JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. Society for the Study of Evolution is collaborating with JSTOR to digitize, preserve and extend access to Evolution.Abstract.-Field studies in South Africa showed that floral spur length in the Disa draconis complex (Orchidaceae) varies enormously between populations in the southern mountains (means = 32-38 mm), lowland sandplain (mean -48 mm), and northern mountains (means = 57-72 mm). We tested the hypothesis that divergence in spur length has resulted from selection exerted through pollinator proboscis length. Short-spurred plants in several southern mountain populations, as well as long-spurred plants in one northern mountain population, were pollinated by a horsefly, Philoliche rostrata (Tabanidae), with a proboscis length that varied from 22 to 35 mm among sites. Longspurred plants on the sandplain were pollinated by the tanglewing fly, Moegistorynchus longirostris (Nemestrinidae), which has a very long proboscis (mean = 57 mm). Selection apparently favors long spurs in sandplain plants, as artificial shortening of spurs resulted in a significant decline in pollen receipt and fruit set, although pollinaria removal was not significantly affected. Fruit set in the study populations was limited by pollen availability, which further suggests that selection on spur length occurs mainly through the female component of reproductive success.

Mimicry is a classic example of adaptation through natural selection. The traditional focus of mimicry research has been on defense in animals (protective mimicry), but there is now also a highly developed and rapidly growing body of research on floral mimicry in plants. Being literally rooted to one spot, plants generally have to use food bribes to cajole animals into acting as couriers for their pollen. Plants that lack these food rewards often deploy elaborate color and scent signals in order to mimic food sources, oviposition sites, or mating partners of particular animals, and thereby exploit these animals for the purposes of pollination. This book addresses the question of whether the evolutionary and ecological principles that were developed for protective mimicry in animals also apply to floral mimicry in plants. Visual, olfactory, and tactile signals can all be important in floral mimicry systems. The traditional focus has been on visual cues, but there is increasing evidence that some forms of mimicry, notably sexual and oviposition-site mimicry, are largely based on chemical cues. The molecular basis for these signals, their role in cognitive misclassification of flowers by pollinators, and the implications of these signals for plant speciation are among the topics covered in the book. The chapters of this book are designed to highlight particular systems of floral mimicry and to integrate them into the broader theory of mimicry.

We present a new algorithm to automatically schedule Halide programs for high-performance image processing and deep learning. We significantly improve upon the performance of previous methods, which considered a limited subset of schedules. We define a parameterization of possible schedules much larger than prior methods and use a variant of beam search to search over it. The search optimizes runtime predicted by a cost model based on a combination of new derived features and machine learning. We train the cost model by generating and featurizing hundreds of thousands of random programs and schedules. We show that this approach operates effectively with or without autotuning. It produces schedules which are on average almost twice as fast as the existing Halide autoscheduler without autotuning, or more than twice as fast with, and is the first automatic scheduling algorithm to significantly outperform human experts on average. CCS Concepts: • Computing methodologies → Image processing; • Software and its engineering → Domain specific languages.

Modern cameras typically use an array of millions of detector pixels to capture images. By contrast, single-pixel cameras use a sequence of mask patterns to filter the scene along with the corresponding measurements of the transmitted intensity which is recorded using a single-pixel detector. This review considers the development of single-pixel cameras from the seminal work of Duarte et al. up to the present state of the art. We cover the variety of hardware configurations, design of mask patterns and the associated reconstruction algorithms, many of which relate to the field of compressed sensing and, more recently, machine learning. Overall, single-pixel cameras lend themselves to imaging at non-visible wavelengths and with precise timing or depth resolution. We discuss the suitability of single-pixel cameras for different application areas, including infrared imaging and 3D situation awareness for autonomous vehicles.

In the nacre or aragonite layer of the mollusk shell there exist proteomes which regulate both the early stages of nucleation and nano-to-mesoscale assembly of nacre tablets from mineral nanoparticle precursors. Several approaches have been developed to understand protein-associated mechanisms of nacre formation, yet we still lack insight into how protein ensembles or proteomes manage nucleation and crystal growth. To provide additional insights we have created a proportionally-defined combinatorial model consisting of two nacre-associated proteins, C-RING AP7 (shell nacre, H. rufescens) and pseudo-EF hand PFMG1 (oyster pearl nacre, P. fucata) whose individual in vitro mineralization functionalities are well-documented and distinct from one another. Using SEM, flow cell STEM, AFM, Ca(II) potentiometric titrations and QCM-D quantitative analyses, we find that both nacre proteins are functionally active within the same mineralization environments, and at 1:1 mole ratios, synergistically create calcium carbonate mesoscale structures with ordered intracrystalline nanoporosities, extensively prolong nucleation times and introduce an additional nucleation event. Further, these two proteins jointly create nanoscale protein aggregates or phases that under mineralization conditions further assemble into protein-mineral PILP-like phases with enhanced ACC stabilization capabilities, and there is evidence for intermolecular interactions between AP7 and PFMG1 under these conditions. Thus, a combinatorial model system consisting of more than one defined biomineralization protein dramatically changes the outcome of the in vitro biomineralization process.