Bot deptlecture20110310(1)

Arctic vegetation along two long bioclimatic transects in North America and Russia Prof. D.A. (Skip) Walker, Fulbright Scholar Alaska Geobotany Center, Institute of Arctic Biology, University of Alaska Fulbright Lecture Masaryk University, 10 March 2011

Overview of talk Introduction: Rationale for the two transects: The international Polar Year and the Greening of the Arctic IPY project. Overview of the two transects: The North America Arctic Transect and the Biocomplexity of Patterned Ground project. The Eurasia Arctic Transect the remote sensing objectives. Toward a synthesis of results from the two projects: Spatial and temporal patterns of circumpolar NDVI and biomass.

The International Polar Year (2007-2009) The International Polar Year is a large scientific programme focused on the Arctic and the Antarctic from March 2007 to March 2009, currently still in its synthesis phase. Organized through the International Council for Science (ICSU) and the World Meteorological Organization (WMO) The fourth polar year. Others were in 1882-3, 1932-3, and 1957-8.

IPY goals An intense scientific campaign to explore new frontiers in polar science. Improve our understanding of the critical role of the polar regions in global processes. Educate the public about the polar regions. Projects were expected to be interdisciplinary in scope; involve a pulse of activity during the IPY period; leave a legacy of infrastructure and data; expand international cooperation; engage the public in polar discovery; and help attract the next generation of scientists and engineers.

Research supported Understanding Environmental Change: Research that advances the understanding of the physical, geological, chemical, human, and biological drivers of environmental change at the poles, their relationship to the climate system, their impact on ecosystems, and their linkages to global processes. Human and Biotic Systems in Polar Regions: Research to address fundamental questions about social, behavioral, and/or natural systems that will increase our understanding of how humans and other organisms function in the extreme environments of the polar regions. 60 countries involved: Czech Republic was one!

Honeycomb chart of IPY projects Over 200 projects, with thousands of scientists from over 60 nations examining a wide range of physical, biological and social research topics at both poles. Greening of the ArctIc

Greening of the Arctic IPY intiative Documenting, mapping and understanding the rapid and dramatic changes to terrestrial vegetation expected across the circumpolar Arctic as a result of a changing climate. Comprised of four projects: North America Arctic Transect Land-cover and land-use changes on the Yamal Peninsula Synthesis of Arctic System Science: Greening of the Arctic Toolik-Arctic Geobotanical Atlas

Two transects through all 5 Arctic bioclimate subzones Bioclimate subzones Sub- Zone MJT Shrubs A 1-3 ˚C none B 3-5 ˚C prostrate dwarf-shrubs C 5-7 ˚C hemi-prostrate dwarf shrubs D 7-9 ˚C erect dwarf-shrubs E 9-12 ˚C low-shrubs Along the tundra bioclimate gradient : 10˚ C change in the MJT, 10-fold change in zonal biomass, 10-fold change in productivity, 5 to 10-fold change in vascular-plant diversity. CAVM Team 2003

Major goal of the Greening of the Arctic project: Link spatial and temporal trends of NDVI observed on AVHRR satellite images to ground observations along both transects. Climate Vegetation Soils Permafrost Spectral properties 5 N-factor Biomass Plant species cover NDVI and LAI Active layer depth Soil characterization Site characterizatiion Permafrost boreholes

Focus of this talk North America Arctic Transect Land-cover and land-use changes on the Yamal Peninsula

Trend in Arctic sea-ice Since 1980, perennial sea ice extent in the Arctic has declined at the rate of 10.1% per decade, and area trend is -11.4% decade. 2007 2007 minimum sea-ice extent, 4.5 million km2. NSIDC. Comiso et al. 2008, Geophysical Research Letters 35: L01703.

The Arctic tundra is a maritime biome 61% of the tundra is within 50 km of sea ice (blue buffer). 80% is within 100 km (magenta and blue buffers). 100% is within 350 km (all colors). Changes in the Arctic ocean sea ice will very likely affect terrestrial ecosystems. Map by Hilmar Maier. Walker, D. A., 2005. The Circumpolar Arctic Vegetation Map. Journal of Vegetation Science.

Normalized Difference Vegetation Index (NDVI): An index of greenness CAVM Team. 2003 In general, land cover with high reflectance in the NIR and low reflectance in the visible portion of the spectrum has dense green vegetation. Reflectance spectra of common ground-cover types Normalized Difference Vegetation Index NDVI = (NIR - VIS) / (NIR + VIS) NIR = spectral reflectance in the near-infrared band (0.7 - 1.1µm), where light scattering from the cell-structure of the leaves dominates. VIS = reflectance in the visible, chlorophyll-absorbing portion of the spectrum (0.4 to 0.7µm). Circumpolar patterns of NDVI

Time series of peak NDVI for northern Alaska (1981-2001) Jia et al. 2003 Geophysical Research Letters. 30: 2067. 17 ± 6% increase in peak NDVI from 1981-2001. Available biomass data indicate that this increase in NDVI corresponds to about a 150 g m-2 increase in biomass. Changes in NDVI show a long term increase and also some correspondence to yearly fluctuations in temperature. NDVI vs. Time in Bioclimate Subzones C, D, and E Temperature vs. Time in Subzones C, D, and E

Time series of peak NDVI anomalies in the tundra and boreal forest (1981-2005) Bunn et al. 2007. EOS. Northern high latitude ecosystems respond to climate change. 88: 333-335. 88% of the region is shows no significant trends in NDVI. 3% have decreasing trends, and 9% have increasing trends. Most of the positive changes are in tundra areas, particularly in North America. Forest areas are showing an overall decline in NDVI. Green: increasing NDVI Red: decreasing NDVI White: no trend

Skepticism regarding the NDVI trends “ Should we believe in the NDVI trend? There are no “ground truth” measurements of photosynthesis at northern high latitudes over the same period, and so the accuracy of the trend cannot be established unambiguously…. It will be a challenge for ecologists to explain how photosynthesis could possibly have increased by approximately 10% from 1981 to 1991.” (Inez Fung 1997) At the time of Inez Fung’s statement the causes of variation of NDVI within the Arctic as well as seasonality of NDVI had not been examined. The relative roles of interacting factors (e.g., climate, sea-ice distribution, elevation, glacial history, substrate chemistry, water regimes) need to be better understood. Is the current trend still partially a long-term response from the last glaciation? There are virtually no long-term biomass data to support or dispute the trends shown by NDVI.

There still is not a lot of direct evidence for change in Arctic vegetation. Photo – M. K. Raynolds Mostly experimental evidence; Green-house experiments (Chapin et al.) ITEX experiments One long-term biomass study of Shaver at Toolik Lake that is suggestive of change but inconclusive. Photo record of shrub cover change in northern AK (Matthew Sturm, Ken Tape, and Chuck Racine): Over 30% increase in alders on some stable valley slopes in the warmest parts of the Arctic. Dramatic increase in shrub cover on river terraces. This is changing. In the past year, there have been several major syntheses from international research projects that are examining change. And these will be published soon.

GOA studies are focused along two Arctic transects. CAVM Team. 2003 Transects in North America and Eurasia through all five bioclimate subzones as portrayed on the CAVM. Two Arctic Transects

NAAT was created during the Biocomplexity of Patterned-Ground project Howe Island, AK. Photo; D.A. Walker How do biological and physical processes interact to form small patterned- ground ecosystems? How do patterned-ground processes vary across the Arctic climate gradient? Walker et al. 2008, JGR

Why focus on these dumb spots?

WELL, BECAUSE: The processes involved in the formation of patterned-ground landscapes are not well understood. The importance of patterned ground with respect to biogeochemical cycling, carbon sequestration and other ecosystem processes is poorly known. They are an ideal natural system to to help predict the consequences of climate change of disturbed and undisturbed tundra across the full Arctic climate gradient.

IPY objectives of the North America Arctic Transect Create a legacy dataset of baseline information along the North American Arctic Transect (NAAT) that represents the full range of zonal vegetation types in the Arctic. Communicate the results of the studies through a three-part education/outreach component that includes an Arctic Field Ecology course, contributions to a new “Arctic Geobotanical Atlas” web site, and a field trip for the 9th International Conference on Permafrost. Biomass, leaf-area, spectral data and other site information were collected from each site to provide a baseline against which to monitor future changes.

North American Arctic Transect Dalton Highway (7 locations) Arctic Bioclimate Subzones Canada Sub-zone MJT (˚C) SWI (˚C mo) A <3 <6 B 3-5 6-9 C 5-7 9-12 D 7-9 12-20 E 9-12 20-35 Forest >12 >35

North American Arctic Transect Activities CALM Grids Active layer Vegetation Snow Climate /permafrost Met station Soil temperatures Frost heave Soils Characterization Nitrogen mineralization Decomposition Vegetation Classification Biomass Mapping Remote sensing NDVI Mapping Modeling Education Isachsen Grid, Subzone A Photo D.A. Walker

Biocomplexity grids Subzone A: Satellite Bay, Canada - 1 Isachsen, Canada - 3 planned Subzone B: Mould Bay, Canada - 2 Subzone C: Howe Island, Alaska - 1 West Dock, Alaska - 1 Green Cabin, Canada - 3 Happy Valley Grid Subzone D: Deadhorse, Alaska - 1 Franklin Bluffs, Alaska - 3 Sagwon MNT, Alaska- 2 Ambarchik, Russian - 1 Subzone E: Sagwon MAT, Alaska - 1 Happy Valley, Alaska - 3 Kurishka, Russia - 1 TOTAL 20 + (3 planned) = 23

Trend in patterned-ground along the Arctic bioclimate gradient Sub- zone: A B C D E Drawings modified from Chernov and Matveyeva 1997 Small non-sorted polygons Non-sorted circles Earth hummocks

Subzone A Isachsen, Ellef Ringnes Island, mean July temperature = 3 ˚C, SWI = 4 ˚C mo

Subzone C Howe Island, Ak and Green Cabin, Banks Island, MJT, 8 ˚C, SWI = 16 ˚C mo

Subzone E Tuktuyaktuk, NWT, Happy Valley, AK, MJT = 12 ˚C, SWI = 30 ˚C mo

The roles of vegetation in patterned formation Plant cover: Insulates the surface decreasing the heat flux and summer soil temperatures. stabilizes cryoturbation and limits needle-ice formation. Promotes nitrogen and carbon inputs to the soil. Bill Steere collecting Bryum wrightii on a frost boil at Prudhoe Bay, July, 1971. N, Matveyeva - Map and drawing of frost boil vegeation on the Taimyr Peninsula, Russia.

Martha Raynolds Anja Kade Vegetation mapping and analysis of of active-layer/heave/vegetation relationships

Maps of vegetation, biomass, active layer and snow depth Raynolds et al. 2008 JGR -Biogeosciences Maps are from representative zonal sites in each bioclimate subzone.

Trends in biomass, snow depth, and thaw on zonal grids Subzone Biomass Active Layer Snow depth Active Layer Subzone Both biomass and snow depth depth increase toward the south and both have major effects on soil temperature regimes. Raynolds et al. 2008, JGR-Biogeosciences

NDVI along the NAAT 2-fold increase of the NDVI on zonal surfaces. 6-fold difference in NDVI on PGFs. 2-fold difference in NDVI between PGFs. Random hand-held NDVI on zonal surfaces along the temperature gradient (no attention to PGFs) NDVI of patterned-ground features and nearby tundra along the gradient Courtesy of Howie Epstein and Alexia Kelley I A I B I C I D I E I

Classification of patterned-ground vegetation along the NAAT Used the Braun-Blanquet appraoch. Low Arctic: Kade, A., Walker, D.A., and Raynolds, M.K., 2005, Plant communities and soils in cryoturbated tundra along a bioclimate gradient in the Low Arctic, Alaska: Phytocoenologia, v. 35, p. 761-820. High Arctic: Vonlanthen, C.M., Walker, D.A., Raynolds, M.K., Kade, A., Kuss, H.P., Daniëls, F.J.A., and Matveyeva, N.V., 2008, Patterned-ground plant communities along a bioclimate gradient in the High Arctic, Canada: Phytocoenologia, v. 38, p. 23-63. Plant communities Soil and site data

Plant species and cover information for each plant community Kade et al. 2005, Plant communities and soils in cryoturbated tundra along a bioclimate gradient in the Low Arctic, Alaska. Phytocoenologia , 35: 761-820. Plant community table (cover) Classification according to Braun-Blanquet approach

Frost-boil plant communities, soil and site information Kade et al. 2005, Plant communities and soils in cryoturbated tundra along a bioclimate gradient in the Low Arctic, Alaska. Phytocoenologia , 35: 761-820. Plant communities Soil and site data

Landscapes in the Low Arctic examined by Anja Kade et al. (2005) Photos by A. Kade Howe Island Subzone C Deadhorse Subzone D Happy Valley Subzone E

Summary of plant communities for the Low Arctic part of the transect Kade et al. (2005) Of these, Anja described 5 communities occurring on patterned-ground features and 4 zonal communities occurring between patterned-ground features or on sites without patterned ground.

Somewhat different approach than previous studies Matveyeva (1998) considered the patterned-ground features and areas between patterned-ground features as part of the same associations. We felt that this violates the site homogeneity requirement of the Braun-Blanquet approach. Besides, we needed to examine the plant community and site characteristics of the patterned-ground features separately to determine biomass, the effects on the microclimate, thermal properties of the soils, and effects on the soils, and permafrost regimes.

Contrast in vegetation on and between frost features Kade et al 2005 Deadhorse Subzone C Braya purpurascens-Puccinellia angustata community Dryas integrifolia-Salix arctica community Nonsorted Circle Between Circles

Contrast in vegetation on and between frost features Kade et al 2005 Franklin Bluffs Subzone D Junco biglumis-Dryadetum integrifoliae typicum subass. nov. Dryado integrifoliae-Caricetum bigelowii Walker et al. 1994 Nonsorted Circle Between Circles

Contrast in vegetation on and between frost features Kade et al 2005 Happy Valley Subzone E Nonsorted Circle Between Circles Cladino-Vaccinietum vitis-idaeae ass. nov. Sphagno-Eriophoretum vaginati Walker et al. 1994

DCA ordination and biplot of Low Arctic vegetation plots Subzone C Subzone E ∂ Subzone D ∂ ∂ ∂ Patterned-ground features Between pgf Kade et al 2005

Species richness and cover Kade et al 2005 Species richness peaks in subzone D in both micro-habitats (155 species in zonal sites and 165 on the pattterned ground features). Cover values were greatest between the patterned ground features except in subzone E.

Selected soil physical properties

Selected soil chemical properties for

Ordination of zonal patterned ground vegetation: controlling environmental gradients along the full gradient NMDS ordination. Clear gradient of vegetation response to cryoturbation within each subzone and clear floristic separation between subzones. But no clear overall controlling factors for the whole data set. Floristic separation between Alaska and Canada portions of the gradient due to different floristic provinces, and substrate differences. Patterned-ground features Between patterned-ground features Intermediate Walker, D.A., Kuss, P., et al., 2011 (in revision), Vegetation and patterned-ground relationships along the Arctic bioclimate gradient in North America Applied Vegetation Science.

Four models of patterned ground formation Coupled Water-Ice-Temperature / ArcVeg model, Daanen et al. 2008 JGR Conceptual model of frost-boil and hummock formation, Shur et al. 2008, NICOP proceedings Differential frost-heave model Peterson and Krantz, JGR 2008 Thermo-mechanical model, Nicolsky et al., JGR 2008 Each provides a unique perspective.

Conclusions from the NAAT: Still a work in progress The structure of the vegetation affects patterned-ground morphology on zonal sites in predictable ways that vary with differences in climate, soil-moistur and soil-texture. Contrasts in the vegetation on and between patterned-ground features is best developed in Subzones C and D. These differences drive the movement of heat and water and the development of frost heave. New and improved models are needed to better understand the interactions the physical and biological processes in patterned-gound formation. More replication is needed and more sampling in different habitats, e.g. wetlands.

Synthesis of information from the North American Arctic Transect Synthesis of the project published in special issue of Journal of Geophysical Research - Biogeosciences: 9 papers from the North American Arctic Transect. 5 papers from the Thule, Greenland biocomplexity project (Welker et al). 21 collaborators. Walker, D. A. et al. 2008. Arctic patterned-ground ecosystems: a synthesis of field studies and models along a North American Arctic Transect. Journal of Geophysical Research - Biogeosciences 113:G03S01, doi10.1029/2007JG000504.

Expeditions and Timeline: Dalton Highway: 2001-2002 Green Cabin: 2003 Mould Bay: 2004 Isachsen: 2005 Synthesis: 2006-2008 2001-02 2003 2004 2005

Field Camp at Green Cabin, Banks Island

The Eurasia Arctic Transect: Vegetation Analysis and Mapping D.A. Walker, H.E. Epstein, H. A. Maier, G.V. Frost, M.K. Raynolds, U.S. Bhatt, J. Comiso, R. Daanen, D.S. Drozdov, B. Forbes, A.A. Gubarkov, G. Jia, E. Kaarlejarvi, O. Khitun, A.V. Khomutov, P. Kuss, M.O. Leibman, G. Matyshak, N.G. Moskalenko, P. Orekhov, J.E. Pinzon, V.E. Romanovsky, C.J. Tucker, N.G. Ukraintseva, Q. Yu Yamal Peninsula, Russia. Photo: D.A. Walker

Overview of EAT effort Hierarchical mapping analysis AVHRR (1 km 12.5 km and 50 km) Landsat ETM+ (15 m and 30 m) GeoEye(40-cm resolution) Hand-held measurements of NDVI Ground observations along the climate gradient Data collected and data report Vegetation analysis Analysis of the linkages between climate change, sea-ice retreat and changes in the terrestrial vegetation.

The Eurasia Arctic Transect About 2000 km from 65˚ 19’ N to 80˚ 38’. Subzone A: Krenkel, Franz Josef Land Subzone B: Ostrov Belyy Subzone C: Kharasavey Subzone D: Vaskiny Dachi Subzone E: Laborovaya Forest-tundra transition: Nadym and Kharp Four expeditions (2007-10).

Much of project focuses on greenness patterns and change using the Normalized Difference Vegetation Index (NDVI) Chlorophyll absorbs red light for photosynthesis and reflects near infrared light. NDVI = (NIR-R)/(NIR + R) . The difference between the reflectance in the NIR and R portions of the spectrum is a measure of the photosynthetic capacity of the surface. The difference is divided by the sum of the reflectances to adjust for variations in the index due to slope and shadows. NDVI is much greater in vegetation with high chlorophyll content. Plants absorb red light and reflect NIR radiation.

A hierarchical approach to examining greenness patterns and change Plant- to plot-scale: Ground measurements of 5 x 5 m plots. Quickbird 60-cm pan-sharpened pixel size. Landscape- to Regional-scale: Yamal 1-km AVHRR data from CAVM. Landsat ETM 30-m pixel size (USGS GLS 1990). Global scale: Global 12.5 km data: NDVI (Pinzon). Global sea-ice, land-temperature and NDVI data: 25-km pixels based on Comiso sea-ice and temperature and cubic convolution of 8-km GIMMS NDVI.

Multiple-scale analysis of Yamal NDVI: 1-km AVHRR NDVI derived from CAVM data set USGS data set used for the CAVM Circumpolar MaxNDVI Calibrated NDVI to biomass using the zonal biomass values from the Yamal. More biomass information is needed from shrublands and cryptogamic tundra areas. Courtesy of M.K. Raynolds. 2010.

Analysis of 1-km NDVI with Landschaft and CAVM map units Loamy uplands have higher NDVI than sandy uplands. Landschaft does not delineate some known sandy areas (e.g. O. Belyy). Broad river channels have highest NDVI despite large amount of lakes in the valleys. 1-km data is not fine enough to resolve the greening patterns within the highly eroded upland areas. Courtesy of M.K. Raynolds. 2010. NDVI with Landschaft boundaries NDVI on marine terrace uplands and drainages of differing soil texture

Enhanced TM (15-m) derived maps of Ostrov Belyy Maier and Walker. 2010. Poster at 2 nd Yamal LCLUC Workshop False-color image 15-m resolution panchromatic band is used to enhance the 30-m resolution TM data. Single ETM+ scene covers all of Ostrov Belyy. Unsupervised classification used 15 spectral clusters. Salt marshes classified separately NDVI map shows clear relationship of productivity with respect soil moisture (predominantly moist loamy soils in the north vs. dry sandy soils in the south). Land-cover map NDVI map

Land-cover mapping with 30-m Landsat TM data Maier and Walker. 2010. Poster at 2nd Yamal LCLUC Workshop GLS-1990 mosaic Land-cover maps of Yamal study locations Landsat mosaic provides intermediate-resolution terrain information of the whole peninsula. Mosaic is composed of many scenes with different acquisition dates (May to September). Difficult to get consistent land-cover classification or MaxNDVI for the whole peninsula. Land-cover maps produced separately for each LCLUC location. May be possible to get consistent classification for whole area by combining all decadal and mid-decadal mosaics to get one coverage displaying MaxNDVI for all pixels.

1. Extensive nutrient-poor surface sands with lichens that are easily overgrazed by reindeer. 2. Underlain by permafrost with massive pure ice. 3. Extensive landslides are rapidly eroding the landscape. 4. This exposes salt-rich and nutrient-rich clays. 5. Complex vegetation succession process that results in willow-shrub tundra and much greener vegetation in the eroded valleys. High-ice Permafrost Landscapes Photos: D.A. Walker and M. Liebman (upper right)

Extensive azonal willow shrublands due to landslide disturbances Greening associated with eroded sandy uplands

Landslides and cryogenic erosion Photos D.A. Walker Strong greening on landslide slopes cover extensive areas of the Yamal. Large effect on patterns of greenness in many areas. Need temporal series of high-resolution satellite images and/or photos in landslide areas to assess the rate of change. Key: A – stable areas B – shear surface C – landslide body Before landslides After landslides Low-willow shrublands develop on landslides during 200-yr succession, greatly changing biomass and NDVI. Ukraintseva and Leibman et al. 2000, 2007, 2008 Biomass 1 – young landslide 2 – old landslide 3 – very old landslide

Snow, hydrology, NDVI relationships in landslide areas RESURS-01 image, spring, unknown date Expansion of the major river valleys and mosaic of smaller drainages is occurring very rapidly. Need better understanding of the rate of growth of the drainage networks. Need models and high-resolution imagery to address expansion of drainage networks

New High Resolultion GeoEye image of Vaskiny Dachi area clearly shows erosion patterns Vaskiny Dachi 0.41 meter resolution!

Erosional patterns are clearly discernable on GeoEye scene. Vaskiny Dachi Should permit analysis of rate of erosion and related greening.

Other factors affecting greenness patterns: (1) 300,000 reindeer on the Yamal Photos: Bruce Forbes. Overgrazing Trampling Grassification Wind erosion Effects on reindeer on NDVI are unknown at present because of lack of control areas to study the effects (exclosures). Potential major effect in sandy areas.

Nenets camp on Yamal in Salix low shrub tundra Reindeer grazing Salix thickets in Nenets Okrug. If they grow over ≈ 2 m high, herders can lose sight of animals. The changes in willow growth are affecting reindeer management. Forbes et al. 2009 PNAS , ENSINOR project Photos courtesy of Bruce Forbes.

Other factors affecting greening patterns: (2) Impacts of gas development Timo Kumpula: Yamal LCLUC Workshop, Moscow, 28-30 Jan 2008 . Locally important but still relatively small extent. Need development scenario models to help predict and plan for expansion of road networks.

Analysis of biomass and NDVI trends across the climate gradient NDVI & LAI Soils Field data collected: Plant Biomass Ground temperatures Active layer Plant Cover

2010 Expedition to Hayes Island, Franz Josef Land Ground-based observations in Bioclimate Subzone A of the Eurasia Arctic Transect. Northern-most permafrost borehole in Russia at 80˚ 37’ N. Completed parallel transect studies in North America and Eurasia.

Typical layout of transects and plots at each site Five 50-m transects Five 5 x 5-m plots (relevés) Biomass harvests in each plot (x) iButtons for n-factor in corner of each plot (•) Soil pit in SW corner Soil pit

EAT data reports Descriptions of each study location General description of the region and study sites Physiography and geology Climate summary Transect data Plant species cover LAI NDVI Thaw depth Photos of transects Relevé data Cover abundance of plant species Soil chemical and physical data Site factors Biomass by plant functional type Photos of relevés Soil pits Descriptions Photos

Spatial patterns of biomass and NDVI: EAT transect biomass Zonal sites show little variation across the peninsula, except in subzone E. Biomass values for zonal site in subzone E is close to tussock tundra values for Alaska (≈ 750 g m^2). Total live biomass values of zonal loamy sites in subzones C and D are close published values for mesic tundra Barrow and Prudhoe Bay (≈ 450 g m^2). Biomass (G m^2) Total live and dead above-ground biomass including trees Total live above-ground biomass Zonal loamy sites Subzone FT E D C B Understory only Zonal loamy sites

Vegetation Analysis: NMS Ordination of tundra study plots based on floristic similarity Colors according to CAVM bioclimate subzones: Blue : Subzone B Green: Subzone C Olive: Subzone D Red : Subzone E Shape of symbols represent soil texture: Sandy Loamy JJ Frost et al. 2010, Yamal LCLUC Workshop

Ordination with biplot arrows showing environmental relationships Biplot arrows show direction and strength of correlations for each measured environmental variable. R^2 cutoff = 0.25. Variables pointing in the horizontal direction are correlated with percent silt: (+): soil nutrients, soil moisture (-): sand, thaw depth. Variables pointing in the vertical direction are more strongly correlated with latitude : (-) Summer warmth, microrelief X and Y axes are interpreted as complex environmental gradients with numerous variables covarying along each axis: X axis: soil texture/ pH/ moisture gradient Y axis: latitude/ climate/ microrelief gradient JJ Frost et al. 2010, Yamal LCLUC Workshop

Ordination space interpreted as an environmental space with clear bioclimate and soil texture gradients First interpretation Subzones B to E: Horizonal axis: Sand to Loams Vertical axis: Warm to cold JJ Frost et al. 2010, Yamal LCLUC Workshop USDA Soil Texture Triangle Subzone E Subzone D Subzone C Subzone B Loamy Sandy

Ordination with bipolot plant-variables and NDVI correlations Plant variables correlated with complex soil texture gradient Pleurocarpous mosses (+) Graminoid cover (+) Crustose lichens (-) Moss thickness Variables correlated with the latitude (-temperature) gradient Bare soil cover (+) Liverwort cover (+) Shrub biomass (-) Evergreen shrubs (-) LAI (-) NDVI and plant production shows strong relationships to both gradients. JJ Frost et al. 2010, Yamal LCLUC Workshop NDVI

Significant relationships between hand-held NDVI measurements and other variables Strongest NDVI relationships are with some plant cover variables (e.g. pct. cover of evergreen shrubs, acrocarpous mosses, deciduous shrubs, LAI, moss height, organic thickness). Also strong environmental correlations (e.g. vol. soil moisture, pct. sand, latitude, SWI) mostly related to soil moisture and warmth gradients. Key variables

Toward a synthesis of the two transects Although the research along the two transects had different objectives. There is a common primary data set from both transects: Vegetation, soils, and site factors from zonal vegetation along the complete Arctic bioclimate gradient. Ground measurements of key plant productivity variables: biomass, LAI, and NDVI. A circumpolar remote-sensing data set that contains vegetation, land temperatures, and NDVI data for both transects and changes in NDVI since 1982. This allows us to compare the spatial and temporal trends in vegetation, biomass, and NDVI between the two transects in response to ongoing changes in climate and land-use.

Field studies along two 1800-km Arctic transects USGS 1-km AVHRR data set used for the CAVM. North America Arctic Transect: 2002-2006 Biocomplexity of Arctic Patterned Ground Ecosystems Project (NSF). Eurasian Arctic Transect: 2007-2010, Greening of Arctic (NASA). Both transects through all five Arctic bioclimate subzones.

Zonal vegetation along both transects Eurasia Transect A - Hayes Island B - Ostrov Belyy C – Kharasavey D - Vaskiny Dachi E - Laborovaya North America transect A - Isachsen B- Mould Bay C - Green Cabin D - Sagwon MNT E - Happy Valley

Summer Warmth Index Vegetation (CAVM Team 2003) Variation in summer land temperature and vegetation along the transects

Differences between the transects: EAT has less biomass in subzone A (Wetter, much colder). EAT has more biomass in subzone C, (Wetter, unglaciated, more nutrients?) NAAT has much more biomass in subzone E (grazing effect?). Plot-level biomass trends along EAT and NAAT NAAT EAT

Fewer evergreen shrubs and lichens along the EAT especially in subzones D and E. (Reindeer?) NAAT EAT Differences between EAT and NAAT

Comparison of EAT and NAAT Hand-held NDVI vs. biomass, and LAI For equivalent amounts of biomass and LAI, the HH-NDVI readings were much higher along the EAT.

Comparison of EAT and NAAT Leaf Area Index vs. Biomass An equivalent amount of biomass has consistently much higher LAI values along the NAAT than along the EAT and the difference increases at higher biomass values. Reflects the different structure of the vegetation along the two transects. Higher proportion of the total biomass is non-green along the NAAT (more wood, taller plants?).

Comparison of EAT and NAAT 1-km AVHRR NDVI & biomass, vs. summer warmth index Greater biomass and NDVI for equivalent temperature along the EAT compared to NAAT. Biomass values are landscape-level averages for zonal landscapes. EAT is greener in equivalent summer climates.

Comparison of EAT and NAAT: 1-km AVHRR NDVI and zonal landscape-scale biomass Very strong correlation between AVHRR NDVI and biomass along both transects and for combined data set.

Use of ground data to calibrate regional AVHRR-NDVI /biomass data

Circumpolar aboveground biomass derived from NDVI Raynolds et al. 2011 submitted, Geophysical Research Letters

Temporal analysis of NDVI change with respect to summer sea ice and land temperatures Arctic Tundra Vegetation March Sea-Ice Extent Sea Ice: http://www.arctic.noaa.gov/reportcard/figures/seaice2009fig1.jpg Vegetation and NDVI: http://www.arcticatlas.org/maps/themes/cp/cpvg 80% of Arctic tundra is within 100 km of ice-covered seas (100% is within 350 km). Models have shown that melting the sea ice will affect land temperatures and permafrost even at great distances from the Arctic Ocean. Changes is summer sea-ice distribution should affect land temperatures and the productivity of tundra. Max NDVI

Study Framework: Division of Arctic Ocean and associated land masses Uma Bhatt, D.A. Walker, M.K. Raynolds, J. Comiso, H.E. Epstein, G.J., Jia, J. Pinzon, and C.J. Tucker, 2010, Earth Interactions. Russian Arctic Atlas for seas. CAVM Florist provinces for land masses. Analysis of 50-km buffers seaward and landward along each sea coast and also for entire non-alpine tundra area.

Changes in summer land-surface temperature (1982-2010)

Temporal patterns of NDVI in relationship to changes in area of summer open water Walker et al. 2011. Bulletin American Meteorological Society. State of the Climate . New analysis based on new GIMMS 3g AVHRR NDVI data by Pinzon et al. 2010 (in progress) and sea ice data by Comiso et al. 2010. Pct. Change MaxNDVI (1982-2010) 26% increase +4% increase Beaufort Kara Beaufort Kara Oceans: % change of May-Aug open water (1982-2010) Tundra land areas: % change in TI-NDVI (1982-2010) From Bhatt et al. 2010, AGU Fall Meeting

Percentage change (based on least squares fit method) in coastal open water, land temperatures, and NDVI and significance of trends In general, areas of enhanced NDVI patterns are corresponding to areas of warmer land temperatures. The connections between land warming and more open water are less clear in 2010 than they were previously. Cooling in the Kara region despite very large increases in open coastal water (more fog, more winter precip. Shorter growing seasson?) Walker et al. 2011, BAMS State of the Climate, in prep.

This study has shown: The feasibility of studying and monitoring zonal landscape-level biomass and NDVI across the full Arctic bioclimate gradient at a circumpolar scale. Broad similarities in biomass between North America and Eurasia along the Arctic temperature gradient, but also major differences related to different disturbance regimes, geology, and precipitation patterns. Very good correlation between AVHRR NDVI and zonal landscape-level biomass. Strong linkages between change in NDVI and summer land temperatures, but weaker linkages with coastal sea-ice concentrations. Good start, but more replication is needed for ground observations in all subzones along both gradients. And more long-term studies of biomass trends are needed.

Collaborations Institutions: University of Alaska University of Virginia Earth Cryosphere Institute (RAS), Arctic Centre, Rovaniemi 2 Funding U.S.: NSF, NASA Finland: ENSINOR Russian Academy of Science Members of 2010 Expedition to Hayes Island Děkuji!

Bot deptlecture20110310(1)

More Related Content

What's hot

Similar to Bot deptlecture20110310(1)

Bot deptlecture20110310(1)

Editor's Notes