Science uses filtering processes --- grant panels and journal peer review, for example --- to ensure quality and rigor. The same processes have another critical function in addition to quality control: Filters also focus effort and attention on the most promising questions and most interesting results. Investigators cannot pursue all their ideas; funding agencies cannot fund all proposals submitted; journals cannot publish all the manuscripts they receive. However, we understand little about how these filters shape the questions that scientists ask and the discoveries that their studies produce. We propose to develop mathematical theory to better understand how funding and publishing filters influence the direction of scientific research activity and the knowledge that this activity generates. We will focus on three questions. First, how do journals' publication decisions shape the ability of their readership to learn from the scientific literature? Second, how do the incentives created by funding and publication structures motivate scientists to work on certain projects and avoid others? Third, what is the best way for peer review to find consensus when expert reviewers disagree? For each question, we will also ask if and how the behavior of peer-review filters differs for basic vs. applied research. We investigate these questions by building and analyzing mathematical models that incorporate elements of information theory, economics, forecasting assessment, and statistics.

This project will train undergraduate and graduate students, together with postdoctoral fellows, in creating mathematical models of biological systems and confronting them with biological data. Combining approaches from applied mathematics and statistics, trainees will learn a wide range of modeling and parameter estimation methodologies, producing cohorts of mathematical scientists that have received an interdisciplinary training and who are versed in cutting-edge modeling and statistical techniques.

We seek support for a 4-y project building on an existing effort addressing the effects of Ocean Acidification (OA) on coral reefs at the Moorea Coral Reef (MCR) LTER. We request $1.xx million for a program of hypothesis-driven investigations contextualized by the results of our 4-y effort and focused on 2 themes: (1) bridging scales of OA effects across space, time, and functional levels, and (2) evolutionary responses to OA, both within- and among- generations. These themes will be addressed using month- and year- long experiments in mesocosms, outdoor flumes, and on the reef. The research will focus on organism-level questions using corals and calcified algae, and community-level questions using constructed and natural communities selected for ecological relevance relative to contemporary reefs of Moorea. The results will be integrated through synthetic activities embedded within each theme. A postdoc, technician, and local technical labor will be supported, and 4 MS-level graduate students will complete their research within this framework. Intellectual Merit: Acidification of the oceans is one of the most serious threats affecting marine ecosystems, with numerous consequences operating in concert with global climate change and local disturbances. In the last 4 y there have been rapid advances in understanding of the organismic effects of high pCO2, and while many taxa have been studied, there is no framework allowing these effects to be scaled up, or to evaluate the capacity for organisms to change through acclimatization or adaptation. These issues contextualize this project and motivate research on: (1) scale-dependence of the effects of OA, and (2) evolutionary aspects of the response to OA. To address these topics, we will build on our recent OA research, and embed our experiments in a framework informed by the results of the MCR-LTER. We will expand our technological and biological capacity through new research infrastructure: (1) shore-based outdoor flumes for year-long experiments, (2) in-water flumes to serve as Proto-FOCE chambers for assessing the response of intact communities to OA, and (3) common gardens for the cultivation of corals and calcified algae for use in experiments to explore genetic and phenotypic capacities to ameliorate the effects of OA. The research infrastructure will be used for experiments using 4 corals and 4 algae as well as constructed and natural communities. We will address 7 questions and integrate the results with 2 synthetic activities: (1) Do results of short-term OA experiments scale up to predict outcomes over a year?, (2) To what extent do coral reef communities exhibit emergent calcification properties, and how are these properties affected by OA?, (3) How do deeper reef communities respond to OA?, (4) How are in situ coral reef communities affected by OA?, (5) What is the potential for genetic variability to yield OA-resistant lineages?, (6) What is the potential for phenotypic plasticity to modulate the response of corals and algae to OA?, and (7) To what extent do the effects of OA on calcification of corals and algae affect fitness? To integrate the results from these experiments we will use integral projection models (IPM) to scale organismic experiments to demographic and population consequences, and the metabolic theory of ecology (MTE) to address scale-dependence of calcification from organisms to communities. Broader Impact: Broader impacts will accrue through a diversity of research and educational activities platformed from the RUI status of our Hispanic-serving primary campus (CSUN). At the broadest level, our research will contribute to societal objectives of better understanding how OA will affect tropical reef environments, particularly in regards to the ability to project reef structure and function into a more acidic future. This objective will be accomplished through mentoring and training of a postdoctoral scholar, technician, and multiple graduate students that will contribute to US efforts to increase preparation for STEM careers. These t

The fungal, bacterial, and viral microbial communities embedded as endosymbionts within all free-living organisms are extremely diverse and encode the vast majority of genes in the biosphere. The community of microbes in a human, for example, has 100 times more genes than the host it inhabits, and similar results are emerging for virtually all free-living organisms. Although disease is the best studied host-microbe interaction, endosymbiotic microbial populations and communities also are responsible for critical functions in their hosts including nutrient uptake (plants), reduction in inflammatory responses (animals), digestion (animals), anti-herbivore defenses (plants), and pathogen resistance. In spite of the overwhelming diversity and functional importance of the microbial biome to free-living organisms, we have no predictive understanding of the biotic and abiotic factors controlling the composition of host microbial communities or the spatial scale at which these ecological factors affect host and microbial community interactions and functions. The proposed research uses experiments of unprecedented scale to examine the abiotic and biotic drivers and functional significance of a host?s fungal, bacterial, and viral microbiome at scales ranging from individual hosts and local host communities to regional and global bioclimatic gradients on four continents. Experiments will address the following questions: 1. Which environmental and host filters are most important in constraining microbial composition within a host across global, continental, regional, and local scales? 2. How does the within-host microbial community interact with abiotic context to influence plant fitness and susceptibility to colonizing microbes? and 3. How does the within-host microbial community shape microbial population phenotypes and feed back to affect microbial transmission and basic reproductive rates? The proposed experiments will be replicated in 30 grasslands around the world where nutrient supply and herbivory are being experimentally manipulated (NutNet). High-throughput sequencing approaches will be used to elucidate the composition of within-host microbial communities at scales ranging from meters to continents. Manipulative experiments and statistical models will clarify the functional significance of microbial community composition for host fecundity and resistance to microbial colonization, and for microbial phenotypes and interactions. Active collaboration on microbial population spatial modeling will facilitate scaling up from the empirical work to make predictions for microbe transmission among plants, sites, and regions. Grassland communities are experimentally tractable, cover 30% of Earth?s ice-free surface, and occur across greatly varying climatic conditions, providing an unparalleled context for clarifying mechanisms driving microbial composition and function within and among multiple spatial, temporal, and taxonomic scales.

Most species host multiple pathogens, yet the processes determining pathogen diversity within a single host and the effects of coinfection on pathogen dynamics in natural systems are rarely examined. Coinfection can affect pathogen transmission and virulence. In turn, coinfection can be regulated within hosts by interactions such as cross-protective immunity or at broader spatial scales via vector distributions. In this research, we will use field and modeling experiments to study the distribution and abundance of five coexisting aphid-vectored, viral pathogens (barley and cereal yellow dwarf luteoviruses and poleroviruses) in a native perennial and invasive annual grasses in both space and time.

Rodent pests cause major economic losses and threaten food security and biodiversity worldwide. The problem is particularly acute on islands where most vertebrate extinctions occur. We propose to test an innovative approach based on genetic engineering. This would also support graduate training in the NCSU Genetic Engineering and Society Center.

The relationship between species diversity and biological productivity is one of the most important, well-studied, and yet controversial topics in of all of ecology. Aside from helping to reveal the inherent controls over community structure, understanding this relationship is of fundamental importance for predicting how ecosystems will respond to human impacts that are altering the biodiversity and productivity of ecological communities around the globe. Researchers have tried to understand this relationship from two fundamentally different angles, either asking how diversity is influenced by variation in ecological productivity, or more recently, asking how diversity itself controls the production of biomass. These contrasting perspectives have engendered a lively debate about the nature and direction of causal pathways that connect species diversity and biological production. Recently, we published a mathematical model that presents one way of fusing the productivity-drives-diversity and diversity-drives-productivity perspectives together into a single conceptual framework. This model postulates a set of pathways by which three distinct variables -- nutrient supply, species richness, and production of community biomass -- influence one another. A novel and key prediction of this model is that species diversity and nutrient supply interact to regulate the production of community biomass. This prediction is important because it suggests we need to understand the joint changes in species diversity and nutrient enrichment imposed on ecosystems if we are to forecast how basic ecological processes like productivity will respond to human impact. Although our model represents an important first step, it only scratches the surface of a complex set of relationships. Here we propose to develop, test, and refine a conceptual framework that clarifies the pathways by which species diversity and productivity influence each other. Our work will consist of laboratory, field and modeling studies, all focused on freshwater ecosystems that are undergoing some of the most rapid changes in productivity and species loss. This ensemble of studies is a promising approach because it allows us to test our hypotheses across the spatial and temporal scales that span the inherent trade-off between experimental simplification and natural complexity. Because our current mathematical model is limited to overly simplistic systems of primary producers, we also propose to take on the difficult, but important task of extending our predictions to more realistic multi-trophic systems.

This multi-investigator, multi-institution grant aims to pioneer the application of bioinformatics to problems in applied insect ecology. Our plan is to use Lygus hesperus feeding on California cotton as a model system to achieve two broad aims. First, we will attempt to demonstrate the utility of a bioinformatics approach for solving real problems in pest management, including key questions of plant compensation, secondary pest outbreaks, and the landscape ecology of pest populations. Second, we will attempt to develop and troubleshoot new data extraction and analysis techniques that will be broadly useful for others working with consultant-generated data sets.

Many of the traditional approaches in ecological theory are based on the paradigm that natural ecosystems have stable equilibria. Thus, traditional approaches to environmental problems draw on the large body of mathematical and statistical techniques for analyzing asymptotic, near-equilibrium behavior in both deterministic and stochastic models. These approaches, however, have clear limitations. Just as ecologists emphasize that long-term experiments yield different results from short-term experiments, theory needs to take the role of time scales into account, and look beyond an emphasis on dynamics around stable equilibria. Not only will this aid in our basic understanding of real ecological systems, it will also refine old and open up new approaches to environmental management. To cope with the often sudden and large impacts that humans can have on ecosystems, there is growing interest in adaptive management, in which management strategies change as managers learn and as an ecosystem is beset by new shocks. Implementing adaptive management, however, requires a far more sophisticated understanding of how short-term, observed dynamics of a system generate long-term patterns that ultimately dictate the sustainability of managed ecosystems. Our goal here is to develop the mathematical and statistical tools to understand the role of transient dynamics in ecological systems.