Saltcedar (Tamarix spp.), a shrub native to Eurasia, is associated with major alterations to wetland and riparian systems in the southwestern United States. Since the 1960s saltcedar has been naturalized in northern states of the U.S. where its growth potential and impacts are not well known. Here, we describe the occurrence, age, size, and relative cover of saltcedar populations in several river basins in central eastern Montana, USA, to identify potential patterns of spread across the region and changes in individual populations as they age. Stands were aged according to the oldest saltcedar individuals and were sampled for dominant plant cover and soil properties. Multiple introductions appear to have occurred in Montana, with the oldest stands occurring on the Bighorn River in southern Montana. Saltcedar absolute and relative cover and stand area increased significantly with stand age, while native tree and shrub relative cover remained low across all stand ages. These results suggest that saltcedar stands establish where woody natives are not abundant and that they persist and expand over time. Although soil salinity remained constant, soil pH decreased with saltcedar stand age, indicating a possible effect of organic matter inputs. An analysis of annual wood increment of saltcedar and sandbar willow (a native with analogous growth form) stems along a latitudinal gradient showed that stem growth of both species did not differ significantly among regions. Stem growth decreased inversely with elevation for both species while growth responses to elevation did not differ between species. Our results show an increase in number of populations and continued viability of these populations. Mechanisms of saltcedar increases in this region are yet to be determined. Anthropogenic influences, such as saltcedar plantings, watershed alterations (e.g., river flow control), and habitat disturbances (e.g., cattle grazing or habitat clearing) may facilitate its spread in similar climates of the Great Plains.

Eared Grebes (Podiceps nigricollis) use the Great Salt Lake (GSL) primarily as a fall staging area. They feed on brine shrimp (Artemia franciscana) to build up energy reserves to complete their southward migration. Whereas it has been speculated that timing of grebe departure from the GSL is due to declining shrimp density, this hypothesis has not been tested. We monitored grebe weight, behavior, and departure dates to test this hypothesis and to determine the effects of declining shrimp densities on Eared Grebes during 1999, 2000, and 2001 when there were different shrimp densities. Despite differences in shrimp densities, grebes left the GSL about the same time each year. Juvenile and subadult mass was significantly lighter (P < 0.001) in 1999 than in 2000 or 2001; however, adult grebe mass was similar in 1999 and 2000, despite the large difference in available of food. While there was no difference (P > 0.05) in dive or surface duration between 1999 and 2000, there were differences in the percentage of time spent foraging. Our results indicated that at shrimp densities <0.1 shrimp · L⁻¹, grebes increased their foraging effort to offset the declining food resource instead of departing the GSL at an earlier date.

From a landscape perspective, riparian corridors can be viewed as mosaics of vegetation patches. We delineated 10 patch types within the floodplain of the San Pedro River (Arizona) on the basis of physiognomy, dominant overstory species, and tree size class; and we assessed differences in hydrogeomorphology, vegetation structure, plant species richness, and soil chemistry and texture. Patches of tamarisk (Tamarix), an introduced species, fell within the continuum of variation shown by other patch types in the landscape mosaic. Among the tree-dominated types, cottonwood-willow (Populus-Salix) and tamarisk patches were inundated more frequently than mesquite (Prosopis) patches, while cottonwood-willow patches had shallower groundwater than tamarisk or mesquite patches. Due to the wetter conditions, cottonwood-willow patches had a high relative abundance of wetland and exotic species in the understory. Tamarisk patches and wet shrublands (Baccharis salicifolia–Salix exigua) had high woody stem densities while cottonwood-willow patches had dense canopy cover. In association with differences in canopy cover, cottonwood-willow patches had low herbaceous species richness but high woody species richness, while tamarisk patches had high herbaceous and low woody species richness. Soil electrical conductivity, silt content, organic matter content, and available phosphorus increased from young to old stands of both tamarisk and cottonwood-willow, often resulting in greater differences between patches of different size/age class than between patches with different dominant species. Surface soil salinity (electrical conductivity) was low in all patches, including those dominated by tamarisk. Nitrate was abundant in soils of tamarisk patches (perhaps reflecting their high clay content) and wet shrubland patches. Dry shrublands (Hymenoclea-Ericameria) and wet shrublands were similar to young forest patches in having coarse soils with little organic matter.

The xystodesmid milliped tribe Pachydesmini is supported by the autapomorphic absence of bold aposematic pigmentations and possibly also by relative somatic inflexibility; it comprises 3 component genera: Pachydesmus Cook, 1895; Dicellarius Chamberlin, 1920; and Thrinaxoria Chamberlin and Hoffman, 1950. Three representatives occur west of the Mississippi River—P. clarus (Chamberlin), P. crassicutis crassicutis (Wood), and T. lampra (Chamberlin). New localities are documented in Louisiana and Texas, and literature records are summarized for these states and Arkansas; records of P. clarus from Cherokee County, Texas, are the westernmost in this state for the east-Nearctic xystodesmid fauna. An apparently allopatric population of P. crassicutis crassicutis occurs in Evangeline and Lafayette Parishes, Louisiana. These are the first records of this species from west of the Mississippi River, and new localities of this form are cited from east of the watercourse. Thrinaxoria lampra, known previously in Texas from only Gregg County, is documented from Bowie, Cass, Harrison, Rusk, Sabine, and Titus Counties plus Webster Parish, Louisiana. Diagnoses and pertinent anatomical illustrations are provided for each species including the first cyphopod drawings for P. clarus and T. lampra; occurrences west of the river are depicted on a map. A new tribal diagnosis is published along with a distribution map.

Our main objective was to improve understanding of herbicide effects on community dynamics to refine the use of technology and advance the development of ecologically based weed management strategies. We hypothesized that native grasslands would exhibit reductions in culturally sensitive forb cover, biomass, and density relative to the rate of application of selective rangeland herbicides, and that hand-removal of sulfur cinquefoil (Potentilla recta L.) would increase indigenous species cover, biomass, density, species richness, and diversity. Treatments consisted of 3 rates each of 2,4-D clopyralid (0.28 kg ai · ha⁻¹ 0.0532 kg ai · ha⁻¹, 0.56 kg ai · ha⁻¹ 0.1064 kg ai · ha⁻¹, 0.84 kg ai · ha⁻¹ 0.1596 kg ai · ha⁻¹); 2,4-D amine (0.532 kg ai · ha⁻¹, 1.064 kg ai · ha⁻¹, 1.596 kg ai · ha⁻¹); metsulfuron (0.0042 kg ai · ha⁻¹, 0.021 kg ai · ha⁻¹, 0.032 kg ai · ha⁻¹); picloram (0.14 kg ai · ha⁻¹, 0.28 kg ai · ha⁻¹, 0.56 kg ai · ha⁻¹); and clopyralid (0.05025 kg ai · ha⁻¹, 0.21 kg ai · ha⁻¹, 0.42 kg ai · ha⁻¹). This experiment was replicated 3 times at 2 late-seral, noninfested sites in southeastern Montana. In a companion study, sulfur cinquefoil was removed adjacent to paired nonremoved controls in 5 replicates at 2 sites in 1-m² plots for 2 growing seasons. Canopy cover, density, and biomass were collected 24 months after initial treatment at all sites. Indigenous perennial grass cover and biomass increased with herbicide application; however, picloram, metsulfuron, and clopyralid reduced native forb density at 1 site, and picloram reduced forb cover at both sites regardless of rate. Effects of herbicides on species richness or diversity were not detected. Hand-removing sulfur cinquefoil increased total plant richness, especially that of native forbs. Restoring species richness and diversity may be difficult using selective broadleaf herbicides because key functional groups, such as forbs, appear to be at risk.

We use macroscopic charcoal analysis to reconstruct fire history in sagebrush (Artemisia tridentata var. wyomingensis and A. tridentata var. tridentata), in Newark Valley, Nevada. We analyzed charcoal at continuous 1-cm intervals (~7–127 years), and pollen at 2- to 10-cm intervals (~70–263 years) in a core spanning the last 5500 cal yr BP (calendar years before present). A charcoal peak in the historic period was associated with a >1400-ha fire dated to 1986 that burned in the watershed. We reconstructed the prehistoric fire history by inferring fires from similar charcoal peaks that were significantly greater than the background charcoal accumulation. Our results suggest the fire regime is climate and fuel driven. During periods of wetter climate, sagebrush increased and fires were more abundant, and during extended dry periods when sagebrush decreased, fires were less frequent. Our method does not allow calculation of a fire-return interval; however, our results support models that estimate a mean fire-return interval of up to a century in Artemisia tridentata var. wyomingensis. The charcoal record indicates that fires have increased within the historic period. This contrasts with pinyon/juniper studies that indicate an expansion of woodland associated with fewer fires in the historic period. We suggest that in the central Great Basin, a regime of frequent fires in sagebrush that limits woodland expansion is true for the sagebrush-woodland ecotone, but in sagebrush-dominated valleys with lower fuel loads, fires have always been less frequent. Protecting sagebrush-dominated valleys from frequent fire would appear to be consistent with the prehistoric fire regime.

This study explores how the relationship between flow and riparian vegetation varies along a montane river. We mapped occurrence of woody riparian plant communities along 58 km of the San Miguel River in southwestern Colorado. We determined the recurrence interval of inundation for each plant community by combining step-backwater hydraulic modeling at 4 representative reaches with Log-Pearson analysis of 4 stream gaging stations. Finally, we mapped bottomland surficial geology and used a Geographic Information System to overlay the coverages of geology and vegetation. Plant communities were distinctly arrayed along the hydrologic gradient. The Salix exigua Nuttall (sand-bar willow) community occurred mostly on surfaces with a recurrence interval of inundation shorter than 2.2 years; the Betula occidentalis Hooker (river birch) community peaked on sites with recurrence intervals of inundation between 2.2 and 4.6 years. The hydrologic position occupied by communities dominated by Populus angustifolia James (narrowleaf cottonwood) was strongly related to age of trees and species composition of understory shrubs. The fraction of riparian vegetation on surfaces historically inundated by the river decreased in the upstream direction from almost 100% near Uravan to <50% along the South Fork of the San Miguel River. In upstream reaches much of the physical disturbance necessary to maintain riparian vegetation is provided by valley-side processes including debris flows, floods from minor tributaries, landslides, and beaver activity. Where valley-side processes are important, prediction of riparian vegetation change based on alterations of river flow will be incomplete.

We present an approach to quantitatively assess nonnative plant invasions at landscape scales from both habitat and species perspectives. Our case study included 34 nonnative species found in 142 plots (0.1 ha) in 14 vegetation types within the Grand Staircase–Escalante National Monument, Utah. A plot invasion index, based on nonnative species richness and cover, showed that only 16 of 142 plots were heavily invaded. A species invasive index, based on frequency, cover, and number of vegetation types invaded, showed that only 7 of 34 plant species were highly invasive. Multiple regressions using habitat characteristics (moisture index, elevation, soil P, native species richness, maximum crust development class, bare ground, and rock) explained 60% of variation in nonnative species richness and 46% of variation in nonnative species cover. Three mesic habitats (aspen, wet meadow, and perennial riparian types) were particularly invaded (31 of 34 nonnative species studied were found in these types). Species-specific logistic regression models for the 7 most invasive species correctly predicted occurrence 89% of the time on average (from 80% for Bromus tectorum, a habitat generalist, to 93% for Tamarix spp., a habitat specialist). Even with such a modest sampling intensity (<0.1% of the landscape), this multiscale sampling scheme was effective at evaluating habitat vulnerability to invasion and the occurrence of the 7 most invasive nonnative species. This approach could be applied in other natural areas to develop strategies to document invasive species and invaded habitats.

We documented distribution, relative abundance, diet composition, and body condition of Sacramento pikeminnow Ptychocheilus grandis during 2001 and 2003 at 5 sites in the Sacramento–San Joaquin Delta, California. Sacramento pikeminnow densities in nearshore habitats were higher in 2003 than 2001. In both years, spatial distribution of beach seine densities was similar. There were no significant differences in density among sampling sites except for the southernmost site where the catch was near zero. Based on rotary screw-trap data from a 6th site, we found relative abundance of Sacramento pikeminnow entering the Delta via an artificial floodplain was positively correlated with flow. Most individuals collected using all 3 gear types were age 1 or older, and appeared to grow quickly based on data from previous studies. Sacramento pikeminnow had diverse diets composed of freshwater and estuarine invertebrate and fish taxa. Incidence of piscivory was only 2% of the diet of individuals <150 mm, but increased to 50% for fish over 150 mm. No salmonids were observed in foregut contents during the study. In both years body condition declined abruptly in July. Our results suggest Sacramento pikeminnow are more common in the turbid, tidal freshwater habitats of the Delta than was previously recognized. Stream flows may play an important role in moving juvenile Sacramento pikeminnow into the Delta from upstream areas. Similar to northern pikeminnow P. oregonensis, but in seeming contrast to endangered Colorado pikeminnow P. lucius, the present study showed that Sacramento pikeminnow can be successful in altered habitats.

We examined stomach contents of preserved specimens of larval Pacific giant salamander (Dicamptodon tenebrosus) and Cope's giant salamander (D. copei) collected from sympatric and allopatric stream populations. The dietary components of these specimens were used to calculate dietary overlap between the 2 species and to determine if changes in overlap existed between sympatric and allopatric populations. To statistically test overlap values, a randomization algorithm was used to construct a simulated data matrix (i.e., null model) in order to compare observed values of dietary overlap to a distribution of overlap values from the null model. Significant levels (P < 0.05) of dietary overlap occurred in all cases of sympatry as well as allopatry. Average dietary overlap in sympatry was significantly lower than in allopatry, suggesting a dietary shift when in sympatry to reduce competition. Diet composition also differed between sympatric and allopatric populations of each species, further suggesting a partitioning of food resources by one or both species when in the presence of its congener.

Using livestock as seed dispersal agents may be an effective method for increasing species diversity on degraded and previously seeded rangelands. We quantified seed passage and recovery rates, and post-passage germinability of Wyoming big sagebrush (Artemisia tridentata Nutt. ssp. wyomingensis Beetle & Young), bottlebrush squirreltail (Elymus elymoides [Raf.] Swezey), and gooseberry globemallow (Sphaeralcea grossulariaefolia [H. & A.] Rydb.) by feeding Holstein heifers seeds of each species at 3 levels (15,000; 30,000; and 60,000 seeds) over a period of 3 weeks. One-kg fecal samples were collected 1, 2, 3, and 4 days after seed ingestion. Undamaged seeds were extracted from the samples and tested for germinability. Globemallow had the highest percentage of recovered, undamaged seed, followed by squirreltail and sagebrush. Globemallow and sagebrush seed passage was highest on Day 1, after which seed numbers dropped sharply. Squirreltail passage and recovery were more consistent through time, with higher seed recovery at lower seed feeding levels. Post-passage germinability was highest for squirreltail and globemallow on Day 1. Sagebrush germination was negligible. Differences in physical seed properties (size, shape, and seed coat) likely influenced interspecies variation in passage, recovery, and germinability. Globemallow and squirreltail seeds may be suited for livestock dispersal, but sagebrush seeds are not.