The negative density dependences due to pathogen or herbivore are considered to be important to maintain the diversity in plant communities, so called Janzen-Connell hypothesis. Especially the effect of species-specific soil fungi on the abundance and species richness has been documented. The key component of Janzen-Connell hypothesis is the distance-dependent mortality, which expects the high mortality of seedlings near the conspecifics. It is expected to lead the low recruitment near conspecifics and to prevent the over-dominance of the common species. Although the role of Janzen-Connell hypothesis on the maintenance of diversity has been extensively studied, little is known about the effect on the spatial pattern formation of plant communities. To explore the role of activity of soil pathogen on the spatial pattern of vegetation as well as the species richness, we constructed a 2-demensional spatially explicit mathematical model which described the dynamics of plant community and species-specific soil pathogen. It was assumed that each plant species had a specialist pathogen which infected on the seedlings of focal species and kill them. The growth and spread of pathogen was described as reaction-diffusion equation. As the results, as the growth rate of soil fungi increased, each individual more sparsely distributed and higher species richness was maintained. There was an optimal value of diffusion coefficient of pathogen, in which the highest species richness was realized and the most sparsely each individual distributed. The results suggest that there is the most effective mobility of pathogen to maintain the diversity in forest ecosystems.

     Gastropods are the largest and the most diversified taxon in the phylum Mollusca. Commonly known as snails and slugs, gastropods are able to live in remarkably diversified habitats through a wide range of adaptation over time. Their shells form an important component for their adaptation because most of them protect their body from physical, physiological, and ecological perturbations with their shells.
     In this talk, I will show two different theoretical approaches for understanding a morphological diversity of gastropod shells. First approach is an example of ultimate explanations of a morphological diversity of gastropod shells. Although shells of gastropods are important component for their adaptation, the shells can become a burden to them because they need to form and bring their shells. I assessed these constraints as efficiency of shell construction and shell balance, respectively. Biomechanical analysis showed that these two functions of shells are conflict in each other in terms of shell height and umbilical width. Most of actual specimens show moderate forms, which do not have extremely ill balanced or inefficient shells. Moreover, the biometric results suggest that land snails are more highly constrained than marine species in achieving a balance between shell balance and efficiency of shell construction.
     Second approach focuses on an aspect of proximal explanations. The shell is formed through an accretionary growth which the epidermis of the mantle secretes shell materials little by little. Most of gastropod shells coil spirally due to asymmetric growth along inner and outer lip of an aperture. I propose a method to quantify such growth pattern along an aperture from actual specimens, and a comparability of the growth pattern with experimental data.

     World population is expected to reach 9.7 billion by 2050. To feed the growing population, farmers must produce 70% more food by 2050; an average increase in production of 44 million metric tons per year is required. To achieve this increase in food production under limited agricultural resources, we need to substantially increase yield potential of crop plants. As a solution for the increment of yield potential of crop plants, genomic selection (GS), a technology for rapid genetic improvement of an organism, has received a lot of attention recently. In GS, the relationship between genome-wide marker genotypes (x) and phenotype (y) is modeled with the function y=f(x). Using this model, we predict the potential of individuals under selection as f(xs) based on their marker genotypes (xs) even without their phenotypic records and select individuals with desirable predicted values. Because with GS we can select individuals without evaluating their phenotypes, an unprecedented breeding system, which is unachievable with conventional selection, can be realized. For instance, even in Japan, we can develop varieties that are adaptable to hostile environments around the world using selection based on the prediction model. In another instance, we can speed-up breeding process largely using GS with a rapid generation advancement technique using an artificial environment. From these reason, theoretical researches of GS and researches for implementation of GS to breeding programs have been conducted in agricultural research institutes over the world.
     Within the context, we have conducted the following researches to implement GS to plant breeding and to accelerate and streamline breeding process: (1) Evaluation of potential of GS based on simulation (Iwata et al. 2011; Yabe et al. 2013; Yabe et al. 2014) and experimental studies (Yabe et al. in prep). (2) Development and accuracy evaluation of GS prediction models in various species (Iwata and Jannink 2011; Iwata et al. 2013a; Onogi et al. 2014; Onogi et al. in press). (3) Development of new modeling methods for GS and GWAS (Iwata et al. 2010; Hayashi and Iwata 2010; Hayashi and Iwata 2013; Iwata et al. submitted; Onogi et al. submitted). (4) Development of new breeding methods with GS models (Iwata et al. 2013b) and computer programs for GS modeling (Onogi et al. submitted; Galliot in prep). In this talk, I will introduce our current researches on genomic selection for accelerating plant breeding. I will also talk about issues of current genomic selection breeding and our challenges for solving the issues.