Sire conception rate

Definition

On-farm sire conception rate is the conception rate of inseminations grouped by service sire (the bull whose semen was used): pregnant services ÷ services with a known outcome, for each sire, in one herd.

It answers “how well has this bull’s semen worked here?” It reflects the bull’s fertility, but also everything else that differs between the cows he was used on: service number, parity, timed AI vs detected heat, technician, season, and whether the semen was sexed. Herds often use some bulls only on heifers or only on repeat breeders, which makes raw sire comparisons unfair.

Not the same as SCR

Sire Conception Rate (SCR) is a national bull-fertility evaluation published by the Council on Dairy Cattle Breeding (CDCB) for US AI bulls. It is a different thing from on-farm CR by sire:

It uses pregnancy outcomes from many herds’ DHI records.
It adjusts for non-bull factors (herd, cow, parity, service number, month, and others) in a statistical model (Kuhn & Hutchison 2008).
It is expressed as a deviation in percentage points from the average bull. An SCR of +2.0 means that bull’s expected CR is about 2 points above the average bull’s, under the same conditions. It is published with a reliability.

When USDA introduced SCR in August 2008 it used up to 7 services per lactation, reported SCR to one decimal as a deviation from the mean (+1.2 = 1.2 percentage points above average), and required Holstein bulls to have at least 300 total breedings, 100 in the most recent 12 months, and breedings in at least 10 herds (AIPL, Changes to evaluation system, August 2008).

Use SCR to choose bulls. Use on-farm CR by sire to spot a problem with a specific semen batch, handling or tank, and only when the sample is large enough.

Formula

Plain text:

Sire CR (%) = pregnant services to sire s / (pregnant + open services to sire s) × 100

Approximate 95% confidence interval (normal approximation):
  CR ± 1.96 × √( CR × (1 − CR) / n )            CR as a fraction, n = services with a known outcome

LaTeX:

\mathrm{CR}_s = \frac{N_{\text{preg},s}}{N_{\text{preg},s} + N_{\text{open},s}} \times 100
\qquad
\mathrm{CI}_{95} = \hat{p} \pm 1.96\sqrt{\frac{\hat{p}(1-\hat{p})}{n}}

For small samples (under about 30) or rates near 0% or 100%, use a Wilson or exact (Clopper–Pearson) interval instead; the normal approximation is too narrow there.

How wide the intervals are

At a true CR of 35%, the approximate 95% interval is:

Services (n)	95% CI (±)	Range
15	±24 points	11–59%
30	±17 points	18–52%
100	±9 points	26–44%
300	±5 points	30–40%

Two sires whose intervals overlap heavily are not shown to differ. With a dozen sires in use, one will look bad by chance alone.

How to pull it in DC305

BREDSUM\S FOR LACT>0           CR by service sire, cows
BREDSUM\YS                     CR by service sire, heifers
BREDSUM\Z FOR LACT>0           CR by breed of sire
BREDSUM\* FOR LACT>0           CR by semen type
BREDSUM\M FOR LACT>0           CR by sire manufacturer (stud)
BREDSUM\XB* FOR LACT>0         cross table: times bred × semen type

\S is service sire. \E is the 21-day pregnancy risk table, not a by-sire table.

Split cows and heifers, and if possible first service vs later services.
Leave embryo-transfer “breedings” out. In DC305 an embryo transfer is often recorded as a BRED event with an embryo breeding-type code or a donor/sire code, and it will appear as a “sire” row with recipient-pregnancy outcomes. See embryo transfer KPIs.
Sexed vs conventional semen is identified by the semen type (\*), sire code, a separate sire entry, or a breeding remark, depending on the herd. Breeding-type and remark codes vary by herd; look them up in the herd’s private dc305-vocab mapping and breeding type codes.
Use a recent window (for example the last 90–180 days of completed services) so the result describes semen currently in the tank.
Pool sires with few services (for example under 10–20) into an “other” row rather than ranking them.

Typical range

Herd context: US Holsteins, freestall or dry-lot, mix of timed AI and detected heats, first pregnancy check at about 28–40 days.

Most AI sires used in one herd fall within about ±5–8 points of the herd’s overall CR once each has 100+ services. Bigger gaps on small samples are usually noise.

Sexed semen: CR is usually about 75–90% of conventional in the same group, i.e. several points lower in heifers and often more in lactating cows. In US on-farm records, first-service CR with sexed semen averaged 47% in Holstein heifers and 53% in Jersey heifers, about 80% of conventional in the same herds (DeJarnette et al. 2009).

SCR values: published SCR for most active AI bulls falls between about −3 and +3. Each point of SCR is one percentage point of expected CR relative to the average bull (AIPL 2008). For comparison, publishable Angus bulls bred to Holstein cows ranged from −5.1 to +4.4 (McWhorter et al. 2020).

Common mistakes

Ranking sires on small samples. 15 services gives a ±24-point interval at 35% CR. Don’t drop a bull on that.
Comparing sires used on different cows. A bull used mostly on heifers or first services will beat one used on repeat breeders. Compare within the same group.
Mixing sexed and conventional semen of the same bull in one row.
Treating on-farm CR as SCR, or SCR as on-farm CR. SCR is a national, adjusted evaluation; on-farm CR is raw herd data.
Including embryo transfers as if the donor or embryo code were a sire.
Counting pending services as open, which penalises the newest bulls.
Missing the real cause. A sire with a sudden drop may reflect one bad semen shipment, thawing or handling, or one technician. Check by technician and by date before blaming the bull.