In China, women constitute about 70 percent of the agricultural labour force and perform more than 70 percent of farm labour. The general pattern is - the poorer the area, the higher women's contribution.

The world over, 70% of the agricultural workers, 80% of food producers, and 10% of those who process basic foodstuffs are women and they also undertake 60 to 90% of the rural marketing; thus making up more than two-third of the workforce in agricultural production. Despite the fact that women produce much of the food in the developing world, they also remain more malnourished than most men.

I could go on with more information on the vital role that women play in feeding the world but I know from previous comments that you do not argue in good faith, or with any discernible knowledge but only with the sole purpose to repetitively express your contempt for women.

Mal,
I have not expressed contempt for women. I think you've made that up.

Well then use census data or occupation - eg doctors versus hospital cleaners. Forbes billionaires are hardly a large or typical sample. If by Trivers-Willard you mean simply that male infant mortality is traditionally higher and higher income, higher status families have lower infant mortality, then you subvert the purported evolutionary benefit as the driver of it.

If you are looking for evolutionary explanation you need to
a) show that there exists or could exist a molecular/genetic mechanisms to underlie the two strategies.
b) Space and time for the two strategies to compete and the genetics underlying one strategy to become more prevalent.
Otherwise you are dealing with nothing more than Just-So stories.

For example there have been suggestions that Hepatitis B +ve mothers might favour male births - the latest evidence from Taiwan suggests there might be a small effect. The mechanism is unknown yet, but at least we can imagine possibilities. Its difficult to construct a mechanism about how a female fetus could come to realisation that its father was a Forbes billionaire and spontaneously abort.

It is also probable that there quite skewed male to female ratios in early contact Australia aboriginal society. So much so that early anthropologists or proto-anthropologists took it for granted that Aboriginal societies engaged in systematic female infanticide. That I think is extremely unlikely, but as to what actually was driving it I can't say. It wasn't Trivers-Willard that is for sure.

Oh dear, where to start.

First passing over the argument from authority "peer reviewed journal", peer reviewed journals are filled with work that can't be reproduced and/or are completely contradicted a few years later. In point of fact pubmed only turns up 44 papers on Trivers-Willard, suggesting it is not a particularly popular concept either with demographers or evolutionists (and of that 44 by no means all support the model - lactating dairy cows being a notable holdout in this prestigious collection of 44 publications. In fairness, the star of the fabulous 44 might well be the Estonian female moose).

Emily Oster did not originate that claim hepatitis B status have influence on sex ratios, it has been around since the late 70s, she simply over-enthusiastically applied it. The recent data from Taiwan (around 3 million mothers) suggests a small but significant effect 106:100 boy/girl ratio to 108:100 ratio - which by the way is about the same magnitude as the effect found with the famine years in China.

Virtually all the data from human populations looking at post birth developments can be explained by the following model.

1. Infant mortality is higher in males than in females - this is commonly observed.
2. Infant mortality is reduced in cohorts of higher status and increased in times of crisis.
3. Better infant mortality outcomes disproportionally effect male infants for the simple reason they have more room for improvement, the opposite is true worse infant mortality outcomes.

Hence, higher status individuals will have relatively more surviving male children than lower status individuals - simply because infant mortality is lower in the higher up the class structure you go.

The benefits in a group selection scenario that would selectively advantage a population for having a HIGHER male infant mortality rate seem to be ephemeral. Whereas selecting for populations that have a lower female infant mortality (non-fetus gestating males are somewhat more expendable) seems to carry obvious benefits.

Of course neither are really scientifically testable in anyway and the slight increase in male infant mortality may simply be haphazard - as evolution can only throw up a small set of possible mutations for selective testing.

But let us know when papers mentioning Trivers-Willard reach 3 figures.

.

1. You can always access Pubmed so long as you have an internet connection, so it disingenuous to pretend otherwise. I accept that T-W is more widespread that the Pubmed search (which I presume you checked!) suggests. Google Scholar results will be inflated by the fact it checks not just abstracts (like Pubmed) but the entire paper including references. Also older abstracts don't get searched by Pubmed.

2. I am not sure what you mean by conspiracy theories. I worked for a period in the lab of your fellow UNSW academic (now emeritus) Professor Barry Milborrow. He told me that 80% of published papers were "wrong" - I always thought that figure a little high, but in fairness perhaps he was basing it on the output of his UNSW colleagues. Suffice to say in all my time at universities I never met any academic who would suggest a paper was beyond criticism or indubitably correct because it appeared in either a peer-reviewed or high ranking journal - and I am sure you wouldn't suggest as such to your students either.

3. Aboriginal sex ratio data I have come across is patchy. Two examples are George Grey and his expedition accounts (which I don't have to hand), the other is "Australian Reciminiscences and Papers of LE Threleld" Volume 1. He was based in the Hunter. On page 138 he provides the following data

1835 (24 returns) Men 904, Women 681, Boys 291, Girls 217
1836 (15 returns) Men 727, Women 461, Boys 225, Girls 169
1837 (16 returns) Men 735, Women 454, Boys 195, Girls 147
There also seems to be a decided gender bias in the numbers of Aborigines moved from Flinders Island to Oyster Cove in Tasmania (I haven't tracked down the data for the original transfer to Flinders Island after the Black Line - admittedly a distorting selection bias in its own right). Early anthropologists took it for granted that female infanticide was practised (some of the nuttier ones like Daisy Bates solemnly informed us Aborigines had babies deliberately in order to eat them), but their reasoning was faulty inasmuch they assumed that more females were a nutritional burden. Modern research tells us that in these type of societies females tended to provide more calories than the males who chased the big ticket items. Hence I became interested in alternative explanations that might explain reported sex ratios (that did seem to show missing girls) and invalidate what seemed poorly evidenced claims of female infanticide. This was not long after Emily Oster was building on Professor BS Blumberg's work regarding Hepatitis B and sex ratios. However, it quickly came clear to me that any effect Hepatitis B might have (and it was present in pre-contact Australia) was not of sufficient magnitude to explain the skewed ratios.

4. I don't much care whether you are interested in open debate or not. You seem not to be able to defend your propositions particularly well to me, but that is by the by. Principally I engage in such discussions in order to encourage myself to have an open mind and do further research - rather than expect an open mind in academics. And yet again it has paid off, as I now find there is some rather interesting work showing in both mice and men that Toxoplasma gondii infection in females tends to skew sex ratios towards males (PMID: 17028886, 17651529). This opens the possibility of a common mechanisms with the small but real Hepatitis B effect.

5. Most zoologists describe the T-W effect as females in poorer condition having more female offspring, while I don't have much problem with that, I think the extension you made to Forbes billionaires is weak and unsupported. In point of fact, the tendency of mothers with hepatitis B or toxoplamosis towards more boys directly contradicts this hypothesis.

6. If the Australian Aboriginal data holds generally, then it would suggest that hunter-gatherer cultures tend to have very skewed ratios towards males. Given the hazards of child birth, that's not entirely surprising. As such any traits, that favoured female survival are going to have far more selective advantage than those traits that promote male survival. This would be sufficient, in my view, to explain all the human data, including longer life-expectancy of women in advanced cultures and the fact that infant males are more susceptible to mortality. One can argue that wealthier families having more sons would be advantageous, but its unlikely that it would have represented an advantage that was ever selected for.

Since the human data can be explained by hunter gatherer societies tending towards skewed male ratios and so a selective advantage is obtained by promoting female robustness in infants, we are able to discard T-W hypothesis as superfluous in this case.

For completeness I will add the rest of the LE Threlkeld data (found on page 283). Data collection ranged from Bateman's Bay to Penrith, to Berrima, to Port Sevens (Threlkeld was himself based at Lake Macquarie)

Sir I herewith enclose the returns for the Aborigines consisting of 102 the number transmitted to me with thanks for the loan of them; but regret that their imperfect state in each year renders it impossible to use them as a criterion to judge of the increase or decrease of the Aborigines, many places not being returned. I have forwarded with them a copy shewing the comparitive [sic] number of the sexes, which is all the available service the returns could afford.

[snip detailed breakdown by district and station]

1838 Men 754, Women 448, Boys 205, Girls 153
1839 Men 737, Women 392, Boys 175, Girls 136
1840 Men 341, Women 208, Boys 130, Girls 100

From my limited reading re 'first contact' of two very different civilizations, is/was there not a tendency for women and children to 'go bush' when strange (male & armed?) outsiders appear? If so reported numbers could be quite distorted. A similar more recent situation would possible be the N.G. highlands. I guess the reports/stories etc from anthropologists/ plantation owners/ miners / missionaries / government patrol officers could assist.