Longterm ecological research LTER the challenge of converting

Long-term ecological research (LTER): the challenge of converting long term monitoring into science Mark Williams, University of Colorado

Outline The need for long-term research Niwot Ridge LTER examples LTER program overview Recommendations for developing long-term research programs

Duration of all observational and experimental studies N = 623 Eighty percent of studies in the ecological literature last less than three years From Tilman, D. 1989. Ecological experimentation: strengths and conceptual problems. pp. 136 -157. In Likens, G. E. (ed). Long-Term Studies in Ecology. Springer-Verlag, New York.

Only 10 percent of studies capture unusual events Variable Unusual events reset systems. Short-term studies initiated before and after a rare event are viewing different system states. Time (yrs)

High-elevation areas are important bellwethers of global change: we need long-term research

Glaciate dvalley SADDLE

External Drivers: Temperature Increasing air temperature since early 1980’s Summer air temps warming fastest Earlier lake ice-out dates 5ºC increase in 25 years

External Drivers: Precipitation Greater precipitation with increasing elevation Increases in the winter months (more snow) Summer drought starting in 2000

External Drivers: N deposition • Increased rates of N deposition (wetfall) • N loading increases, despite drought

ARIKAREE GLACIER ROCK GLACIER D 1 CLIMATE GREEN LAKE 4 3, 600 m

Earlier Snowmelt 2 -6 days decade-1

Arikaree glacier is dying Drought Tipping point Arikaree Glacier: Mass balance (Bn), cm water equiv

N dep + warming T = N saturation N Critical load: Aquatic 4 Kg N/ ha/ yr Annual VWM concentrations of nitrate increase at all stream sites

Stoichiometric controls on Ncycling Scatterplot of NO 3 - vs. DOC: NO 3 - ratio for eight sites in Green Lakes Valley.

Chytrids Microbes in barren soils NWT similar to Himalayas Freeman et al. 2009

Saddle site

Saddle Grid 88 points every 50 meters -biweekly snow depth -annual NPP -annual species composition and coverage

Results: snow-based variables and vegetation changes 90 -97 Wilcoxon test Dir P-val Cover (%) Snowbed Moist meadows Sibbaldia procumbens NS 0. 109 Deschampsia cespitosa NS 0. 663 Geum rossii Kobresia Dry meadows myosuroides Fellfields Selaginella densa 97 -06 Trend Sen slope Snow period Wilcoxon test Dir Longer P-val Cover (%) Increase Decrease 0. 000 Trend Sen slope 0. 006 0. 009 0. 000 Snowmelt / Snowperiod Earlier / shorter Decrease 0. 000 Longer Increase 0. 000 Earlier / shorter Decrease 0. 025 Longer NS 0. 383 Earlier / shorter Randin talk earlier

how much N input does it take to produce a change in species composition? (= N critical load using biotic response) Addressed experimentally in species rich dry meadow, using additions of 2, 4, 6 g N/m 2/yr

species composition response: treatment x year P < 0. 01 Carex rupestris Treatment specific rate of change in cover similar response for Trisetum spicatum

Empirical estimation of N critical load for plant species responses in alpine dry meadows N Critical load: 10 -40 Kg N/ ha/ yr Whole community response

Woody Willow Encroachment

Feedbacks between willow encroachment and snow, temperature, and N Adapted from Sturm et al. 2001, Journal of Climate

Willow Encroachment Experiment Factorial manipulation of N (fertilizer), snow (snowfence), summer temperature (opentopped chambers), in all possible combinations + Addition of Salix glauca seedlings

Feedbacks to ecosystem processes (mg N/g soil/30 days) increases litter, N availability, and snow depth. These effects may accelerate encroachment Net N mineralization Willow encroachment Absent Present Willows

Advantages of long-term research • • Slow processes or transients Episodic or infrequent events Trends Multi-factor responses Processes with major time lags Leverage of experiments with long-term data Sites become research platforms – Attract other research projects/funding