I just added one yesterday: “Meta-analysis at the intersection of evolutionary ecology and conservation.” You’ve spotted the trend, right? I was planning to write about the whole bunch of them together, a mega-post on the latest thinking on the relationship between biodiversity on the one hand and ecosystem health on the other. They’ve been there for months. I just haven’t been able to get round to them, what with one thing and another. Like work, mainly. And maybe a bit of laziness.
But there’s an upside to prevarication. You wait long enough to do something, if the thing is really important, you’ll find someone does it for you. And so it has proved on this occasion, because “Biodiversity loss and its impact on humanity” has just come out in Nature, and it provides a comprehensive review of the sort of papers that have been sitting in that corner of my desk, lots of them, going back years.
Which means all I need to do here is further summarize the already admirably succinct synthesis that the authors provide. 1 And that I think I can do in a few bullet-points:
Loss of biodiversity (really loss of diversity in functional traits) decreases the efficiency and stability of ecosystems.
The impacts of biodiversity loss on ecosystem functioning are big, accelerating and predictable.
Biodiversity is predictably positively correlated with the provisioning of some ecosystem services, but the data in the case of other services is either mixed, insufficient or runs counter to expectation.
And yes, the dataset included crops, and here’s the snippet of the summary table that deals with agrobiodiversity and ecosystem services: 2
No doubt about the importance of genetic diversity to yield, though surprisingly mixed results for species diversity. But look at the numbers of data points involved (N): 575 data syntheses (DS) for genetic diversity and 100 for species diversity. Makes that pile of papers I’ve been avoiding look rather puny. And me not just a bit lazy.
The internet has been a bit a-flutter recently on the subject of declining nutrients, particularly in fruit and veg. We nibbled as much a week or so ago, and I asked whether the data had been published. 3 Indeed it had, as Amanda Rose graciously pointed out. As for reasons, Amanda said:
An editor of Organic Gardening magazine suggested that use of chemical fertilizers and subsequent decline in soil minerals was the cause. A peer reviewed study of the data provides another explanation: commercial cultivation of seed has traded nutrient density for yield, pest resistance, and transportability, factors critical to the commercial success of any crop.
Off, then, to the peer reviewed study. The best I can say is: it’s complex. Don Davis, of the University of Texas, and his colleagues used existing USDA data on the nutrient content of 43 garden crops, mostly vegetables, to compare quantities from the 1950 edition of the USDA’s “Composition of Foods” with those in the 1999 edition. Many pitfalls await the unwary in such an exercise, not least of which is the lack of basic information about some of the sample sizes and distribution of the results. Davis and his colleagues seem to have thought of them all, and worked round them as far as possible in various ways. 4 In essence, Davis et. al compare median values in 1950 with medians in 1999; less than 1 represents a decrease, more than 1 an increase. And the bottom line:
As a group, the 43 foods show apparent, statistically reliable declines (R < 1) for 6 nutrients (protein, Ca, P, Fe, riboflavin and ascorbic acid), but no statistically reliable changes for 7 other nutrients. Declines in the medians range from 6% for protein to 38% for riboflavin. When evaluated for individual foods and nutrients, R-values are usually not distinguishable from 1 with current data. Depending on whether we use low or high estimates of the 1950 SEs, respectively 33% or 20% of the apparent R-values differ reliably from 1. Significantly, about 28% of these R-values exceed 1.
That needs unpacking. First, there’s the question of “apparent” declines. Davis points out that while his statistical approaches eliminate random uncertainties, it is always possible to postulate a systematic error affecting any particular nutrient in either direction. Iron, for example, is tens of thousands of times more abundant in soil than in crops; merely changing the way samples are washed, to remove more soil in the 1999 samples, would result in a large decrease in iron.
That is why I call these R values “apparent”.
But there’s a kicker:
However, it would seem scarcely credible to attribute all the statistically significant median R < 1 to multiple systematic errors, each one operating in only one direction.
So the overall decline, considering all the crops and all the nutrients, is real enough, even though for each crop and each nutrient most of the differences (between 67% and 80%) are not statistically significant.
What, no vitamin A?
Some nutrients did increase, most notably vitamin A and riboflavin. Which makes it plenty delicious that in all the fluttering about declining nutrients, Scientific American (no less) chose to major on carrots, 5 and to blame it all on “soil depletion”.
You mean it isn’t soil depletion?
Davis comes down firmly against the “organic” idea that “chemical fertilizers and subsequent decline in soil minerals” is the problem. The fact that some nutrients definitely increased, coupled with the observation that protein (mostly nitrogen, N), phosphorus (P) and ash (mostly potassium, K) each declined from 1950 to 1999 – despite the fact that “chemical fertilizers” are largely N, P and K – puts paid to that idea.
Instead, Davis et al. focus on two different “dilution” effects. Environmental dilution is an idea that has been around a while. In essence, plants in well-fertilized, well-watered soil grow larger but take up the same total nutrients. So the nutrients are spread through a larger crop, giving a lower concentration per gram of dry matter.
Davis also identified another effect that he called genetic dilution. The bulk of the dry weight yield of most fruits, vegetables and grains is carbohydrate. When breeders select for high yield, they are effectively selecting for high carbohydrates …
… with no assurance that dozens of other nutrients and thousands of phytochemicals will all increase in proportion to yield. Thus, genetic dilution effects seem unsurprising.
The best evidence for genetic dilution comes from contemporary studies in which a time-series of varieties is grown side by side, eliminating all the problems of historical data, agronomic practices, etc etc. Davis cites a few such studies, and more are being published. In general, they do show declines in nutrient density, which can be ascribed to breeder selection.
So what about the question Luigi posed a while back? Is modern plant breeding bad for your health? Possibly, a little, but not in the way most people think. The big problem isn’t that nutrients in fruit and vegetables have decreased; despite the decline, they remain among the most nutrient dense foods you can eat. So eat them. Choosing one growing regime rather than another isn’t going to make any difference. The big problem is that we’re still eating too little fruit and veg and too much refined staples. Last word to Don Davis:
Our findings give one more reason to eat more vegetables and fruits, because for nearly all nutrients they remain our most nutrient-dense foods. Our findings also give one more reason to eat fewer refined foods (added sugars, added fats and oils, and white flour and rice), because their refining causes much deeper and broader nutrient losses than the declines we find for garden crops.
Technology should allow us to increase selected nutrient concentrations. 6 But will we learn 20 or 40 years later that there were new, unintended side effects? Another question looms large: Is it wise, in the era of technology, to keep crop size (or even the concentrations of a few, selected nutrients) as our primary measure of farming success?
What I should have mentioned in my recent post on online information resources for wheat varieties, as Jeremy then pointed out to me, is that those linkages between the different databases that I was hoping for are going to be much easier to make and maintain if we had a system of digital object identifiers (DOI) in place for wheat varieties. A DOI is a string that specifies a unique object within a particular system. The object could be a scientific publication. Or indeed a scientist. It is probably about time we implemented such a system for genebank accessions in general, and crop varieties in particular. No?
But anyway. Say you’re reading about rattans in West Africa and you get interested in, say, Eremospatha barendii, which is kind of rare and probably perhaps sort of endangered, maybe. So you head on over to GBIF to get a better idea of its geographic distribution, but you come up blank. But then you do some more digging and you realize that there’s a new taxonomic revision, describing in minute detail the morphology and ecology of all the African species, with nice drawings and lists of specimens and identification keys. 7 And, by golly, it has pretty maps too.
Pretty, but unusable. The damn things are in a PDF, not the nice Google Earth files you would have got from GBIF. But the coordinates of all the specimens 8 are given in the text, so you extract them from the PDF and plonk them into Google Maps. It’s the little green arrow in southern Cameroon shown below.
But you also want to know to what extent that area is threatened. So you head on over to the Global Forest Disturbance Alert System and you look at the latest data on where there is forest disturbance happening in Cameroon, which are the little red dots here.
Now you can see if that population of yours is perhaps threatened. Well no, you can’t do that now, not easily, because there’s no way to export the little red dots to Google Maps, or import the little green arrow into the Global Forest Disturbance Alert System. But I’m sure you will be able to do that one day. And then you could actually visit 3.066667 N 10.716667 E, and check out that Eremospatha barendii population, assuming it is still there, what with you spending so long mashing up the data and all, and also ground-truth any disturbance that the Global Forest Disturbance Alert System might have, er, alerted you to; maybe even describe its causes. And annotate the little green arrow and the little red dots with your observations.
Phylogeography of Asian wild rice, Oryza rufipogon: a genome-wide view. Fancy markers come through where lesser breeds caused confusion. Two groups, clinally arranged, with the China-Indochina group close to indica, neither close to japonica. So one, Chinese, domestication event, yada yada.
