Search Results

Evolutionary dynamics may help us understand the structure and dynamics of food webs. The theoretical understanding of empirical food web patterns faces a dilemma, as it is difficult to account simultaneously for dynamical components (demography, evolution) and for the complexity of these systems (species number, connectance). Current knowledge of food web structures is dominated by many-species models that do not incorporate any dynamical aspects, and by models that detail demographic or evolutionary dynamics of species but consider communities composed of few species. Community evolution models incorporate both the dynamical components of food webs and the complexity that is necessary to understand empirical food web data.

Popularized by scientists in the 1970s, adaptive management is an integrative, multi-disciplinary approach to managing landscapes and natural resources. Despite its broad appeal many critics complain that adaptive management rarely works in practice as prescribed in theory. This chapter traces the history and evolution of the concept and assess its implementation challenges. One reason adaptive management has not always delivered on its promise to make natural resource management more “rational” is that in the real world of policymaking scientists and natural resource managers must contend with advocates that have conflicting values and goals. Scientists and managers also operate in the context of institutions that create particular constraints and opportunities, and are generally inflexible and resistant to change. In recognition of these sociopolitical realities, the focus of much adaptive management practice and scholarship has shifted to governance, particularly collaboration with stakeholders, transformation of the institutions responsible for management, and the process of social learning.

This paper considers the task of building efficient regression models for sparse multivariate analysis of high dimensional data sets, in particular it focuses on cases where the numbers q of responses Y = ( y _k ,1 ≤ k ≤ q) and p of predictors X = ( x _j , 1 ≤ j ≤ p) to analyse jointly are both large with respect to the sample size n, a challenging bi‐directional task. The analysis of such data sets arise commonly in genetical genomics, with X linked to the DNA characteristics and Y corresponding to measurements of fundamental biological processes such as transcription, protein or metabolite production. Building on the Bayesian variable selection set‐up for the linear model and associated efficient MCMC algorithms developed for single responses, we discuss the generic framework of hierarchical related sparse regressions, where parallel regressions of y _k on the set of covariates X are linked in a hierarchical fashion, in particular through the prior model of the variable selection indicators γ _kj , which indicate among the covariates x _j those which are associated to the response y _k in each multivariate regression. Structures for the joint model of the γ _kj , which correspond to different compromises between the aims of controlling sparsity and that of enhancing the detection of predictors that are associated with many responses (“hot spots”), will be discussed and a new multiplicative model for the probability structure of the γ _kj will be presented. To perform inference for these models in high dimensional set‐ups, novel adaptive MCMC algorithms are needed. As sparsity is paramount and most of the associations expected to be zero, new algorithms that progressively focus on part of the space where the most interesting associations occur are of great interest. We shall discuss their formulation and theoretical properties, and demonstrate their use on simulated and real data from genomics.

Finding the integrated likelihood of a model given the data requires the integration of a nonnegative function over the parameter space. Classical Monte Carlo methods for numerical integration require a bound or estimate of the variance in order to determine the quality of the output. The method called the product estimator does not require knowledge of the variance in order to produce a result of guaranteed quality, but requires a cooling schedule that must have certain strict properties. Finding a cooling schedule can be difficult, and finding an optimal cooling schedule is usually computationally out of reach. TPA is a method that solves this difficulty, creating an optimal cooling schedule automatically as it is run. This method has its own set of requirements; here it is shown how to meet these requirements for problems arising in Bayesian inference. This gives guaranteed accuracy for integrated likelihoods and posterior means of nonnegative parameters.

This chapter provides details about order effects in neural networks, a commonly used modeling approach. It examines two networks in detail, the Adaptive Resonance Theory (ART) architecture and Jeff Elman's recurrent networks. The ART model shows that about a 25% difference in recognition rates can arise from using different orders. Elman's recurrent network shows that, with the wrong order, a task might not even be learnable. The chapter also discusses why these effects arise, which is important for understanding the impact and claims of computational models.

Sewall Wright originally conceived of his Adaptive Landscape as a visual device to capture the consequences of non-linear (epistatic) interactions between genes. A useful way to visualise a multivariate non-linear function is through a ‘landscape’, an important factor to consider when applying Adaptive Landscape models to questions about the evolution of development. This chapter examines how a phenotype landscape (also known as phenotypic landscape or developmental landscape) can explicitly map genetic and developmental traits to the phenotypic traits upon which selection acts. After outlining the basic properties of phenotype landscapes, it considers how they are used in concert with an Adaptive Landscape to study the evolution of development. It then describes the formal theory for evolution on phenotype landscapes and how it generalises the quantitative genetic approaches that are often applied to Adaptive Landscapes. The chapter concludes by illustrating how phenotype landscape theory can be used to study the evolution of genetic covariance, heritability, and novelty.

Sewall Wright’s classic Adaptive Landscape has been a highly successful metaphor and scientific concept in evolutionary biology. It has influenced many different research subdisciplines in evolutionary biology and inspired generations of researchers, even though it has also sparked deep scientific and philosophical controversies. Among such subdisciplines are population genetics, evolutionary ecology, quantitative genetics, experimental evolution, conservation biology, speciation and macroevolutionary dynamics, mimicry, saltational evolution, behavioural ecology, molecular biology, protein networks, and theoretical studies on development.

Adaptive procedures are developed to reduce the burden of data collection in psychophysics by creating more efficient experimental test designs and methods of estimating either statistics or parameters. In some cases, these adaptive procedures may reduce the amount of testing by as much as 80% to 90%. This chapter begins with a description of classical staircase procedures for estimating the threshold and/or slope of the psychometric function, followed by a description of modern Bayesian adaptive methods for optimizing psychophysical tests. We introduce applications of Bayesian adaptive procedures for the estimation of psychophysically measured functions and surfaces. The bias, precision and efficiency of estimates is considered. Each method is accompanied by an illustrative example and sample results and a discussion of the practical requirements of the procedure.

Three general steps move civic ecology practices from small local innovations to broader policy innovations: giving a label to the phenomenon (in our case “civic ecology”); becoming more effective as local providers of ecosystem services and contributors to community well-being through partnerships with scientists; and government and larger NGOs formulating policies that allow civic ecology practices to spread. Civic ecology practices are small social or “social-ecological innovations,” whereas larger NGOs and government agencies are policy entrepreneurs who shape the policy environment. Policy entrepreneurs can also bridge between multiple civic ecology practices and larger management initiatives to form regional adaptive and collaborative resource management systems.

This paper explores a variety of approaches to the assemblage of knowledges from differing traditions without subsuming them into the overarching suppositions and prerequisite standards of any given tradition. The paper is in two parts; the first draws on recent thinking about complex adaptive systems, and about narrative and movement in the construction of knowledge, place and space. The second part describes the work of TASSIT (Trails and Storied Spaces in Time), a project working to design digital “third spaces” in which the narratological re-conception of knowledges allows for the creation of an interactive commons useful for teaching and research at universities as well as for exhibition and collections data-basing at museums. These spaces attempt to re-contextualize accounts that may have first appeared as scholarly and scientific papers, elders’ stories, social and cultural interactions, maps, khipu, hieroglyphic inscription, dance, structures, textiles, rock art, and numerous other forms in which knowledge may be fixed, implied or performed.

The adaptive cycle describes changes in ecosystems and social-ecological systems over time. It expands on older notions of ecosystems as moving from a pioneer or exploitation phase to a steady state or so-called climax or conservation phase. The adaptive cycle includes two additional phases—the release or chaotic phase after a catastrophic disturbance pushes a social-ecological system beyond a threshold, and the subsequent rebuilding or reorganization phase. Social-ecological systems resilience refers to the ability of systems to adapt or respond to small changes in the conservation (steady state) phase, and to transform once a system has crossed a critical threshold. Civic ecology practices can play a role in both adaptation and transformation. Panarchy refers to multiple adaptive cycles embedded in and impacting each other. For example, at the small-scale, a successful civic ecology practice may spur a municipal government to make policies supporting civic ecology practices in a process called “revolt.” And in a process called remembrance, a government that places restrictions on land use may limit the ability of civic ecology practices to be successful.

Chapter 4 provides further examples of RECtification, this time with the aim of showing how REC can fruitfully ally with and strengthen two prominent nonrepresentational E-approaches to cognition—Autopoietic-Adaptive Enactivism and Ecological Dynamics. These examples of RECtification reveal REC’s capacity to marshal and combine powerful resources for explaining basic minds in naturalistic terms. The chapter concludes by discussing the need to show how to basic, contentless minds can meet contentful minds in REC terms – namely, to explicate how REConnecting is possible. Doing so is necessary in light of REC’s commitment to two ideas: that some cognition is content-involving and that organisms become capable of content-involving cognition by mastering special sociocultural practices.

During the 1980's and 1990's Fodor, McLaughlin, and Pylyshyn claimed that thought is in various respects systematic. Further, they argued that so-called “Classical” syntactically and semantically combinatorial representations provide a better explanation of the systematicity of thought than do non-combinatorial representations or non-Classical combinatorial representations. During the 1990’s, part of what made the systematicity arguments problematic was the subtlety of the idea of providing a better explanation. In what sense is the Classical account better than its rivals? During what we might call the Post-Connectionist era of roughly the last ten years, however, theoretical shifts have made it even more difficult to bring considerations of the systematicity of thought to bear on the nature of cognition. Post-Connectionist cognitive science has come to focus less on cognition. This chapter reviews these changes in the cognitive science landscape regarding systematicity.

This chapter describes metareasoning in an image interpretation architecture called GRAVA (Grounded Reflective Adaptive Vision Architecture), where the goal is to produce good image interpretations under a wide range of environmental conditions. GRAVA employs metareasoning because it provides the mechanisms necessary to support two of the core problems of self-adaptive software—a mechanism for reasoning about the state of the computational system and a mechanism for making changes to it. These are referred to as introspective monitoring and metalevel control, respectively.

The authors conclude that four well-established domain-general mechanisms of human memory—memorial benefits caused by (1) arousal, (2) item congruity, (3) richness of encoding, and (4) the combination of single-item and relational processing—have proven successful in accounting for many results on moderators and mediators of the SPE. In contrast, hypotheses attributing the SPE to the operation of domain-specific, evolutionarily acquired “cognitive modules” tailored to solve specific adaptive problems our ancestors faced in Pleistocene environments (i.e., survival modules or planning modules) were less successful.