Participant Abstracts

Click here for Participant References

ALLEN, CRAIG D., U.S. Geological Survey, Los Alamos, NM 87544
Disturbance Histories and Threshold Dynamics as Determinants of BioComplexity in Piņon-Juniper Woodlands

The biocomplexity of Southwestern landscapes is shaped in part by interactions among multiple disturbances through time. Interactive changes in human land uses, natural disturbances, and landscape patterns are described over the past 200 years for semiarid forests within the Bandelier Wilderness in New Mexico using dendrochronology, historical photography, and field studies. Findings include: (1) open forests were heavily grazed by domestic livestock (1880s to 1930s) and subsequently by feral burros (ca. 1935 to 1980), diminishing herbaceous cover and causing decreased fire frequency; (2) active fire suppression (after 1910) further reduced fire frequency; (3) piņon and juniper, which are adept at extracting shallow moisture, established under ponderosa pine forest overstory in response to the reduced fire frequency and herbaceous cover; (4) ponderosa pine sensitivity to drought and beetle infestation was heightened due to competitive interactions with piņon and juniper, resulting in a large-scale loss of ponderosa pine (and presumably herbaceous understory) during the severe 1950s drought; and (5) as thresholds of bare soil were exceeded by reductions in vegetation cover, high erosion rates were triggered that persist today. Physical processes now dominate the desertified interspaces between piņon and juniper trees. Long-term monitoring data show the persistence of these patterns since the 1970s. These results illustrate how disturbances can interact to trigger rapid threshold responses among a variety of ecosystem phenomena in piņon-juniper woodlands. This case study also highlights the need to better understand rapid, mortality-driven, ecosystem dynamics in assessments of the effects of climate variability on biocomplexity.

ALLEN, MICHAEL F., Center for Conservation Biology, University of California, Riverside, CA 94607
Mycorrhizal Dynamics in Pinyon-Juniper Woodland

We have focus work at the Goat Draw area of the Sevilleta LTER site. We have monitored root and fungal dynamics under pinyon and juniper using a combination of minirhizotron and coring. In particular, we are studying the role of mycorrhizal types in root and fungal dynamics. Root and mycorrhizal dynamics have been monitored using minirhizotron since the fall of 1997. An N fertilization experiment was also incorporated (100 kg/ha). To date, we have found that N fertilization does not change the % of mycorrhizal infection. However, it does change the relative root lengths, leading to an altered total amount of mycorrhizae, with a greater total amount of mycorrhizae in the control plots. Anti-sera techniques indicate that all genera of arbuscular mycorrhizal fungi are present in association with juniper, but sporulation with N fertilization was restricted to Glomus. EM richness and diversity is currently being evaluated. We have also estimated the life spans of rhizomorphs. Under pinyon, where most rhizomorphs are likely EM, no effect of fertilization was found. Under juniper, an AM plant, the rhizomorphs (likely more saprobic) life-span declined significantly. Individual rhizomorphs live for several months.

BALDA, RUSS, Department of Biology, Northern Arizona University, Flagstaff, AZ 86011
Bird Dynamics in Pinyon-Juniper Woodlands

BETANCOURT , JULIO L., U.S. Geological Survey, Tucson, AZ 85745, (in collaboration with Tom Van Devender, Kate Rylander, Steve Jackson, Mark Lyford, Steve Gray)
Long-term History and Dynamics of North American Pinyon-Juniper Woodlands

The late Quaternary history of North American pinyon-juniper woodlands is better known than for any other vegetation type worldwide. In large part, this is due to the rich plant macrofossil record in > 2000 packrat middens dated and analyzed from British Columbia to Baja California during the past 40 years. Most of the pinyon and juniper species can be readily identified as plant macrofossils in middens. The high taxonomic resolution afforded in middens has permitted a species-specific history of past distributions for pinyon and juniper species, including zones of past and present hybridization. During the last glacial, what are now desert lowlands (below ~1500m) harbored Pinus remota and P. edulis in the Chihuahuan Desert, Pinus californiarum var. fallax in the Sonoran Desert, P. monophylla in the Mohave Desert, and P. juarensis in central Baja California. Hybrids of P. juarensis x californiarum var. fallax occurred in what is now Joshua Tree National Monument. The northern limits of P. edulis, the dominant species today in the southern Rockies and Colorado Plateau, were along the Little Colorado River Valley. Possible hybrids of P. edulis x monophylla occurred in the Lower Grand Canyon, and of P. edulis x califoniarum var. fallax on the south slope of the Santa Catalina Mountains near Tucson. For junipers in the Chihuahuan Desert, some of the highlights include J. ashei, J. pinchotii0 and J. coahuilensis in the lowlands, J. scopulorum down to ~1500 m, and the absence of J. deppeana, which must have been displaced south of the U.S.-Mexican border. J. scopulorum was the dominant juniper in the foothills of the southern and central Rockies, the Colorado Plateau and the mid-elevations of the Great Basin. J. scopulorum was found as low as 975 m in the Ajo Mountains in Organ Pipe Cactus National Monument, 270 km from the nearest populations today. A similar displacement for J. occidentalis is recorded at Cataviņa (~30°S) in central Baja California ~450 km south of its southernmost stands today in the San Bernadino Mts. In the lowest elevations of the Southwest, dominant junipers included J. californica in the Lower Colorado River Valley and along the Arizona-Sonora border and J. osteosperma in what is now the Sonoran Desert uplands, the Mohave Desert, and the southern Great Basin. The Holocene record of these pinyons and junipers is complex. At desert elevations, pinyons were extirpated from desert elevations by ~11 14C kyr B.P. (radiocarbon kiloyears before present. Pinus remota retreated upslope to a few mountain refugia in west Texas and Coahuila, and P. californiarum var. fallax wedged itself between 1250-1800 m and below P. edulis at the base of the Mogollon Rim. Some junipers persisted at desert elevations until ~8 14C kyr B.P. The Holocene migrational histories of pinyon and juniper species are dynamical, modulated by dispersal, climate and physiography. For example, after 11 14C kyr B.P. regional warming (and drying) lifted upper limits 500-800 m and pinyon (P. edulis, P. monophylla) and juniper (J. osteosperma, J. monosperma) species quickly spread northward into low elevation (1500-2200 m), rocky habitats of the southern Rockies, the Colorado Plateau and the Great Basin. J. deppeana also spread north from the Sierra Madre into the mountains of west Texas, southern Arizona and southern New Mexico. The climate space for these pinyon and junipers species narrows considerably to the north, so that secondary climatic changes during the Holocene have either retarded or accelerated northward migration. At their northernmost sites, pinyons and junipers have been spreading into progressively narrower elevational wedges constrained by relatively high base levels at the bottom and cooler summers and competitive exclusion by higher elevation conifers at the top. These wedges become available for invasion only during brief (~1-2 kyr) warm and dry episodes. For example, J. osteosperma abutted against the Rockies in northeastern Utah/southwestern Wyoming by 9 14C kyr B.P., jumped over them to reach its northernmost sites in southern Montana by ~5 14C kyr B.P., occupied only a fraction of its present range in Wyoming/southern Montana during the wet Neoglacial (~4-2.5 14C kyr B.P.), and then exploded across the Wyoming Basins in the last 2 kyr. Founder events in the last 1000 years explain northern disjuncts of P. monophylla near Reno and P. edulis at Flaming Gorge National Recreational Area and near Ft. Collins. In the northern peripheries of pinyon and juniper distributions, population expansion and infilling associated with ongoing, natural invasion may account for much of the encroachment and increasing stem density presently attributed to post-settlement grazing and fire suppression.

BRESHEARS, DAVID D., Los Alamos National Laboratory, Los Alamos, NM 87545
The Canopy/Intercanopy Mosaic in Pinyon-Juniper Woodlands: A Framework for Assessing Biocomplexity along the Grassland-Forest Continuum

Pinyon-juniper woodlands are part of more general gradients in semiarid environments that range along a continuum ranging from grassland to forest. Sites along this continuum can be viewed as a two-phase mosaic of vegetation, comprised of individual or aggregated woody plants on the one hand and the intercanopy locations on the other. These two patch types differ with respect to most major components of water and energy budgets, and hence quantifying the heterogeneity and connectivity between these two patch types is essential for addressing large-scale problems and understanding biocomplexity. In recent studies of semiarid pinyon-juniper woodlands of northern New Mexico, at sites that are intermediate along the grassland-forest continuum, we quantified heterogeneity and connectivity between canopy and intercanopy patches with respect to soil water content, soil temperature, soil evaporation, plant water use, and runoff. We used these patch-scale results to assess processes and patterns at larger scales. Our results demonstrate how small change in herbaceous ground cover can trigger a transition from low to high erosion rates, and how small variations in the cover, height, and spatial pattern of canopy patches effects near-ground solar radiation. Collectively these results highlight the importance of ecological and hydrological processes at the patch scale in determining larger-scale patterns and processes for sites along the grassland-forest continuum. More generally, these results indicate the importance of studying ecological and hydrological processes in concert in assessing environmental biocomplexity along gradients.

CAREY, BERN, Center for Data Insight, Northern Arizona University, Flagstaff, AZ 86011
Knowledge Discovery in Very Large Databases

In the past 20 years, digital computers have provided us with the ability to generate very large amounts of raw data, gigabytes to terabytes, about businesses, economic markets, manufacturing processes and areas of scientific study. We are unable to examine or analyze such large datasets either manually or with conventional methods. During this period advanced database and analytical software has also been developed to permit us to semi-automatically manipulate and analyze these enormous data sets in order to discover meaningful patterns and relationships in the data to help us solve business, technical and scientific problems. This talk will present an overview of these analytical methods for very large datasets.

COLE, KEN, U.S. Geological Survey, Forest and Rangeland Ecosystem Science, Colorado Plateau Field Station, Northern Arizona University, Flagstaff, AZ 86011
Disassociation of Pine and Juniper Distributions Through Time.

Pinyon-Juniper woodland and Ponderosa Pine Forest are currently recognized as ecologically significant phytogeographic entities of the Colorado Plateau. The current abundance and distinctness of these units is challenged by the historical distributions of these plant species. During the latest Pleistocene, the distributions of Pinyon (Pinus edulis) and Ponderosa Pine (Pinus ponderosa) were perched somewhere near the southern rim of the Colorado Plateau. As the climate warmed at the close of the Pleistocene, between about 13,000 and 10,000 years ago, these species raced northward into their current ranges. By 10,000 years ago both pinyon and ponderosa pine then became abundant across the southern Colorado Plateau. Ponderosa pine went from a complete absence on the Colorado Plateau to spreading over an elevational range that was somewhat greater than at present as it may have benefited from increased amounts of summer precipitation between 8000 and 10,000 years ago. During the Pleistocene, isolated records of Pinyon occurred as far north as Springerville and the western Grand Canyon. By 12,000 years ago, Pinyon was found at Wupatki, and by 10,000 years it filled its present range within the eastern Grand Canyon. The distributions of juniper species stand in sharp contrast to the current pines of the southern Colorado Plateau. During the Pleistocene, Utah Juniper (Juniperus osteosperma) occurred over a much greater elevational range in Arizona than currently. Up until about 9000 years ago it was present in most habitats above 400 m elevation. It's range may have extended nearly as high as today, although it seems to have been far less abundant in the communities above 1500 m which were then dominated by limber pine (Pinus flexilis) and spruce (Picea sp). Utah juniper remained in the low deserts until about 8500 years ago when its lower limit retreated upslope to just below that of pinyon. The Pleistocene northern limit of Pinyon and Ponderosa pine in central Arizona may have been analogous to their modern northern limits in central and northern Utah where Utah juniper currently extends further northward. Extreme cold at high elevations may have combined with summer drought at low elevations in preventing reproduction north of certain latitude. Late Holocene records of Pinyon show that it is still expanding its range, especially in the state of Colorado. It is tempting to rely on modern-day phytogeographic entities to explain changes through time; and, this concept can be used in explaining broad latitudinal shifts. But when doing so, it must be realized that the reliable species groupings that might be found in one place break down through time and space, as is the case with the association between Pinyon Pine and Utah Juniper during the Pleistocene of northern Arizona and the modern distributions in northern Utah. Because of the disassociation between climate variables which will probably be caused by global warming processes, it is unlikely that we will be able to project future plant distributions relying only on today's phytogeographic units.

DIAL, ROMAN, Alaska Pacific University, Anchorage, AK 99508
A Theoretical Framework for Species Diversity.

A simple model of species richness relates the number of species, S, to four variables: (1) The quality of the environment for supporting life as measured by the total abundance of individuals of all species, A. (2) The riskiness of the environment as measured by an abundance threshold, M, such that if any species population drops below this, it is lost from the community. (3) A measure of niche space availability within the environment, 0 < n < 1, that measures how unconstrained the environment is (n=0 is very constrained and n=1 is unconstrained). And (4) a historical constant, h. The relationship is S=h(A/m)^n. The effect of an environmental change on species richness could be predicted by its anticipated effects on total abundance, risk, and niche. Some examples include eutrophication, generalized over-harvest, and habitat destruction. Habitat destruction reduces A: species richness should fall as a concave down, decreasing curve in response to increasing habitat loss. Generalized over-harvesting reduces A and increases M: species richness should drop steeply at first, then level off and drop steeply again as the level of over-harvest increases. Eutrophication increases A but reduces n: this should result in a hump shaped relation between eutrophication and richness.

DIRZO, RUDOLFO, Instituo de Ecologia UNAM. Mexico 04510
Diversity of Pinyon Pines in Mexico: Threats to Their Habitats and Ecological Consequences.

In this presentation I would like to highlight the diversity of pinyon pine species in Mexico, a region which concentrates the greatest diversity of pines in the world. A brief description of the 11-13 pinyon pine species of Mexico and their associated herbivores will be given. In addition I will present results of simulation models of climate change to predict possible scenarios of pinyon pine habitat coverage. I will compare these climate change threat with that expected for other forest types in Mexico. Finally, I will speculate on some of the possible consequences of global environmental change on the ecological interactions of pinyon pines and their associated herbivores. For example assuming that CO2-enriched atmospheres may increase the levels of herbivory on pinyon pines the reproductive potential of the plants may be reduced, but this effect may be more pronounced for the male reproductive function.

GEHRING, KITTY, Northern Arizona University, Flagstaff, AZ 86011 (in collaboration with Thomas Whitham, Tad Theimer, Neil Cobb, Randy Swaty, Paul Keim, Teresa Del Vecchio-Lane, Kris Haskins)
Interactions Between Abiotic Stress, Herbivores and Mycorrhizae in Pinyon Pine

Mycorrhizal associations are widespread and important relationships that can influence plant performance, plant and fungal distributions, and ecosystem function. Our studies indicate that both abiotic and biotic stresses have significant, complex effects on the mycorrhizal fungal mutualists associated with the roots of pinyon pine (Pinus edulis) in northern Arizona. The ectomycorrhizal fungi of pinyons growing in nutrient and moisture-poor cinder soils differ in abundance, community composition and responsiveness to environmental changes relative to the ectomycorrhizal fungi of pinyons growing in more nutrient and moisture-rich sandy-loam soils. In addition, herbivores that occur in high densities on pinyons growing in cinder soils alter the abundance and community composition of ectomycorrhizal fungi. Ectomycorrhizal fungi are highly diverse (>5000 species) and vary in their anatomy, physiology, ecology and relationships with host plants. It is thus likely that changes in the abundance and community composition of these fungi will affect plant performance as well as patterns of ecosystem nutrient dynamics. Our long-term goals are to understand how global changes will affect mycorrhizal associations and the importance of these mutualisms relative to other abiotic and biotic interactions in pinyon-juniper woodlands.

GOSZ, JIM, University of New Mexico, Albuquerque, NM 87131
Spatial and Temporal Dynamics of P-J Ecotones

Ecotones have been described as sensitive indicators of environmental change; however, few studies have detailed the complex behaviors of the ecosystems associated with these transitions. Environmental gradients that create transitions among lifeforms (e.g., from Pinon to Juniper to grassland) may be especially valuable in identifying common properties of ecotones that can be generalized to other areas. One example is the mosaic pattern of patch sizes of the lifeforms in an ecotone. Analyses from the Sevilleta biome transition area in New Mexico will demonstrate the robustness of these patterns and hypotheses will be presented on the biotic and abiotic factors that must be tested to develop the generality needed to extend the results. Also needed is recognition of site-specific abiotic factors and unique characteristics of the species in the transition versus the self-organizing properties of the biological entities that may be common to all ecotones. Quantitative studies and experiments to understand the spatial and temporal dynamics of mosaics associated with various ecotones are necessary. The key in this evaluation is the mosaic pattern of patch types: initiation of new patches, maintenance, expansion, and coalescence of patches, disintegration of patches (i.e., conversion of one type to another), and the juxtaposition of different patch types. We hypothesize that patch dynamics depend on: 1) patch size. Linear or non-linear responses are possible depending on the ecosystem variable of interest. Non-linear, threshold responses result from positive or negative feedbacks among plants, soils, animals, and the microclimate within and among patches of different sizes; 2) matrix type. Juniper patches within a Pinon matrix are hypothesized to function differently than Juniper patches in a grassland matrix; 3) patch type. Pinon patches within a Juniper matrix are hypothesized to function differently than grass patches within a Juniper matrix; 4) patch distance from matrix. Juniper patches located near a large patch or stand of Juniper are hypothesized differently than patches of similar size located at large distances from the matrix stand of Juniper; 5) patch configuration. Juniper patches located adjacent to Pinon patches are hypothesized to function differently than Juniper patches located adjacent to grassland patches.

HART, STEVE AND GEORGE KOCH, Northern Arizona University, Flagstaff, AZ 86011
The Role of Keystone Herbivores in Regulating Ecosystem Processes in Piņon-Juniper Woodlands

JACOBS, BRIAN, Bandelier National Monument, NM 87544
Woodland Restoration Studies at Bandelier National Monument, New Mexico

Woodland communities at Bandelier are characterized by high densities of young trees (J. monosperma and P. edulis) with a suppressed herbaceous understory, suggesting recent tree colonization. Scant herbaceous plant cover and the resulting exposed soils have provided opportunities for accelerated erosion and collateral damage to pre-historic cultural resources. Efforts are underway to characterize woodland systems (i.e. age-class, erosion rates) and assess the impacts of pre-historic (i.e. fuel-wooding and agriculture) and historic land-uses (i.e. livestock grazing and associated loss of fire regime). Mechanical thinning techniques have yielded impressive increases in herbaceous plant cover and decreases in sediment production. Prescribed fire has subsequently been applied (i.e. at five years post-thinning) as a savanna maintenance treatment. Response to mechanical and fire treatments is being documented using a variety of biotic (i.e. plant cover and biomass, diversity and abundance of arthropods, birds, and butterfly's) and abiotic measures (i.e. soil moisture and soil micro topography). Still unresolved are issues concerning management intervention in designated wilderness and structural versus functional targets in defining desired future conditions. Manipulation of woodland systems provides opportunities to characterize controlling ecological factors including: scale effects, threshold responses, and disturbance regimes.

KUSKE, CHERYL R., Los Alamos National Laboratory, Los Alamos, NM 87544
Exploring the Diversity and Abundance of Soil Microorganisms at Different Levels of Resolution.

Soil microorganisms play important roles in nutrient cycling, soil development, and plant survival and productivity. We have compared the abundance and diversity of soil microorganisms associated with pinyon rhizospheres and between-plant interspaces in two pinyon-juniper woodlands. Comparisons between the soil microflora present in an extreme cinders environment at Sunset Crater, AZ, and an adjacent sandy loam site have allowed us to test the hypothesis that soil microbial diversity is reduced in the extreme cinder environment relative to the sandy loam soil, and that the presence of the pinyon rhizosphere affects diversity and composition of the soil bacterial community. Diversity comparisons were conducted across different levels of resolution, and included both DNA-based and culture-based analyses. Analysis of total bacterial diversity using 16S rDNA sequences indicated that the level of diversity was extremely high (over 21 different divisions in the two soils), that most of the bacteria in the soils were previously undescribed, novel species and that most of them were members of completely novel divisions. A method based on 16S-TRF analysis to "fingerprint" community diversity and composition in complex soil communities resolved differences between the cinders and soil samples. Populations of heterotrophic bacteria and humate-degrading actinomycetes that represented decomposer guilds were correlated with soil fertility, while the fungal heterotrophs and Pseudomonas spp. bacteria exhibited close associations with pinyon rhizospheres, regardless of the status of soil development or other soil characteristics. Populations of Pseudomonas spp. were extremely diverse and were more abundant on physiologically stressed trees than on healthy trees.

LEAVITT, STEVE, University of Arizona-Laboratory of Tree-Ring Research, Tucson, AZ 85745
Multifaceted Studies of Isotopic Variability in Leaves and Wood of Southwestern Pinyon and Juniper

LUO, YIQUI, Department of Botany and Microbiology, University of Oklahoma, Oklahoma City, OK 73107
Inverse Analysis in Ecology: A Novel Approach to Untangling Biocomplexity

Ecologists are dealing with complex systems all the time. But we lack of a general approach to untangling biocomplexity. My talk will show that inverse analysis may be a very powerful approach to studying complex systems. Inverse analysis has been widely used in other scientific disciplines but has not been applied in ecology. The approaches fundamentally focus on data analysis for pattern recognition and mechanism identification. Its counterpart is the forward analysis, which is usually accomplished using simulation models. The latter predicts systems behavior with given model structure and a set of prescribed parameter values. Generally speaking, the forward analysis asks what a model can tell us about the biocomplexity of a system in question whereas the inverse analysis asks what the data can tell us about the same system. Combining the two approaches, we are able to probe mechanisms underlying the system. In this talk, I will use the Duke Forest free-air CO2 enrichment (FACE) study as an example to illustrate (1) the need of inverse analysis, (2) inverse analysis for characterization of carbon structure in the rhizosphere; and (3) inverse analysis for parameter estimation. Applications of the inverse analysis to the Duke FACE project have offered insights into several long-standing issues in ecosystems ecology, such as quantification of root exudation, root turnover, and sustainability of carbon sequestration in terrestrial ecosystems.

MAST, JOY NYSTROM, Department of Geography and Public Planning, Northern Arizona University, Flagstaff, AZ 86011
Tree Invasion Within a Pine/Grassland Ecotone: An Approach with Historic Aerial Photography and GIS Modeling

In previous studies, evidence of tree invasion into grasslands has mainly been through comparison of historical terrestrial photographs and/or tree age data. The goal of this research is to provide a quantitative description of the tree invasion process at a landscape scale using historical aerial photography, image processing and geographic information systems (GIS) approaches. Various map interpretive techniques provided evidence of shifts in the pine-grassland ecotone along the Colorado Front Range since the late 1930s. Historical aerial photos were digitally scanned and the outlines of tree invasions into the grassland were determined based on gray tone density slicing. Image processing of digitized aerial photography identified areas of change in tree cover and quantified locations and total hectares of tree invasions into grassland areas. Overall, the results clearly show an increase in woodland areas where there formerly existed grasslands. GIS modeling was used to relate tree invasion patterns to topographic orientation and changes in settlement patterns. The importance of terrain aspect on rate of tree invasion is clearly shown by the greater rate of tree invasion on north-facing slopes (generally moister with less heat stress) versus south-facing slopes. The most dramatic change in the controls of vegetation patterns over the past one or two centuries has been the decline in fire frequency due to fire suppression policy since ca. 1920. However, changes in grazing regimes may also have played an important role. When comparing these results to the instrumental climate record of the area, periods of favorable climatic conditions for seedling establishment generally correspond to periods of increased rate of tree invasion into grassland areas.

MILLS, JAMES, Center for Disease Control, National Center for Infectious Diseases, Atlanta, Georgia 30333
The Rodent-Borne Hemorrhagic Fevers: Host-Virus-Habitat Relationships

Because of the severity and the dramatic nature of the diseases they cause, the rodent-borne hemorrhagic fever viruses have received considerable attention recently from ecologists and health scientists. During the past five years researchers have identified at least 30 "new" hantaviruses and arenaviruses, all associated with murid rodents, and coevolutionary theory suggests that many additional host-virus associations await discovery. At least three hantaviruses are associated with rodents that frequent piņon-juniper habitats. Sin Nombre virus is associated with Peromyscus maniculatus, Limestone Canyon virus with P. boylii, and El Moro Canyon virus with Reithrodontomys megalotis. A cross sectional survey of all major biomes in the southwestern U.S. found that piņon-juniper woodland supported the highest prevalences of hantavirus infection in host populations. Whitewater Arroyo arenavirus associated with Neotoma species was only recently shown to be a human pathogen associated with fatal human disease in California. The extent of infection by this virus in woodrat populations is poorly known. The epidemiology of human disease caused by rodent-borne hantaviruses and arenaviruses is tightly linked to host population dynamics which, in turn, are tied to environmental factors such as rainfall and habitat quality. A major goal of CDC longitudinal studies of host populations is to collect and integrate knowledge of host-environment interactions toward the development of mathematical models that predict the risk of human disease at specific times and places.

MITTON, JEFFRY B., University of Colorado at Boulder, Boulder, CO 80309
Genetic Variation in Pinyon Pine for Studying Historical Processes and Adaptation

Genetic variation for allozymes, mtDNA and cpDNA are available for studies of pinyon pine. MtDNA markers will be particularly useful for studying historical events, particularly for inferring information about glacial refugia, and movements out of the glacial refugia. Some allozymes are associated with resistance to insect herbivores, while others are associated with moisture stress. Genetic markers in the mtDNA of pines are ideal for detecting ancient events of differentiation and dispersal. The nuclear, chloroplast and mitochondria components of the genomes of most conifers have different modes of inheritance, disparate potentials for gene flow, and heterogeneous degrees of population structure. For most species in the family Pinaceae, mtDNA is maternally inherited, and is dispersed only by the movement of seeds, whereas cpDNA is inherited paternally, and is dispersed first by wind-borne pollen, then by the movement of fertilized seeds. Because wind-borne pine pollen typically moves much further than pine seeds, the potentials for gene flow are high for cpDNA and nuclear genes, but low for mtDNA. The utility of mtDNA for discovering glacial refugia was demonstrated by a study of limber pine, Pinus flexilis. The range-wide population structure was examined with eight size variants in the second intron of nad1. Haplotypes were determined for 704 trees from 40 localities. The majority of the mtDNA haplotypes are restricted to minor portions of the geographic range. The data are consistent with mtDNA differentiation among seven glacial refugia, followed by dispersal out of those refugia. Allozyme heterozygosity is associated with resistance to herbivores in pinyon pines on the cinder soils around Sunset Crater, near Flagstaff, AZ. These pines experience chronic water and nutrient stress, and also sustain higher densities of herbivores than do pines living nearby on normal soils. However, the impact of the stress and herbivory varies, for some trees suffer little or no damage from herbivores, whereas others are trimmed so regularly by the stem moth, Dioryctria albovitella, that they assume a different growth form. Resistant trees were significantly more heterozygous than susceptible trees at two of four allozyme loci. In addition, in both susceptible and resistant trees, older trees were more heterozygous than younger trees, revealing viability differentials favoring heterozygotes. The glycerate dehydrogenase polymorphism is associated with water stress in pinyon pine. The frequency of allele 3 increases on arid sites, and the 33 homozygote has superior viability on arid sites. Trees within one of the dry sites were examined to test the hypothesis that stomatal sizes and densities are heterogeneous among genotypes. Heterozygotes had the longest and widest stomata; the stomatal area of heterozygotes was 28% greater than the stomatal area of homozygotes. The stomates of the 33 homozygotes were relatively small and significantly narrower than the stomates of other genotypes. These results suggest that stomatal size and shape may play a role in adapting pinyon to heterogeneity in soil moisture.

NEILSON, RONALD P., USDA Forest Service, Corvallis, OR 97339
Perspectives on Patterns of Climate and Pinyon-Juniper in the Interior West

Pinyon and Juniper distributions in the West are clearly related to the large-scale distribution and temporal dynamics of regional airmass gradients. The distributional patterns of pinyon species also bear interesting relationships to those of Ponderosa Pine and various species of oak. The pines and oaks appear to have similar ecological constraints, but differ in that the oaks have the capability to reproduce asexually (clone) in addition sexual reproduction. I will review work on the regional and elevational patterns of distribution of some of the pines and oaks and their relationship to regional airmass gradients and the mode of reproduction. Implications with regard to local to regional diversity patterns will also be discussed. Simulations with a new biogeography model, MAPSS, are able to re-iterate many of the spatial patterns of both horizontal and elevational ecotones in the region under the present climate. Simulations under global warming scenarios will be discussed in terms of the expected climate shifts and their similarities to mid-Holocene climate shifts and the possible implications to regional biotic complexity at a range of scales.

OGLE, KIONA, Duke University, Durham, NC 27708
Drought-induced Pinyon Mortality: The Role of Environmental Stress, Tree Age, and Recent Growth

Drought events are common in the American Southwest and could become more frequent and severe under climatic change. Altered drought patterns could have repercussions on species composition and distribution. The record drought of 1996 in northern Arizona caused widespread pinyon mortality and offered a novel opportunity to examine how drought and environmental stress interact to effect this dominant tree. Mortality rates were compared between two sites designated as "low" and "medium" stress; contrary to expectations, rates were 2´ greater in the low stress site (2.3% vs. 5.4%). The odds of dying increased by 25% for every 100 m drop in elevation in the low stress site, but were independent of elevation in the medium stress site. Unpredictable across-site mortality patterns may be associated with heterogeneous precipitation patterns or drought preconditioning in the medium stress site. However, within a site, where rainfall is more homogenous, mortality rates were predictable; pinyons growing on cinder cones suffered higher mortality than less-exposed populations. In addition to site quality, we also investigated size-dependent mortality. Large trees were 2´ and 3´ more likely to die than medium and small trees, respectively. A model of recent ring-width deviations vs. drought severity showed a loss of "climatic sensitivity" with age in dead trees. Within their respective age (size) classes, young (small) sensitive and old (large) insensitive trees are more likely to die during a drought. Overall, old (large) insensitive trees are especially predisposed to die. As the Southwest has a dynamic climate typified by severe droughts, it is important to understand how droughts act as bottleneck events to affect a dominant tree in a major vegetation type of the U.S.

SWETNAM, THOMAS W., University of Arizona-Laboratory of Tree-Ring Research, Tucson, AZ 85745 (in colloraboration with Julio L Betancourt and Gerald Gottfried)
Climatic Effects on the Regional and Long-term Demography of Southwestern Pinyon-Juniper Woodlands

Pinyon-juniper woodlands cover ca. 20 million ha of mostly public lands in 9 states. These public lands are managed by the USDA-Forest Service, Department of Interior (BLM, NPS, F&W, BIA), Department of Defense, and state land agencies. Although pinyon-juniper woodlands comprise the third largest vegetation type in the U.S., we know little about their regional and long-term behavior. Regional demographic histories and trends have not been identified for pinyon-juniper woodlands or other western forest types despite their obvious utility in setting restoration and conservation targets, identifying and predicting responses to changing land use and climate, or estimating sustainable uses. The U.S. Geological Survey, the USDA-Forest Service and the University of Arizona are collaborating on developing regionally-scaled, high-resolution demographic histories essential to understanding and managing for the dynamic behavior of pinyon-juniper woodlands. This pilot study was conducted at the Sevilleta Wildlife Refuge, one of twenty-one Long-term Ecological Research (LTER) Sites funded by the National Science Foundation. The Sevilleta LTER encompasses vast pinyon-juniper woodlands, is a protected site, has a rich context of long-term monitoring and research, and offers and ideal platform for discussing and promoting ecological advances. A possible outgrowth of the Sevilleta pinyon study is a multiagency effort to identify regional demographic trends at other sites and for other tree-ring species (e.g., ponderosa pine) prevalent in the West. At the Sevilleta LTER, dendrochronological methods were used to crossdate pinyon births and deaths from the late 1500s to the present across a series of plots. Because both false and double rings occur commonly, ring counts are not a suitable measure of age. Size is also a poor proxy for age. Crossdating was necessary to determine not only establishment age of living trees, but also recruitment and mortality dates for dead trees. For seedlings (<5 cm), both ring and internode counts can be used with relatively good success. All live and dead trees were aged in five 50 x 100 m (0.5 m ha) plots at different sites and then aggregated (summed up) for both births and deaths. Independent tree-ring reconstructions of climate help explain past and present trends in terms of climate variability. Establishment ages for dead trees permits reconstruction of the dynamical rather than the usual static age structure. Recruitment episodes are strongly episodic, with major events represented at two or more plots in 1630-1670, 1710-1730, 1770-1790, and 1900-1920 (Fig. 9). Based on the seedling and sapling data, there also appears to be a pulse in recruitment since 1976 (Fig. 4, 5, 9). The period from 1942-1956, arguably the worst drought of the last few hundred years, accounts for most of the mortality evident in the vast number of dead trees at the Sevilleta and throughout New Mexico. This event eliminated most trees established before 1850, shifted the age distribution to younger trees, and decreased the proportion of pinyons to junipers at all of the plots. Below the litter layer, there were no differences in soil carbon and nitrogen between living trees and trees that have been dead for 50 to 100 years. How much of this carbon and nitrogen is actually bioavailable is not known. Regional demographic trends were compared against regional climate reconstructions from tree rings. This comparison suggests the following model for the long-term behavior of these woodlands. First, broadscale mortality during catastrophic droughts, such as 1575-1595, 1667-1680, 1730-1750, 1890-1904, and 1942-1957 release existing seedlings and saplings from competition for light, water and nutrients and open up niches for recruitment. A preferred niche might be the footprint of a dead tree (a pool of soil and carbon and nitrogen, some bioavailable), which persists for at least 100 years. Recruitment pulses tend to occur in the first sustained wet period following the drought and the associated mortality. An example of this process may be the pulse in recruitment since 1976, on the heels of the mid-century drought and sustained by a subsequent string of warm, wet springs. The post-1976 shift to warm, wet springs produced both anomalous tree growth at high-elevation sites and accelerated tree recruitment at lower elevations. These trends may be related to the Greenhouse Effect, producing longer growing seasons, and enigmatic warming of the tropical Pacific, promoting more frequent incursions of winter storms typical of El Niņo years. Similar surges in post-1976 recruitment have been noted for junipers, ponderosa pine and other woody species. Most recently, the seedling demography was expanded to other areas in the middle Rio Grande Basin to allow discrimination between climate and grazing effects (e.g., the Sevilleta LTER was fenced in 1976). Preliminary results from a series of protected sites, with different histories of grazing removal, also show the same pulse in recruitment since 1976. The Sevilleta study underscores the importance of reconstructing, observing, and assessing demographic trends at the appropriate scales, i.e., mesoscales and centuries. Ecological synchroneity at these scales is the hallmark of climatic effects on ecosystems and is a key to separating cultural from natural causes of environmental change shows that age structure is very much a transient property of pinyon-juniper woodlands, a function of the cohort development and time since the last catastrophic drought. This historical perspective will be useful for setting conservation and restoration targets, or in discouraging or justifying management actions. For example, the pulse in pinyon recruitment since 1976 could easily have been mistaken for the effects of fire suppression and/or grazing in the absence of long-term data.

WAHL, EUGENE R, University of Minnesota, St. Paul, MN 55108
Issues and Methods in Pollen-Based Paleoclimate and Paleovegetation Reconstruction: The Modern Analog Technique Criteria for Selecting Analogs to Modern Vegetation Testing Climate Reconstructions Efficiency: How Many Grains to Count?

The modern analog technique is often used to provide calibration of fossil pollen assemblages in terms of vegetation types and climate parameters. The method relies on the assumption that close similarity ("analogy") between modern and ancient pollen assemblages means that similar vegetation produced the assemblages (which is verified to an extent by relationships between modern pollen and vegetation), and further, that similar modern and ancient vegetation imply similar modern and ancient climates. The purpose of this paper is not to question these assumptions, but rather to examine three important issues involved in implementing the modern analog technique: determining optimal thresholds of similarity between pollen assemblages; validating climate reconstruction methods to determine their likelihood to give good paleo-reconstructions; and determining minimum numbers of pollen grains that can be counted while ensuring good quality in climate and vegetation reconstruction. Optimal thresholds of similarity between pollen assemblages have typically been examined by evaluating different "cutoff" levels for the square chord distance (SCD) dissimilarity statistic, a non-Euclidean multivariate distance measure with signal-to-noise enhancing characteristics. The primary concern has been focused on finding maximum cutoff levels that are able to reliably distinguish pollen assemblages from differing kinds of vegetation with similar representation of major pollen types (e.g., distinguishing southeastern U.S. pine forests from northeastern and Great Lakes pine forests). Failure to distinguish such pollen assemblages as coming from different kinds of vegetation is called "false positive" error. The analysis presented in this paper provides a quantitative method for determining cutoff levels that jointly minimize both false positive error and the "false negative" error that occurs when pollen assemblages from like vegetation are incorrectly considered dissimilar at a specific cutoff level. Results from a newly analyzed set of modern pollen samples in southern California show that a range of SCD cutoff values between ~ 0.20 and 0.25 jointly minimizes both kinds of error. False negative errors increase sharply at lower cutoff values, whereas false positive errors increase less rapidly at higher cutoff levels. The method used is general, and can be applied to other modern pollen calibration sets. A method for reconstructing the paleoclimates indicated by fossil pollen assemblages uses averages of the climate parameters associated with the suite of modern analogs sites "selected" by the fossil pollen. These averages are weighed by the inverse of the SCD statistic, so that the climate associated with very close analogs is given high weight in the reconstruction process. A validation of this procedure can be done by reconstructing the modern climate associated with each site of a modern analog reference set, and comparing these reconstructions with climate values assigned to the sites based on instrumental data. Results from the new southern California reference set show that unbiased and relatively precise reconstructions can be achieved with this method. In addition, these results show that the best reconstructions occur at the same levels of the SCD cutoff value, as those determined to be optimal for identifying like vegetation types. Determining the appropriate number of pollen grains to count highlights the tradeoff between efficiency (the amount of time spent counting or the number of samples counted per unit time) and reliability (the variation of pollen assemblages in relation to count size). Up to now, this issue has been examined in terms of the number of grains needed to ensure that relatively rare types have a high probability of being encountered--leading to a rule of thumb of at least 300 grains per sample, often analysts count to 500 or more. Since rare pollen types are typically not included in the multivariate measures used in the modern analog technique, a different way of examining the efficiency/reliability tradeoff is called for in situations where the analog technique is being used. I and a colleague (David Lytle) have developed a Monte Carlo random simulation method to examine this issue, generating sets of 1000 simulated pollen assemblages at each of a number of count sizes, from 75 to 1000. The probabilities used in these simulations are based on the pollen proportions in a set of actual fossil assemblages from the eastern United States that were counted to over 1500 grains. Analog-derived climate and vegetation reconstructions based on the simulated counts were compared to reconstructions based on the reference fossil samples. The simulations show that good quality climate reconstructions can be achieved with counts sizes of only 200 grains; the distribution of anomalies from the reference sample-based reconstructions at this count size is very similar to the distribution of anomalies at a count size of 1000. Similarly, vegetation reconstructions based on simulated counts of 200 grains are often nearly as accurate as those based on higher count sizes--with the twin caveats that vegetation reconstruction is sensitive to the kind of voting scheme used in the reconstructions and that highly mixed vegetation probably requires larger count sizes to be reliably reconstructed.

THOMAS G. WHITHAM, Northern Arizona University, (in collaboration with Neil S. Cobb, Catherine A. Gehring, Kerry M.Christensen, Ken Charters Susan Mopper, Jeffry B. Mitton)
Environmental Stress and Plant Genetics affect Community Structure: Interactions of Pinyon Pine, Insects, Birds and Mammals.

Since the eruption of Sunset Crater in 1064AD, colonizing pinyon pines have encountered a new stressful environment for which they are not well adapted. These stressful conditions have resulted in increased susceptibility to keystone insect herbivores (the stem-boring moth, Dioryctria albovittella, and the scale insect, Matsucoccus acalyptus). At Sunset Crater, resistant and susceptible trees to these insects exhibit pronounced differences in tree architecture, reproduction and resin defenses. Long-term removal experiments show that these insects affect the rest of the community including mycorrhiza, seed dispersing birds and mammals. Because resistance and susceptibility traits to insect attack are in part genetically based and interact with environmental stress, we aim to quantify the environmental and genetic components of community structure and biodiversity. With increasing climatic stress, our findings argue that herbivores will exert greater influence on the pinyon community, biodiversity will decline, and many mutualistic associations will be altered.