Effects of Short- and Long-Term Disturbance Resulting from Military Maneuvers on Vegetation and Soils in a Mixed Prairie Area

Loss of grassland species resulting from activities such as off-road vehicle use increases the need for models that predict effects of anthropogenic disturbance. The relationship of disturbance by military training to plant species richness and composition on two soils (Foard and Lawton) in a mixed prairie area was investigated. Track cover (cover of vehicle disturbance to the soil) and soil organic carbon were selected as measures of short- and long-term disturbance, respectively. Soil and vegetation data, collected in 1-m² quadrats, were analyzed at three spatial scales (60, 10, and 1 m²). Plant species richness peaked at intermediate levels of soil organic carbon at the 10-m² and 1-m² spatial scales on both the Lawton and Foard soils, and at intermediate levels of track cover at all three spatial scales on the Foard soil. Species composition differed across the disturbance gradient on the Foard soil but not on the Lawton soil. Disturbance increased total plant species richness on the Foard soil. The authors conclude that disturbance up to intermediate levels can be used to maintain biodiversity by enriching the plant species pool.     

Native Perennial Forb Variation Between Mountain Big Sagebrush and Wyoming Big Sagebrush Plant Communities

Big sagebrush (Artemisia tridentata Nutt.) occupies large portions of the western United States and provides valuable wildlife habitat. However, information is lacking quantifying differences in native perennial forb characteristics between mountain big sagebrush [A. tridentata spp. vaseyana (Rydb.) Beetle] and Wyoming big sagebrush [A. tridentata spp. wyomingensis (Beetle & A. Young) S.L. Welsh] plant communities. This information is critical to accurately evaluate the quality of habitat and forage that these communities can produce because many wildlife species consume large quantities of native perennial forbs and depend on them for hiding cover. To compare native perennial forb characteristics on sites dominated by these two subspecies of big sagebrush, we sampled 106 intact big sagebrush plant communities. Mountain big sagebrush plant communities produced almost 4.5-fold more native perennial forb biomass and had greater native perennial forb species richness and diversity compared to Wyoming big sagebrush plant communities (P < 0.001). Nonmetric multidimensional scaling (NMS) and the multiple-response permutation procedure (MRPP) demonstrated that native perennial forb composition varied between these plant communities (P < 0.001). Native perennial forb composition was more similar within plant communities grouped by big sagebrush subspecies than expected by chance (A = 0.112) and composition varied between community groups (P < 0.001). Indicator analysis did not identify any perennial forbs that were completely exclusive and faithful, but did identify several perennial forbs that were relatively good indicators of either mountain big sagebrush or Wyoming big sagebrush plant communities. Our results suggest that management plans and habitat guidelines should recognize differences in native perennial forb characteristics between mountain and Wyoming big sagebrush plant communities.     

Measuring Plant Diversity in the Tall Threetip Sagebrush Steppe: Influence of Previous Grazing Management Practices

In July 2000, a 490-ha wildfire burned a portion of a long-term grazing study that had been established in 1924 at the US Sheep Experiment Station north of Dubois, Idaho, USA. Earlier vegetation measurements in this tall threetip sagebrush (Artemisia tripartita spp. tripartita) bunchgrass plant community documented significant changes in vegetation due to grazing and the timing of grazing by sheep. A study was initiated in May 2001 using 12 multiscale modified Whittaker plots to determine the consequences of previous grazing practices on postfire vegetation composition. Because there was only one wildfire and it did not burn all of the original plots, the treatments are not replicated in time or space. We reduce the potential effects of psuedoreplication by confining our discussion to the sample area only. There were a total of 84 species in the sampled areas with 69 in the spring-grazed area and 70 each in the fall- and ungrazed areas. Vegetation within plots was equally rich and even with similar numbers of abundant species. The spring-grazed plots, however, had half as much plant cover as the fall- and ungrazed plots and the spring-grazed plots had the largest proportion of plant cover composed of introduced (27%) and annual (34%) plants. The fall-grazed plots had the highest proportion of native perennial grasses (43%) and the lowest proportion of native annual forbs (1%). The ungrazed plots had the lowest proportion of introduced plants (4%) and the highest proportion of native perennial forbs (66%). The vegetation of spring-grazed plots is in a degraded condition for the environment and further degradation may continue, with or without continued grazing or some other disturbance. If ecosystem condition was based solely on plant diversity and only a count of species numbers was used to determine plant diversity, this research would have falsely concluded that grazing and timing of grazing did not impact the condition of the ecosystem.     

Identifying Preservation and Restoration Priority Areas for Desert Fishes in an Increasingly Invaded World

A commonly overlooked aspect of conservation planning assessments is that wildlife managers are increasingly focused on habitats that contain non-native species. We examine this management challenge in the Gila River basin (150,730 km²), and present a new planning strategy for fish conservation. By applying a hierarchical prioritization algorithm to >850,000 fish records in 27,181 sub-watersheds we first identified high priority areas (PAs) termed “preservation PAs” with high native fish richness and low non-native richness; these represent traditional conservation targets. Second, we identified “restoration PAs” with high native fish richness that also contained high numbers of non-native species; these represent less traditional conservation targets. The top 10 % of preservation and restoration PAs contained common native species (e.g., Catostomus clarkii, desert sucker; Catostomus insignis, Sonora sucker) in addition to native species with limited distributions (i.e., Xyrauchen texanus, razorback sucker; Oncorhynchus gilae apache, Apache trout). The top preservation and restoration PAs overlapped by 42 %, indicating areas with high native fish richness range from minimally to highly invaded. Areas exclusively identified as restoration PAs also encompassed a greater percentage of native species ranges than would be expected by the random addition of an equivalent basin area. Restoration PAs identified an additional 19.0 and 26.6 % of the total ranges of two federally endangered species—Meda fulgida (spikedace) and Gila intermedia (Gila chub), respectively, compared to top preservation PAs alone—despite adding only 5.8 % of basin area. We contend that in addition to preservation PAs, restoration PAs are well suited for complementary management activities benefiting native fishes.     

An evaluation of woodland reclamation on strip-mined lands in east Texas

We compared the composition and structural characteristics of reclaimed and native woody plant communities near Fairfield, Texas, to evaluate the effectiveness of woodland reclamation 3–11 years since establishment. Species composition, foliage density, canopy cover, and woody plant densities were recorded in plots randomly placed along transects bisecting blocks of reclaimed and native woodlands. During summer, vertical foliage densities at heights ≤2 m were similar among native and reclaimed areas. Foliage density and canopy cover declined in reclaimed blocks during winter, but remained relatively constant in native woodlands, where evergreens and vines were more common. Canopy cover was absent in reclaimed woodlands <6 years old but increased with age in 6 to 11-year-old blocks. These data indicated that approximately 27 years will be needed before trees in reclaimed blocks will achieve the stature of canopy trees in native woodlands. Reclaimed woodlands contained different woody plant species and had lower woody stem densities compared to native woodlands. On average, stem densities in reclaimed blocks were six times lower than densities in native woodlands. Comparisons with planting records indicate that survival of most commonly planted woody species was low. Only green ash(Fraxinus pennsylvanica), Russian oliver(Elaeagnus commutata), smooth sumac(Rhus glabra), and redbud(Cercis canadensis) had estimated survival rates >50%. Reclamation procedures used at Big Brown Mine (BBM) during 1981–1988 have not produced woodland habitats with vegetative characteristics comparable to premined woodlands and may not be providing the cover needed to encourage use by certain wildlife species. Procedures for improving woodland reclamation are recommended.     

A Comprehensive Approach to Identifying Monitoring Priorities of Small Landbirds on Military Installations

Military installations provide important native habitat for songbirds, including many species that have experienced population declines in recent decades. As part of the Land Condition Trend Analysis (LCTA) program to monitor animal populations on military lands, we surveyed small (<250 g) breeding landbirds on 60 permanent plots on the Fort Riley Military Installation in northeastern Kansas from 1991 to 2002. During this period, species richness averaged 39.0 species (SE 0.9)/year and mean species richness per plot ranged from 3.6 species (SE=0.2)/plot (1999) to 7.5 species (SE=0.3)/plot (1992). Turnover (the appearance and disappearance of species on all plots from one year to the next) ranged from 5 species (2000–2001) to 16 species (1992–1993) and was driven primarily by turnover of woodland species. We developed an index of relative difference (C) to evaluate relative trends of local populations and found that 25 species declined, 15 species increased, and 7 did not change. Based on migration assemblages, more resident species (6 of 10) and more short-distant migrants (9 of 12) decreased than long-distance migrants (10 vs. 11). Our analysis of major vegetation communities on plots showed few changes in the quantity of habitats (grassland vs. woodlands) during the study. Our results indicate that Fort Riley provides important habitats for many landbirds, particularly those that require grasslands for breeding. Several species exhibited local declines when compared to the regional Breeding Bird Survey routes. We offer an approach that evaluates population changes of small landbirds and provides objective inputs for conservation directives. These can be adopted easily for use on military installations (that use LCTA), parks, and wildlife refuges that have data from annual breeding bird surveys.     

Effects of Alien Plants on Ecosystem Structure and Functioning and Implications for Restoration: Insights from Three Degraded Sites in South African Fynbos

We investigated the type and extent of degradation at three sites on the Agulhas Plain, South Africa: an old field dominated by the alien grass Pennisetum clandestinum Pers. (kikuyu), an abandoned Eucalyptus plantation, and a natural fynbos community invaded by nitrogen fixing—Australian Acacia species. These forms of degradation are representative of many areas in the region. By identifying the nature and degree of ecosystem degradation we aimed to determine appropriate strategies for restoration in this biodiversity hotspot. Vegetation surveys were conducted at degraded sites and carefully selected reference sites. Soil-stored propagule seed banks and macro- and micro-soil nutrients were determined. Species richness, diversity and native cover under Eucalyptus were extremely low compared to the reference site and alterations of the soil nutrients were the most severe. The cover of indigenous species under Acacia did not differ significantly from that in reference sites, but species richness was lower under Acacia and soils were considerably enriched. Native species richness was much lower in the kikuyu site, but soil nutrient status was similar to the reference site. Removal of the alien species alone may be sufficient to re-initiate ecosystem recovery at the kikuyu site, whereas active restoration is required to restore functioning ecosystems dominated by native species in the Acacia thicket and the Eucalyptus plantation. To restore native plant communities we suggest burning, mulching with sawdust and sowing of native species.     

Prairie grass establishment on calcareous reclaimed mine soil

Reclaimed Appalachian surface mined lands have difficulty in sustaining native deciduous forest communities. Establishing prairie communities could increase ecosystem function; however, a native model system does not exist. We evaluated establishment of 15 North American prairie grasses as monocultures on reclaimed mine soil in southeast Ohio in four randomized complete blocks planted May 2005 and 2006. Population density was assessed 30 d after planting (30 DAP) and in October of the planting year (YR1) and second year following planting (YR2) and expressed as percentage of viable seeds sown (PVSS). Canopy cover of nonnative species reestablishing in the plots was measured in 2007. Eastern gamagrass ( L.) population was >50 PVSS in all censuses. Western wheatgrass [ (Rydb.) A. L?ve] was initially 7 PVSS at 30 DAP, but increased to 154 PVSS by YR2 from rhizomes spreading into gaps. Big bluestem ( Vitman) was 7 PVSS at 30 DAP and 4 PVSS at YR2. Blue grama [ (Willd. ex Kunth) Lag. ex Griffiths] and sideoats grama [ (Michx.) Torr.] did not survive past YR1. Gaps left from poor stand establishment were primarily recolonized by nonnative Kentucky bluegrass ( L.) in the 2005 planting and birdsfoot trefoil ( L.) in the 2006 planting, but was least in eastern gamagrass and tall dropseed [ (P. Beauv.) Kunth]. This research demonstrates the potential for increasing diversity and species richness on mine soil habitats with regionally native grasses that could increase functional quality through ecological resilience.     

The Effects of Long-Term Grazing Exclosures on Range Plants in the Central Anatolian Region of Turkey

Over the last fifty years, almost half of the steppe rangeland in the Central Anatolian Region of Turkey (CAR) has been converted to cropland without an equivalent reduction in grazing animals. This shift has led to heavy grazing pressure on rangeland vegetation. A study was initiated in June 2003 using 6 multiscale Modified-Whittaker plots to determine differences in plant composition between areas that have not been grazed in 27 years with neighboring grazed plant communities. A total of 113 plant species were identified in the study area with the ungrazed plots containing 32 plants more than the grazed plots. The major species were Astragalus acicularis, Bromus tomentellus, Festuca valesiaca, Genista albida, Globularia orientalis, Poa bulbosa, and Thymus spyleus ssp rosulans. Grazing impacts on forbs were more pronounced than for grasses and shrubs. Based on Jaccard’s index, there was only a 37% similarity of plant species between the two treatments. Our study led to four generalizations about the current grazing regime and long-term exclosures in the steppe rangeland around the study area: (1) exclosures will increase species richness, (2) heavy grazing may have removed some plant species, (3) complete protection from grazing for a prolonged period of time after a long history of grazing disturbance may not lead to an increase in desirable plant species with a concomitant improvement in range condition, and (4) research needs to be conducted to determine how these rangelands can be improved.     

Persistent Impacts of Trace Metals from Mining on Floodplain Grass Communities Along Soda Butte Creek, Yellowstone National Park

Phleum pratense and Poa pratensis were significantly lower (P ≤ 0.001) on plots with more than 250 ppm copper. Above-ground biomass of Phleum pratense was also significantly lower on plots with copper levels above 250 ppm. Decreased mean grass density was found on plots with pH < 6.4, but the only statistically significant difference was for Juncus balticus, which had increased density on plots with pH < 6.4. In contrast to the clear impacts of trace metals and pH on vegetation, other site characteristics did not alter measured vegetation characteristics.     

Establishment and Growth of Experimental Grass Species Mixtures on Coal Mine Sites Reclaimed with Municipal Biosolids

The Surface Mining Control and Reclamation Act of 1977 requires that coal mine sites in the United States be reclaimed to establish vegetative cover that is diverse, native, and capable of plant succession. However, there is a question as to whether vegetation established on coal mine sites reclaimed with biosolids is diverse and capable of plant succession. The influx of nutrients with the addition of biosolids leads to long-term dominance by early-successional species, most notably grasses, and consequently, a low establishment of woody and volunteer species. Additionally, many grass species commonly planted in reclamation have aggressive growth habits that lead to their dominance in coal mine plant communities. The establishment and growth of selected grass mixes was evaluated to determine whether alternative grass mixes would be less competitive with woody and volunteer species as compared to commonly used grass mixes. Percent vegetative cover, species richness, and the survival of direct-seeded woody species were assessed for each treatment grass mixture. It was found that Poa compressa and a mixture of P. compressa, Panicum virgatum, and Trifolium repens provided adequate coverage while still allowing the highest species richness and survival of woody species. Use of these species mixtures in coal mine reclamation with biosolids in the eastern United States would likely lead to establishment of a more species-rich plant community with a greater woody species component while still providing erosion control and site protection.     

Erosional Consequence of Saltcedar Control

Removal of nonnative riparian trees is accelerating to conserve water and improve habitat for native species. Widespread control of dominant species, however, can lead to unintended erosion. Helicopter herbicide application in 2003 along a 12-km reach of the Rio Puerco, New Mexico, eliminated the target invasive species saltcedar (Tamarix spp.), which dominated the floodplain, as well as the native species sandbar willow (Salix exigua Nuttall), which occurred as a fringe along the channel. Herbicide application initiated a natural experiment testing the importance of riparian vegetation for bank stability along this data-rich river. A flood three years later eroded about 680,000 m³ of sediment, increasing mean channel width of the sprayed reach by 84%. Erosion upstream and downstream from the sprayed reach during this flood was inconsequential. Sand eroded from channel banks was transported an average of 5 km downstream and deposited on the floodplain and channel bed. Although vegetation was killed across the floodplain in the sprayed reach, erosion was almost entirely confined to the channel banks. The absence of dense, flexible woody stems on the banks reduced drag on the flow, leading to high shear stress at the toe of the banks, fluvial erosion, bank undercutting, and mass failure. The potential for increased erosion must be included in consideration of phreatophyte control projects.     

Revegetation of serpentine substrates: Response to phosphate application

Revegetation was studied on stockpiled serpentine substrate. The native vegetation surrounding the revegetation site is annual grassland. The seed mixture applied to both subsoil and topsoil plots was largely ineffective for revegetation. No growth occurred in the subsoil plots and most of the growth in the topsoil plots was from indigenous seed. Phosphate application (100 kg P ha^–1 as NaH₂PO₄ · H₂O) to the topsoil plots resulted in a significant increase in total above-ground productivity. Annual legumes (mostlyLotus subpinnatus Lag.) and, to a lesser degree,Plantago erecta Morris responded to the added phosphate with an increased above-ground productivity. Other annual forbs and annual grasses showed no significant response. The legumes also increased in abundance. Mycorrhizal root colonization forPlantago was not significantly affected by phosphate application, but was lower in this disturbed serpentine site compared to other undisturbed serpentine annual grassland sites nearby.     

The effects of military tank traffic on prairie: A management model

The effect of various frequencies and seasons of military tank traffic on native mixed-grass prairie was examined in a randomized and replicated field experiment. Vegetation (in 10×10 m plots) was subjected to tank traffic at the following rates: (a) one pass per day of training from May until August; (b) one pass per day in May and June; (c) one pass per day in July and August; (d) one pass every three weeks from May until August; (e) zero (control). Species composition and the amount of bare ground were found to vary significantly with traffic frequency. Plant species alien to North America invaded plots subjected to spring driving. Regression analysis showed spring driving to produce more bare ground than summer driving. The regression models suggested that much higher intensities of training could be conducted without damage if spring driving were avoided. Regression models were also used to estimate the frequency of traffic associated with a significant change in species composition, where species composition was expressed as the ratio ofBouteloua gracilis toStipa spartea, an indicator of disturbance-induced change in prairie vegetation. This relationship predicted the capacity of the vegetation of a training area of any given width to support tank traffic without changing species composition. The predictive ability of the model was tested by comparing predicted traffic capacities with the amount of traffic actually applied to two training areas in 1986. Where traffic capacity was exceeded, the model successfully predicted a significantly higher frequency of bare ground and ratio ofBouteloua gracilis toStipa spartea.     

Laying the foundation for a comprehensive program of restoration for wildlife habitat in a riparian floodplain

We analyzed the past and current distribution and abundance of vegetation and wildlife to develop a wildlife habitat restoration plan for the Sweetwater Regional Park, San Diego County, California. Overall, there has been a substantial loss of native amphibians and reptiles, including four amphibians, three lizards, and 11 snake species. The small-mammal community was depauperate and dominated by the exotic house mouse (Mus musculus) and the native western harvest mouse (Reithrodontomys megalotis). It appeared that either house mice are exerting a negative influence on most native species or that they are responding positively to habitat degradation. There has apparently been a net loss of 13 mammal species, including nine insectivores and rodents, a rabbit, and three large mammals. Willow (Salix) cover and density and cottonwoods (Populus fremontii) had the highest number of positive correlations with bird abundance. There has been an overall net loss of 12 breeding bird species; this includes an absolute loss of 18 species and a gain of six species. A restoration plan is described that provides for creation and maintenance of willow riparian, riparian woodland, and coastal sage scrub vegetation types; guides for separation of human activities and wildlife habitats; and management of feral and exotic species of plants and animals.     

Predicting River Floodplain and Lateral Channel Migration for Salmon Habitat Conservation1

Abstract: In this article, we describe a method for predicting floodplain locations and potential lateral channel migration across 82,900 km (491 km² by bankfull area) of streams in the Columbia River basin. Predictions are based on channel confinement, channel slope, bankfull width, and bankfull depth derived from digital elevation and precipitation data. Half of the 367 km² (47,900 km by length) of low‐gradient channels (≤ 4% channel slope) were classified as floodplain channels with a high likelihood of lateral channel migration (182 km², 50%). Classification agreement between modeled and field‐measured floodplain confinement was 85% (κ = 0.46, p < 0.001) with the largest source of error being the misclassification of unconfined channels as confined (55% omission error). Classification agreement between predicted channel migration and lateral migration determined from aerial photographs was 76% (κ = 0.53, p < 0.001) with the largest source of error being the misclassification of laterally migrating channels as non‐migrating (35% omission error). On average, more salmon populations were associated with laterally migrating channels and floodplains than with confined or nonmigrating channels. These data are useful for many river basin planning applications, including identification of land use impacts to floodplain habitats and locations with restoration potential for listed salmonids or other species of concern.     

Recovery in Naturally Dynamic Environments: A Case Study from the Sperrgebiet,Southern African Arid Succulent Karoo

Little is known about the process of vegetation recovery and associated time frames in the Succulent Karoo Biome of southern Africa. This study investigated the recovery of vegetation on sites impacted by mining (different types of dumps and mined areas) in the arid succulent karoo. The main aim of this study was to determine the state of recovery, time frames, successional stages, and the influence of environmental factors on recovery of coastal dune and sand plain plant communities. For this purpose, vegetation was recorded on some 121 sites throughout a coastal strip of approximately 100 × 3 km in Namibia’s restricted diamond area (Sperrgebiet). Using the species pool concept to derive vegetation reference sites and dominance-diversity curves, recovery of vegetation (measured in terms of species richness and cover) in these altered landscapes reached about 46% on the oldest, 51-year-old mine dumps. However, based on species richness, richness levels similar to the undisturbed reference sites were recorded after 30 years, following a logarithmic trend. Successional stages of natural recovery were indicated in this dynamic coastal environment and Cladoraphis cyperoides and Galenia fruticosa appear to be early successional species. Scaling up of studies to landscape level and developing a target community using the species pool concept are discussed as means to measure recovery in dynamic biological communities. On these altered, man-made landforms, the availability of seed may be the bottleneck to achieve vegetation cover comparable to undisturbed vegetation in the surrounding. Hence, restoration efforts should focus on this aspect.     

Urban Impacts on Streams are Scale‐Dependent With Nonlinear Influences on Their Physical and Biotic Recovery in Vermont,United States1

Fitzgerald, Evan P., William B. Bowden, Samuel P. Parker, and Michael L. Kline, 2012. Urban Impacts on Streams Are Scale‐Dependent with Nonlinear Influences on Their Physical and Biotic Recovery in Vermont, United States. Journal of the American Water Resources Association (JAWRA) 48(4): 679‐697. DOI: 10.1111/j.1752‐1688.2012.00639.x Abstract: The physical and biological conditions of stream reaches in 16 watersheds within the Lake Champlain Basin of Vermont, United States, were assessed and analyzed for a response to total impervious area (TIA) at multiple spatial scales. Natural gradients (e.g., channel slope) and human impacts to channel boundary conditions (e.g., bank armoring) were considered to ensure a robust test of the Impervious Cover Model for upslope TIA. The response of geomorphic stability and sensitive macroinvertebrates to TIA was nonlinear and significant (p < 0.001), decreasing rapidly at 5% TIA. The effect of urbanization on stream condition was shown to interact significantly with drainage area and channel slope using the analysis of covariance (ANCOVA) (p < 0.05). Hydraulic geometry regressions for urban and rural watersheds and ANCOVA were used to describe a significant watershed scale‐dependent response of channel width to urbanization (p = 0.001). The analysis of macroinvertebrate data from reaches in different stages of channel evolution indicated that stable reaches supported greater richness of pollution intolerant species (p < 0.001) and overall taxa richness (p < 0.01) than unstable reaches, and that biotic integrity improves as channels regain stability during their evolution into a state of quasi‐equilibrium. We conclude that macroinvertebrate communities can respond positively to channel evolution processes leading to natural channel restabilization.     

Recent Land Cover History and Nutrient Retention in Riparian Wetlands

Wetland ecosystems are profoundly affected by altered nutrient and sediment loads received from anthropogenic activity in their surrounding watersheds. Our objective was to compare a gradient of agricultural and urban land cover history during the period from 1949 to 1997, with plant and soil nutrient concentrations in, and sediment deposition to, riparian wetlands in a rapidly urbanizing landscape. We observed that recent agricultural land cover was associated with increases in Nitrogen (N) and Phosphorus (P) concentrations in a native wetland plant species. Conversely, recent urban land cover appeared to alter receiving wetland environmental conditions by increasing the relative availability of P versus N, as reflected in an invasive, but not a native, plant species. In addition, increases in surface soil Fe content suggests recent inputs of terrestrial sediments associated specifically with increasing urban land cover. The observed correlation between urban land cover and riparian wetland plant tissue and surface soil nutrient concentrations and sediment deposition, suggest that urbanization specifically enhances the suitability of riparian wetland habitats for the invasive species Japanese stiltgrass [Microstegium vimenium (Trinius) A. Camus].     

Weed manipulation for insect pest management in corn

Populations of insect pests and associated predaceous arthropods were sampled by direct observation and other relative methods in simple and diversified corn habitats at two sites in north Florida during 1978 and 1979. Through various cultural manipulations, characteristic weed communities were established selectively in alternate rows within corn plots.Fall armyworm (Spodoptera frugiperda J. E. Smith) incidence was consistently higher in the weed-free habitats than in the corn habitats containing natural weed complexes or selected weed associations. Corn earworm (Heliothis zea Boddie) damage was similar in all weed-free and weedy treatments, suggesting that this insect is not affected greatly by weed diversity. Only the diversification of corn with a strip of soybean significantly reduced corn earworm damage.In one site, distance between plots was reduced. Because predators moved freely between habitats, it was difficult to identify between-treatment differences in the composition of predator communities. In the other site, increased distances between plots minimized such migrations, resulting in greater population densities and diversity of common foliage insect predators in the weed-manipulated corn systems than in the weed-free plots. Trophic relationships in the weedy habitats were more complex than food webs in monocultures.Predator diversity (measured as mean number of species per area) and predator density was higher in com plots surrounded by mature, complex vegetation than at those surrounded by annual crops. This suggests that diverse adjacent areas to crops provide refuge for predators, thus acting as colonization sources.