A deeply cynical skeptic, this claim immediately throws up questions in my brain, like:
- “Is this true? What is it based on? Who did the study?” and;
- “How much more milk do they produce? Is it enough to matter?”
Particularly since I often see this factoid displayed in sexed semen advertising, and just because it seems so darn convenient, I decided to investigate the science behind it.
What I found surprised me. The claim is based firmly upon data. The study behind the factoid is legitimate, independent, and not funded by commercial interests. It’s also published in the Public Library of Science (PLOS) journal, and is therefore both peer-reviewed and open-access (read: free).
Anyone can read it. In fact, here it is! Holsteins Favor Heifers, Not Bulls: Biased Milk Production Programmed during Pregnancy as a Function of Fetal Sex.
Are you as shocked as I am? For those of you who are interested, let’s walk through the study together.
I’d like to start with looking at who did the study. This is important because we want to know if it was funded independently or if it was done to support a commercial product. In this case, the main authors are Assistant Professor Katie Hinde (Department of Human Evolutionary biology, Harvard University) and Associate Professor Barry J. Bradford (Department of Animal Sciences and Industry, Kansas State University) – highly reputable research institutions. If it was commissioned or performed by a company, this doesn’t mean that we automatically throw out the paper – but it does send up flags that will make us look long and hard at how the study was designed.Let’s focus on Dr Hinde’s work to start off with. Unlike a lot of the science that gets featured on this blog, Katie has nothing to do with dairy or agriculture. Instead, her work focuses on how mother’s milk influences infant development in rhesus monkeys. To understand her interest, let’s have a brief lesson in reproductive biology. (Bear with me on this one).
Evolution is about survival of the fittest, where ‘fit’ means ‘able to successfully reproduce’. Depending on the species, a particular sex may be more or less likely to be able to pass on their genes. For example, a male lion may be able to pass his genes to multiple females – therefore, this system has a male sex bias. On the other hand, a male seahorse provides the majority of care to his offspring, allowing females to reproduce with greater abandon.
(This is only one model of reproductive success amongst many other competing models, and we won’t go into those too much here. Models in use tends to be species-specific.)
(Also, try not to think too deeply about how this may apply to human behaviour. Just trust me on this one.)Parents want to spread their genes as much as possible. There are two alternatives in the game – have a son, or have a daughter. In a system based on male competition, sons tend to be good gambles – you can end up with a ‘hero’ who spreads his genes wide and far, or a ‘zero’ who does not spread his genes at all. Daughters are more reliable, but limited in how many offspring they can produce.
The theory is that under certain environmental conditions, mothers of some species can adjust their offspring sex ratio. If a mother is dominant or experiencing good conditions, she may take the gamble to produce a son, providing she has available resources to maximise his chances of success. If she is weak or in poor condition, she produces a daughter to ensure that her genes propagate instead of risk producing a weak and evolutionary dead end son.Now, this is where cows come in. While some species can influence the sex ratio of their offspring (scientists have identified 198 of these in a single study), others may be able to adjust their investment in maternal resources. Since this is a dairy blog after all… let’s be clear: we’re talking about milk. The hypothesis that Katie Hinde was testing is that in sex-biased milk production, a mother is able to sink more resources (her nutritional energy) into nourishing sons, as they need to be competitive early to become successful. This is an extension of the Trivers-Willard hypothesis (read more about it here).
There is some evidence to support this theory. Studies show that Iberian red deer and Rhesus macaques alter their milk composition depending on gender of offspring. Male offspring of deer mothers are favoured with higher protein content milk, while macaque mothers provide higher fat content to sons. This is not affected by environmental conditions (as hypothesized), but is affected by parity: differences are greater in smaller, younger mothers compared to those who have had multiple births.
Okay, so how does this apply to dairy cows?
Unfortunately, studying milk composition in wild animals is very difficult. It’s hard to get a large enough sample of animals, it’s hard to extract the milk (if you think it’s tricky getting cups on a cow, imagine milking a monkey!), and you don’t know much about the history of each individual. In other words, there just isn’t enough data to draw meaningful conclusions.This is where our second study author, Dr Bradford, stepped in. After reading this post on Dr Hinde’s blog (amusingly named ‘Mammals Suck’) lamenting the need for more lactation data, Barry realised that the dairy industry is an information goldmine. Farmers collect huge amounts of data on cows and cow lactations – not only things like composition, but also records across multiple lactations, management decisions, and histories.
Using the combined powers of scientific collaboration and social media, Katie and Barry were then able to 1) improve our knowledge of mammalian lactation and 2) investigate the potential economic effects of sex-bias on the dairy industry in one fell swoop.
And that’s where this paper came from.
Barry provided access to herd records through US Dairy Records Management Systems (something a little bit similar to ADHIS). Katie provided her knowledge of reproductive biology. Together, their research team retrieved every lactation record from 1995 to 1999, cleaning them up to get rid of incomplete or duplicated entries. They ended up with 2.39 million lactations, covering 1.49 million cows – a huge sample size compared to the tiny numbers that animal behaviour scientists normally settle for. (As a comparison: Katie’s rhesus monkey study includes 106 mothers, and the Iberian deer study covers 60 hinds. It took researchers years to gather this much information, whereas millions of dairy industry records could be accessed in a fraction of this time).
By having such a vast database to begin with, the researchers could be ruthless in excluding unwanted data. They controlled for the effects of dystocia (which reduces lifetime lactation and is associated with male calves). They stopped at the year 1999 to avoid sexed semen, which started seeing use in America in the 2000’s. They sorted through information about which cows were treated with bST – a hormone that is injected into cows to increase production (not legal in Australia). Finally, they only examined a single breed (Holsteins) and removed lactations influenced by abortions and twins.*
The result? Well, this is where the claims are coming from.
- First-time heifers giving birth to a daughter produce 142 ± 5.4kg more milk over a 305 day lactation period than those giving birth to sons.
- This represents a 1.3% increase in whole-lactation production, without any change in fat or protein composition.
- The effect is similar but marginally smaller in cows during their second to fifth lactations.
Interestingly, they also discovered that the sequence of genders affects production.
- Having a daughter first is ‘protective’ – having a son in subsequent births does not significantly decrease production. (Cows with two daughters produced 7954 ± 12.6kg of milk, and cows with a daughter and then a son produced 7940 ± 12.3kg on average).
- Having a son first results in decreased production. However, this is partially improved by having a daughter in the next pregnancy. (Cows with two sons produced 7768 ± 11.4kg of milk, and cows with a son then a daughter produced 7876 ± 12.2kg on average).
See the graph below for a better visual representation of these numbers (this was sourced directly from the paper itself).This is the opposite result to what the reproductive biologists were expecting – cows favour heifers instead of bulls. But while biologists may weep, dairy farmers rejoice! The results fit neatly into our preferred system.
So, is a 1.3% increase in whole-lactation production significant? This research shows that the cumulative increase in milk yield over two lactations of heifers (compared to bulls) adds up to about 445kg of extra milk. For a quick back-of-the-envelope calculation:
- 445kg multiplied by the current milk price of 43 cents/L = $191.35
- The natural birth sex-ratio for dairy cattle is 53% male. Out of 100 pregnant cows, 53 will have bull calves. Out of those 53 cows, 28 go on to have a second bull calf.
- 28 cows each seeing a benefit of 445kg extra milk across two lactations would have totalled 12,460kg – a $5,357.80 shortfall in potential income.
- If we use sexed semen, the birth sex-ratio decreases to 10% male. Out of 100 cows, 10 have bull calves. Out of these 10 cows, 5 go on to have a second bull calf (presuming sexed semen is used on heifers only). This results in only a $956.75 shortfall in potential income.
- The difference between the two scenarios is $5,357.80 – 956.75 = $4,401.05 extra income for 100 pregnant cows across two lactations using sexed semen on heifers.**
This is a very rough calculation – for example, it doesn’t take into account the feed costs of extra milk production. It also doesn’t account for the fewer days in milk due to lowered sexed semen conception rates (however, the introduction of SexedULTRA™ semen onto the Australian market has the potential to significantly mitigate this effect).
For farmers already using sexed semen having weighed up the costs and benefits (see here for a quick list), the extra milk income is a nice little bonus. For farmers currently on the fence about sexed semen – does this information change your mind about using it?
*It’s also important to note that these results were obtained in the United States, using American Holstein cows. A study was recently performed on New Zealand Holstein Friesian and Jersey cows which showed similar results, but with lower milk yields (in accordance with lower overall production). This study can be read here.
**This erroneously used $/kg MS instead of c/L before – thanks to several eagle-eyed readers for pointing this out! This extra income is still good, but definitely no longer in the ‘wait, I could fill several Olympic-sized swimming pools with this kind of money’ category… unfortunately.