David A. Andow教授のセミナー（近畿地区）のご案内
(2010年3月 2日公開)
国際交流基金で招聘する David A. Andow教授（University of Minnesota）には、第54回大会におけるシンポジウムでご講演いただきますが、近畿地区でも2回のセミナーを行っていただきますので、ご案内いたします。内容は、大会シンポジウムとはそれぞれ異なりますので、ふるってご参加ください。
（事務長 日本典秀）
日時：2010年4月1日（木）15:00-17:00
会場：〒631-8505奈良市中町3327-204
近畿大学 農学部 教室棟207教室
交通の便は下記の近畿大学農学部HP参照
http://nara-kindai.unv.jp/index.html
演題：Assessment of the effects of GE crops on biological diversity: the process and the data
Commercial use of genetically engineered (GE, also known as “genetically modified”) organisms have tended to generate controversy wherever they are proposed around the world. One part of this has been the potential ecological effects of these organisms. There are three kinds of potential ecological effects that require investigation: 1) gene flow and its consequences, 2) effects on species, ecological communities and ecosystems, and 3) evolutionary effects. In this talk, I will focus on the second of these.
Based on experiences with GE crops, seven main kinds of potential adverse effects on species, ecological communities and/or ecosystems have been identified. For GE crops, these are: a) adverse effects on crop production, b) reduced soil health or quality, c) reduced value of non-crop economic activities, d) reduced cultural value, e) increased conservation concern, f) reduced environmental quality, and g) increased human disease via environmental change. Analysis of each of these has been uneven across regulatory regimes. Here I present a detailed analysis of 80 laboratory studies of the effects of Cry toxins and proteinase inhibitors (PIs) on natural enemies that were published in the peer-reviewed literature. Natural enemies have been studied because of concerns that GE crops could disrupt naturally occurring biological control, thereby giving rise to secondary pest outbreaks. They are tested first in the laboratory because under some regulatory regimes, these are the required tier 1 tests. These tier 1 tests play a key role in regulatory decision-making. If a natural enemy is significantly affected in these tier 1 tests, then tier 2 tests must be initiated. If it is not affected in the tier 1 test, then the toxin is considered ecologically safe, and no other tests are needed. Thus, proper determination of the outcome of these laboratory tests is essential for the risk assessment process to proceed with confidence.
Using a simple meta-analysis, we examined the 1648 responses in the papers. Only Cry1Ab, Cry1Ac, and GNA have received sufficient study that general conclusions might be drawn. Most of the responses have been reported from Europe and east- and austral-Asia, which are not the geographic locations where most of the GE crops are used. Five species account for nearly 1/3 of the reported responses. There has been an overemphasis on these five species when some taxa have been rarely examined. We found widespread evidence that both Cry toxins and PIs have effects on natural enemies, both directly and indirectly. These effects are both positive and negative, and occur across many different kinds of responses. Our work has been publicly criticized, and we have responded to all of these criticisms. I summarize these exchanges, and conclude that there are effects on natural enemies that were not identified by the tier 1 tests. These tier 1 tests do not reliably indicate the presence/ absence of an effect and should not be relied on to lend confidence to a risk assessment.
日時：2010年4月2日（金）15:00-16:30
会場：京都大学 農学・生命科学研究棟セミナー室（1）（農学部総合館の北側にある建物の1階です）
演題：Managing the rate of evolution of virulence in insects to host plant resistance
The repeated and rapid evolution of virulence in arthropod herbivores to methods to control them is one of the empirical linchpins underlying evolutionary theory. During the past 15 years, the managing the rate of virulence evolution has been stimulated by the advent of transgenic Bt crops. Starting with a simple directional selection model first developed by Comins in 1976, my colleagues and I have been exploring the rich variation in evolutionary rates that simple modifications of the model produce. We start with the simplifying assumptions that virulence is determined by a single recessive autosomal allele, selection is very intense, and the herbivore population is regulated by some density-dependent factor. Spatially heterogeneous selection (a refuge without selection) greatly changes the rate of virulence evolution, although the reasons for this are not immediately obvious. By separately examining dispersal of males and females separately, and allowing variation in dispersal from habitats with selection and those without selection, we found that: a) dispersal of males and females from the habitat with selection has little effect on the rate of evolution; b) dispersal of males from the non-selected habitat always decreases the rate; and c) dispersal of females from the non-selected habitat always increases the rate. Viability selection dominates the process, and sexual selection and assortative mating have little effect, even though assortative mating is frequently cited as the reason for changes in the rate of evolution. Thus, virulence evolution is delayed when males are much more dispersive than females, and virulence and host range expansion might be expected when males are much less dispersive than females. We also examined the case of two independent plant resistance alleles with two corresponding independent virulence loci, each with one recessive allele for virulence to one of the two resistance alleles. We found that there are two modes of resistance evolution. When populations of purely susceptible insects can persist, the rate of virulence evolution changes slowly with the proportion of non-selected habitat. However, once this proportion crosses a threshold below which a susceptible population cannot persist, the rate of virulence evolution increases rapidly. We compared evolution in fine-grained environments with these results from coarse-grained environments. As found for single locus resistance, fine-grained environments increase virulence evolution if larvae move among plants. The dynamics of the two locus system are driven by the same factors as the single locus system with viability selection dominating the process.
To use this theory to guide management, pedigreed screens can be used to estimate the frequency of rare, recessive virulence alleles. An F2 screen has been used in many countries to monitor virulence evolution to Bt crops. Other pedigreed screens are being implemented for monitoring virulence in rice gall midge to the 11 resistance loci in rice. I review these data and the evidence that virulence is evolving in several of the target herbivores of Bt crops, and describe the present state globally of virulence to Bt crops.