Nature © Macmillan Publishers Ltd 1998

8

correspondence

Sir — I appreciate the timely appearance ofthe article and editorial on the support ofUS agricultural science (Nature 394, 207 &210–211; 1998). They should serve toinform the scientific public about howscience at the US Department ofAgriculture (USDA) is currently fundedand the uncertain future of the authorizednew monies that would provide additionalfunding for plant and animal genomeresearch (among other topics). The articlealso stressed the beleaguered state of theNational Research Initiative (NRI), thedepartment’s competitive grants arm,where I am chief scientist on leave from theUniversity of Missouri.

I want to emphasize that the NRI doesmore than support plant science. There are29 panels, covering all aspects ofagricultural research, from economics togenomics. Yes, the plant people have a

tough time compared with the typicalinvestigator funded through the USNational Institutes of Health (NIH), but,with ingenuity, they can get support fromthe National Science Foundation or eventhe Department of Energy. There is stillsome plant science funded by the NIH, ifthe work is properly dressed up to be healthrelated.

There are, however, certain areas wherethe NRI is about the only source of federalfunds. Consider, for example, theindividual working on an agriculturalproject to study the interaction betweengrazing, plants, soils and water, oruniversity scientists studying animaldiseases that are not ‘models’ useful inhuman medicine. There are also broadareas of animal science that areunderstandably unpalatable to NIH;ruminant nutrition or lactation don’t

exactly appeal to study sections, and who isgoing to fund competitive grants forgenomics of agriculturally importantanimal species, if not the NRI?

I am on temporary (60%) assignment tothe USDA. I can afford to do this onlybecause of the indulgence of my universityand because I am well funded by NIH onprojects that, in a more logical setting,would be of at least equal interest toagriculture as to human health.

A sorry footnote is that the NRI maylose a further $5 million of its already smallappropriation because of an amendment topay for a fiscal obligation elsewhere in theUSDA unrelated to competitive grants.R. Michael Roberts158 Animal Sciences Research Center,University of Missouri, Columbia,Missouri 65211, USAe-mail: robertsrm@missouri.edu

NATURE | VOL 394 | 6 AUGUST 1998 517

Finding the funding for agriculture

101 uses for a naturalhistory museumSir — The Briefing about natural historymuseums was very informative but theeditorial, “101 uses for a dead bird”,promotes a number of misconceptions(Nature 394, 105 & 115–119; 1998).

Museum collections are, as yourecognize, the ultimate basis of most if notall biodiversity studies, including those thatpermit the decline in biodiversity to bemonitored and controlled. Principal inputis through the discipline of systematics, thatis, the recognition of species, their variationand their relationships. Systematicscondenses the vast amounts of disparateinformation in large numbers of specimensfrom many localities and puts it incomprehensible and usable form. Museumsalso act as repositories for other dataincluding those on species distribution.

Systematics and related data collectionhave been pursued with increasing pace innatural history museums since theirfoundation, often in the nineteenth or even eighteenth centuries. Strong internalcultures within museums have helpedensure continued systematic output, even in conditions of considerableadversity. Any implication that museumsand similar institutions have not previously been conducting what areessentially biodiversity studies is wrong;they are the very places where suchinvestigations have been predominantlycarried out.

You also state that museums have beenparochial and “can no longer afford to work

in isolation”. In the vast majority of cases,they never were and never did. There hasbeen extensive global cooperation betweenlarge museums for nearly 200 years. Forexample, George Boulenger, the lonecurator of lower vertebrates at the NaturalHistory Museum in London, exchangedinformation in ten years around the turn ofthe century with some 700 correspondents,many in museums in other countries,including 20 in Italy alone. Such large-scaleinteraction is far more extensive today. Notonly information but study specimensconstantly pass between museums aroundthe world and have done so for many years.In the Natural History Museum, manyhundreds of loans involving tens ofthousands of specimens are dispatchedannually.

There is already a great deal ofcooperation between relatively rich andpoor nations, including the case youmention of Mexico’s National Commissionof Knowledge and Use of Biodiversity,which sent scientists to museumsworldwide to assemble data on Mexicanspecimens. The commission is to becongratulated on producing cheaply abiodiversity management system by makinguse of these data, but it must beremembered that the information wasreadily available because museumscooperated willingly and actively in theproject. More importantly, they hadpreviously put a far greater amount of effortand expense into accumulating andcurating the material, painstakinglyidentifying it and subjecting it to furthersystematic analysis, resulting in thedescription of many new species, and

making the data easily accessible. After allthis, putting the information into adatabase as the Mexicans did wascomparatively easy.Nicholas ArnoldDepartment of Zoology, Natural History Museum,Cromwell Road, London SW7 5BD, UK

German ambivalence togenetic engineeringSir — You reported the results of a Germanproject on perceptions of biotechnology,saying that Germans are deeply suspiciousof genetic engineering and thatexpectations are predominantly negative(Nature 393, 299; 1998). But our studyshows that the predominant attitude ofGermans towards genetic engineering is infact ambivalence.

Half the population thinks that geneticengineering is neither good nor bad. Alongwith this ambivalent perception of geneticengineering in general, there are ratherclear, and differentiated, reactions tospecific applications of genetic engineering.Fewer than 2% of the sample either agreewith or reject all applications.

Medical and pharmaceuticalapplications are accepted at the same levelas other positively evaluated, uncontestedtechnologies, for example computer andinformation technologies. Agriculturalapplications are still not accepted.Jürgen HampelCenter of Technology Assessment in Baden-Württemberg, Industriestrasse 5,D-70565 Stuttgart, Germany