Fireweed in bloom, Lake Clark National Park and Preserve

Overview

From calving glaciers to rumbling volcanoes, the Southwest Alaska Network is located in one of the most geologically active regions on the continent. Ice and permanent snowfields blanket approximately one-fifth of the land area in the Southwest Alaska Network. Valley and tidewater glaciers radiate from massive snowfields in the coastal mountains of Kenai Fjords National Park, Lake Clark National Park and Preserve, and Katmai National Park and Preserve, while rapidly retreating glaciers leave ribbons of till in their wake. Rivers and streams fill with salmon in the summer and fall, feeding bears and eagles that congregate along their banks. Spruce woodlands and open tundra stretch across the land in between.

The intersection of these geologic and biological features makes the SWAN an ideal laboratory for long-term ecological studies. Here, in a region where approximately half of the land area has a mean annual temperature of 0 °C, Coastal Aleutian, low Arctic, interior-boreal, and Pacific coastal flora and fauna converge. Relatively small changes in temperature are expected to have broad-scale effects on seasonal snow and ice dynamics, growing season length, and myriad other ecological processes. Colonization by new species due to disturbance and/or climate change, changes in the distribution of existing species, or changes in the timing of critical life stages or migration could have important implications for park management and resource protection.

Background

Glaciers cover approximately one-eighth of the combined area of Katmai, Kenai Fjords, and Lake Clark National Parks. Changes to these glaciers have both global and local consequences with impacts ranging from their contribution to global sea level rise to the transformation of scenic and recreational values for park visitors. Alaska glaciers have experienced widespread retreat and thinning since the end of the Little Ice Age (approximately 1880) resulting in significant changes in land cover.

Monitoring Objectives

Map the extent of glacier ice in the SWAN parks on a decadal scale.

Methods

The use of satellite imagery is recognized as a simple and effective means for documenting change in glacier extent. Suitable satellite imagery is available from the 1970s to the present and this imagery is being used to map glacier extent on a decadal scale. Ice fields and outlying glaciers are outlined using a combination of methods including heads-up digitizing, supervised classification, and band ratio techniques.

Current and Future Work Efforts

Glacier extent mapping has been completed for the following parks:

*Arendt et al. 2012. **LeBris et al. 2011.

The Alaska national park glaciers status and trends project, which will be completed at the end of 2013, is documenting glacier extent, surface elevation changes, and studies of focus glaciers in the nine national park units in Alaska that contain glacier ice. Glacier extent is mapped for the “baseline” period (1950s, based on topographic maps) and the “modern” period (late 2000s, based on satellite imagery). Other data generated from the mapping component of this project include individual glacier margins, area, and hypsometry (area by elevation bands).

Glacial extent mapping in SWAN parks (2012), red areas show ice loss and blue areas show ice gain

References

• LeBris, R., F. Paul, H. Frey, and T. Bolch. 2011. A new satellite-derived glacier inventory for western Alaska. Annals of Glaciology 52(59):135-143.
• Arendt, A., C. Larsen, M. Loso, N. Murphy, and J. Rich. 2012. Alaskan National Park glaciers: Status and trends – second progress report: March 30, 2012. National Park Service Unpublished Report, Anchorage, Alaska.

Fall senescence in Lake Clark National Park and Preserve (Hardenburg Bay, 2009)

Overview

‘Landscape processes‘ focus on the seasonality of physical and biological processes: lake ice formation and breakup; the timing and duration of snowpack; and growing season length. The SWAN is developing methods for using remotely-sensed data to describe seasonal change through collaboration with the USGS-Earth Resources Observation Systems (EROS) Data Center, the University of Alaska Fairbanks-Geographic Information Network of Alaska (GINA), and the PhenoCam Network at Harvard Forest.

Moderate Resolution Imaging Spectroradiometer (MODIS) data will be used to monitor phenology, lake ice, and snow extent at the landscape scale. Time-lapse cameras installed at remote weather stations will be used to gather site-specific data on the timing of snowfall, green-up, and senescence.

Sampling Design and Objectives

Monitoring objectives are to document long-term changes in start (freeze-up) and end of season (break-up) dates for lake ice; the start and end of the snow season; and the start and end of the growing season, as indicated by the normalized vegetation difference index (NDVI).

Current and Future Work

As of 2012, growing season (NDVI) metrics have been developed for twelve years of composited eMODIS data (2001-2012). We are working with GINA to develop comparable metrics for Advanced Very High Resolution Radiometer (AVHRR) data, which span a much longer period (1978-2012). Start and end of season dates for lake ice have been summarized for 17 large lakes in the SWAN (2001-2010) using manual interpretation of daily MODIS true-color images. Snow season metrics are under development using the MODIS snow product developed by the National Snow and Ice Data Center.

Above (L-R): MODIS time series of breakup of lake ice on Lake Iliamna, April 15 - May 10, 2005. (1) April 15; (2) April 17; (3) May 8; (4) May 10.

A time-series of daily photographs taken from cameras mounted on remote weather stations has been compiled for four sites in the SWAN (2010-2012). The photos have been archived with the PhenoCam network, and processed to generate curves showing changes in “greenness” (Green chromatic coordinates, an index of the spectral quality of foliage) through the growing season. We will use these data to explore methods for identifying start and end of season dates, and for manually interpreting growing season and snow on/snow off dates that can be used to validate the MODIS-derived metrics. An example from Contact Creek, in Katmai National Park and Preserve, is shown below:

Preliminary Results

Preliminary interpretations of the MODIS data (2001-2007) highlight the variation inherent in these seasonal events (Reed et al. 2009). The 7-year period that we reviewed was too short to detect changes in seasonality, but did provide a window into the variation we could expect to see. During that time, the length of the snow season was stable within the study area, but snow season start and end dates were variable, and frequent cloud cover made it difficult to pinpoint those dates. A strong El Niño year (2002-2003) resulted in anomalously low snowpack and a short to nonexistent ice season on many lakes. Growing season dates were less variable, but tended to track snowmelt dates.

As we build on these data sets, including AVHRR, we will revisit long-term trends in seasonality.

Lower Twin Lake, Lake Clark National Park and Preserve

Vegetation Composition and Structure

Vegetation is integral to ecosystem function, energy transfer and element cycling. It drives ecosystem productivity, provides habitat and forage for wildlife, and is sensitive to environmental change.

Remotely-sensed data can effectively address many monitoring needs at the landscape scale, providing wall-to-wall coverage at a relatively low cost. Ground-based monitoring is desirable for validation of the remotely-sensed data, and for addressing questions of structure and composition at a finer spatial and temporal scale. Both remote sensing and ground-based monitoring approaches are being used to track changes in vegetation in focal plant communities, including sensitive communities.

Monitoring Objectives

• Map and quantify long-term, landscape-level changes in the distribution and extent of major land cover classes using satellite imagery and/or aerial photographs.
• Estimate long-term changes in species richness, cover and diversity in focal ecosystems.
• Estimate long-term changes in vegetation structure (physiognomy), including changes in the demography and density of woody species.

Current and Future Work

Landscape-level Change: Remote Sensing Approaches

Between 2006 and 2010, the SWAN worked with collaborators at the USFS-Pacific Northwest Research Station to adapt methods used for change detection in the Pacific Northwest to the parks in southwest Alaska. The approach (LandTrendr) relies on Landsat scenes collected annually from an area of interest and applies algorithms to identify periods of stability and change in the time-series. This process of trajectory segmentation, performed for each pixel, results in a “change” signal. The approach, though promising, was difficult to apply in the SWAN due to high cloud cover and missing data, both of which limited our ability to compile and interpret a time series. We have since suspended the project.

A second approach, using manual interpretation of air photos, is under development with collaborators at Saint Mary’s University. High resolution scans of aerial photos from the 1950s and 1980s, orthorectified to an IKONOS base image and digital elevation map , are being used to interpret change in areas of interest in the SWAN parks.

Repeat photography has been used to document change on the landscape over time-frames ranging from less than 20 years to greater than 100 years.

Lastly, see the NASA image gallery for spectacular views of the dynamic SWAN landscape, including ash plumes from the Valley of Ten Thousand Smokes in 2003; an outburst flood at Bear Glacier in 2005; the eruption of Mt. Redoubt in 2009; and sea ice in Bristol Bay in 2012.

Vegetation and Composition Structure: Field-based Approaches

Since 2008, we have established 105 monitoring plots in Lake Clark National Park and Preserve and Katmai National Park and Preserve. Forest monitoring plots will be added in Kenai Fjords in 2013. Field sites have been selected using a spatially balanced probabilistic sampling framework (GRTS design), weighted by accessibility and landscape attributes, including elevation and aspect. Focal vegetation types include mid- and high-elevation tundra as well as low elevation spruce woodland. Sensitive community types monitored using the same plot design include high-elevation alpine and nunatak communities, treeline, and low-elevation mesic spruce forest.

Salt marsh communities are monitored under a separate protocol, described under ‘Sensitive Communities.‘

Preliminary Results

Nonvascular species are an important component of all vegetation types sampled in the SWAN, and in most cases they comprise 65% or more of the species recorded at our sites. Species of conservation concern collected by the vegetation monitoring program include the rare mosses Rhytidiopsis robusta, Iwatsukiella leucotricha, Buxbaumia aphylla, and Schistostega pennata, and lichens Erioderma pedicillatum, Cetrelia alaskana, and Hypogymnia pulverata. For a view of these species and others, see the photo gallery. Beginning in 2013-15, SWAN will be conducting lichen inventories in Kenai Fjords, Katmai, and Lake Clark.

We have also been using tree-ring data to examine the influence of climate (temperature, precipitation) on tree growth, and have found that trees are responding favorably to warmer temperatures in recent decades, particularly at more interior sites. Forest measurements include estimates of basal area, mortality, and seedling densities. In 2012, we began sampling epiphytic lichens in forested plots using the US Forest Service-Forest Inventory and Analysis (FIA) program’s off-frame lichen plots. The data will be used to track changes in species composition that could be due to changes in climate or air quality.

Brown bears foraging in a salt marsh on the edge of Hallo Bay, Katmai National Park and Preserve

Overview

Plant communities that occur at the margins of their environmental tolerance can serve as early indicators of change on the landscape. These plant communities are strongly controlled by climate and other physical factors (e.g., hydrology) and are expected to be sensitive to climate change. Salt marshes and nunataks, featured here, host a suite of sensitive community types targeted for monitoring by the SWAN.

Nunataks, or exposed mountain ridges surrounded by ice, are of interest for monitoring because they may have supported plant populations that survived the Last Glacial Maximum, approximately 20,000 years ago. These fragile alpine environments are expected to be sensitive to warming, and may play an important role in the re-establishment of plant and animal species in recently deglaciated areas. Salt marshes and tidal flats along upper Cook Inlet provide critical feeding and resting habitat for brown bears, waterfowl, and shorebirds. Hydrology, salinity, sedimentation, and erosion shape the distribution and abundance of plant communities in these systems. Monitoring data are expected to support a broader understanding of change in the marine nearshore ecosystem.

Sampling Design and Objectives

Monitoring objectives for nunatak communities are to estimate long-term changes in species richness, abundance, and cover using the plot design outlined in the protocol for Vegetation Composition and Structure. A separate protocol for salt marshes includes standard operating procedures for measuring topographic gradients, soil characteristics (salinity, pH), and vegetation (species richness, cover). Monitoring objectives include documenting change in environmental variables and in geomorphic and/or landscape-level features.

Current and Future Work

In 2005, the SWAN inventoried nunataks in Kenai Fjords National Park and Lake Clark National Park and Preserve and established eight monitoring sites. In 2007-2008, the SWAN established three monitoring sites in coastal meadows of Lake Clark and Katmai National Parks and Preserves. We sampled 130 vegetation plots, and installed data loggers to monitor soil temperature and pressure transducers to track water level, temperature, and salinity. Air photo interpretation was used to document changes to the landscape that occurred between 1950 and 2005.

All monitoring sites will be revisited every 10 years.

Preliminary Results

The nunatak inventory supplemented a more extensive vascular plant inventory and yielded a number of rare and/or disjunct alpine species. We collected over eighty vascular taxa, five of which were new records for the parks. Rare species included Arnica diversifolia (rayless arnica), Carex phaeocephala (dunhead sedge), Douglasia alaskana (Alaska rock jasmine), Papaver alboroseum (pale poppy), Potentilla drummondii (Drummond’s cinquefoil), Thlaspi arcticum (arctic pennycress). An additional species, Boechera lemmonii (Lemmon’s rockcress), a member of the mustard family and a Rocky Mountain disjunct, was found more than 550 km from the nearest collections near the Alaska-Yukon border, suggesting that it may have persisted through the Last Glacial Maximum. During the inventory, we also collected leaf tissue from several common species for DNA analysis (Marr et al. 2008, Marr et al. 2012) as part of a larger study by the University of Victoria, Canada.

In salt marshes, air photos showed a general expansion of the shoreline seaward since the 1950s, due to a combination of uplift following the 1964 earthquake and sediment deposition. Photo interpretation and field sampling indicated slight decreases in sedge meadow (1-2%) and infilling of ponds by sedges and salt-tolerant vegetation. Spruce expansion onto beach berms and dunes accounted for roughly 2% of land cover change. Changes at one site in Lake Clark (Chinitna Bay) are visible in the following image pair:

Aerial photographs show changes in Chinitna Bay salt marsh between 1957 and 2004

A summary of changes (mean ± 95% CI) calculated across the three sites are shown below.

Graph comparing percent area change in Chinitna Bay between 1957 and 2004

Spruce beetle-related tree mortality, Bay of Isles, Katmai National Park and Preserve

Overview

Over the last two decades, a massive spruce beetle (Dendroctonus rufipennis) outbreak has killed 1.5 million hectares of forest in south-central and southwest Alaska. Smaller, more localized irruptions of native defoliators have affected alder, willow and dwarf birch. To better understand forest disturbance dynamics, the SWAN has used tree-ring data to examine the frequency of historic beetle outbreaks in Lake Clark and Katmai, and the relationship between climate and disturbance regionally. In addition, we are investigating the environmental conditions that predispose trees to beetle attack through a series of targeted studies.

Current and Future Work

Shrub Defoliators in Lake Clark and Katmai

A handful of native geometrid and noctuid moths are thought to be responsible for the widespread defoliation of dwarf birch and willow that occurred in interior Lake Clark between 2009-2012. The willow tortrix, Epinotia cruciana, was the most abundant moth found in 2011, and is likely responsible for the recent damage on willows. Damage to dwarf birch (Betula nana) occurred at the beginning of the outbreak, and it remains unclear which defoliators used birch a host. Additional moths identified in 2011 include Pararctia subnebulosa, Grammia quenseli, and Lasionycta taigata, a new recrod for Alaska. In Katmai, a widespread irruption of a native noctuid moth, Sunira verberata, between 2000-2006 led to extensive defoliation and dieback of alder and willow. Aerial survey data indicated that the moth caused damage as far west as Dillingham and Aleknagik (US Forest Service 2006); in Katmai, it was concentrated primarily around Naknek Lake, Lake Coville and Lake Grosvenor. Many areas of damaged alder still remain and have been slow to recover.

Shrub defoliation at Snipe Lake, Lake Clark National Park and Preserve

Spruce beetle-related mortality in Lake Clark and Katmai

Major spruce beetle outbreaks have occurred in south-central Alaska at least four times over the last two centuries. Tree-ring data indicate that these outbreaks have tended to occur one or more years following warm-phase El Niño Southern Oscillation (El Niño) years and drier-than-average summer conditions (Sherriff et al. 2011). In a companion study, we’re using stable isotopes (d13C, d18O) in cellulose to examine the response of trees that were killed by, or survived, the recent spruce beetle outbreak. When completed, data from these projects will be posted to the International Tree-ring Data Bank.

Spruce beetle-related tree mortality, Brooks, Katmai National Park and Preserve

Preliminary Results

In stands affected by the recent spruce beetle outbreak, trees that died often showed greater sensitivity to temperature in the decades before beetle attack than trees that initially survived. A moving correlation between temperature and growth for a site on the Valley of 10,000 Smokes Road shows that trees that succumbed early to the spruce beetle (right) were those that had been most sensitive to warming. After 1990, tree growth became decoupled from temperature in the trees that died, whereas growth was enhanced by warming in trees that survived (left). These results suggest that warming, and presumably drought stress, can predispose some trees to beetle attack.

Figure depicting correlation between temperature and annual growth

Orange Hawkweed (Hieracium aurantiacum), Kenai Peninsula

Overview

Exotic plant species can outcompete native species for limited resources and change the structure and function of ecosystems. Throughout Alaska, over 280 non-native plant species have been documented--accounting for approximately 13.5% of the flora. National Park Service lands have so far avoided invasion by many exotic species found elsewhere in North America. However, the likelihood that exotic plants will become established is increasing due to a warming climate, increases in construction-related disturbance, and tourism.

Current and Future Work

The National Park Service’s Alaska Exotic Plant Management Team has conducted invasive plant surveys in three of the SWAN parks: Katmai National Park and Preserve (KATM), Kenai Fjords National Park (KEFJ), and Lake Clark National Park and Preserve (LACL). The greatest invasive plant management efforts have been invested at KEFJ, where annual surveys and/or eradication efforts have occurred since 2004. KATM and LACL were surveyed in 2005, and KATM has had control work performed subsequently in 2007. The remaining SWAN units, Alagnak Wild River (ALAG) and Aniakchak National Monument and Preserve (ANIA), have not been surveyed.

Preliminary Results

Between 5-9 hectares of inventoried area in each park were found to have invasive species during a 2008 inventory (Rapp 2008).

Table showing invasive species in each SWAN park unit

Sampling Design and Objectives

SWAN does not currently monitor exotic plant populations.