Buying semen is one of the cheapest, highest-leverage decisions on a dairy farm: a single straw costs a few hundred shillings but its genetics stay in your herd for generations. Yet most farmers choose on price or a familiar bull name because the numbers look like alphabet soup. They aren't — once you know what each one means, you can pick a bull that genuinely improves your herd.
First: buy from an authorised source In Kenya, buy semen from KAGRC (Kenya Animal Genetic Resources Centre) or licensed distributors, and have it stored and inseminated by a trained AI technician. Semen is shipped and stored in liquid nitrogen at −196 °C; a straw that has been allowed to warm — even briefly — may be dead, and you won't know until the cow returns to heat 21 days later. Cheap semen from an unknown source, poorly stored, is the most expensive mistake in this whole article.
What's on the straw and in the catalogue A straw is labelled with the bull's name, breed, and a registration/stud code (in North-American catalogues a NAAB code). The catalogue or app entry then lists the bull's proof — his genetic evaluation. That proof is the real product you're buying.
Breeding values (PTA / EBV) — the core numbers A breeding value estimates how much of a trait a bull will *pass on* to his offspring, expressed as a deviation from the breed average. You'll see it as PTA (Predicted Transmitting Ability) or EBV (Estimated Breeding Value) depending on the country.
- A bull with +400 kg milk PTA is expected to raise his daughters' milk by ~400 kg per lactation above the average bull. (PTA is "per transmitting", roughly half his own genetic superiority, because each parent passes on half their genes.) - Positive is better for production traits (milk, fat, protein). For some traits — somatic cell score, calving difficulty — lower/negative is better. Always check which direction is good.
The traits that matter (don't just chase milk) Production: - Milk (kg) — volume. - Fat and protein (kg and %) — what you're actually paid for; protein especially. A bull who adds litres of watery milk is worth less than one who adds protein.
Functional traits — these decide whether a cow stays profitable or breaks down: - Daughter fertility — gets back in calf on time. The single most important trait for a smallholder; a cow that won't conceive is a cow you're feeding for nothing. - Calving ease / calving difficulty — critical when breeding heifers; a hard calving can kill the heifer and the calf. - Somatic cell score (SCC) / udder health — lower is better; protects against mastitis. - Productive life / longevity — how many lactations she'll last. - Udder and feet & leg conformation (type) — a well-attached udder and sound feet keep her milking. - Temperament — handling ease, which matters in a hand-milked smallholder system.
Reliability (%) — how much to trust the number Every breeding value comes with a reliability (or "Rel %"). It tells you how much daughter data backs the proof: - Proven (daughter-proven) bulls — many milking daughters, reliability often 90 %+. The number is trustworthy. - Genomic young bulls — evaluated from a DNA test before they have daughters, reliability ~50–70 %. Often cheaper and genetically exciting, but more of a gamble on any one bull. A sound strategy is to use a *team* of several genomic bulls across the herd so the average comes through, and proven bulls on your best cows and heifers where you want predictability.
Indexes — the one-number shortcut Selecting on a dozen traits is hard, so each country publishes an economic index that weighs the traits by their value: Net Merit (NM$) in the US, PLI in the UK, and so on. A higher index = more profit per daughter *in the system that index was built for*. Use the index to make a shortlist, then look at the individual traits that matter most for your system.
The tropical / smallholder reality This is where farmers go wrong copying temperate-country choices. A bull bred to top the charts in a Wisconsin free-stall barn on a total mixed ration may sire daughters that are too big, too hungry, and too fragile for a zero-grazing smallholding on Napier and dairy meal. For most Kenyan systems: - Weight fertility, longevity, and udder health over peak milk. A modest cow that calves every year and lasts six lactations earns far more than a high-peak cow that's culled barren after two. - Match the bull to your system, not the catalogue cover. Consider heat tolerance and the demands the daughters will place on your feed. - Crossbreeding with a hardy breed (Sahiwal, or a Jersey cross for higher solids and a smaller, more efficient cow) is often a better fit than chasing a pure high-yield Friesian you can't feed.
Matching a bull to a cow (mating, not just buying) - Correct her weaknesses. Use a bull strong where she is weak (e.g. a poor-uddered cow → a bull with a high udder score). Don't pile strength on strength. - Avoid inbreeding. Don't use a bull closely related to the cow; check pedigrees. Inbreeding depresses fertility and vigour — the opposite of what you want. - Use calving-ease (easy-calving) bulls on heifers and small cows, always.
The honest bottom line The cheapest straw is rarely the best value, and the highest-milk bull is rarely the best choice. Pick for fertility, longevity, udder health, and fit to your feed and climate, use the economic index to shortlist, respect reliability, and buy from an authorised, properly-stored source. Genetics compound: a few good decisions now quietly raise your whole herd for years.
Sources KAGRC — AI Services & Bull Catalogue (Kenya). ICAR / Interbull — Genetic Evaluation Standards. KALRO / ILRI — Dairy Cattle Breeding for Smallholder Systems. DAGRIS — Crossbred Dairy Performance in East Africa.