Episodic Recruitment and Climate Analysis of Ponderosa Pine on the Palmer Divide, Eastern Colorado

University of Denver Digital Commons @ DU

Electronic Theses and Dissertations Graduate Studies

11-1-2012

William Henry Brenton Jr. University of Denver

Follow this and additional works at: https://digitalcommons.du.edu/etd

Part of the Climate Commons, and the Geography Commons

Recommended Citation Brenton, William Henry Jr., "Episodic Recruitment and Climate Analysis of Ponderosa Pine on the Palmer Divide, Eastern Colorado" (2012). Electronic Theses and Dissertations. 85. https://digitalcommons.du.edu/etd/85

This Thesis is brought to you for free and open access by the Graduate Studies at Digital Commons @ DU. It has been accepted for inclusion in Electronic Theses and Dissertations by an authorized administrator of Digital Commons @ DU. For more information, please contact [email protected],[email protected].

EPISODIC RECRUITMENT AND CLIMATE ANALYSIS OF PONDEROSA PINE ON THE PALMER DIVIDE, EASTERN COLORADO

______

A Thesis

Presented to

the Faculty of Natural Sciences and Mathematics

University of Denver

______

In Partial Fulfillment

Requirements for the Degree

Master of Arts

______

William Henry Brenton, Jr.

November 2012 Advisor: Donald G. Sullivan

Author: William H. Brenton, Jr. Title: EPISODIC RECRUITMENT AND CLIMATE ANALYSIS OF PONDEROSA PINE ON THE PALMER DIVIDE, EASTERN COLORADO Advisor: Donald G. Sullivan Degree Date: November 2012

Abstract

Previous recruitment studies on populations of ponderosa pine (Pinus ponderosa

P. & C. Lawson) in Colorado have been limited to the western mountains and the Front

Range. In this research I used tree-ring data to reconstruct recruitment for ponderosa pine near the eastern limits of its distribution at two sites on the Palmer Divide, Eastern

Colorado, to determine the relative climate sensitivity of the two sites, and the extent to which climate or other factors may have influenced recruitment at the sites. The results of the tree-ring analysis suggest that ponderosa pines in more the easterly site lower elevation population are more sensitive to climatic factors than are in the more westerly site. Both higher elevation populations showed a history of episodic recruitment that only partially reflects local climatic conditions. Climate is probably the most important factor in recruitment pulses, but only if there is opportunity within the stand for recruitment. Although favorable climate patterns were present at both sites at the same time, the opportunity for recruitment within each stand was different at separate times.

While climate conditions must be favorable for seedling establishment and growth, conditions of stability and competition within the stand dictate the ultimate recruitment success.

Key Words: Dendroecology, Episodic Recruitment, Climate, and Palmer Divide

ii Acknowledgements

This thesis would not have been possible unless for the assistance of many people.

I would especially like to thank my thesis advisor Donald Sullivan for his patience, guidance, critiques, and mentoring. I would like to thank the other members of my committee, Hillary Hamann, Michael Daniels, and Martin Quigley for their suggestions, making my project better. John Sakulich shared his time and tree-ring expertise and this thesis benefitted greatly from his advice. Charles Guetz for allowed me access to his land for my eastern study site. Douglas County park ranger Scott McEldowney allowed me access to Spruce Mountain, my western study site. My mother for funding this project. I give thanks to my Undergraduate advisor James Doerner for introducing me to Donald

Sullivan and to the University of Denver. My wife Angela and our sons William and

Emory deserve special thanks for their love and patience.

iii Table of Contents

Chapter One: Introduction……………………………………………………………….1 Literature Review………………………………………………………………..2 Research Questions…………………………………………………………...…4 Hypotheses………………………………………………………………………4 Palmer Divide.………………………………………………………………..…5 Topography…………………………………………………………………...…5 Climate…………………………………………………………………………..6 Vegetation………………………………………………………………..……...6 Chapter Two: Methods…………………………………………………………………..8 Site Selection…………………………………………………………………….8 Field Methods……………………………………………………………………9 Laboratory Methods………………………………………………………..……9 Climate Analysis………………………………………………………………..11 Episodic Recruitment Analysis…………………………………………………12 Chapter Three: Results…………………………………………………………………14 COFECHA Results…………………………………………………………….14 ARSTAN Results………………………………………………………………15 Episodic Recruitment Results………………………………………………….16 Climate Results………………………………………………………………...17 Chapter Four: Discussion………………………………………………………………18 Climate Sensitivity……………………………………………………………..18 Recruitment…………………………………………………………………….19 Potential Alternative Hypotheses………………………………………………22 Logging………………………………………………………………………...22 Cattle Grazing………………………………………………………………….23 Fire……………………………………………………………………………..23 Chapter Five: Conclusions…………………………………………………………..…25 Bibliography……………………………………………………………………………27 Figures………………………………………………………………………………….32 Tables…………………………………………………………………………………...44 Appendices……………………………………………………………………………...47 Appendix A…………………………………………………………………….47 Appendix B…………………………………………………………………….51

iv List of Figures Figure 1……………………………………………………………………………….33 Figure 2……………………………………………………………………………….34 Figure 3…………………………………………………………………………….…35 Figure 4……………………………………………………………………………….36 Figure 5…………………………………………………………………………….…36 Figure 6…………………………………………………………………………….…37 Figure 7……………………………………………………………………………….37 Figure 8…………………………………………………………………………….…38 Figure 9……………………………………………………………………………….38 Figure 10………………………………………………………………………………39 Figure 11………………………………………………………………………………39 Figure 12………………………………………………………………………………40 Figure 13………………………………………………………………………………41 Figure 14………………………………………………………………………………41 Figure 15………………………………………………………………………………42 Figure 16………………………………………………………………………………42 Figure 17………………………………………………………………………………43 Figure 18………………………………………………………………………………43 Figure 19………………………………………………………………………………44 Figure 20………………………………………………………………………………44

CHAPTER ONE: INTRODUCTION Wildfire history and climate reconstruction for the southwestern United States have been extensively studied using tree-rings as a proxy (Grissino-Mayer and Swetnam

2000, Esper et al. 2002, Westerling et al. 2006, Kitzberger et al. 2007). Numerous dendrochronology studies have examined a multitude of subjects in the Colorado Front

Range including wildfire history (Veblen et al. 2000, Kaufmann et al. 2003, Baker et al.

2007), stream flow (Woodhouse 2001), age structure (Veblen 1986), and climate reconstruction (Villalba et al. 1994, Woodhouse and Overpeck 1998). Several researchers have studied recruitment of ponderosa pine in the Western United States

(Allen et al. 2002, Keyes et al. 2007, Peters and Sala 2008, Mooney et al. 2011), particularly in Colorado (Brown and Wu 2005), and on the Colorado Front Range

(Linhart and Mitton 1985, Mast et al. 1997, Mast et al. 1998, League and Veblen 2006,

Baker et al. 2007). Woodhouse and Brown (2001) and Woodhouse et al. (2002) examined old and new tree-ring chronologies to reconstruct climate, looking at drought trends as part of a larger study of the Western Great Plains.

In spite of these studies, little is known about the episodic recruitment and climate response of ponderosa pine near the eastern limits of its distribution on the Palmer Divide in Colorado (Fig.1). Previous research involving tree recruitment in Colorado has been limited to areas that are either in the mountains or along the Front Range. None of these studies have examined tree recruitment on the eastern end of the Palmer Divide, where 1 the stands of trees exist on montane islands surrounded by grasslands with insufficient precipitation to support tree growth (Thompson et al. 2000).

Understanding past climate variation and its effect on vegetation is vitally important to researchers, land managers, and policy makers. Unfortunately, many parts of the western U.S. have less than a century of reliable instrumental climate records.

Proxy data such as tree rings have filled the gap in knowledge of past climate in many parts of the region.

Ponderosa pine (Pinus ponderosa P. & C. Lawson) is the dominant species in the montane forest at the lower elevation in the Colorado Front Range (Schubert 1974). The elevational range of the montane forest is approximately 1850-2770 m (Huckaby et al.

2003). The Palmer Divide is a dissected upland south of Denver and north of Colorado

Springs that separates the drainages of the South Platte and Arkansas Rivers. Elevations on the divide range from about 1700-2350 m. Stands of ponderosa pine are found in several locations along this broadly sloping ridge. In this project I analyze the climatic sensitivity of the tree-ring record, and the conditions that influence the recruitment of ponderosa pines at two sites, one located near the base of the Front Range, and the other at the easternmost ponderosa pine stand on the Palmer Divide.

Literature Review

Examining old and newly acquired tree-ring chronologies, Woodhouse and

Brown (2001), and Woodhouse et al. (2002) found a strong correlation between spring and summer Palmer Drought Severity Index (PDSI)(Palmer 1965) and tree-ring widths

2 from the eastern plains of Colorado. Previous research involving tree recruitment in

Colorado has been largely limited to study areas in the Front Range. Mast et al. (1998), in a study in Rocky Mountain National Park and Boulder County, identified a correlation between recruitment and climate. Mast et al. (1998) cored trees at uniform (1.4m) height and assumed that trees all reached that height at the same age. However, coring at uniform heights does not necessarily translate to consistent establishment years; since trees may grow at different rates, so their results suffer from necessarily coarse chronological resolution. Despite problems associated with determining the recruitment year, the authors determined that successful ponderosa pine recruitment was positively linked to above average precipitation in the spring and early summer.

Similar results were reported by Brown and Wu (2005) in a study of ponderosa pines on Archuleta Mesa near Pagosa Springs, Colorado; they studied germination success as it relates to El Niño Southern Oscillation (ENSO). This study used uniform coring heights to estimate germination dates. Results showed an increase in germination during wet springs and early summers of El Niño years above that of La Niña years, which were dryer and had a higher occurrence of fires. Brown and Wu (2005) compared climate factors, including ENSO, with fire chronologies in the Four Corners region and recruitment dates for ponderosa pine and other tree species. They found that pulses of recruitment coincided with periods of low fire and moister conditions. Major recruitment pulses occurred around 1700-1715, 1825-1840, 1860-1900, and 1905-1920, summarized in Fig 2.

3 League and Veblen (2006) studied recruitment of ponderosa pines on the Front

Range by “destructively sampling” seedlings in an effort to effectively date the establishment year to annual resolution. Seedlings, including the root ball, were removed to more precisely date the year of germination.

Research Questions

1. Do the tree-ring data sets from the eastern and western extremes of the Palmer

Divide differ in their sensitivity to climate, particularly precipitation?

2. To what extent can tree-ring data be used to determine if there are differences

between annual recruitment in stands of Pinus ponderosa (ponderosa pine)

growing near its extreme eastern distribution limit, and at higher elevations near

the west side of the Palmer Divide?

3. Can any differences in recruitment be attributed to differences in precipitation? If

so, do the data indicate a relationship with broader, hemispheric scale anomalies

such as sea-surface temperatures (SST) such as El Niño-Southern Oscillation

(ENSO) or Pacific Decadal Oscillation (PDO)?

Hypotheses

1. Eastern and western stands of ponderosa pine on the Palmer Divide provide

sufficiently sensitive dendroecological information to reflect local changes in

precipitation.

4 2. Successful recruitment in the eastern stands of ponderosa pine is limited to years

with favorable climatic factors such as wet springs or early summers.

3. Successful recruitment in the western stands is more or less randomly distributed

in wet and dry years.

4. Successful recruitment in the eastern stand is related to warm phases of PDO

during El Niño years, which are generally related to wetter years.

Palmer Divide

Topography

The Palmer Divide is a remnant of the mid-Tertiary high plains surface that extends east from the Colorado Front Range to the eastern plains near Limon, Colorado

(von Ahlefeldt 1992, Skinner 1996, Gilmore 2004, Hays 2008). Since the mid-Tertiary, tributary streams of the South Platte and Arkansas Rivers have down cut into an older high plains surface, producing a landscape of isolated buttes and ridges on the Palmer

Divide whose summit elevations decline from the west side of the divide to the east. At the extreme east end the divide grades into the rolling surface of the high plains. The

“wedge shaped” Palmer Divide ridge is an area of great transition in elevation (Trimble

1980, Hays 2008, von Ahlefeldt 1992), from the Rocky Mountains to the high plains, and separates the South Platte River basin to the north from the Arkansas River basin to the south. The highest elevation on the Palmer Divide is at the western end near the Front

Range at Palmer Lake on Ben Lomand Mountain (2,336 meters), and the lowest elevation is on the eastern end of the divide is near Cedar Point at an elevation of 1,700 meters,

(Skinner 1996). The bedrock of much of the Palmer Divide is Dawson Arkose. 5 The buttes, mesas, and ridges on the divide are usually capped by Castle Rock

Conglomerate, especially on the west side of the divide (Skinner 1996).

Climate

Due to the relatively higher elevation on the Palmer Divide, mean annual temperatures are lower and mean annual precipitation is higher than in the lowlands to the north, south, and east of the divide. At various times throughout the year three major air masses may converge on the Palmer divide: the polar, Pacific, and Gulf air masses

(von Ahlefeldt 1992).

Precipitation decreases easterly towards the plains. For example, Palmer Lake, near the west side of the divide, receives an average of 510 mm of annual precipitation;

Kiowa, farther to the east receives an average of 435 mm of precipitation; and Limon, near the eastern end of the divide averages 395 mm of annual precipitation (Skinner

1996). All of these sites receive more moisture than Denver (355 mm annually) (Skinner

1996). The elevation of the highest portions of the divide are ca. 700 meters higher than

Denver, creating an environment unique from the surrounding plains with sufficient moisture to support stands of montane forest.

Vegetation

The vegetation of the Palmer Divide is a mosaic of montane conifer forest, gamble oak, and short grass prairie. The cooler, wetter conditions on the divide permit stands of ponderosa pine and Douglas fir to establish on favorable sites. These stands are larger and more common on the higher elevations along the western and central parts of 6 the Palmer Divide, and restricted to the ridge tops on the eastern side of the Divide. On the extreme eastern side of the Palmer Divide the ponderosa pines are limited to east facing slopes near the tops of isolated ridges.

CHAPTER TWO: METHODS

Site Selection

Because this study focuses on analyzing the recruitment at elevation extremes on the Palmer Divide, we identified study sites at the easternmost stand of ponderosa pine on the Divide and on the westernmost end of the Divide. Two field sites were used in this study (Fig.3). The eastern site (39.3°N, -104.14°W, elevation 1,860 meters) is located on a north-south trending ridge 40 km west of Limon, Colorado. I cored the trees at this site,

East Ridge (ER) in April 2006 and August 2007. It is likely that moisture levels are insufficient to support tree growth at lower elevations east of the eastern study site. The property on which the stand is located is owned by Charles Guetz, and the site is used for low intensity grazing of cattle in the summer months.

The western site, Spruce Mountain (SM) (39.156°N, -104.882°W, elevation 2,286 meters), is located 3.5 km southwest of Greenland, Colorado, and 3 km east of the base of the Colorado Front Range. I cored the trees at this site in September 2009. This site was selected as a control for the eastern site, being at the westernmost edge of the Palmer

Divide, but not part of the Colorado Front Range. Spruce Mountain is part of the

Douglas County Open Space.

8 Field Methods

Because episodic recruitment was a main focus of the study, random sampling rather than targeted sampling methods were used (Speer 2010). The methods for both study sites were similar: The objective of the study was 50 trees for both East Ridge and

Spruce Mountain. 100 m transects were randomly placed and all standing ponderosa pines > 4 cm d.b.h. (Mast et al. 1998) within 5 meters of the transect line were cored.

The eastern site was sampled in two transects, one in April 2006 and one in August 2007.

Sampling at the western site was along one transect in September 2009.

Data recorded for each tree sampled included its location along the transect tape, diameter at breast height (d.b.h.), GPS coordinates, and overall tree condition. All trees were cored with an increment borer and an effort was made to capture the pith. Two cores were removed, from opposite sides of each tree, parallel to the slope contours.

Cores in this study were collected as close to the base of the tree as possible

(approximately 23 cm height, half the length of the increment borers) and parallel to the ground in order to maximize estimation of the recruitment years and allow for analysis of climate. Extracted cores were placed into individual paper straws for transport back to the lab.

Laboratory Methods

After drying, cores were removed from straws and glued, with the cross-sectional view facing up, into prefabricated wooden coremounts. The cores were held in place by masking tape while the glue dried. The site identification, tree number, sample number, species, and date sampled, were written on the side of the wooden blocks. A belt sander was used to sand each core with progressively finer belts starting with ANSI-60 and 9 ending with ANSI 400-grit. Each core was then hand sanded with ANSI-600 grit finishing film. The cores were visually counted using a low-powered stereo-microscope to mark the ring years (Stokes and Smiley 1996, Speer 2010).

A combination of memorization and list methods (Douglass 1941, Phipps 1985,

Yamaguchi 1991, Speer 2010) were used to crossdate the cores. Annual ring widths of the crossdated cores from the Palmer Divide were measured to the nearest 0.001 mm. using a Velmex linear-measurement encoder stage with Measure J2X software. The output files were then run through the measuring and crossdating verification program

COFECHA (Holmes 1983, Grissino-Mayer 2001).

Three verified ponderosa pine chronologies from the International Tree-Ring Data

Bank (ITRDB), part of the National Climatic Data Center, were used to identify and produce a regional list of narrow ring years. “Black Forest East” (ITRDB 2012),

“Jefferson County Colorado Recollection” (ITRDB 2012), and “Ridge Road” (ITRDB

2012) were combined in COFECHA to form a regional chronology I refer to as

“region1”. The tree ring measurements from the Palmer Divide study sites were compared in COFECHA with “region1”. A listing of narrow rings was produced and used to assist in crossdating the cores from the Palmer Divide.

Once the crossdating of the Palmer Divide study sites was verified by

COFECHA, the measurements were processed with ARSTAN (Cook 1985, Cook and

Krusic 2005). ARSTAN removes the growth trend, converts the tree-ring measurements into indices, and offers a selection of detrending options to aid in removing the noise in the chronology, to assist in climate reconstruction analysis (Cook 1985, Cook and

Holmes 1984). Output includes three versions of chronologies with different levels of 10 detrending. It is customary to evaluate the different chronologies against climate variables to identify the best signal for further analysis.

Climate Analysis

Three ARSTAN-produced indices (‘STNDRD’, ‘RESID’, and ‘ARSTAN’) from both Palmer Divide study sites were correlated in Excel against summer (JJA) Palmer

Drought Severity Index, (PDSI) (Palmer 1965) from Colorado Climate Division 1

(ftp://ftp.ncdc.noaa.gov/pub/data/cirs/).

Analysis of climate data was made using data from 1932-2007 for both sample sites. Prior to 1930, the Spruce Mountain chronology lacks a robust sample depth desired for the analysis. Post 1930, the methods and standardization of procedures were more uniform collecting instrumental climate data (H. Grissino-Mayer, per. comm.).

All monthly climate data were examined for significant correlations with the tree- ring widths. Correlation analysis was preformed using the climate analysis program

DENDROCLIM 2002 (Biondi and Waikul 2004) using bootstrapping (Guiot 1990,

1991), with replacement at the 95% confidence level to analyze climate and teleconnection variables. The program correlates (running each equation 1,000 times) yearly tree indices to 24-months of climate from the previous year of the subject year through the current year. Variables examined were Colorado Climate Division

1(ftp://ftp.ncdc.noaa.gov/pub/data/cirs/) monthly mean temperature (1895-2007), monthly precipitation (1895-2007), monthly PDSI (1895-2007), monthly El Nino-

Southern Oscillation (ENSO) (1931-2007)

(http://jisao.washington.edu/data/globalsstenso/) (data is from the International

11 Comprehensive Ocean-Atmosphere DataSet version 2.5 (ICOADS)), and monthly Pacific

Decadal Oscillation (PDO) (1930-2007)

(ftp://ftp.ncdc.noaa.gov/pub/data/paleo/treering/reconstructions/pdo.txt)(data is from

UKMO Historical SST data set for 1932-81, Reynold’s Optimally Interpolated SST (V1) for 1982-2001, and OI SST Version 2 (V2) beginning 2002). Correlation analysis was run on Excel between tree-ring widths and summer seasonal (JJA) PDSI (1895-2007), and reconstructed summer seasonal (JJA) PDSI (Cook et al. 1994) (1800-1978).

Episodic Recruitment Analysis

All cores were accurately dated identifying the pith year or the oldest year on each core. Since two cores were collected from each tree, the core containing the pith was used for the analysis. Where neither core contained the pith, the core containing the oldest year was retained for the analysis.

For well-dated cores that did not contain the pith, we used a pith estimating procedure (Applequist 1958, Speer 2010), in which a collection of five various sized rings with uniform concentric circles was copied into a transparency and placed over the ends of the cores to reach an estimated pith date. There were two circumstances for which the pith estimator was not used: if the oldest core of a given tree did not accurately crossdate to the earliest part of the core, another core from the same tree was used in its place; and if a given tree contained no cores that produced an accurately-dated early year, the tree was not included in the episodic recruitment analysis.

Basic statistics were produced for each study site including average pith year, average pith year including estimated pith years, oldest pith year, oldest estimated pith

12 year (if older), youngest trees, and histograms of pith/estimated pith dates. Figures were produced identifying annual pith year for each study site.

CHAPTER THREE: RESULTS A total of 117 trees from the East Ridge and Spruce Mountain sites were cored.

From these, master chronologies were constructed for each site, and statistical analyses were conducted to determine the climate sensitivity of each site. Pith years were determined on 103 trees and these were used to construct histograms for each site. Pith years were plotted against d.b.h. for each site (Fig. 4 and 5).

COFECHA Results At the East Ridge (ER) study site I sampled 63 trees, collecting two cores per tree.

53 trees were used for the episodic recruitment study. The ER master chronology (Table

1) consisted of 24 cores with no problem segments, yielding a series intercorrelation of

0.771, and average mean sensitivity of 0.468. Mean length of series was 118 years. The longest series for the climate analysis was 157 years (1851-2007).

At the Spruce Mountain (SM) study site I sampled 54 trees, collecting two cores per tree. 50 trees were used for the episodic recruitment study. The SM master chronology (Table 1) consisted of 22 cores with no problem segments, a series intercorrelation of 0.756, and average mean sensitivity of 0.433. Mean length of series was 86.1 years. The longest series for the climate analysis was 122 years (1888-2009).

ARSTAN Results ARSTAN graphically displays each of the four chronologies: raw, STNDRD,

RESID, and ARSTAN with a mean line, and sample depth chart (Figs. 6-12). As is customary, all three chronologies are run against climate variables to identify which has the best correlation.

After running correlation equations between the three ARSTAN outputs and two versions (instrumental (IPDSI) and reconstructed (RPDSI)) from tree-ring chronologies with seasonal (June, July, and August or JJA) Palmer Severe Drought Index (PDSI) the

‘ARSTAN’ chronology was selected over both ‘STDRD’ and ‘RESID’, due to a better correlation value with the PSDI data (Figs. 13-16). The ER correlation between

ARSTAN and IPDSI was r = 0.561, ARSTAN and RPDSI was r = 0.553. The SM correlation between ARSTAN and RPDSI was r = 0.479, ARSTAN and RPDSI was r = 0.718. A comparison of the two sites is shown in Table 2.

As identified in the methods section, a combination of memorization and list methods were used to crossdate the cores (Douglass 1941, Phipps 1985, Yamaguchi

1991, Speer 2010). As an aid to crossdating, three verified chronologies from the

International Tree Ring Data Bank (ITRDB 2012) were used to help establish narrow ring years. The data from the three chronologies, ‘Black Forest East’, ‘Ridge Road’, and

‘Jefferson County Update’ were downloaded from the ITRBD website as .rwl files and processed individually and in combination with COFECHA. ‘Black Forest East’ (Tables

1 and 3), CO564.rwl located at 39.30°N, -104.13°W, elevation 1,800m, has ring dates

15 between 1709-1997, with 33 cores, no problem segments, series intercorrelation of 0.775, and has an average mean sensitivity of 0.242. ‘Jefferson County Update’ (Tables 1 and

3), CO608.rwl located at 39.14°N, -105.12°W, elevation 1965m, has ring dates between

1487-2003, with 28 cores, no problem segments, series intercorrelation of 0.757, and has an average mean sensitivity of 0.399. ‘Ridge Road’ (Tables 1 and 3), CO565 located at

39.23°N, -104.12°W, elevation 1850m, has ring dates between 1779 and 1998, with 16 cores, no problem segments, series intercorrelation of 0.629, and has an average mean sensitivity of 0.394.

The three ITRDB chronologies were combined to form a single chronology referred to here as region1. This single chronology (series intercorrelation of 0.67 and average mean sensitivity 0.407) was used to identify the following narrow ring years:

2006, 2004, 2002, 1989, 1966, 1963, 1956, 1954, 1934, 1925, 1916, 1911, 1908, 1899,

1880, 1870, and 1863. (The years in bold indicate a missing ring in either one of the

Palmer Divide chronologies and at least one of the reference chronologies).

Episodic Recruitment Results

In order to estimate recruitment years, pith dates at 23cm height were determined for 53 trees at ER and 50 trees at SM (Figs. 17 and 18, respectively). Pith dates for ER span the period 1836-1960. The greatest concentration of years falls between 1850 and

1864 (21 trees, 39.6% of the total). Pith dates at SM span the period 1860- 1945, with the greatest concentration of pith years between 1905 and 1916 (25 trees, half of the total).

16 Climate Results

All monthly climate data were examined for significant correlations with the tree- ring widths. The most significant correlations were charted (Appendix B) with both chronologies. East Ridge has higher correlations with all tested variables than Spruce

Mountain. The highest correlations are for Monthly and Summer Season PDSI.

Using the ARSTAN data presented in Figures 6 and &, and assuming that dry conditions are particularly highlighted in the data, several notable dry periods can be identified from the ER and SM data sets.

For the ER site (1850-2007), notable dry conditions can be identified for periods prior to 1860, 1928-1940, 1950-58, and 1962-1970. Wetter conditions indicated for

1876-1886, 1904-1910, 1918-1926, 1942-1950, and in the mid 1980s.

For the SM site (1888-2008), notable dry conditions are indicated for 1894-1904,

1928-1938, 1950-58, 1962-1970, and 2000-08. Wetter conditions were 1918-1924,

1946-1950, and 1982-88.

CHAPTER FOUR: DISCUSSION

In this section the new data presented in the previous sections are interpreted and these results are compared with those of previous researchers. While both ER and SM sites showed sensitivity to climate variables and episodic recruitment pulses, the two sites differed from one another in a number of ways.

Pith Years and D.B.H.

The pith years vs. d.b.h. (Figs. 4 and 5) illustrated differences between the two sites.

Trees at East Ridge showed a moderately strong correlation between pith year and d.b.h.

(r = -0.562). In general, trees at the ER site had a greater diameter than those at the SM site. The largest d.b.h. at ER exceeded 50 cm.

Trees at Spruce Mountain showed a more random relationship between age and diameter (r = -0.316), and were generally smaller than those at ER. The greatest d.b.h.s at SM were less than 30 cm. The stand of trees on Spruce Mountain is on the top of the mountain, where thinner soils may hold less moisture than the soils at the ER site

(Skinner 1996), accounting for the slower growth at SM.

Climate sensitivity

The East Ridge site showed greater sensitivity in COFECHA statistics interseries correlation, average mean sensitivity and all tested climate variables including 18 precipitation, mean temperature, PDSI, ENSO and PDO (Figs. 13-16)(Table 2). These results are not surprising since the eastern site is located at a lower elevation with drier conditions. But the western site is also sensitive, probably because it is still near the low elevation limit for ponderosa pine despite being located at higher elevation than the eastern site. The series intercorrelation and average mean sensitivity at the ER and SM sites compares favorably with those reported for the Black Forest East, Jefferson County

Colorado Recollection and Ridge Road reference chronologies.

Recruitment

I hypothesized that recruitment at the eastern site would be more influenced by climate, where wet phases of ENSO and PDO would be a factor, while the western site would have more or less consistent recruitment not as heavily influenced by climate. The data (Figs. 17 and 18) show two pulses of episodic recruitment for East Ridge and Spruce

Mountain. These pulses occurred at different times. High pith year recruitment occurred at East Ridge 1850-1864 and 1870-74. High pith year recruitment occurred at Spruce

Mountain 1895-1900 and 1903-1915. Using the pith dates from the trees with the smallest d.b.h. (Figs. 4 and 5) and calculating approximate growth rates at both study sites, I estimated an approximate 5-year lag between the pith date at 23 cm height, and the actual establishment year. Adjusting for the lag, the years with the high numbers of tree establishment are ca. 1845-1859 and 1865-69 for East Ridge, and ca. 1890-95 and

1898-1910 for Spruce Mountain.

I compared these periods with the data reported by Woodhouse and Brown (2001) and Woodhouse et al. (2002) and Brown and Wu (2005). The pulses for both sites are 19 synchronous with periods of higher moisture, lower fire frequency, increased recruitment, and wet phases of ENSO identified by Brown and Wu (2005), and shown in Fig. 2.

Brown and Wu (2005) showed that tree recruitment in southwest Colorado tended to reach highest values when climate conditions were favorable combined with periods of fire quiescence. In other words, while drought conditions corresponded with low recruitment, wetter years did not necessarily produce high recruitment, unless those

(Fig.2) wetter years also coincided with low fire frequency—favorable climate conditions served as a prerequisite for strong recruitment, but only when other local conditions were also favorable.

The Black Forest East (BFE) data indicate pulses of recruitment in the early

1700s and 1859-1870 (Fig 20) and with the one of the pulses of establishment at ER.

These dates compare well with two of the pulses reported by Brown and Wu 2005.

However, the BFE site does not show a recruitment peak corresponding with Brown and

Wu’s 1825-1840 pulse, despite evidence that climatic factors were favorable. In addition the BFE site shows a pulse of recruitment around 1775-1790 that is not reflected in the

Brown and Wu 2005 data, although this period also corresponds with a moist interval in the Four Corners region.

The recruitment chronology for SM extends from the mid 1800s to 1940. There are two periods of peak establishment in this record, around 1890-95 and 1898-1910.

These dates correspond with the last two pulses of recruitment shown by Brown and Wu.

Brown and Wu note that these pulses occurred during periods of fire exclusion after

Euro-American settlement (Euro-American settlement in Colorado began in the 1860s, 20 but much of the eastern plains of Colorado was not settled until near the end of the century).

The ER recruitment chronology includes pulses of recruitment for the period

1845-1859, and a major pulse beginning around 1865 and extending through 1870.

These were both wet periods in the Four Corners region. The ER site does not show a large pulse of recruitment in the late 1800s or early 1900s.

The SM and ER sites have periods of establishment that correspond with some of the high recruitment years identified by Brown and Wu 2005, but neither site agrees entirely with the data presented by these authors. I suggest that this indicates that while moister conditions may set the stage for successful recruitment, other local factors, including stand density, fire history, etc., may inhibit germination and growth. ER had no recruitment from 1880-1890, and 1907-1925 despite wet conditions indicated with the

ER climate data and Brown and Wu (2005). SM had no recruitment from 1873-1890, and 1915-1930, also despite wet conditions. Brown and Wu 2005 showed recruitment during these times, leading me to the conclusion that although wet conditions are necessary for recruitment the condition at the site ultimately controlling recruitment.

The oldest ring dates with BFE data were not reported as pith dates, but the trees were cored by experienced researchers, and we assume that most cores intersected the pith of the tree. Assuming a 10-15 year lag between the BFE pith dates and the actual establishment years, the 1845-1859 recruitment pulse on East Ridge appears to be approximately synchronous with the high recruitment shown in the graph of oldest ring dates on the cores from Black Forest East (BFE) (Figs. 19 and 20). The BFE record also 21 indicates two small pulses in the 1700’s with little establishment between pulses.

Examining the three study sites (BFE, ER, and SM), the data suggest that although climatic conditions must be favorable for recruitment to occur, conditions of stand stability and competition within the stand after a pulse of recruitment may prevent further recruitment until some disturbance occurs.

Potential Alternative Hypotheses

In this section we review potential alternative hypotheses for the recruitment record seen in the ER and SM data, including the possible roles of logging, grazing, and fire.

Logging

The south branch of the Smoky Hill Trail passed near East Ridge from 1859-

1871. It is conceivable that pioneers using the trail could have impacted the tree stands in the area (fire, cutting trees, etc.). There is no evidence of fire scars on the standing trees or in the cores. There is a pulse of recruitment between 1860 and 1870 that could reflect an opening of the stand as a result of logging, but this may also be attributable to a period of wetter conditions.

On the west side of the Palmer Divide, the Denver and Rio Grande Western

Railroad began in Denver in 1870 and crossed the Palmer Divide not far from Spruce

Mountain in 1871-72. There is anecdotal evidence and unpublished data suggesting that larger trees were cut along the rail route for ties and construction (D. Sullivan, pers. comm.). Smaller trees (those less than about 20 years old) were evidently too small, and 22 were not cut. If the forest on Spruce Mountain had been logged by the railroad during that time, trees germinated during the 2-3 decades prior to 1870 would probably have been left standing. Had the railroad logged older trees in the 1870s, it might be expected that there could also have been a pulse of new recruitment in the 1870s- early 1890’s in the Spruce Mountain area. However, the SM data (Fig. 18) indicate no significant recruitment during this time. This suggests that if any logging had occurred at Spruce

Mountain in the late 1800s, it was not extensive, and played no important role in the recruitment story.

Cattle Grazing

The impact of cattle grazing on ponderosa pine mortality has been studied by several authors (Rummell 1951, Madany and West 1983, and Kingery and Graham

1991). These authors concluded that cattle grazing had no significant impact on ponderosa pine mortality (Knight and Graham 1991), or actually benefitted ponderosa pine establishment and higher stand density (Rummell 1951, Madany and West 1983).

This would suggest that grazing had little negative impact on the establishment of pines at the two sites, nor is there evidence of a positive grazing impact.

Fire

Ponderosa pine forests evolved with a regime of frequent, low intensity fires.

Fire suppression policies over the last century have allowed Douglas fir trees

(Pseudotsuga menziesii) to encroach on ponderosa pine stands in many areas. The presence of Douglas fir trees increases the chances of stand-replacing, high intensity fires

23 in ponderosa pine stands (Colorado State Forest Service, n.d.), as the Douglas fir trees allow a fire to “ladder” into the canopy.

There are currently no Douglas fir trees in the East Ridge stand. A few Douglas fir trees are scattered through the Spruce Mountain stand, occupying moister locations.

There was no evidence of fire scars in any of the 117 trees (234 cores) sampled for this study, nor was any evidence of charred stumps observed. It seems unlikely that even low intensity burns would have passed through these stands without leaving some evidence.

While the impact of fire on the recruitment pulses at the two stands cannot entirely be ruled out at this point, it does not appear to have played a significant role.

CHAPTER FIVE: CONCLUSION

In this research I attempted to establish recruitment data for ponderosa pine trees at two sites on the Palmer Divide, to determine the relative sensitivity of the two sites, and the extent to which climate factors or other factors may have influenced recruitment at the sites.

The results of the tree-ring analysis suggest that the East Ridge site is more sensitive to climatic factors than the Spruce Mountain site, as hypothesized.

Both the East Ridge site and Spruce Mountain site showed a history of episodic recruitment. The data suggesting episodic recruitment in the west (Spruce Mountain) is a departure from the original hypothesis which suggested that recruitment at the western site would be more or less consistent, and that climate would play a minor role, if any, in recruitment. In fact, climate apparently plays a significant role, although, perhaps not the largest, in the recruitment pulses at both sites. Although the two sites correlate well together (r = 0.551) in response to climate, the pulses of recruitment occur at different periods at each site.

Three potential sources of disturbance, logging, cattle grazing, and fire were examined as possible alternative explanations for the dissimilar periods of recruitment

25 between the sites. My conclusion is that none of these stand disturbances were likely the cause of the differences in recruitment in the two stands.

Climate is probably the most important factor in recruitment pulses, but only if there is opportunity within the stand for recruitment. Both sites are near the low elevation limit for ponderosa pine in eastern Colorado, and intra-stand competition for resources is strong. Although favorable patterns in climate have been present for both sites at the same time, the opportunity for recruitment within each stand has been different at different times.

I would argue that at the local stand level, this is because seedling germination and growth are more likely to be hindered by shading, competition, human activity, etc. even when climate conditions are favorable. While climate conditions must be favorable for seedling establishment and growth, conditions of stability and competition within the stand dictate the ultimate recruitment success.

26 Bibliography

Allen, C.D., Savage, M., Falk, D.A., Suckling, K.F., Swetnam, T.W., Schulke, T., Stacey, P.B., Morgan, P., Hoffman, M., and Klinge, J.T. 2002. Ecological restoration of southwestern ponderosa pine ecosystems: A broad perspective. Ecological Applications 12(5): 1418-1433.

Applequist, M.B. 1958. A simple pith locator for use with off-center increment cores. Journal of Forestry 56:141.

Baker, W.M., Veblen, T.T., Sherriff, R.L. 2007. Fire, Fuels and Restoration of ponderosa pine Douglas fir forests in the Rocky Mountains, USA. Journal of Biogeography 34: 251-269.

Biondi, F. and Waikul, K. 2004. DENDROCLIM2002: A C++ program for statistical calibration of climate signals in tree-ring chronologies. Computers and Geosciences 30:303-311.

Brown, P.M. and Wu, R. 2005. Climate and disturbance forcing of episodic tree recruitment in a southwestern ponderosa pine landscape. Ecology 86(11): 3030- 3038.

Colorado State Forest Service, n.d. csfs.colostate.edu/pages/forest-types-ponderosa- pine.html

Cook, E.R. 1985. A time series analysis approach to tree ring standardization. Ph.D. dissertation, The University of Arizona, Tucson. 171 pp.

Cook, E. R. and Holmes, R.L. 1984. User manual for Program ARSTAN. Laboratory of Tree-Ring Research, University of Arizona, Tucson, Arizona, USA.

Cook, E., and Kairiukstis, L. editors. 1990. Methods of dendrochronology. Kluwer, Amsterdam, The Netherlands.

Cook, E.R. and Krusic, P.J. 2005. A Tree-Ring Standardization Program Based on Detrending and Autoregressive Time Series Modeling, with Interactive Graphics. Tree-Ring Laboratory Lamont Doherty Earth Observatory of Columbia University, Palisades, New York.

Douglass, A.E. 1941. Crossdating in dendrochronology. Journal of Forestry 39(10):825- 831.

27 Esper, J., Cook, H.R., Schweingruber, F.H. 2002. Low-frequency signals in long-term chronologies for reconstructing past temperature variability. Science 295(5563): 2250-2253.

Gilmore, K.P. 2004. Analysis of the artifact collection from the Jarre Creek site (5DA541), a terminal early ceramic period occupation on the Palmer Divide, Colorado. Southwestern Lore Journal of Colorado Archaeology 70(2): 1-32.

Grissino-Mayer, H.D. 2001. Research Report Evaluating Crossdating Accuracy: A Manual and Tutorial for the Computer Program COFECHA. Tree-Ring Research 57(2): 205-221.

Grissino-Mayer, H.D. and Swetnam, T.W. 2000. Century-scale climate forcing of fire regimes in the American Southwest. The Holocene 10: 213-220. Guiot, J., 1990. Methods of calibration. In: Cook, E.R., Kairiukstis, L. (Eds.), Methods of Dendrochronology. Kluwer, Dordrecht, The Netherlands, pp. 165–178.

Guiot, J., 1991. The bootstrapped response function. Tree-Ring Bulletin 51, 39–41.

Hayes, H.L.G. 2008. Palmer Divide as a vital prehistoric place: interpreting and understanding the prehistoric occupation of the Palmer Divide region in Colorado. Masters Thesis. University of Denver, Denver, Colorado: 376p.

Holmes, R.L. 1983. Computer-Assisted Quality Control in Tree-Ring Dating and Measurement. Tree-Ring Bulletin Vol. 44, 69-75.

Huckaby, L.S., Kaufmann, M.R., Fornwalt, P.J., Stoker, J.M., Dennis, C. 2003. Identification and ecology of old ponderosa pine trees in the Colorado Front Range. Gen. Tech. Rep. RMRS-GTR-110. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 47 p.

International Tree-Ring Data Bank. 2012.

Kaufmann, M.R., Huckaby, L.S., Fornwalt, P.J., Stoker, J.M., Romme, W.H. 2003. Using tree recruitment patterns and fire history to guide restoration of an unlogged ponderosa pine/Douglas-fir landscape in the southern Rocky Mountains after a century of fire suppression. Forestry 76 (2) 231-241.

Keyes, C.R., Maguire, D.A., Tappeiner, J.C. 2007. Observed dynamics of ponderosa pine (Pinus ponderosa var. ponderosa Dougl. Ex Laws.) seedling recruitment in the Cascade Range, USA. New Forests 34: 95-105.

28 Kitzberger, T., Brown, P.M., Heyerdahl, E.K., Swetnam, T.W., Veblen, T.T. 2007. Contingent Pacific-Atlantic Ocean influence on multicentury wildfire synchrony over western North America. Proceedings of the National Academy of Sciences of the United States of America 104: 543–548.

Kingery, J.L., Graham, R.T. 1991. The effect of cattle grazing on ponderosa pine regeneration. The Forest Chronicle. 67 (3) 245-248

League, K., Veblen, T. 2006. Climatic variability and episodic Pinus ponderosa establishment along the forest-grassland ecotones of Colorado. Forest Ecology and Management 228:98-107.

Linhart, Y.B., and Mitton, J.B. 1985. Relationships among reproduction, growth rates, and protein heterozygosity in ponderosa pine. American Journal of Botany 72(2): 181-184.

Madany, M.H. and West N.E. 1983. Livestock grazing-fire regime interactions within montane forests of Zion National Park, Utah. Ecology 64 (4) 661-667.

Mast, J.N., Veblen, T.T., Hodgson, M.E. 1997. Tree invasion within a pine / grassland ecotones: an approach with historic aerial photography and GIS modeling. Forest Ecology and Management 93: 181-194.

Mast, J.N., Veblen, T.T., Linhart, Y.B. 1998. Disturbance and climatic influences on age structure of ponderosa pine at the pine/grassland ecotones, Colorado Front Range. Journal of Biogeography 25: 743-755.

Mooney, K.A., Linhart, Y.B., and Snyder, M.A. 2011. Masting in ponderosa pine: comparisons of pollen and seed over space and time. Oecologia 165: 651-661.

Palmer, W.C. 1965. Meteorological drought. United States Weather Bureau Research Paper 45, United States Department of Commerce: 58p.

Peters, G. and Sala, A. 2008. Reproductive output of ponderosa pine in response to thinning and prescribed burning in western Montana. Canadian Journal of Forest Research (38) 844-850.

Phipps, R.L. 1985. Collecting, Preparing, Crossdating, and Measuring Tree Increment Cores. Water-Resources Investigations Report 85- 4148. United States Geological Survey.

Rummell, R.S. 1951. Some effects of livestock grazing on ponderosa pine forest and range in Central Washington. Ecology 32 (4) 594-607.

29 Skinner, C.P. 1996. The role of soil characteristics and seasonal soil moisture in the distribution of vegetation on the Palmer Divide, Colorado. Masters Thesis. University of Denver, Denver Colorado: 203p.

Schubert, G.H. 1974. Siliviculture of southwestern ponderosa pine: the status of our knowledge. USDAF.S. Res. Pap. RM, 123, 71 p.

Speer, J.H. 2010. Fundamentals of Tree-Ring Research. The University of Arizona Press. Tucson. 333p.

Stokes, M.A. and Smiley, T.L. 1996. An Introduction to Tree-Ring Dating. Tucson, Arizona. The University of Arizona Press. 73p.

Thompson, R.S., Anderson, K.H., and Bartlein, P.J. 2000. Atlas of relations between climatic parameters and distributions of important trees and shrubs in North America. United States Geologic Survey professional Paper 1650-A.

Trimble, D.E.1980. The Geologic Story of the Great Plains. Geological Survey Bulletin 1493,United States Government Printing Office, Washington.

Veblen, T.T. 1986. Age and size structure of subalpine forests in the Colorado Front Range. Bulletin of the Torrey Botanical Club 113(3) 225-240.

Veblen, T.T., Kitzberger, T., and Donnegan J. 2000. Climatic and human influences on fire regimes in ponderosa pine forests in the Colorado Front Range. Ecological Applications 10(4) 1178-1195.

Villalba, R., Veblen, T.T., Ogden, J. 1994. Climatic influences on the growth of subalpine trees in the Colorado Front Range. Ecology 75(5) 1450-1462. von Ahlefeldt, J.G.P. 1992. The landscape ecology of the Palmer Divide, central Colorado. Ph.D. dissertation, Colorado State University: Fort Collins, Colorado: 371p.

Westerling, A.J., Hidalgo, H.G, Cayan, D.R., and Swetnam, T.W. 2006. Warming and earlier spring increase Western U.S. forest wildfire activity. Science. 313: 940- 943.

Woodhouse, C.A., 2001. Tree-ring reconstruction of mean annual streamflow for Middle Boulder Creek, Colorado, USA. Journal American Water Resources Association: 37 p 561–570.

Woodhouse, C.A., and Brown, P.M. 2001. Tree-ring evidence for Great Plains drought. Tree-Ring Research: 59 89–103.

Woodhouse, C.A., Lukas, J.J., Brown, P.M. 2002. Drought in the western Great Plains, 1845-56: impacts and implications. American Meteorological Society: 1485-1493.

Woodhouse, C.A. and Overpeck, J.T. 1998. 2000 years of drought variability in the central United States. Bulletin of American Meteorology Society 79: 2693–2714.

Yamaguchi, D.K. 1991. A simple method for cross-dating increment cores from living trees. Canadian Journal of Forest Research 21: 414-416.

Figure 1. North American distribution map of ponderosa pine (Source: http://esp.cr.usgs.gov/data/atlas/little/pinupond.pdf) 32

Figure 2. Comparison of ENSO, hydroclimate, fire-year, and tree recruitment chronologies. (a) ENSO time series (blue, Niño3 SST index [Cook 2000]; red, SOI [Stahle et al. 1998]). SOI is reversed to be consistent with other moisture indices. Heavy lines are annual series smoothed with 20-yr cubic smoothing splines. Years of significant triennial wet/dry oscillations (yt − yt−2) identified by superposed epoch analyses (SEA) in Fig. 3c, d are shown by arrows centered on the drought years. Biennial oscillations (yt − yt−1) also were tested and found to be largely absent during the fire-quiescent periods of 1684– 1724 and 1818–1851. (b) Reconstructed hydoclimate time series, smoothed with 20-yr cubic splines. The solid line shows the annual precipitation in northeastern New Mexico (Grissino-Mayer 1996), and the dashed line shows the Palmer drought severity index for the four-corners area (Cook et al. 2004). (c) Fire-year chronology for Archuleta Mesa. Horizontal lines mark time spans of individual trees, with fire scars designated by inverted triangles. Fire years at bottom are those recorded on ≥2 trees and used for SEA in Fig. 3. (d, e) Tree recruitment dates by 5-yr periods for (d) ponderosa pine and (e) other tree species. Recruitment dates are truncated toward the present, since we only collected data on trees >20 cm. The red vertical bar marks the 1580s megadrought, green bars mark fire-quiescent periods, the blue bar marks recent decades of fire exclusion after Euro-American settlement, and the purple bar marks the early 20th-century wet period and pulse of recruitment in ponderosa pine centered on 1919 (Savage et al. 1996) (Brown and Wu 2005).

Figure 3. Map of the Palmer Divide region. The East Ridge (yellow) and Spruce Mountain (blue) sampling sites are indicated.

34 East Ridge D.B.H. Values r = -0.562

DBH values

30 DBH in cm

0 1820 1840 1860 1880 1900 1920 1940 1960 1980 Pith Year

Figure 4. East Ridge comparison of pith year to d.b.h. r =-0.562

Spruce Mountain D.B.H. r = -0.316

Dbh

15 Dbh Values

0 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 Pith Years

Figure 5. Spruce Mountain comparison of pith year to d.b.h. r = -0.316. 35

Figure 6. ARSTAN output charts showing raw measurement chronology and standard chronology along with sample depth chart for East Ridge.

Figure 7. ARSTAN output charts showing residual chronology and ARSTAN chronology along with sample depth chart for East Ridge. 36

Figure 8. ARSTAN output charts showing raw measurement chronology and standard chronology along with sample depth chart for Spruce Mountain.

Figure 9. ARSTAN output charts showing residual chronology and ARSTAN chronology along with sample depth chart for Spruce Mountain. 37 East Ridge ARSTAN Chronology

2.5

1.5 ER ARSTAN Index

0.5

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Figure 10. East Ridge ARSTAN chronology shown in red.

Spruce Mountain ARSTAN Chronology

2.5

1.5 SM ARSTAN Index

0.5

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Figure 11. Spruce Mountain ARSTAN chronology shown in blue. 38 East Ridge and Spruce Mountain ARSTAN Chronology

2.5

ER ARSTAN 1.5 SM ARSTAN Index

0.5

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Figure 12. Overlapping East Ridge and Spruce Mountain ARSTAN chronology r = 0.551

39 Correlations between East Ridge ARSTAN Chronology and Climate

0.8

0.6

0.4

0.2 Precip Temp PDSI 0 Significant Correlations -0.2

-0.4

-0.6 Prev. Prev. Prev Jan Feb March April May June July Aug Sept Oct Nov Dec Oct Nov Dec Months

Figure 13. Significant correlations between East Ridge and monthly climate variables (r values).

Correlations between Spruce Mountain ARSTAN Chronology and Climate

0.8

0.6

0.4

0.2 Precip Temp PDSI 0 Significant Correlations -0.2

-0.4

-0.6 Prev. Prev. Prev Jan Feb March April May June July Aug Sept Oct Nov Dec Oct Nov Dec Months

Figure 14. Significant correlations between Spruce Mountain and monthly climate variables (r values).

40 Correlations between East Ridge ARSTAN Chronology and ENSO/PDO

0.4

0.35

0.3

0.25

ENSO 0.2 PDO

0.15 Significant Correlations 0.1

0.05

Jan Feb April May June July Aug Sept March Prev. JanPrev. Feb Prev. Oct Prev Dec Prev. AprilPrev. MayPrev. JunePrev. JulyPrev. AugPrev. Sept Prev. Nov Prev. March Months

Figure 15. Significant correlations between East Ridge and monthly ENSO/PDO (r values).

Correlations between Spruce Mountain ARSTAN chronology and Climate

0.4

0.35

0.3

0.25

ENSO 0.2 PDO

0.15 Significant Correlations 0.1

0.05

Jan Feb April May June July Aug Sept March Prev. JanPrev. Feb Prev. Oct Prev Dec Prev. AprilPrev. MayPrev. JunePrev. JulyPrev. AugPrev. Sept Prev. Nov Prev. March Months

Figure 16. Significant correlations between Spruce Mountain and monthly ENSO/PDO (r values).

41 East Ridge Pith Years

2 trees Number of Trees

183618401844184818521856186018641868187218761880188418881892189619001904190819121916192019241928193219361940194419481952 Year .

Figure 17. Annual East Ridge tree recruitment, showing concentrations of pith years in 1840s, and 1852-64

Spruce Mountain Pith Years

Trees

3 Number of Trees

0 1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 Year

Figure 18. Annual Spruce Mountain tree recruitment, showing concentrations of pith years between 1895- 1900 and 1903-1917. 42 East Ridge Pith Years

2 trees Number of Trees

183618401844184818521856186018641868187218761880188418881892189619001904190819121916192019241928193219361940194419481952 Year

Figure 19. East Ridge tree recruitment by year

Oldest Ring Year Black Forest East 1709-1885

Oldest Year Number of Trees 1

0 1709 1715 1721 1727 1733 1739 1745 1751 1757 1763 1769 1775 1781 1787 1793 1799 1805 1811 1817 1823 1829 1835 1841 1847 1853 1859 1865 1871 1877 1883 Year

Figure 20. Black Forest East oldest ring year. (Woodhouse et al. 2002) Common surges in recuitment for both sites between 1856-1871.

Table 1. COFECHA summary ranking and comparing series intercorrelation and average mean sensitivity between East Ridge and Spruce Mountain and the three reference chronologies. (Source: Woodhouse and Brown 2001, Woodhouse et al. 2002) Note that series intercorrelation and average mean sensitivity of both East Ridge and Spruce Mountain are comparable to the three reference chronologies used in this study.

44 Site Sensitivity comparison using COFECHA Statistics and Climate Variables

ER SM Series Intercorrelation 0.771 0.756 Ave Mean Sensitivity 0.468 0.433 Monthly Precipitation 0.44 0.366 Monthly Mean Temperature -0.435 -0.407 Monthly PDSI 0.633 0.55 Monthly ENSO 0.37 0.322 Monthly PDO 0.322 0.296 Seasonal (Summer JJA) PDSI 0.561 0.479

Table 2. East Ridge and Spruce Mountain comparison of COFECHA statistics and climate variables- results are correlation coefficients (r values).

Table 3. Comparison of East Ridge, Spruce Mountain, three other regional chronologies, and ITRDB averages demonstrating similar statistics to experienced researchers (results are correlation coefficients (r values).

46 Appendix A

Selected Terms

47 Master Chronology is a record of ring widths representing the stand level signal. The List Method is a technique used to develop the chronology of marker rings without the added steps of plotting on graph paper. Memorization Method is a method of chronology building that uses an established set of verified chronologies, and can be used with the list method. The International Tree-Ring Data Bank (ITRDB) is a repository of worldwide tree- ring chronologies and maintained by the NOAA Paleoclimatology Program and World Data Center for Paleoclimatology. Ring-Width Chronology: The averaged standardized ring-width indices from a number of trees sampled in a particular site that can be used for crossdating. COFECHA is a quality-control program used to check the crossdating and overall quality of tree-ring chronologies. Series Intercorrelation is a measure of the strength of the signal (typically the climate signal) common to all the sampled trees at the site. It is the average correlation of each series with the master chronology derived from all other series. Average Mean Sensitivity is a measure of the relative change in ring-width from one year to the next in a give series. Mean sensitivity is positively correlated with series intercorrelation; trees with stronger common signals tend to be more sensitive. The default Segment Length in COFECHA is 50 years. Number Problem Segments: The number of 50-year segments that were flagged by COFECHA (includes A and B flags). A flag issued indicates lower signal strength in those segments and can be used to identify potential dating errors. Percent Problem Segments: The number of problem segments divided by the total number of segments x 100. About 25% of all ITRDB chronologies have no problem segments; most chronologies have less than 10% problem segments. An “A” Flag indicates that a segment has a correlation with the master chronology of less than 0.328, the 99% significance level (one-tailed), but the segment does not correlate better with the master chronology in any of the other shifted positions. A “B” Flag means that COFECHA has found a better correlation for that segments in a 20-year window of -10 to +10 years from the place where it is currently dated. The program ARSTAN produces chronologies from tree-ring measurement series by detrending and indexing (standardizing) the series, then applying a robust estimation of the mean value function to remove effects of endogenous stand disturbances. Autocorrelation is the correlation of a variable with itself over successive time intervals.

48 Standardization: The raw ring width data verified by COFECHA are standardized and the results are averaged into a site chronology. The standardization process involves fitting a curve to the ring-width series, and then dividing each ring-width value by the corresponding value to generate a series of growth indices. Standardization removes age- related growth trends and other long-term variability that can be considered noise, and produces a series mean equal to 1.0. ARSTAN creates four chronologies: raw, STNDRD (standardized), RESID (Residual), and ARSTAN. Index Values (Indices) are unitless with a nearly stable mean and variance, allowing indices from numerous trees to be averaged into a site chronology. The Raw chronology is a simple average of the raw ring widths (i.e. no standardization has been performed on the series). STANDRD chronology: Detrended (standardized) tree-ring series are combined into a mean value function of all series. This chronology does not have the autocorrelation removed. The RESID (residual) is similar to the standard chronology except it has had all of the autocorrelation removed from the series. The ARSTAN chronology is calculated by removing the autocorrelation from the series, modeling it, and reintroducing a stand-level autocorrelation back into the chronology. DENDROCLIM 2002 is a program specifically written for dendrochonological applications to identify correlations with bootstrapping (with replacement) technique between tree-ring and meteorological data. Bootstrapping is a statistical technique that is used to determine the significance of any statistic even when the data are over correlated, not normally distributed, or when the data set is small. National Climatic Data Center (NCDC) is the world’s largest active archive of weather data. Monthly Average Temperature- Colorado Division 1 (1895-present) from the NCDC Monthly Precipitation- Colorado Division 1 (1895-present) from the NCDC. Monthly Palmer Drought Severity Index (PDSI) (1900-2003) is a measurement of recent dryness based on precipitation and temperature. Reconstructed PDSI- (Cook et al. 2000) (1700-1978) reconstructed from tree-ring chronologies EL NINO-Southern Oscillation (ENSO) International Comprehensive Ocean- Atmosphere DataSet version 2.5 (ICOADS) 49 Pacific Decadal Oscillation (PDO) UKMO Historical SST data set for 1932-81, Reynold’s Optimally Interpolated SST (V1) for 1982-2001, and OI SST Version 2 (V2) beginning 200

50 Appendix B

East Ridge and Spruce Mountain Chronologies Compared with Selected Climate Indexes

51 East Ridge and March Precip

2.5

East Ridge 1.5 March Precip Index/Inches

0.5

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

East Ridge and current monthly March precipitation. r = 0.44

Spruce Mountain and June Precip

2 6

5 1.5

Spruce Mountain 3 June Precip

Index/Inches 0.5

0 1

-0.5 0

19321935193819411944194719501953195619591962196519681971197419771980198319861989199219951998200120042007 Year

Spruce Mountain and current monthly June precipitation. r = 0.366 52 East Ridge and Prev. October Temp.

3 60

2.5

ER ARSTAN 1.5 50 Prev.Oct. Temp Index/ Degrees 1

0.5

0 40

193219351938 1941194419471950195319561959196219651968 197119741977198019831986198919921995 1998200120042007 Year

East Ridge and previous monthly October temperatures. r = -0.435

Spruce Mountain and Prev. October Temp.

2 60

1.5 55

SM ARSTAN 50 Prev. Oct. Temp.

0.5 Index/ Degrees

45 0

-0.5 40

19321935193819411944194719501953195619591962196519681971197419771980198319861989199219951998200120042007 Year

Spruce Mountain and previous monthly October temperatures. r = –0.407 53 East Ridge and June PDSI

3 10

8 2.5 6

2 4

2 ER ARSTAN 1.5 June PDSI Index 0

1 -2

-4 0.5 -6

0 -8

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

East Ridge and current monthly June PDSI. r = 0.633

Spruce Mountain and June PDSI

2 10

1.5 6

1 2 SM ARSTAN June PDSI Index 0 0.5

-2

-4 0

-6

-0.5 -8

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Spruce Mountain and current monthly June PDSI. r = 0.55 54 East Ridge and Prev. July ENSO

3 60

2.5 40

2 20

ER ARSTAN 1.5 0 July ENSO Index

1 -20

0.5 -40

0 -60

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

East Ridge and previous monthly July ENSO. r = 0.37

Spruce Mountain and Prev. March ENSO

2 80

60 1.5

SM ARSTAN 20 March ENSO Index

0.5

0 -20

-0.5 -40

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Spruce Mountain and previous monthly March ENSO. r = 0.322 55 East Ridge and March PDO

3 3

2.5 2

2 1

ER ARSTAN 1.5 0 March PDO Index

1 -1

0.5 -2

0 -3

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

East Ridge and current monthly March PDO. r = 0.322

Spruce Mountain and March PDO

2 3

2 1.5

SM ARSTAN 0 March PDO Index

0.5

-1

0 -2

-0.5 -3

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Year

Spruce Mountain and current monthly March PDO. r = 0.296 56 East Ridge and Summer (JJA) PDSI

3 6

4 2.5

2 2

0 ER ARSTAN 1.5 JJA PDSI Index -2

1 -4

0.5 -6

0 -8

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 Year

East Ridge and summer (JJA) PDSI. r = 0.561

Spruce Mountain and Summer (JJA) PDSI

1.8 6

1.6 4

1.4

2 1.2

0 1 SM ARSTAN JJA PDSI Index 0.8 -2

0.6 -4

0.4

-6 0.2

0 -8

1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 Year

Spruce Mountain and summer (JJA) PDSI. r = 0.479 57 East Ridge and Reconstructed Summer (JJA) PDSI

2 4

1.8 3

1.6 2

1.4 1

1.2 0 ER ARSTAN 1 RPDSI Index -1 0.8

-2 0.6

-3 0.4

0.2 -4

0 -5

1932 1934 1936 1938 1940 1942 1944 1946 1948 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 Year

East Ridge and summer (JJA) RPDSI. r = 0.553

Spruce Mountain and Reconstructed Summer (JJA) PDSI

1.8 4

1.6 3

1.4 2

1.2 1

1 0 SM ARSTAN RPDSI Index 0.8 -1

0.6 -2

0.4 -3

0.2 -4

0 -5

1932 1934 1936 1938 1940 1942 1944 1946 1948 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 Year

Spruce Mountain and summer (JJA) RPDSI. r = 0.718 58