No abstract

This study documents the scale and intensity of drying over the last half century in the Kenai Lowlands of south-central Alaska. Using historical aerial photos and field sampling of wetlands, including muskegs, kettle ponds, and closed and open basin lakes, we present data on drying and successional changes in woody vegetation between 1950 and 1996. The results of this study suggest that the Kenai Peninsula is becoming both woodier in its vegetation and drier. A regional analysis of 1113 random points indicated increased forest cover and decreased open and wet areas in both burned and unburned areas between 1950 and 1996. A census of water bodies in three subregions indicates that almost two-thirds of water bodies visited show some level of decrease in spatial area. Over 80% of field sites visited have experienced some level of drying, where vegetation transects indicate substantial invasion into former lake beds by facultative upland plants. These results are consistent with a regional change in climate that is both warming and drying as documented in Kenai and Anchorage weather records.

Abstract.-Theoretical models and computer simulations of the genetic structure of a continuous population predict the existence of patches of highly inbred individuals when gene flow within the population is limited. A map of the three genotypes of a two-allele locus is expected to exhibit patches of homozygotes embedded in a matrix of heter ozygotes, when gene flow is limited. A search for such patch structure was made on a 160 X 160 m plot within a continuous 60+ ha old-growth stand of Qiiercus laevis (turkey oak). Approximately 3400 trees were genotyped for 9 polymorphic loci using starch-gel electrophoresis, and the genetic structure was analyzed with spatial autocorrelation (both nominal and interval), hierarchical E statistics, and number-of-alleles-in-common. Adults (diameter at breast height > 0) and juveniles were analyzed separately but showed similar structure. While no distinct patch structure was found, a greater degree of relatedness was observed on a scale of 5 m-10 m than at greater distances, probably because of the limited acorn dispersal from maternal trees and a small amount of cloning by root sprouts. A computer simulation of a 10,000 tree forest breeding for 10,000 yr indicates that the effective neighborhood sizes (of randomly drawn seed-and pollen-donors) are both in excess of 440 individuals. The model thus cannot distinguish the observed data from panmictic mating.Key words.-Allozyme, F statistics, gene flow, genetic structure, isolation-by-distance, neighborhood size, numberof-alleles-in-common, Qiierciis laevis, randomization, spatial autocorrelation.Received December 22, 1992. Accepted March 30, 1994 Plant species and their populations are rarely randomly distributed within communities. Similarly, the spatial distri bution of individuals within populations is often nonrandom, as are the distributions of genotypes and alleles. Genetic structure can result from several factors, including isolation in small patches, limited pollen and seed dispersal, and se lection acting at the microhabitat level. Sewall Wright (1943aWright ( , 1946 argued that reproductive isolation can take place in a continuous population simply by means of limited gene flow. He described this process as " isolation by dis tance" and proposed that limited gene flow would generate patches of genetically similar individuals, which he called neighborhoods or demes. According to Wright's (1982) " shifting-balance" theory of evolution, if demes become ecologically adapted, then genetic structure can be enhanced. When environmental conditions change, some demes adapt to new conditions and are thus selected, whereas others be come extinct. It is plausible that deme-structure can arise in physically isolated populations, but it is less easy to visualize its creation in a continuously distributed population.Wright (1943b) used his isolation-by-distance model to describe the genetic differentiation of patches of white and blue flowers of the diminutive annual Linanthiis parryae in the Mojave Desert. He divided the 130-km long...

We used tree ring data (AD 1601-2007) to examine the occurrence of and climatic influences on spruce beetle (Dendroctonus rufipennis) outbreaks in south-central and southwest Alaska and found evidence of regional-scale outbreaks dating from the mid-1700s, related to climate variability at multiple temporal scales. Over interannual time scales (approximately 1-3 years), El Niño years, combined with severe late-summer drought, appeared to contribute significantly to spruce beetle outbreaks in the study area. Over multidecadal time scales (up to approximately 40 years), cool-phase Pacific Decadal Oscillation (PDO) conditions tended to precede beetle outbreaks, regardless of the phase of El Niño-Southern Oscillation (ENSO). All sites showed low-severity disturbances attributed to spruce beetle damage, most notably during the 1810s. During other major periods of disturbance (i.e., 1870s, 1910s, 1970s), the effects of spruce beetle outbreaks were of moderate or higher severity. The highly synchronized timing of spruce beetle outbreaks at interannual to multidecadal scales, and particularly the association between cool-phase PDO conditions and beetle disturbance, suggests that climate (i.e., temperature, precipitation) is a primary driver of outbreaks in the study area. Our disturbance chronologies (mid-1700s to present) suggest that recent irruptions (1990s to present) in south-central and southwest Alaska are within the historical geographic range, but that outbreaks since the 1990s show greater spatiotemporal synchrony (i.e., more sites record high-severity infestations) than at any other time in the past approximatly 250 years.

We document accelerating invasion of woody vegetation into wetlands on the western Kenai Peninsula lowlands. Historical aerial photography for 11 wetland sites showed that herbaceous area shrank 6.2%/decade from 1951 to 1968, and 11.1%/decade from 1968 to 1996. Corresponding rates for converting herbaceous area to shrubland were 11.5% and 13.7%/decade, respectively, and, for converting nonforest to forest, were 7.8% and 8.3%/decade, respectively. Black spruce (Picea mariana (Mill.) BSP) forests on three wetland perimeters established since the Little Ice Age concluded in the 1850s. Dwarf birch shrubs at three wetland sites showed median apparent tree-ring age of 13 years, indicating recent shrub colonization at these sites. Peat cores at 24 wetland sites (basal peat ages 1840 -18 740 calibrated years before present) indicated that these peatlands originated as wet Sphagnum-sedge fens with very little woody vegetation. Local meteorological records show a 55% decline in available water since 1968, of which one-third is due to higher summer temperatures and increased evapotranspiration and two-thirds is due to lower annual precipitation. These results suggest that wet Sphagnum-sedge fens initiating since the end of the Wisconsin glaciation began to dry in the 1850s and that this drying has greatly accelerated since the 1970s.Résumé : Nous documentons l'invasion accélérée de la végétation ligneuse dans les zones humides de l'ouest des basses terres de la péninsule de Kenai. D'anciennes photographies aériennes de 11 stations humides montrent que la superficie occupée par les plantes herbacées a diminué de 6,2 % par décennie entre 1951 et 1968 et de 11,1 % par décennie entre 1968 et 1996. Les taux correspondants de conversion de la superficie occupée par les plantes herbacées en prairie boisée étaient de 11,5 et 13,7 % alors que les taux de conversion des aires non forestières en aires forestières étaient de 7,8 et 8,3 % par décennie. Des forêts d'épinette noire se sont établies en périphérie de trois zones humides depuis la fin du Petit  ge glaciaire dans les années 1850. Des bouleaux nains établis dans trois stations humides avaient un âge médian apparent de 13 ans d'après le décompte des cernes annuels; ce qui indique que la colonisation de ces stations par les arbustes est ré-cente. Des carottes de tourbe provenant de 24 stations humides (âge de la tourbe basale allant de 1840 à 18 740 années calibrées BP) indiquent qu'elles étaient initialement des tourbières humides dominées par des sphaignes et des graminées avec très peu de végétation ligneuse. Des archives météorologiques locales montrent une diminution de 55 % de l'eau disponible depuis 1968, due pour un tiers à l'augmentation des températures estivales et de l'évapotranspiration et pour les deux tiers à la diminution des précipitations annuelles. Ces résultats indiquent que les tourbières dominées par les sphaignes et les graminées formées depuis la fin de la glaciation du Wisconsin ont commencé à s'assécher durant les années 1850 et que cet assèchement s'e...

Tall shrubs and trees are advancing into many tundra and wetland ecosystems but at a rate that often falls short of that predicted due to climate change. For forest, tall shrub, and tundra ecosystems in two pristine mountain ranges of Alaska, we apply a Bayesian, error-propagated calculation of expected elevational rise (climate velocity), observed rise (biotic velocity), and their difference (biotic inertia). We show a sensitive dependence of climate velocity on lapse rate and derive biotic velocity as a rigid elevational shift. Ecosystem presence identified from recent and historic orthophotos ~50 years apart was regressed on elevation. Biotic velocity was estimated as the difference between critical point elevations of recent and historic logistic fits divided by time between imagery. For both mountain ranges, the 95% highest posterior density of climate velocity enclosed the posterior distributions of all biotic velocities. In the Kenai Mountains, mean tall shrub and climate velocities were both 2.8 m y(-1). In the better sampled Chugach Mountains, mean tundra retreat was 1.2 m y(-1) and climate velocity 1.3 m y(-1). In each mountain range, the posterior mode of tall woody vegetation velocity (the complement of tundra) matched climate velocity better than either forest or tall shrub alone, suggesting competitive compensation can be important. Forest velocity was consistently low at 0.1-1.1 m y(-1), indicating treeline is advancing slowly. We hypothesize that the high biotic inertia of forest ecosystems in south-central Alaska may be due to competition with tall shrubs and/or more complex climate controls on the elevational limits of trees than tall shrubs. Among tall shrubs, those that disperse farthest had lowest inertia. Finally, the rapid upward advance of woody vegetation may be contributing to regional declines in Dall's sheep (Ovis dalli), a poorly dispersing alpine specialist herbivore with substantial biotic inertia due to dispersal reluctance.

Several studies have noted a relationship between vegetation type and fire frequency, yet despite the importance of ecosystem processes such as fire the long-term relationships between disturbance, climate and vegetation type are incompletely understood. We analysed pollen, plant macrofossils and sedimentary charcoal from three lakes within the Kenai lowlands to determine postglacial relationships between disturbance, climate and vegetation for the Boreal forest of southwest Alaska. An herb tundra was established in the lowlands following deglaciation by 13 000 cal. BP. Salix, Alnus and probably Betula kenaica, expanded in the area after 10 700 cal. BP, followed by Picea glauca by 8500 cal. BP. Picea mariana became established by 4600 cal. BP. The early Holocene was probably the driest time during the postglacial, as determined by aquatic plant macrofossils and climate models. Lake levels reached nearmodern conditions by at least 8000 cal. BP. Mean Fire Intervals (MFI) were longest during the shrub Áherb tundra phase (1389/65 yr), decreased after expansion of B. kenaica, Salix and Populus (779/49 yr) and Picea glauca (819/41 yr), and increased again with the arrival of P. mariana (1309/66 yr). Unlike previous studies, our data demonstrate the highest fire frequencies during the early to mid-Holocene and less frequent fire during the late Holocene when P. mariana forests dominated the lowlands. Early Holocene forests of P. glauca and B. kenaica existed in summers that were longer and drier than today, while the increasingly wetter and cooler climates of the late Holocene probably hindered forest fire around Paradox Lake, perhaps because of less frequent summer drought.

No abstract