertown, WI) Advantage Progeny Testing Program, reported that 43 per cent of herds used a clean-up bull.

Non-pregnant cows were moved to the clean-up pen after 6 failed AIs or 232 d postpartum. (Caraviello et al., 2006). A common perception among dairy producers that use NS is that more cows are bred by NS compared to AI because human errors in estrus detection are avoided when bulls are used. These producers also contend that a NS breeding program is cheaper and easier to manage than an AI program.

Bulls used for NS are under the influence of nutritional, environmental and management factors that may affect their ability to impregnate cows. Whether or not NS bulls classified as satisfactory potential breeder maintain reproductive soundness during the time that they are used for breeding is not known. With the understanding that bull fertility in an integration of biological as well as management factors, the potential for deviation in reproductive potential is possible.

Pregnancy rate is defined as the number of pregnant cows eligible to become pregnant over a specified period of time (every 21 days) and is the product of estrus detection and conception rate. A high PR at the end of the postpartum voluntary waiting period results in more cows pregnant earlier in lactation, maximizing farm income (Risco et al., 1998). Poor estrus detection is a major factor that contributes to low PR on many dairy farms in Florida and throughout the US. The use of NS and timed artificial insemination (TAI; cows are AI at a fixed time without being detected in estrus) are two options that minimize problems with estrus detection. This paper discusses research that compares reproductive performance between TAI and NS breeding systems.

NS vs. AI at detected estrus

Several studies have compared reproductive performance between AI at detected estrus and NS breeding systems. Seasonal effect on AI and NS fertility in dairy herds was evaluated under field conditions using Dairy Comp 305® (Valley Agricultural Software, Tulare, CA) (Niles et al., 2002). During periods of heat stress (summer), overall PR dropped for cows bred by either AI or NS, and no difference in PR was found between NS vs. AI bred cows during the cool season. In herds with poor estrous detection, NS resulted in a higher PR (Niles et al., 2002).

The effects of four combinations of AI and NS breeding systems (BS) on production and reproduction responses were evaluated using Dairy Herd Improvement Association herd summary information (Smith et al., 2004). Herds were assigned to BS by percentage of NS usage as follows: 1) 0 per cent, 2) 1 to 20 per cent, 3) 21 to 89 per cent, and 4) 90 to 100 per cent. Actual calving interval was shorter in herds that used mostly NS (BS4) compared with other systems.

However, herds using a combination of AI and NS or mostly NS had longer dry periods than herds using all AI. Days dry and the percentage of dry periods greater than 60 d were less for herds that used all AI breeding. Overall efficiency assessed by the percentage of cows in milk and herd milk yield was greater for herds that used all AI and declined as the percentage of NS increased. The effects of AI and NS BS on PR by stage of lactation and season over an 8 year time period showed that the use of NS bulls did not result in meaningful advantages or disadvantages in terms of PR over time (De Vries et al., 2005).

In contrast to the previously cited studies, a California study that compared calving to conception intervals for cows in AI pens with cows exposed to NS sires found that cows AI had a higher risk for pregnancy across all days in milk (DIM) (Overton and Sischo, 2005).

NS vs. timed AI: a field observation

A field observation is presented that demonstrates the ability of TAI and resynchronized timed TAI to enhance herd PR in a large commercial dairy herd that utilized extensively NS (Thatcher et al., 2006). The herd was comprised of 2000 cows with approximately 1200 calving per year. Cows were housed in covered barns with self-locking stanchions and free-stalls.

Bulls used for NS underwent a breeding soundness exam and entered the NS program if classified as a satisfactory potential breeder according to the guidelines for the SFT (Chenoweth, 1992). Breeding soundness exams were repeated every 6 months and bulls that graded unsatisfactory were replaced. The BCR was one bull per twenty cows. The ratio in each pen was maintained based upon a monthly diagnosis of non-pregnant cow numbers. Bulls were rested for 14 d after 14 d of cow exposure. Cows more than 127 dim and diagnosed non-pregnant were identified.

A decision was made to implement a timed AI program for these cows that were considered infertile (not pregnant by ample time exposure to bulls) in the NS program as described. The average day in milk was 356 for the 245 cows enrolled during the period from January 11 to June 21, 2006. The timed AI program entailed a presynchronization (i.e., two injections of PGF2á given 14 d apart), followed by Ovsynch (i.e., GnRH given 14 d after the 2nd PGF2á injection of presynchronization, followed by PGF2á given 7 d after the first GnRH injection, and a second GnRH injection given 2 d after PGF2á with a timed insemination between 16 to 20 h after the second GnRH), and a resynchronized timed AI with the Ovsynch protocol repeated in cows diagnosed open by ulstrasound at 32 d after the previous timed AI.

Resynchronization was repeated twice in which cows went through three possible timed AI. This reproductive management sequence coordinated animal handling to 3 d of the week. As a population of infertile cows to NS, a cumulative PR of 56 per cent was obtained based upon ultrasound diagnosis at 32 d for the first three TAIs. At 60 d after insemination, 37.8 per cent of the inseminated cows were diagnosed pregnant via palpation per rectum.

Pregnancy losses between d 32 and 60 were 32.5 per cent, which is substantial and reflects that the cows as a group were indeed sub fertile. Nevertheless, 37.8 per cent of the cows were pregnant that had not conceived through NS management. Although mean PR declined from first to third service (i.e., 37.5, 28.7 and 14.4 per cent, respectively), overall PR (i.e., either at day 32 or 60), did not differ for the intervals from 200 to 700 dim. The take home message from this field experience is that reproductive management centered on timed AI resulted in 38 per cent of subfertile cows becoming pregnant.

Economic comparison NS vs. timed AI

An attempt was made to compare the costs and revenues of the NS program to the timed AI program in the direct comparison study. Partial budgets are in Tables 2 and 3. Labor costs and pregnancy rates in both programs were assumed equal based upon experimental results described above. The net cost of the NS program was a $92.29 per slot per year. For the timed AI program, the net cost was $51.61 per slot per year. Therefore, the NS program cost $40.68 per slot per year more, including an opportunity cost of $13.67 for less genetic progress. Without considering genetic progress, the NS program would cost $27.02 more per slot per year.

Overton (2005) calculated an extra cost of $10.27 per slot per year for a NS program compared to an AI program including 30per cent timed AI in large western Holstein dairy farms. He also assumed equal pregnancy rates. Overton assumed that for every 2 NS bulls, 1 extra cow could enter the herd. Thus, his AI program allowed for more cows than the NS program. When the number of cows in both programs was assumed equal, the extra cost per slot per year for a NS program was reduced to $3.61.

If there were differences in pregnancy rates between both programs as evaluated in this experiment, they could be easily incorporated into the partial budgets. An increase in pregnancy rate of 1 percentage point (e.g. 15per cent to 16per cent) is worth approximately $15 to $20 per slot per year with lower values at higher pregnancy rates (De Vries, 2007).


Cows sired by proven AI sires produced 1400 kg more herd lifetime actual milk and were $148 more profitable when compared to daughters of non-AI sires (Cassel et al., 2002). Despite this economic advantage of AI over NS, use of NS remains popular in many dairy herds. Studies evaluating reproductive performance between AI and NS do not show a clear disadvantage or advantage for using NS over AI. In herds with low PR related to poor estrous detection, uses of TAI or NS are viable options. Both of these breeding systems require strict attention to management compliance in order to optimize reproductive performance. Natural service breeding programs are expensive when direct and indirect costs are considered. Economic analysis within the content of this study showed that TAI is less expensive than NS and allows for immediate submission and a more rapid PR of all animals at the designated waiting period.

Further Reading

- You can view the full conference proceedings by clicking here.

April 2008

© 2000 - 2022 - Global Ag Media. All Rights Reserved | No part of this site may be reproduced without permission.