Ecosystem services have received increasing attention in life sciences, but only a limited amount of quantitative data are available concerning the ability of weeds to provide these services. Following an expert focus group on this topic, a systematic search for articles displaying evidence of weeds providing regulating ecosystem services was performed, resulting in 129 articles. The most common service found was pest control and the prevailing mechanism was that weeds provide a suitable habitat for natural enemies. Other articles showed that weeds improved soil nutrient content, soil physical properties and crop pollinator abundance. Weeds were found to provide some important ecosystem services for agriculture, but only a small number of studies presented data on crop yield. Experimental approaches are proposed that can: (i) disentangle the benefits obtained from ecosystem services provisioning from the costs due to weed competition and (ii) quantify the contribution of diverse weed communities in reducing crop competition and in providing ecosystem services. Existing vegetation databases can be used to select weed species with functional traits facilitating ecosystem service provisioning while having a lower competitive capacity. However, for services such as pest control, there are hardly any specific plant traits that have been identified and more fundamental research is needed.

Insect herbivores often respond to leaf texture in making oviposition or feeding choices. This study examined the importance of leaf pubescence for an assemblage of generalist caterpillars (Lepidoptera: Limacodidae) feeding on oaks (Quercus spp.) and a variety of other tree species in eastern North America. Based on 10 yr of larval sampling on canopy and understory black and white oak (Quercus velutina and Q. alba, respectively) in the Ozark Mountains of Missouri, larval density of slug caterpillars (14 species as a group and 3 individual species) was higher on glabrous canopy leaves of Q. velutina than on highly pubescent understory leaves of that species. In contrast, there was no effect of stratum on overall density for Q. alba, which has glabrous leaves in both microenvironments. Individually, stratum effects for Q. alba were significant for five species, four of which were more abundant in the understory. Additional censusing of larvae on 20 tree species varying in leaf pubescence found that, as a group, slug caterpillar density was negatively correlated with leaf hair density. Finally, feeding trials confirmed that slug caterpillars prefer canopy over understory leaves of Q. velutina and vice versa for Q. alba leaves. When hairs were experimentally removed from one side of Q. velutina understory leaves, caterpillars preferred the side from which hairs were removed over the intact side. Together, these results indicate that leaf pubescence influences patterns of host plant use by these generalist herbivores, and in so doing, contributes to the structuring of local herbivore communities.

Studies of the effects of logging on Lepidoptera rarely address landscape-level effects or effects on larval, leaf-feeding stages. We examined the impacts of uneven-aged and even-aged logging on the abundance, richness, and community structure of leaf-chewing insects of white (Quercus alba L.) and black (Q. velutina L.) oak trees remaining in unharvested areas by sampling 3 years before and 7 years after harvest. After harvest, white oaks in uneven-aged sites had 32% fewer species of leaf-chewing insects than control sites. This reduction in species richness may have resulted from changes in microclimate (reducing plant quality and/or changing leaf phenology) that affected a much larger total area of each site than did even-aged cuts. For black oak after harvest, species richness in uneven- and even-aged sites increased relative to levels before harvest. Harvesting did not alter total insect density or community structure in the unlogged habitat for either oak species with one exception: insect density on black oak increased in the oldest forest block. Community structure of herbivores of black and white oaks in clearcut gaps differed from that of oaks in intact areas of even-aged sites. Furthermore, both richness and total insect density of black oaks were reduced in clearcut gaps. We suggest that low-level harvests alter herbivore species richness at the landscape level. Treatment effects were subtle because we sampled untreated areas of logged landscapes, only one harvest had occurred, and large temporal and spatial variation in abundance and richness existed. Although the effects of logging were greater in uneven-aged sites, the effects of even-aged management are likely to increase as harvesting continues.

In plants that produce seeds with contrasting genetic background (selfed versus outcrossed), the question arises whether the ecological function of the two types of progeny differ. This paper addresses this issue for the ant-dispersed Calathea micans by introducing a novel application of the Neubert–Caswell model for analysis of wave speed for structured populations. Because dispersal as well as vital rates are structured, the model allows for distinct dispersal kernels for different types of progeny and thus permits comparisons of the sensitivity to changes in demographic and dispersal parameters of in situ population growth rate versus population spread across space. The study site was a lowland, evergreen tropical rain forest at La Selva Biological station, Costa Rica, where the species is commonly found throughout the forest. In C. micans, seeds produced by open flowers (potentially outcrossed) or by closed flowers (selfed) bear oily arils and are dispersed by ants. Five life-history stages were used to characterize the population: seedlings originating from seeds produced by open flowers, seedlings originating from seeds produced by closed flowers, juvenile vegetative plants, reproductive plants without new shoots and reproductive plants with new shoots. Demography varied seasonally. Transitions were estimated from marking and following the fate of plants (N=400) in a natural population over a dry and a wet season. The population dynamics was described by a 10×10 matrix, with five life-history stages and two habitat states. The habitat states cycle repeatedly, dry–wet–dry–wet. To estimate dispersal kernels for each seed type, individual seeds (N=225 and 306 seeds produced by open and closed flowers, respectively) were color-coded and placed in depots, allowing the ants to redistribute them. Five months later, seedlings with an attached seed coat bearing the intact color-coding, were surveyed around the depots. Radial distances and angles were recorded for each seedling (N=67 and 81 seedlings arising from open and closed flowers, respectively). The results of the model give an asymptotic growth rate of 1.06 per season and an asymptotic rate of spread of 8.36cm per season. There is a high correlation (r=0.99) between elasticity of growth rate and elasticity of rate of spread of the population. Both rates are most sensitive to changes in stasis of juveniles during the dry season. However, most interesting is the analysis that revealed that population spread is more sensitive than in situ population growth to demographic rates of seedlings arising from open flowers. The analysis suggests a new way of thinking about ecological functions of multiple modes of reproduction.

The repeated use of copper (Cu) fungicides to control vine downy mildew, caused by Plasmopara viticola, has been responsible for the heavy increase of Cu concentration in the upper layers of vineyard soils. To determine the effects of elevated soil Cu on plant development, we created an artificial soil gradient with Cu enrichments ranging from 0 to 400 mg kg-1. On this gradient, and for five ruderal plant species commonly found in vineyards in southern France (Poa annua L., Dactylis glomerata L., Senecio vulgaris L., Hypochoeris radicata L., and Andryala integriflolia L.), we quantified survival, growth, and reproduction throughout one flowering season. High concentrations of Cu in the soil resulted in low survival, low total plant biomass, delay in flowering and fruiting, and low seed set. However, the effects differed among species. Furthermore, high soil Cu concentrations had contrasting effects on patterns of resource allocation depending on the plant species.

We report the effects of two timber harvest methods, even-aged and uneven-aged harvest, versus no harvest on species accumulation curves for leaf-chewing herbivores of Quercus alba and Q. velutina in the Missouri Ozarks. The study was part of a larger project, the Missouri Ozark Forest Ecosystem Project (MOFEP). Herbivores were sampled four times during the year (early May, June, July, and late August) for each of 4 years after cutting. Species accumulation curves were generated by plotting the total number of species recorded per leaf area sampled in all stands (N=6 stands/site) within a site in May 1997, the first census of the first year following cutting, and then adding the number of new species encountered in each subsequent census through the end of 2000. Treatment effects first became apparent in 1998. Uneven-aged management tended to reduce the rate of species accumulation across years for Q. alba compared to no harvest (control) and even-aged management, although marginally so. In contrast, even-aged management significantly increased the rate of species accumulation on Q. velutina compared to no harvest, with uneven-aged curves lower than no harvest. The May and June censuses contributed most to the treatment effect for Q. alba and Q. velutina, respectively. We interpret these results to mean that the treatments either increased or decreased population size relative to controls, thus resulting in an increased or decreased probability, respectively, of sampling a species. Species accumulation curves were lower for younger forests, suggesting that continued cutting, regardless of harvesting method, will reduce species richness of this herbivore fauna.

Many federal and state management agencies have shifted from commodity-based management systems to multiple resource-based management systems that emphasize sustainable ecosystem management. Long-term sustainability of ecosystem functions and processes is at the core of ecosystem management, but a blueprint for assessing sustainability under different management strategies does not exist. Using the Missouri Ozark Forest Ecosystem Project (MOFEP) as a case study, we present one approach to evaluating the landscape-scale, short-term (one and two years posttreatment) consequences of even-aged and uneven-aged forest management treatments on community-level biological diversity. We chose changes in density of ecological species groups, representing groups of species with similar resource requirements, as our response variable. Changes in density are detectable before species completely disappear from an area, and these changes may be an early indicator of significant alterations to community structure and ecosystem function. Meta-analysis was used to statistically combine changes in densities across multiple species groups and assess the overall impacts of management treatments on the animal community. We also separately examined changes in density for each ecological species group. Our findings demonstrated that, in the short-term, even-aged and uneven-aged forest management treatments caused changes in animal community density in Missouri Ozark forests. Even-aged management sites showed greater changes than uneven-aged management sites after harvesting, and changes in species' densities were larger two years posttreatment (1998) than one year posttreatment (1997). Evaluation of treatment effects on individual ecological groups revealed that toads, forest interior birds, and edge/early successional birds were significantly affected by management treatments. We did not expect most species groups to exhibit treatment effects because relatively little forest biomass was removed per experimental site (only 10%), forest cover at the regional landscape level did not change and was generally high during the study, and the time scale was relatively short. The challenges facing ecosystem management evaluation parallel the challenges of ecological science in general: identifying appropriate variables, spatial and temporal scales, and experimental/management treatments. The integrative approach demonstrated in this paper is a first step toward the analysis of the effects of management treatments on multiple organisms within an ecosystem.

To determine whether the probability of parasitoid attack differed across years, seasons, sites, host plant species, and herbivore feeding modes, leaf-chewing insects (Lepidoptera) that feed on Quercus alba L. and Q. velutina Lam. were collected four times in the Missouri Ozarks (May, June. July, and August-September) in each of the years 1993-1995 and reared in the laboratory. Parasitism rates at a given census were relatively constant from year to year and decreased as the season progressed from May to September: 30% of the herbivores collected in May were parasitized, whereas <15% were attacked in the summer and fall. Similarly, parasitism rates were predictable based on host plant (higher on Q, velutina than on Q, alba) and feeding guild (higher for leaf miners than for external feeders, leaf tiers and leaf webbers, in the summer and fall). Leaf rollers encountered in May had the highest probability of being parasitized. There was also significant variation in parasitism rates among sites. Community composition during the spring was very similar among spring censuses across years and different from all other censuses. In contrast, in the summer and fall, the species assemblage of the parasitoid community remained relatively constant within a year despite the fact that the herbivore community composition changed across censuses.

Host/parasitoid records are provided for 32 species of braconid wasps attacking a large complex of caterpillars that feed upon five species of oaks (Quercus spp.) in the Missouri Ozarks. Forty of the 62 host records are new for the given species of braconids.

Summary1. From July 1994 to September 1995, at six censuses, the herbivore community associated with understorey (< 2.5 m height) and canopy (15–20 m) leaves of Quercus alba and Q. velutina was sampled in south‐eastern Missouri, U.S.A.2. Across all censuses, herbivore densities were not significantly different between canopy and understorey for Q. alba and Q. velutina, except in August 1994 when herbivore densities were 60% higher in the canopy on Q. alba. Little significant spatial variation in herbivore densities or community composition was found during the study years.3. The herbivore community was diverse, consisting of 138 species of leaf‐chewing insects. Species richness was significantly greater (by 5–20%) in the understorey than in the canopy for both tree species, and the relative abundance of the main families, different feeding guilds, and most common species differed significantly between understorey and canopy.4. To determine the extent to which leaf quality explained the observed patterns, percentage nitrogen and protein binding capacity were measured in canopy and understorey leaves of Q. alba and Q. velutina. Per cent nitrogen was higher in canopy leaves for Q. velutina while protein binding capacity was higher in canopy leaves for Q. alba.5. These results suggest that the herbivore community associated with these two species of Quercus comprises species that appear to respond individually to environmental and biological conditions encountered in the understorey and the canopy.

We describe spatial and temporal variation in the insect herbivore communities associated with the MOFEP, prior to application of contrasting cutting regimes. No pre-treatment differences were found in total insect density on either black (Quercus velutina) or white oak (Q. alba) during 1993-1995. There was great seasonal variation in insect abundance on both host species as well as high variation across years for white oak, White oak on north- and east-facing slopes tended to have more insects than white oak on south- and west-facing slopes. Sites under the auspices of the Missouri Department of Conservation for a longer time had higher insect densities than more recently acquired sites.

The population level correlates of contrasting dispersal syndromes in closely related species are largely unknown. A family of tropical understory herbs, the Marantaceae, provides the opportunity for comparative studies as it contains many species with contrasting dispersal syndromes. As part of a study of the comparative population biology of ant- and bird-dispersed species, we test the hypothesis that spatial scale and dispersion pattern are related to dispersal type, proposing that bird-dispersed species will have a larger spatial scale than ant-dispersed species, and that, among bird-dispersed species, the scale will vary among three distinct dispersal types. We also propose that ant-dispersed species will show a more clumped dispersion pattern than the bird-dispersed species. Two types of spatial scale are examined: the amount of space occupied by individuals (maximum and actual) and the spacing among individuals within populations. The maximum size of plants showed a trend in the predicted direction. However, this trend was only of marginal statistical significance. The actual distribution of plant sizes, as measured by total leaf area and classified logarithmically, differed significantly among the dispersal types in the predicted direction. Spacing among individuals as measured by nearest neighbor distances also varied significantly by dispersal type, but not entirely in the predicted direction. Dispersion pattern analysis indicated that most study species in most populations had significantly clumped spatial patterning. The aggregation index varied 20-fold among study plots, but did not vary significantly by dispersal type. We conclude that while spatial scale was related generally to type of dispersal, dispersion pattern was not.