Paleoecological reconstruction has been of pinnacle interest in recent years. Archived records stretching as far back as pre-industrial times allows for the observation of natural climate variability and anthropogenic influence. Montane regions are of considerable interest when it comes to research efforts, as these are vulnerable systems under much human and natural influence. How the magnitude and severity of disturbances affects the homogeneity and stabilization of forested landscape composition is still ambiguous. To determine how shifts in vegetation within a montane/subalpine ecotone influence fire activity, high-resolution data will be collected from the Gunnison region of Colorado at Lily Pond. Nearby site locations with archived paleoclimate data will determine climatic characteristics throughout a ~4000 year period. Archeological pollen and charcoal collection will provide insight to vegetational shifts and if corresponding fire activity occurs as a response. Ultimately, climatic anomalies are observed throughout the timeframe of interest proving to affect vegetation directly. Following forest composition alteration, positive and negative feedbacks in fire activity are detected. Variability in fire activity suggests fuel type is a strong influencer for such occurrences.
Climatic factors are the primary forces that drive tree species distribution within the Rocky Mountain forests of Colorado. Transition from a summer to a winter wet stage has significant influence on moisture availability during annual time frames affecting vegetation growth. Climate abnormalities such as the Medieval Climate Anomaly and Little Ice Age show consistency among western United States records to influence vegetation growth. The forces behind driving these climatic shifts can be attributed to expansive climate dynamics; El Niño Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO). These oscillating patterns typically occur on interannual and decadal timescales. The attributing precipitation patterns mirror latitudinal trade-offs of fluctuating temperatures between summer and winter months. Atmospheric waves produced by ENSO and PDO shift westerly storm trajectories latitudinally, creating a North-South dipole in North America . Even small directional changes that bias one direction of change, or the other, have the ability to produce considerably long-term differences among sites.
Climate holds the ability to indirectly influence fire occurrences in forested areas. The shaping of fuel composition and forest structure are both directly affected on annual through millennial time scales. The indirect pathway corresponding to the amount of fuel in an ecosystem mediates direct linkage to fire occurrences in response to climate. Unique vegetation adaptations have developed over time scales to aid in regeneration after high disturbance events. Following a stand destroying crown fire, Pinus contorta has the ability to establish itself from seed showing little evidence of successional change, ultimately leading to a dominate community over extended periods of time. Interestingly enough, Pinus contorta understory fires do not burn as severely as a brush fire within a Picea-Abies forest, due to lower fuel amounts. However, after such an event, Pinus contorta is the successional taxa. Picea engelmannii are known to have thin bark, which make them vulnerable to fire activity. Due to this physical characteristic, the ability to withstand large fire events is rare. It is also important to note that while Picea engelmannii have a lesser recruitment rate, they house greater longevity.
Therefore, presence of this taxa infers a mature forest with perhaps a longer fire-return interval. Conversely, Artemisia struggle, if are not incapable, to resprout following fire events and must rely on wind-blown seeds from neighboring areas to recolonize.MLily Pond is a unique ecosystem, as it is a closed water basin located at high elevations suitable for climate reconstruction efforts. Fire events are of considerable interest to researchers, park managers, and emergency responders. Tracking long-term trends within vegetation structure and composition in relation to shifting climatic events will aid in corresponding trends in fire severity and fire return intervals. Pollen and charcoal data collected from Lily Pond and climate reconstruction studies from nearby locations will allow for observation to answer pressing questions such as how does vegetation composition shift throughout the last ~4000 years? If there are distinct vegetation shifts are there also corresponding changes in fire activity (Fire return interval, fire severity, frequency, and magnitude) relative to these changes? And does a shift in forest composition ultimately increase biomass and consequently an increase in fire severity?
Lily Pond is located in central Colorado within the Gunnison National Forest, situated in the Taylor Park area, ~3200 meters in elevation (Figure 1). Formed near the end of the last ice age, Lily Pond is deemed a kettle lake. This body of water is situated uniquely within the transition zone of a subalpine/montane ecotone. Today, the surrounding vegetation is mostly dominated by Artemisia, (sagebrush), Picea engelmannii (Englemann Spruce), and Pinus contorta (Lodgepole pine).
Using a D-section corer, two sediment cores were collected from the southeast side of Lily Pond. Approximately two meters of each core were recovered upon extraction. Hand written data was recorded including observation of organic matter content, stratigraphy, and distinct transition in color and texture. Each core was wrapped in plastic wrap, placed in PVC piping, labeled accordingly, and transported back to the University of Colorado Denver Paleoecology and Palynology Lab upon investigation of further paleoecologic analyses. To quantify vegetation, charcoal, and ecosystem change over time scales, a reliable chronology was established (using how many samples? And sent where?). High-resolution radiocarbon dates on plant remains were taken, along with 210Pb dates on the upper 50 cm of the cores. From this, an age-depth model was constructed in R using Bacon reaching back to ~5127 cal. year B.P (Figure 2). Loss-on-ignition weights (1cm3) were measured along each centimeter and burned in an oven at 550°C for 2 hours to develop a general understanding of the ecosystem’s organic autochothonous production (Figure 3i).
Samples (1cm3) along intervals that are expected to show significant changes were taken (n=33). For pollen extraction, procedures were carried out using protocol from Faegri & Iversen (1989). Per each sample, pollen species (≥ 300 grains) were counted under 400x magnification. Past vegetation pollen percentages were compared to modern pollen data. A known amount of exotic tracers (Lycopodium spores) were added to each sample to aid in concentration calculations. Measure of forest canopy cover was calculated with arboreal pollen to non-arboreal pollen (AP/NAP) ratio (Figure 3h and Figure 4). In order to deduce accumulation rate for each individual pollen taxon, pollen concentrations were then divided by deposition time of sediment sample. Each core was subsampled at half-centimeter intervals. During the procedure, 1cm3 was removed for further processing and analysis. The subsamples were then soaked in 25 ml of 6% hydrogen peroxide and heated in a drying oven at 50°C for 24 hours, following methodology from Calder (2015).
Each of the samples were washed through a 125-µm sieve and deposited into a petri dish. Grid patterned petri dishes were used to ensure charcoal pieces were not counted more than once. The pieces collected are deemed “macroscopic charcoal” particles. All charcoal pieces were then counted under a microscope at 10-40x magnifications. Local fires in a given area will produce and deposit macroscopic pieces >60-µm and charcoal particles found to be >125-µm most likely is representative of a crown fire event. The CharAnalysis program was used to calculate additional statistical parameters (charcoal accumulation rate, peak magnitude, fire frequency, signal-to-noise index, and fire return interval. Charcoal concentration and sedimentation rate were multiplied together to determine accumulation rates into the system. Ultimately, measurement of accumulation peaks in charcoal aided in determining fire severity (amount of biomass burned). A smoothing window of 800 years was used to background levels of charcoal. To assess fire frequency, a weighted threshold of 95% is used; any event showing positive deviations is deemed a fire episode.
Results and discussion
Lily Pond and its local ecosystem response to climatic shifts and fire occurrence are observed in the following sections: Summer Wet (>5130—2400 B.P), Early Winter Wet (2400—1200 B.P), Medieval Climate Anomaly (1200—850 B.P), Little Ice Age (550—250 B.P), and Modern Age (250 B.P—Present). Forest structure and composition at Lily Pond offer a glimpse into the past ~4000 years and how forest characteristics relate to fire activity. Various fuel types within the ecosystem influence fire regimes and behavior. Therefore, the existing pollen record will allow for interpretation as to how fire activity persisted through time. Oxygen isotopic records retrieved from study sites across the southern Rocky Mountain region support similar climatic trends observed at Lily Pond. Starting at around ~3000 cal. year B.P, winter precipitation began to increase at Bison Lake located near Glenwood Springs, CO (Figure 3e) (Anderson, 2012).
Similarly, in southern New Mexico, 18O extracted from speleothems (Figure 3f) in Pink Panther Cave indicate the same precipitation shift (summer to winter) around ~2700 cal. year B.P. Both direct and indirect millennial time scale shifts from climatic variability in aridity and precipitation balance shift crown and stand-fire severity and fire return intervals. As suggested by Higuera (2014), the combination of a warmer, heavier precipitated summer and reduced snowfall may explain fire activity and subalpine forest expansion, which mirrors modern day. Climate is a direct driver for vegetation shift and indirect driver in fire activity. During the middle Holocene, these changes may have favored these events. The presence of warmer summers coupled with earlier snowmelt contributed to lesser fuel moisture. Due to this, changes in fuel would have been provoked, consequently increasing fire intensity (van Wagner, 1977).