The structured choruses produced by rhythmically signalling males in many species of acoustic animals have long-captured the imagination of evolutionary biologists. Though various hypotheses have been forwarded to explain the adaptive signi¢cance of such chorusing, none have withstood empirical scrutiny. We suggest instead that alternating and synchronous choruses represent collective epiphenomena resulting from individual males competing to jam each other's signals. These competitions originate in psychoacoustic precedence e¡ects wherein females only orient toward the ¢rst call of a sequence, thus selectively favouring males who produce leading calls. Given this perceptual bias, our modelling con¢rms that a resetting of signal rhythm by neighbours' signals, which generates either alternation or synchrony, is evolutionarily stable provided that resetting includes a relativity adjustment for the velocity of signal transmission and selective attention toward only a subset of signalling neighbours. Signalling strategies in chorusing insects and anurans are consistent with these predicted features.

Summary1 Small, isolated populations at species' borders have been postulated to be less likely to have specialist pathogens and predators. Field and herbarium surveys were thus used to determine if two pathogens (a smut and a rust) and a pre-dispersal seed predator were less common at the western range limit of the forest sedge Carex blanda in Kansas, USA . 2 Host plant size, reproduction and density did not decline at the western border of the range. In fact, plants at two western sites had unusually large size and seed production. 3 Host populations at the edge of the range were more likely to be disease-free or lack the pre-dispersal seed predator. Where the smut, seed predator and rust were found, the proportion of infected or infested plants was not related to longitude, latitude or percentage forest cover. 4 More of the peripheral populations lacked the smut than the rust, as expected given the more localized nature of smut spore dispersal and the limited period when smut infection can occur. 5 In the adjacent, more highly forested state of Missouri, there were no geographical patterns in the incidence of the smut or seed predator in herbarium data. 6 The smut and rust increased in frequency over the 129-year span of herbarium collections. 7 Although field and herbarium distributional data were not identical (for example, smut infection was found much farther west in the field than in the herbarium data), the qualitative agreement between the two data sets suggests herbarium data can be used more broadly for studies of natural enemy distributions. 8 Limited dispersal by pathogens and seed predators is probably the reason why small, isolated western populations were less likely to have natural enemies. Peripheral host populations may thus have different ecological and evolutionary trajectories compared with more central populations. This conclusion, as well as the considerable variation among peripheral populations, is relevant to geographical studies of co-evolution and to research on climatic effects on plants inhabiting ecotonal regions.

No abstract

No abstract

Summary1. The success of locally applied treatments for exotic weed control depends on the effectiveness of the method used and workers' abilities to find plants within infested sites. Detectability of exotic plants, however, depends on aspects of a plant's spatial distribution such as the number and size of patches. We lack an explicit examination of how incomplete detection affects the spread of exotic weeds under a range of realistic field conditions. 2. We developed a model of spatial spread of the exotic plant, Lespedeza cuneata, based on 3 years of data in a Kansas, USA, grassland. We then expanded and generalized the model to study how treatment intensity, spatial distribution of stems and detectability influenced control efforts. 3. When left untreated, occupancy and abundance were higher with randomly distributed infestations than with patchy distributions. Control treatments slowed spread, but only the most intense treatments reduced occupancy and abundance. 4. The greatest spread occurred when low detectability was accompanied with low treatment intensity because many small patches were not treated; this phenomenon was particularly common with random spatial distributions that initially lacked large, easily detectable patches. In contrast, high treatment intensity led to similarly slow spread for both spatial patterns and a range of detectability functions. 5. Synthesis and applications. We developed a model to simultaneously manipulate the spatial distribution of the invading plant, the intensity of control methods used to manage the population and the detectability of occupied areas at a site; different combinations of these three factors led to very different rates of exotic spread. Managers will reasonably try to implement intensive weed control. However, poor timing of treatments, for example, could lead to variation in treatment effectiveness. To maximize success, managers should explicitly consider detectability, especially where small patches are scattered throughout a site. To quantify the probability of detection, managers could perform multiple observer surveys; such information could help to determine the effort needed for effective control. Creating weed maps may increase the detection of existing patches from year to year, but workers would still need to search the entire site for new or previously undetected patches.