I compiled the first vascular flora of the Tucson Mountains of southeastern Arizona, part of the Arizona Upland subdivision of the Sonoran Desert. The study area consisted of over 40,000 hectares, with two principal biotic communities: desertscrub and desert-grassland. Over 300 km of trails, ridges, washes, canyons, and peaks were surveyed in a three year period, covering all elevations and habitats. This was supplemented by examination of herbarium records. The area is unusually rich compared to other regional floras in the state of Arizona, including a vascular flora of 605 specific and infraspecific taxa in 331 genera and 80 families. The distribution of life forms is: herbaceous annuals and perennials (73%), shrubs (9%), subshrubs (8%), succulents (4%), and trees (1%). Obligate annuals make up 45% of the flora, of which 62% are obligate spring annuals, 33% obligate summer annuals, and 5% either spring or summer annuals. Historical records from the first part of the century are reviewed and compared with the present day flora; 27 species that were found in the Tucson Mountains prior to 1950 have not been collected since, in spite of increased sampling effort. A disproportionate number of these locally-extirpated species are winter annuals.

Biting has been observed to be an important component of male mating behavior in several elasmobranch species. I observed male biting during courtship and mating in a population of Urolophus halleri, the round stingray, in the Sea of Cortez. Females allow males to bite the posterior and medial edge of their pectoral fin during courtship yet often appear to struggle to dislodge the male after they have been bitten. During mating the male bites the anterior edge of the females' pectoral fin and the female is passive. In response to this biting behavior females have relatively thicker discs than males and males have sexually dimorphic dentition. Larger adult males have relatively smaller yet more sharply curved teeth than smaller males that may allow them to hold on to females better during courtship. Therefore there is the potential for assortative mating based on male dentition and his ability to hold a female.

The effects of sunflower (Helianthus annuus) and groundcherry (Physalis pubescens), two hosts of the tobacco budworm (Heliothis virescens), on predation by paper wasps (Polistes arizonensis) were examined. Initially it was found that P. arizonensis foragers took more caterpillars from H. annuus plants than P. pubescens in paired-choice tests. Regarding physical aspects of the plants, wasps took longer to locate prey on artificial plants with higher leaf surface areas and gave up searching faster on artificial plants with more complex architectures. In paired-choice tests for chemical effects, wasps chose more caterpillars within H. annuus odor than in P. pubescens odor. They also chose more H. annuus-reared larvae than P. pubescens-reared larvae when they were presented without plants. Predator handling times were affected by the amount of food in the caterpillar gut, with a slight trend depending on what plant species was present in the caterpillar gut.

Spawning flannelmouth sucker, Catostomus latipinnis, in the Paria River averaged 478 mm (n = 246) total length (TL). This was 53 mm longer (p < 0.001) than the mean length of spawning fish taken from this same location in 1981 (425 mm, TL, n = 286). Sub adult flannelmouth were common in the Paria in 1981 but no post-larval fish < 379 mm, TL were caught in 1992 or 1993. There is no evidence that juvenile flannelmouth have reared in the Paria River/Glen Canyon Area in the last 12 years. However, some adult fish appear to enter the population from downstream locations. In 1992 and 1993, spawning occurred throughout the lower 10 kilometers of the Paria. Young-of-year were seen in 1992 but could not be found shortly after hatching. No young-of-year were seen in 1993. Growth of adult sized fish is very slow. Based on extrapolations from recaptures, longevity may approach 30 years. Recaptures from fish marked in other studies were originally tagged as far as 229 km downstream from the mouth of the Paria.

Boiga irregularis has caused the extinction of several native vertebrate species on the island of Guam. Information on movement is critical to the management of the species. B. irregularis is active at night and spends daylight hours in secluded refuges (from 10 m up in the forest canopy to 1 m underground). There appears to be no preferences for any particular refuge type. Nighttime movement (distance) was related to distances between daytime refugia. The average net movement ranged from 26.9-97.7 m/day (N = 11). The maximum distance moved from the first daytime location following release to subsequent locations ranged from 142.8-1809.4 m. The activity area (minimum convex polygon) ranged from 1.9-99 ha; the cumulative activity area periodically increased and did not reach an asymptote. Direction of snake movements were random. The data suggest that Boiga lacks a defined activity area, and moves randomly and continuously searching for resources.

Understanding plant community structure is fundamental to successful restoration and conservation of riparian ecosystems. High plant species diversity is often cited as an important characteristic of riparian areas. Graphical summaries, principal components analysis, and analysis of variance were used on species composition and abundance data to determine community patterns in riparian areas and surrounding uplands. Woody plant composition varied with relative elevation above the primary channel, but herbaceous composition appeared unaffected by proximity to the riparian area. Alpha diversity indices, richness, Shannon's H', and Simpson's D, were not always higher within the riparian area. Alpha diversity of woody plants demonstrated no consistent pattern, however vertical structural diversity was highest near the stream. Herbaceous plants showed increased diversity in floodplain sites and decreased diversity in hillslopes sites, suggesting that topographical profile influenced diversity. Herbaceous richness was better predicted from overstory characteristics than herbaceous biomass and environmental variables.

After 6-10 years of mark-recapture observations, I studied seven desert tortoises by radio-telemetry during 1990-1992 in a desert-grassland community in Pinal County, Arizona. Six estimated home-range areas averaged 14.7 ha. Winter-spring (Nov-Jun) use areas (overline x=0.7 ha) were significantly smaller (P = 0.002) than summer-fall (Jul-Oct) use areas (overline x=10.7 ha). A correction formula inflated 1-2 summer-fall use areas of five tortoises 4-41% larger than their corrected home-range areas. Extended movements by females were significantly more frequent (P = 0.0001) than those of males during Mar-Jul, significantly less frequent (P = 0.0057) than males during Aug-Oct, and most frequent by both sexes in September. Use of two slopes and terraces was not season-dependent (P = 0.9159). Tortoises variably used four shelter types (rock, soil burrow, wood rat nest, vegetation), significantly with south-facing entrance aspects (P 0.0005). Hibernaculum structure and location varied. Hibernation ranged from 88-315 days. Radio-equipped tortoises included reuse of mark-recapture locations.

The sediment rating curve (SRC) method was used to study the effects of vegetative manipulations on sediment concentrations on the Beaver Creek watersheds of north-central Arizona. A linear model was best in the analysis of streamflow-sediment relationships. The corrected linear model eliminated some bias of the linear model and improved prediction. Partitioning data by streamflow-generating events and hydrograph stages produced significantly different SRCs and increased the descriptive and predictive powers. Due to insufficient data at the high streamflow range, SRCs for the pinyon-juniper watersheds had low values of ra SRCs for the ponderosa pine watersheds produced better results. More overstory removal resulted in higher sediment concentrations.

Growth at high temperature of wild isolates of three species of Bacillus was analyzed to assess potential responses to global warming. Experimental populations were grown at temperatures from 32° to 60° C. The higher temperatures include ones near and above maxima previously reported for laboratory strains. Summer soil temperatures, three centimeters below the ground surface, were recorded at the same site from which the wild isolates came, show that temperatures in the Sonoran Desert often reach 50° to 60° C. The growth data show that the desert isolates of B. subtilis and B. licheniformis have thermal maxima close to those reported by Gordon et al. (1973), while B. megaterium grew well at 2-3°C above the reported maximum. Global Climate Models predict a rise of 1° to 4.5°C over the next 60-100 years. Such a rise could shorten periods of active growth and nutrient cycling by Bacillus decomposers.

<p> Land use change to meet economic and societal demands has negatively impacted the current state of our world's ecosystems and biodiversity. The alarming rate of landscape change and degradation has prompted ecologists to conserve intact areas and restore historical habitats in the hopes of mitigating the potential damage. As a result, restoration ecology was developed and has grown dramatically within the past couple decades. Historically, the Mississippi Alluvial Valley, provided services such as water enhancement and nutrient cycling, and the characteristic bottomland hardwood forests dominated the land cover and were suitable habitat for many organisms, especially neotropical migrants. As the Upper Mississippi Alluvial Valley was converted into agricultural fields and locks and dams were installed, many neotropical migrants had steady population declines. More recent surveys have noted that approximately two-thirds of the neotropical migrants have been declining for the past half-century. For the past two decades, the U.S. Army Corp of Engineers (USACE) has been responsible for the restoration of bottomland hardwood forests in the Upper Mississippi Valley. Using birds as model organisms, this project assessed the restoration success of this site by (1) comparing avian richness, diversity, and density among a chronosequence of forest restoration age classes to determine any significant differences in species assemblage in the chronosequence and (2) creating a predictive habitat model that assists in the formulation of appropriate recommendations for restoration planning and management in the future. I hypothesize that species richness and diversity is highest at intermediate sites since they will have habitat characteristics of both grasslands and forests and that my habitat models identifies structural variables, such vegetation height, and tree height and ground cover as significant determinants of avian presence. Our study included a chronosequence of nine restoration sites and two references in which bird surveys and vegetation surveys were completed. Avian surveys consisted of 25-m fixed radius point counts at five plots per site. Sites were visited three times for the spring and fall data collections. Spring data will continue May 2014 due to unusual flooding conditions. Within the 25-m radius, a 17.84-m radius vegetation plot was created. Tree species, tree health variables, diameter at breast height (DBH), and abundance were recorded, and the tree was tagged. Shrub stem count data was recorded in four belts within each vegetation plot, with the condition that a shrub is at least 1-m in height. The plot was divided into four belts, and five quadrats of 0.5 m2 were formed along each belt to record forb and grass cover and litter depth. Species richness, Antilog Shannon-Weiner's Diversity Index, and Simpson's Diversity Index was calculated, and the vegetation data was used in the predictive habitat modeling. I recorded 52 bird species during the spring and 79 bird species during that fall that were within the 25-m radius. Bird species richness and diversity differed among the sites, and richness and diversity generally increased as the age of the site increased. Grassland and shrubland birds did not have any significant trends for vegetation parameters. However, most open woodland birds had a significant association with DBH, and a majority of forest birds had significant associations with tree height and crown size. The observed richness and diversity values contradict several studies but may be attributed to the productivity rate of the sites chosen. Density estimates can also be enhanced through new surveying methods which would address detectability issues for naturally skittish birds. Future restoration efforts should include the planting of fast-growing trees and shrubs in order to increase structural diversity at faster rates. In addition, more sites should be acquired and restored in order to create close patches for neotropical migrants to facilitate travel. Continuing efforts in restoration and appropriate management should emphasize the causes of declines for land birds and identify them early enough to develop an appropriate management plan.</p>