Assessment of the Ponderosa Woodlands in Nebraska's Wildcat

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Dissertations & Theses in Natural Resources Natural Resources, School of

7-2018 Assessment of the Ponderosa Woodlands in Nebraska's Wildcat Hills: Implications for Juniperus Encroachment and Management Allie Victoria Schiltmeyer University of Nebraska-Lincoln, [email protected]

Follow this and additional works at: http://digitalcommons.unl.edu/natresdiss Part of the Forest Biology Commons, Forest Management Commons, Hydrology Commons, Natural Resources and Conservation Commons, Natural Resources Management and Policy Commons, Other Environmental Sciences Commons, Other Forestry and Forest Sciences Commons, and the Water Resource Management Commons

Schiltmeyer, Allie Victoria, "Assessment of the Ponderosa Woodlands in Nebraska's Wildcat Hills: Implications for Juniperus Encroachment and Management" (2018). Dissertations & Theses in Natural Resources. 272. http://digitalcommons.unl.edu/natresdiss/272

This Article is brought to you for free and open access by the Natural Resources, School of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Dissertations & Theses in Natural Resources by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.

ASSESSMENT OF THE PONDEROSA PINE WOODLANDS IN NEBRASKA’S WILDCAT HILLS: IMPLICATIONS FOR JUNIPERUS ENCROACHMENT AND MANAGEMENT

Allie Victoria Schiltmeyer

A THESIS

Presented to the faculty of The Graduate College at the University of Nebraska In partial fulfillment of the requirements For the degree Master of Science

Major: Natural Resource Sciences

Under the supervision of Professors David A. Wedin and Dirac Twidwell Jr.

Lincoln, Nebraska July, 2018

ASSESSMENT OF THE PONDEROSA PINE WOODLANDS IN NEBRASKA’S WILDCAT HILLS: IMPLICATIONS FOR JUNIPERUS ENCROACHMENT AND MANAGEMENT

Allie Victoria Schiltmeyer, M.S. University of Nebraska, 2018

Advisors: David A Wedin and Dirac Twidwell Jr.

Ponderosa pine (Pinus ponderosa) is a dominant tree species across western

North America. Its eastern distribution includes three populations in western

Nebraska. This study assesses the distribution, structure and age of ponderosa pine woodlands in one of those regions, the Wildcat Hills. The Wildcat Hills have escaped severe wildfires seen in recent decades in other ponderosa pine regions. Nevertheless, the Wildcat Hills woodlands face multiple threats including climate change, wildfire, drought, pine beetles, and invasive species. Key to these threats is the stand structure of pine woodlands, which have increased in density across much of ponderosa pine’s range. These changes in stand density are associated with high recruitment of young pines or the encroachment of other woody species, e.g. eastern redcedar (Juniperus virginiana). This study examined whether these changes are occurring in the Wildcat

Hills by conducting an inventory of trees, regeneration, and understory vegetation in 63 plots across a 630-hectare study area. 51 of the ponderosa pines were aged using dendrochronological techniques. The study found that 65% of the study area has open

(<70 trees/hectare) or savanna (<450 trees/hectare) with few or no juniper trees and seedlings. In contrast, 35% of the study area was classified as woodland. Over 50% of the trees in woodlands were small junipers. Identifying and understanding thresholds for woodland resilience and management in the Wildcat Hills provides insights into their current functional dynamics and directions for future research regarding their fate. iii DEDICATION:

For Bob & Dee

iv ACKNOWLEDGEMENTS:

Many, many, thanks to my primary advisor, David Wedin, for guiding me through this project and stealing me away from the forest service to take me on this journey of research and science. Thank you to Dirac Twidwell and Eric North for your advice, brilliant ideas, and expert committee member style.

I would like to especially thank Bob Smith and Bruce Rolls with Platte River Basin Environments and Chris Becker with the Nebraska Game and Parks Commission. Without your guidance and knowledge of the Wildcat Hills, this study would not have been successful.

Thank you to The Nature Conservancy for selecting me to receive the J.E. Weaver Grant in 2017. The funds from that grant secured my second field season in the Wildcat Hills with an awesome field crew, conference travel, and equipment that was extremely beneficial to the project.

Thank you to the Nebraska Forest Service for collaborating on the 2018 Growth & Drain Study and providing me with the data.

Thank you to everyone at the Nebraska Forest Service and the Department of Agronomy and Horticulture for supporting me to leave and pursue graduate school full time in SNR. I’m looking at you Eric Berg, Justin Evertson, Christina Hoyt, & Kim Todd.

Thank you to Brian Waters for teaching me how to #writescientific.

So many thanks to my fearless female friends and colleagues in science: Hannah B., Jessica B., Heather N., Tori D., and Amanda H. Each of you has inspired me, given me advice (statistics, writing, ArcGIS), let me cry in your kitchen, or helped me in one way or another throughout this entire process.

Thank you, Hugh. Your unconditional love and support throughout this last year and a half has been incredible and so, so welcomed. Thanks for bringing me food and hugs at the office when I was in the thick of thesis preparation. Thank you also for delivering the final examination forms to graduate studies. (You’re literally making this happen!)

Thank you to my family and close friends for always giving me a thumbs up/nod for continuing my education. Someday, I’ll figure out a way to explain why I do what I do. Linsey, thank you for questioning me, supporting me, and drinking beers with me when I was deciding to quit my cushy job and move away to go back to graduate school.

Thank you to my new lab at Colorado State University for your patience and support while I worked through my final thesis revisions: Troy, Scott, and Dan.

And thank you to a few people who will never see this: Hope Jaren, Oprah, and the #metoo movement. v TABLE OF CONTENTS: LIST OF FIGURES.…………………………………………………………………..…..vi LIST OF TABLES….…………………………………………………………...... …viii

CHAPTER ONE: PONDEROSA PINE WOODLAND STRUCTURE AND FUNCTIONAL THRESHOLDS IN THE WILDCAT HILLS …………………………..1 Introduction………………………………………………………………………..1 Methods….………………………………………………………………………...5 Results……………………………………………………………………………10 Discussion…..……………………………………………………………………22

CHAPTER TWO: A DENDROECOLOGICAL ANALYSIS OF PONDEROSA PINE IN THE WILDCAT HILLS AND RESPONSE TO RECENT DROUGHT………………..27 Introduction….…………………………………………………………………...27 Methods…………………………………………………………………………..28 Results……………………………………………………………………………32 Discussion………………………………………………………………………..43

CHAPTER 3: MANAGEMENT IMPLICATIONS & FUTURE RESEARCH……………………...………………………………………………………46

REFERENCES……………..……………………………………………………………47

APPENDIX 1: UNDERSTORY VEGETATION PERCENT COVER AND FOREST FLOOR DEPTH……..…………………………………………………………………..56

APPENDIX 2: FUEL STRATA GAP…………………………………………………...58

vi LIST OF FIGURES:

CHAPTER 1- Figure 1: Distribution of ponderosa pine in North America. Figure 2: 200 survey points at various management locations in the Wildcat Hills. Figure 3: 63 plot points chosen out of the 200 random stratified points.’ Figure 4: Classification system using TPH and basal area (m2/ha) for 60 plots. Figure 5: Classification systems using remote sensing and basal area. Figure 6: Juniper and ponderosa pine basal area in Wildcat Hills classes. Figure 7: Juniper and ponderosa pine density in Wildcat Hills classes. Figure 8: Histogram tables of distribution of diameter at breast height of juniper and ponderosa pine. Figure 9: Mean DBH for juniper and ponderosa pine across the Wildcat Hills classes. Figure 10: Mean regeneration of ponderosa pine and juniper seedlings and saplings per hectare across Wildcat Hills classes. Figure 11: Understory biomass (herbaceous, forest floor, and total) ungrazed and grazed plots across open, savanna, and woodland classes.

CHAPTER 2- Figure 1: 200 survey points at various management locations in the Wildcat Hills. Figure 2: PDSI values from 1902-2017. Figure 3: Histogram showing range of ages of trees measured. Figure 4: Diameter at breast height vs age across measured ponderosa pine trees. Figure 5: Ponderosa pine trees (<10 cm diameter) within each vegetation classification by age. Figure 6: Illustration of low and high density plot effect on tree growth for ponderosa pine trees <10 cm diameter. Figure 7: Age distribution of ponderosa pine within high-density woodland plots. Figure 8: Age distribution of ponderosa pine within low-density open and savanna plots. vii Figure 9: Mean BAI across 50 ponderosa pines versus the age of the tree when a particular tree ring was formed. Figure 10: Mean basal area increment for trees >20 cm diameter for low and high densities from 2000-2016. Figure 11: Mean basal area increment for trees <10 cm diameter for low and high densities from 2000-2016.

viii LIST OF TABLES:

CHAPTER 1- Table 1: Classification systems of NLCD, LANDFIRE, Nebraska Forest Service, and Wildcat Hills with the data collected in 2016 and 2017. Table 2: Regeneration rates of juniper and ponderosa pine within each of the Wildcat Hills classes. Table 3: ANOVA model results showing response of ponderosa pine and juniper regeneration against vegetation classifications and grazing “treatments.” Table 4: Effect test of Vegetation Class (3 levels), Grazing (2 levels) and their interaction on three aspects of understory biomass (forest floor, herbaceous, and total).

CHAPTER 2- Table 1: ANOVA model including DBH, density, and interaction for density and DBH. Table 2: Linear regression results for the response of tree growth (BAI) to climatic variables. Table 3: Mixed model examining the response of BAI to tree size, tree age, stand density, gpdsi, and the interaction of density and gpdsi. 1 CHAPTER ONE: PONDEROSA PINE WOODLAND STRUCTURE AND FUNCTIONAL THRESHOLDS IN THE WILDCAT HILLS

Introduction -

Identifying functional thresholds in natural landscapes is critical for producing realistic landscape management objectives (Chapin et al., 2002; Lloret et al., 2011;

Twidwell et al., 2013). Management, conservation, and restoration projects across the range of ponderosa pine (Pinus ponderosa) assume resilience (the amount of disturbance a system can withstand before shifting to an alternative state (Holling 1973)) is dependent on current horizontal and vertical stand structure dynamics (Helms 1998). Land managers and foresters identify increases in stand density as a threat to ponderosa pine across its range of distribution (Kaye et al., 2010; Erickson & Waring, 2014). Increases in stand density are often the result of high ponderosa pine recruitment following fire exclusion, grazing, and logging in the late 19th and early 20th centuries (Brown et al., 2015). Other drivers of increased stand density are the recruitment of other tree and understory species, particularly juniper (Juniperus spp.), within ponderosa pine stands (Allen et al., 2002;

Larson & Churchill, 2012). Increasing stand density may drive landscapes across key thresholds (e.g. moving from grasslands to savanna or from savanna to forest) and put the ponderosa pine system at risk.

As forested stands become “too dense,” negative impacts begin to arise and threaten the resilience of the system (Allen et al., 2002). Higher intensity disturbances

(i.e. fire, insects, pathogens, and drought) become more frequent in high density stands and thus alter existing stand structure and shift toward less resilient systems (Chapin et al., 2002). Although there is a general consensus that many ponderosa pine woodlands 2 are “too dense,” surprisingly few quantitative guidelines exist for management and restoration or for the classification of ponderosa pine woodlands that reflect key thresholds for ecological functioning. Because of the lack of quantitative guidelines, much debate still exists over the meaning of “too dense” in ponderosa pine stands (Allen et al., 2002; Moore et al., 2013; Bollenbacher et al., 2014). Contributing to this confusion are contrasting opinions on optimal stand structure from the perspectives of silviculture

(e.g. producing quality timber), forest health (e.g. minimizing insect pests and diseases), wildland fire (e.g. fuels reduction), wildlife habitat (e.g. stand structures including snags for wildlife species of concern), and aesthetics of rural and peri-urban land owners.

Agencies and scientists that have proposed density thresholds, reflect their perspectives and definitions based on region and management goals. For instance, ponderosa pine woodland conservation specialists in Arizona (e.g. the Ecological Restoration Institute at

Northern Arizona University) recommend a maximum basal area of 5-10 m2/ha

(Roccaforte et al., 2014), while pine beetle experts have suggested 17 m2/ha as the basal area above which beetle outbreaks are likely (Negron & Popp 2004), and in another region, the Nebraska Forest Service (NFS) considers a ponderosa pine stand “fully stocked” or at “maximum density” for timber production at approximately 20 m2/ha basal area (NFS Growth & Drain Study, 2018). Confusions over establishing density thresholds also reflect confusion over the term “density”. Foresters and silviculturists generally equate density with stocking rate, which is an indication of stand volume and is often approximated with stand basal area (m2/ha or ft2/acre). In contrast, forest ecologists follow an ecological definition of density, which is number of individuals per unit area

(trees per hectare or trees per acre). A complete assessment of woodland stand structures 3 should integrate both basal area, which is influenced by the stand’s maximum tree diameter and the number of large trees, and density (trees per hectare), which is heavily influenced by the number of small trees, which contribute little to basal area but will play a large role in the future of the stand. Proposed thresholds for ponderosa pine stand density should also have both stand-level and landscape perspectives. Small dense patches of closed-canopy ponderosa pine pose a different threat for both fire and insects if they are isolated in an otherwise open landscape in contrast to connected ponderosa pine patches in a heavily wooded landscape.

Although identifying the management thresholds across ponderosa pine landscapes depend on the location and level and type of threat to the system, few land managers have the time or resources to collect detailed stand inventory data across woodland landscapes. Thus, land managers and forest scientists have turned to remote sensing efforts and subsequent land cover classifications from public agencies, academics and non-governmental organizations. In Nebraska, a 2018 assessment of ponderosa pine stands in the Pine Ridge and Niobrara Valley (NFS Growth & Drain Study, 2018; Fig. 1) concludes that these two landscapes are distinct, and both possibly differ from the

Wildcat Hills (Fig. 1). The ponderosa pine landscapes of Nebraska are also functionally different than the ponderosa stands of the Rocky Mountain Front Range (Colorado and

Wyoming) or the Black Hills (South Dakota). There are few management guidelines or conceptual models for stand structure from these neighboring regions that, a priori, appear to be directly relevant to the Wildcat Hills, including timber management, fuels reduction, and prescribed fire recommendations. 4

Fig 1. Ponderosa pine (Pinus ponderosa) is the dominant tree species in foothills and low elevation montane regions throughout western North America. Its eastern distribution reaches into the Great Plains, with a large population in the Black Hills of South Dakota, and smaller populations in western Nebraska.

The Wildcat Hills have not encountered the recent high intensity fires that the

Pine Ridge and Niobrara Valley have faced. Are the Wildcat Hills less prone to large, intense crown fires because of current stand structure and distribution, or has the landscape avoided stand-replacing fires in recent decades because of other aspect of fire management or chance? In the former case, stand management (e.g., fuels reduction and prescribed fire) may not be necessary, while in the latter case, management may be critical in preventing future woodland losses. Although there have been assessments of the plant communities of the Wildcat Hills (e.g., Plant Community Survey of the Wildcat

Hills, Robert F. Steinauer, 2007), simple land cover classifications provided by the NFS and the National Land Cover Database (NLCD, USGS) are the only data currently available for guiding management and conservation decisions across the region. In 5 addition, ponderosa pine and juniper regeneration are poorly documented in Nebraska’s ponderosa pine woodlands, especially the Wildcat Hills. In this study, we assess the stand structure of woodlands in two representative areas of the Wildcat Hills, including stand species composition, basal area and density, as well as documenting size distributions for pine and juniper trees and quantifying pine and juniper regeneration.

METHODS:

Data collection-

Data collection plots were established via a stratified random sampling based on existing land classification data (NLCD, Homer et al. 2015) for two areas in the Wildcat

Hills: Carter Canyon Ranch (owned by Platte River Basin Environments) and Cedar

Canyon Wildlife Management Area (owned by Nebraska Game and Parks Commission)

(Scott, 1998) (Fig. 2). The southern block of the study area is 4 km x 1.4 km (560 hectares) with edges on the N-S and E-W at 41°44'3.88"N, 103°47'58.09"W (Fig. 2). The smaller northern block is 1 km x 0.7 km (70 hectares) at 41°45'22.48"N, 103°49'9.64"W

(Fig. 2). The total study area is approximately 630 hectares or 1560 acres. 6

0 0.375 0.75 1.5 Km

0 0.25 0.5 1 Miles

Fig. 2. Two-hundred randomly selected survey points at the Carter Canyon Ranch (upper area) and Cedar Canyon Wildlife Management Area (lower area) in Nebraska’s Wildcat Hills.

To ensure the equal sampling of various vegetation types and a range of topography, 200 points in the study area were randomly selected. The points closely matched the proportion of the landscape classified as wooded versus open by NLCD.

Eighty-nine percent of the points were in the larger southern block and eleven percent in the northern block (Fig. 2). After selection of the 200 random points, 63 random points were located in the field based on physical accessibility (Fig. 3). At each point sampled, tree species, number and size of tress, seedling and sapling numbers, biomass, and percent cover estimates of understory vegetation, woody debris and forest floor were collected. 7

Fig. 3. Sixty-three sampling plots chosen from 200 random stratified points. These 63 points were located and sampled during the preliminary field season in 2016 and full field season in 2017. Forty sites were sampled both years, with an additional 20 sites added in 2017.

Preliminary sampling occurred in July 2016. Forty-one of the preselected random points were located by GPS. Flags were placed 5 meters from the point center in four directions. The 10m diameter circle was inventoried for seedlings and saplings of all tree species. The distance to and heights of the 10 nearest trees within 100 m of the plot center were measured (Trimble, LaserAce 1000 Rangefinder). Almost all trees sampled with either ponderosa pine or juniper. We did not attempt to differentiate Rocky Mountain

Juniper from Eastern Redcedar; both occur at the site. Diameter at breast height (1.3 m) was measured manually with a DBH tape. Thus, the area sampled in 2016 to determine woodland structure (trees per hectare and basal area in m2/ha) varied with tree density, with plot areas sampled ranging from less than 0.01 to 0.7 hectares. Data on trees in the 8 2016 plots was remeasured in summer 2017 with plots of fixed size (see below) and 22 new plots were also added in 2017.

For sampling in 2017, flags were placed 16.9 meters from the point center in the four cardinal directions. The 33.85 m diameter circle (0.09 hectare) was wholly inventoried for seedlings and saplings of ponderosa pine, juniper, and other tree species.

Any tree in the 0.09 ha plot was measured and inventoried. Seedlings were counted if they were less than 10 cm in height and saplings were measured if they were between 10 and 200 cm in height. We combined the two regeneration categories (seedlings and saplings) and converted to a per hectare basis for each plot.

Percent ground cover of woody and herbaceous materials were assessed using 4 randomly positioned quadrats (1m x 1m) within each sampling location for non- destructive (percent cover) estimates of grasses, forbs, tall shrubs, short shrubs, bare soil, duff, and dead woody material. Percent cover was classified by categories representing a log-transformed scale: 1= 0-5%, 2= 6-16%, 3=16-25%, 4=26-50%, 5=51-75%, and 6=76-

100%. Within each quadrat, a 50cm x 50 cm quadrat was clipped and sorted in the field into 5 categories: tall shrubs, short shrubs, herbaceous (standing live and dead), small branches (<7.62 cm diameter) and duff (plant litter on the surface above the mineral soil).

For analyses, the small branches and duff categories were combined to estimate the forest floor biomass. Large woody debris (> 7.62cm diameter) biomass was not measured but estimated as percent cover. Samples were weighed in the field and discarded. All sampling was done under dry conditions with low relative humidity. In 2017, the depth of the duff layer was also measured in each of the 1m x 1m quadrats.

9 Statistical analysis-

Three of the initial sixty-three plots were dropped from analysis due to missing data. Stand density, seedlings, tree density, basal area were all standardized to per hectare basis. Regression analysis was used to determine the response of pine and juniper density and size, regeneration, and understory vegetation to the continuous gradient from open to densely wooded vegetation formed by the 60 sample areas. We also analyzed this gradient categorically by placing the plots into three vegetation classes. The 60 plots were divided into three vegetation classes (Fig.4; Open, Savanna, and Woodland) based on natural breaks in the data for stand density and basal area. Each class contained approximately 20 plots. We contrasted this classification using our stand data with classifications of our plots based on NCLD (Homer et al., 2015), LANDFIRE, and NFS

(NFS Growth & Drain Study, 2018). Details of these classification systems are provided under Results.

Because part of the Cedar Canyon WMA was grazed, and the remaining study areas were not, we were able to ask if grazing impacted results for regeneration or understory biomass. We ran two-way ANOVA models for regeneration (Table 3) and for understory biomass (forest floor, herbaceous biomass, and total understory biomass) against vegetation class, grazing, and the interaction of vegetation and grazing. All statistical analysis was completed using JMP by SAS and R. (JMP®, Version Pro 13. SAS

Institute Inc., Cary, NC, 1989-2007; RStudio Team (2015). RStudio: Integrated

Development for R. RStudio, Inc., Boston, MA).

10 RESULTS:

Vegetation Classification-

During the field seasons of 2016 and 2017, we measured a total of 2,552 trees.

This included 2 Prunus spp., which were eliminated from analysis. The rest of the trees were pines and junipers. We summed juniper and ponderosa pine trees measured in the

60 plots to evaluate total basal area (m2 / ha) and the number of trees per hectare (TPH).

The breaking points in the data represent meaningful ecological or management thresholds that merit a comparison with existing vegetation classifications. Within the study area, we found that there were clear thresholds between Open (grassland with no or few individual trees), Savanna (scattered trees with some understory woody vegetation), and Woodland (medium and high-density woodland with some or all closed canopy).

Using TPH basis to define the break points, the Open class had <70 TPH, Savanna class had >70 TPH and <450 TPH, and Woodland class had >450 TPH (Fig. 4).

Fig. 4- Classification system using TPH and basal area (m2/ha) for 60 Wildcat Hills plots surveyed. 11 Alternative Vegetation Classification Systems-

National Land Cover Database (NLCD) classes for my 60 plots included

“Evergreen Forest” and “Grassland/ Herbaceous” categories (Fig. 5 A). We extracted these classifications from each plot’s location from NLCD raster files in ArcGIS.

Evergreen forests are described by the NLCD as “areas dominated by trees generally greater than 5 meters tall, and greater than 30% of total vegetation cover. More than

75% of the tree species maintain their leaves all year. Canopy is never without green foliage.” The Grassland/Herbaceous category is described as "areas dominated by graminoid or herbaceous vegetation, generally greater than 80% of total vegetation.

These areas are not subject to intensive management such as tilling but can be utilized for grazing.”

LANDFIRE (LF) classes for the 60 plots consisted of “Herbaceous- grassland”,

“Open tree canopy”, and “Shrubland” (Fig. 5 B). We extracted the class for each plot from raster files from the LF Existing Vegetation Type (EVT) database. LF classes are determined by decision tree models, field data, Landsat imagery, elevation, and biophysical gradient data. The “open tree canopy” category is coded within LF classes as

3054: Southern Rocky Mountain Ponderosa Pine Woodland, shrubland as 3212: Western

Great Plains Sandhill Grassland, and herbaceous-grassland as 3149: Western Great Plains

Short Grass Prairie. Classifications, like NLCD and LF, are typically derived from 30m resolution remote sensing imagery.

The Nebraska Forest Service (NFS) classification system was used in Western

Nebraska Timber Supply Study Report (2018), where three classes were produced to provide data across stand types for information on volume, growth, and regeneration. The 12 three classes created from the NFS study based on basal area are Savanna (<6.89 m2/ha),

Light Density (>6.89 m2/ha and <16.06 m2/ha), and High Density (>16.06m2/ha) (Fig. 5

C). In terms of silvicultural management, NFS district foresters for the Pine Ridge region describe ~20.66 m2/ha as fully stocked on relatively flat landscapers and generally north facing slopes, whereas steeper south facing slopes are ~9.18 m2/ha to be considered fully stocked. To classify our plots using the NFS system, they were assigned to a class based on plot basal area. 13

Fig. 5- Classification systems with the 60 plots sampled in the Wildcat Hills applied with remote sensing (A: NLCD , B: Landfire) and basal area (C: NFS) classification parameters.

14 Stand structure-

Of the trees measured, 43.1% were juniper. and 56.8% were ponderosa pine. The average combined juniper and ponderosa pine basal areas for each class within our

Wildcat Hills classification (Fig 4) were 0.77 m2/ha for Open, 8.99 m2/ha for Savanna, and

19.05 m2/ha for Woodland (Table 1). The average combined TPH of juniper and ponderosa pine for each class were 18.21 tress/ha in Open, 205.85 trees/ha in Savanna, and 999.77 trees/ha in Woodland (Table 1). Average stand structure results for the

NLCD, LANDFIRE, and NFS classifications are given in Table 1.

Table 1- The percentage of plots and average stand structure for 60 Wildcat Hills Plots within three preexisting classification systems (NLCD, LANDFIRE, NFS), and the classification created by this study (Wildcat Hills). TPH % of plots Mean BA m2/ha (Density) measured NLCD Grass/Herbaceous 5.94 164.72 61.7% Evergreen Forest 16.65 850.04 38.3% LANDFIRE Herbaceous- grassland 3.41 79.92 28.3% Shrubland 6.19 222.89 23.3% Open tree canopy 15.8 729.89 48.3% NFS Savanna 1.36 44.3 38.3% Light Density 10.8 372.18 35.0% High Density 21.53 1050.69 26.7% Wildcat Hills Open 0.77 18.21 30.0% Savanna 8.99 205.85 35.0% Woodland 19.05 999.77 35.0%

Tree species composition-

Mean basal areas of juniper in Open, Savanna, and Woodland plots were 0.059 m2/ha, 0.56 m2/ha, and 5.56 m2/ha, respectively (Fig. 6). Mean ponderosa pine basal areas 15 in Open, Savanna, and Woodlands plots were 0.71 m2/ha, 8.41 m2/ha, and 13.49 m2/ha, respectively (Fig. 6).

Pine and Juniper Basal Area (m2/ha) vs. Classiﬁcation 15 Mean(PineBA(m2/ha)all) Mean(JuniperBA(m2/ha)all)

10 Basal Area (m2/ha) Basal Area

0 Open Savanna Woodland Class17

Fig. 6- Juniper and ponderosa pine basal area (m2/ha) within the Wildcat Hills classification systems.

Mean density for ponderosa pine in Open plots was 14.43 TPH, 160.67 TPH in

Savannas, and 464.27 TPH in Woodlands (Fig. 7). Mean density for juniper for Open plots was 3.78 TPH, 45.17 TPH in Savannas, and 535.5 TPH in Woodlands (Fig. 7). In

Woodland plots, 53% of the trees, on average, were junipers.

16 Pine Density and Juniper Density vs. Classiﬁcation

600 Mean(PineDensity(ha)) Mean(JuniperDensity(ha))

500

400

300 Trees per Hectare Trees

200

100

0 Open Savanna Woodland Class17

Fig. 7- Juniper and ponderosa pine density (TPH) within the Wildcat Hills classification systems.

Size distribution of pine and juniper-

Most juniper trees were small (DBH < 7.5 cm) with a few trees exceeding 30cm

DBH (Fig. 8, A). Ponderosa pine showed a bimodal distribution of DBH, with peaks around 10cm and 40cm. The distributions long right-hand tail indicates a low number of large pines (>50cm) (Fig. 8, B). Ponderosa pine had a larger mean DBH than juniper in all vegetation classes, with larger pines on average in Open plots than in Savanna or

Woodland plots (Fig. 9). 17

Fig. 8- Histogram showing the size distribution of (DBH) of A) Juniperus spp. and B)

Pinus ponderosa measured in 60 Wildcat Hills plots. 18

Pine DBH and Juniper DBH by Classiﬁcation

30 Mean(MNpineDBH(cm)) Mean(MNjuniperDBH(cm))

15 Mean DBH

0 Open Savanna Woodland Class17 Fig. 9- The mean DBH for juniper and ponderosa pine across the three Wildcat Hills vegetation classes.

Regeneration-

In Open plots, the mean ponderosa pine density for seedlings and saplings was

36.6 per hectare (SS/ha), 55.03 SS/ha in Savanna plots, and 155.54 SS/ha in Woodland plots (Fig. 10). No juniper regeneration was observed in Open plots. Within the Savanna plots, juniper was slightly lower than ponderosa pine with 41.8 SS/ha. Juniper regeneration in Woodland plots, 202.04 SS/ha, exceeded that observed for pines (Fig.

10). For juniper, 83% of the regeneration we observed was in Woodland plots (Table 2).

Regeneration was also most common for pine in Woodland plots, although roughly 40% 19 of its seedlings and saplings were observed in the two more open vegetation classes

(Table 2).

Table 2- The percentage of regeneration for juniper and ponderosa pine within each of the Wildcat Hills vegetation classes. Open Savanna Woodland Juniperus spp. 0% 17% 83% Pinus ponderosa 15% 22% 63%

Pine and Juniper Regeneration vs. Classiﬁcation

250 Mean(1617PineRegen(ha)) Mean(1617JuniperRegen(ha))

200

150

100

Pine and Juniper Regeneration (Seedlings+Saplings)/ha 50

0 Open Savanna Woodland Class17

Fig. 10- Mean regeneration (seedlings and saplings per hectare) of ponderosa pine and juniper across the three Wildcat Hills vegetation classes.

20 Although the effect of Vegetation class on regeneration was significant for both pine and juniper (Fig. 10, Table 3), neither the effect of grazing nor the grazing by vegetation interaction had a significant effect for either pine or juniper regeneration.

Table 3- ANOVA model results showing the response of ponderosa pine regeneration and juniper regeneration against vegetation classifications and presence or absence of grazing. Response Model Predictors Vegetation (3 levels) Grazing (2 levels) Vegetation x Grazing Pine Regen F 3.74 P=0.0300* F 0.98 P=0.3300 F 2.22 P=0.12 Juniperus Regen F 10.9 P<0.0001* F 0.15 P=0.6900 F 1.70 P=0.19

Understory Vegetation-

Understory vegetation was sorted and weighted in 5 categories (tall shrubs, short shrubs, herbaceous, woody debris (branches <7.6 cm and cones), and duff), but we simplified this to three categories for analyses. Tall and short shrubs were generally missing, so were added together with the other three categories in total understory biomass. Woody debris and duff were combined to create the forest floor category. The herbaceous category contains standing live and dead grasses and forbs. The effect of vegetation class, grazing and their interaction on each of the three biomass categories was tested with a separate two-way ANOVA (Table 4).

The dominant pattern for understory biomass was the significant decrease in herbaceous biomass (P<0.0001) and the significant increase in forest floor biomass

(P=0.0035) across the density gradent from Open to Savanna to Woodland plots (Figure

11, Table 4). Herbaceous biomass averaged 223 g/m2 in Open plots, 186 g/m2 in

Savanna plots and 133 g/m2 in Woodland plots. There was a significant effect of grazing 21 on herbaceous biomass (P=0.02). Although the interaction of grazing and vegetation class on herbaceous biomass was not significant, the effect of grazing decreased from Open to

Savanna to Woodland plots.

Forest floor biomass was more heterogeneous than herbaceous biomass, but increased significantly across the vegetation classes from an average of 175 g/m2 in

Open plots to 397 g/m2 in Savanna plots to 582 in Woodland plots (Fig.11). Neither grazing nor the grazing by vegetation class interaction had a significant effect on forest floor biomass (Table 4).

Fig. 11- Three aspects of understory biomass (herbaceous, forest floor, and total) ungrazed and grazed plots across the three vegetation classes (means and standard errors, see Table 4 for two-way ANOVA results).

22 Table 4 – Probabilities (P-values) associated with F statistics testing the effect of Vegetation Class (3 levels), Grazing (2 levels) and their interaction on three aspects of understory biomass (forest floor, herbaceous, and total). See Fig.11 for results. Significance Levels Vegetation x Response Vegetation Class Grazing Grazing forest floor 0.0035* 0.87 0.72 herbaceous biomass 0.0001* 0.02* 0.0533 total understory biomass 0.03* 0.61 0.56

DISCUSSION:

Most of the Wildcat Hills landscape is open grassland with scattered ponderosa pine. Although names and criteria for vegetation classes differ among classification systems, our 630-hectare study area was classified with remote sensing imagery as predominantly open or low density by both NLCD (62% Grass/Herbaceous) and

LANDFIRE (52% for Herbaceous-Grassland and Shrubland classes combined). These two landcover classifications correctly distinguished between high- and low-density woodlands for about 80-85% of our ground-based plots. However, neither NLCD nor

LANDFIRE were useful in distinguishing between plots we classified as Open (<70

TPH) or Savanna (<450 TPH). When the NFS criteria, which are based on basal area, were applied to our plots, 73% of the study area was classified as either Savanna or

Light-Density. Using our own classification based on TPH and basal area, 65% of the study area fell into the two vegetation classes with low tree density (Open and Savanna).

Note the confusion regarding the term savanna; this study’s Savanna class is roughly equivalent to NFS’ Light-Density class, and NFS’ Savanna class parallels this study’s

Open class. 23 With the goal of understanding if the study area was representative of the larger

Wildcat Hills landscape, we explored the NFS vegetation classification system (NFS

Growth & Drain Study, 2018) which was used for the entire 44,500-hectare Wildcat Hills region (cropland excluded, see NFS Growth & Drain Study, 2018 for details). While 27% of the study area fell in NFS’ High-Density classification, only 10% of the Wildcat Hills region was classified by NFS as High-Density. Thus, we may assume that vegetation composition and structure within vegetation classes is representative of comparable sites across the Wildcat Hills, however at the landscape scale, our study area appears to be more heavily wooded than the Wildcat Hills on average.

Across ponderosa pine’s distribution, including portions of the Pine Ridge and

Black Hills in the Great Plains, natural resource managers have noted that pine stands are

“too dense” (i.e., beyond the natural range of variability for ponderosa pine) because of high rates of ponderosa pine recruitment early in the 20th century after logging and/or fire exclusion.

Considering ponderosa pine alone (excluding juniper), basal area ranged from 0.7 m2/ha on average in Open plots, to 8.4 m2/ha in Savanna plots, and 13.5 m2/ha in

Woodland plots. However, pine stands that are relatively dense in the Wildcat Hills still have low density (as either TPH or Basal Area) when compared to proposed ecological thresholds for pine density based on other regions. Only 10% percent of this study’s plots would be considered dense enough to support pine beetle outbreaks by Negron and Popp

(2004), who proposed a 17 m2/ha threshold. Within Woodland plots, 29% of stands appear susceptible to pine beetle outbreaks. Using the NFS (2018) silviculture criteria for fully stocked ponderosa pine stands (20 m2/ha), only 5% of this study’s stands would be 24 considered over-stocked for timber production. Thus, if portions of the Wildcat Hills woodlands are considered too dense or over-stocked, it is generally not because of ponderosa pine alone.

Alternatively, ponderosa pine woodlands may be threatened by fire, insects or drought because of crowding from encroaching tree and shrubs, particularly junipers. In this study, junipers contributed only 6% of the basal area in Open and Savanna plots, but

29% of the basal area and 54% of the TPA, on average, in Woodland plots. Using a threshold of 17 m2/ha (Negron and Popp, 2004), 20% of the study area is at risk for beetle outbreaks based on combined pine and juniper data, compared to 10% when considering pines alone.

Approximately 70% of the Woodland stands in this study had >15 m2/ha basal area and >10% juniper basal area, i.e. relatively high basal area with a significant juniper component. Considering that junipers in the Wildcat Hills do not self-prune and provide excellent ladder fuels under pines, juniper is likely contributing to increased risk of intense fire in most Woodland plots. Determining a threshold for fire severity in response to both pine and juniper abundance in the Wildcat Hills requires further field-work and modeling at the stand level. Just as important to fire behavior, is the distribution and structure of woodland stands across the landscape. If we assume that 70% of the

Woodland stands are at risk of intense fire behavior (a guestimate unsupported by models), then 25% of the study area would be at risk. However, less than 10% of the

Wildcat Hills landscape as a whole would be at risk (based on the NFS regional land cover classification). Given the spatial behavior of fire, these would be two distinct fire scenarios. 25 Other ponderosa pine regions in the Great Plains have showed that juniper encroachment under ponderosa pine can cause crown fires resulting in >90% pine mortality (Hefner & Wedin, unpublished data). Where fuel stratum gaps are low or near

0 cm, the chance of crown fire and ponderosa pine mortality increases (Scott &

Reinhardt, 2001; Cruz et al., 2003). Our results show that an overwhelming majority of the plots surveyed in the Wildcat Hills landscape have a fuel strata gap near 0 cm, with the fuel loads being predominantly juniper (Appendix 2). The general lack of a fuel strata gap in the Open and Savanna plots suggest the role of fire has been either minimal or excluded, at least in recorded 20th century history in the Wildcat Hills.

Juniper trees in our Woodland plots were generally small and dense, suggesting recent encroachment in dense pine areas. In contrast, juniper recruitment in Open and

Savanna plots was either low or absent (Fig. 10). This differs from the behavior of juniper in most of Nebraska, where juniper encroachment into grasslands appears to be accelerating. This includes ponderosa pine woodlands of the Niobrara Valley. This lack of juniper encroachment into the Wildcat Hills’ open landscapes cannot be explained by natural or prescribed fires, which have not occurred in our study areas for decades, or the presence or absence of grazing, which had no effect on juniper recruitment in this study.

Neither does dense herbaceous vegetation appear to preclude juniper establishment in

Wildcat Hills grasslands. The herbaceous biomass levels observed in this study are low compared to typical values for the Pine Ridge, the Niobrara Valley, and most of eastern and central Nebraska (NRCS Web Soil Survey). Juniper recruitment may be reduced in

Wildcat Hills grasslands by an environment too hot and dry to support seedling establishment. However, considering that other juniper species readily recruit in even 26 drier landscapes across the western US, that explanation is unsatisfying. Further research is needed on the current and future status of juniper in the Wildcat Hills’ open habitats, which comprise the majority of the Wildcat Hills landscape. In contrast to juniper, ponderosa pines have a relatively balanced size distribution indicating persistence and recruitment across the landscape. Together the seedling and tree size distribution data suggest that the low density of pine in the Open and Savanna areas have been a long-term feature on the Wildcat Hills landscape.

Our results emphasize the need for ground-based data collection when analyzing stand structure for the future resilience of woodland systems; especially poorly understood areas like the ponderosa pine woodlands of the Great Plains. We found that preexisting vegetation classes such as NLCD and LANDFIRE did a relatively poor job distinguishing between high- and low-density pine woodlands in the Wildcat Hills and thus were unable to distinguish between stands with and without significant juniper encroachment. Reconciling thresholds for woodland functioning identified by labor- intensive ground-based studies with the power and scope of regional classifications based on remote sensing remains a research priority.

27 CHAPTER TWO: A DENDROECOLOGICAL ANALYSIS OF PONDEROSA PINE IN THE WILDCAT HILLS AND RESPONSE TO RECENT DROUGHT

Dendrochronology is an important tool in assessing the health of a woodland through analysis of annual growth and observation of individual tree or stand response to climate variability (Pohl et al., 2006; Evans et al., 2017). Combining data on stand structure and annual growth allow for a more complete understanding of individual tree and stand growth patterns for future modeling scenarios including higher intensity drought events. Tree ring data provides a rich source of information to develop past climate scenarios and have traditionally been collected in areas that were climate- sensitive (Evans et al., 2017).

Climate change will eventually impact the geographic distribution of wooded and forested landscapes in the Western United States with increased drought, disease/insect occurrences, and higher intensity fires (Allen 2010, Williams et al. 2012; Bonan 2008;

Fettig et al., 2013). The status of forest stands with major disturbances like these are expected to reduce biomass significantly in the next century (Allen 2010; Williams et al.,

2012; Fettig et al., 2013; Allen et al., 2015; Millar and Stephenson, 2015). In Nebraska, ponderosa pine (Pinus ponderosa) stands have been threatened with encroaching juniper

(Juniperus spp.) and increases in stand density. For example, the 2012 drought and subsequent wildfires through the Niobrara Valley in North central Nebraska and Pine

Ridge in Northwestern Nebraska came along with complete tree mortality in areas with high stand density and juniper encroachment (Juniperus virginiana) in the understory.

The Wildcat Hills area in Nebraska has not seen this magnitude of tree mortality in the last century, but it has encountered the same drought events. As the Wildcat Hills are an 28 ecologically unique landscape and ponderosa pine is on its eastern most native boundary in Nebraska, it is imperative to develop practical landscape management decisions to prevent or delay losses. A dendroecological analysis of disturbance events and responses to stand density in the Wildcat Hills of Nebraska is a reasonable technique given the changing climate along with the threat of an aggressive native juniper encroaching native ponderosa pine woodlands (Brown 2006; Liu et al., 2009; Johnstone et al., 2016;

Symstad & Leis 2017).

The influence and response of stand density as well as climatic variables (e.g. drought) are poorly understood on a per tree or even per stand basis in the Wildcat Hills and within Nebraska’s ponderosa pine range in general. Understanding the current status of the ponderosa pine trees in the Wildcat Hills and how they’ve responded to past climatic events will help us understand how these ecosystem dynamics have been shaped over time. We attempt to understand the following objectives: 1) the age distribution of ponderosa pine trees measured 2) how growth of ponderosa pine trees responds to climate variability in open/savanna (low density) and woodland areas (high density).

METHODS:

A subset of ponderosa pine trees was selected for dendrochronological analyses from over 2,000 pines inventoried in a study of woodland structure in Nebraska’s Wildcat

Hills (Chapter 1). The two study areas, totaling 630 hectares, were located on Carter

Canyon Ranch (managed by Platte River Basin Environments) and Cedar Canyon

Wildlife Management Area (managed by Nebraska Game and Parks Commission) (Scott,

1998) (Fig. 1). 59 individual pines were selected to span the range of sizes observed for 29 pines in six of the plots used in Chapter 1. Four of these plots were classified as low density (classified as open or savanna in the previous study) and two plots were classified as high density (previously classified as woodland).

0 0.375 0.75 1.5 Km

0 0.25 0.5 1 Miles

Fig. 1. 200 survey points at the Carter Canyon Ranch (Platte River Basin Environments) and Cedar Canyon Wildlife Management Area (Nebraska Game and Parks Commission) in the Wildcat Hills.

Seventy-one ponderosa pine increment cores or tree cross sections were collected for analyses in October 2016. The 13 small trees or saplings (diameter < 10cm) were sampled with complete cross sections cut near ground level. The remaining trees were sampled by collecting cores (71 total) at approximately 1.3 meters height (standard height for measuring DBH) with a 5.15 mm increment borer (Forestry Suppliers Inc., Haglöf

Increment Borers).

Cores were dried, mounted on wooden mounts, and sanded with increasingly finer sand paper to a smooth surface. The cores were used to age ponderosa pine trees using standard dendrochronology methods (Brown et al., 2006; Johnson & Abrams 2009; 30 Torbenson et al., 2016). Tree ring width was measured using a Velmex Tree-Ring measurement system with accompanying J2X Tree Ring Measuring software to the nearest 0.001 mm. Four cores lacked accompanying data on size (DBH) and were removed from analyses. Ring widths for trees with multiple cores were averaged by year.

A dataset of tree rings for 51 trees was used in the final analysis.

Tree age correction was based on the average ring growth of a tree in its first decade, using a model by Fraver et al., 2011. The age correction for our ponderosa pines ranged from 5.6 to 18 years (mean 11.0 years). After the age was counted and corrected, we created a mixed model with diameter as a continuous variable and stand density as a categorical variable. This resulted in a regression model for both low and high density stands to compare age versus diameter. Trees that were dead or multi-stem were removed from this portion of the analysis. The two regression equations combined with size distributions (Chapter 1) allowed us to model the age distributions for low versus high density stands.

Analysis of climatic variability and tree growth-

The Palmer Drought Severity Index (PDSI) was used to analyze tree growth response to soil moisture over time (National Oceanic and Atmospheric Administration

(NOAA, 2018)) (Fig 2). The Palmer Drought Severity Index is created from available temperature and precipitation data to estimate relative dryness (NOAA, 2018). The PDSI is a commonly used as a standard in dendrochronology studies and widely accepted in the tree ring community (NCDC, 2018). A PDSI value below –2 is considered a moderate drought, -3 a severe drought, and -4 an extreme drought. 31

Fig. 2- Annual Palmer Drought Severity Index values over the past 120 years for the Western Panhandle in Nebraska to include the Wildcat Hills region (NOAA, 2018).

In addition to annual PDSI, we imported data for precipitation, mean temperature, and maximum temperature from the PRISM database (PRISM Climate Group, Oregon

State University). Monthly climate data was downloaded for the period December 2016 to January 1895 for my study area. We then averaged each year’s growing season (April-

September) and dormant season (October-March) values to derive 8 climatic variables

(PDSI, mean temperature, maximum temperature, and precipitation). Climate analyses in this study focused on the period 2000 – 2016, which had large fluctuations between drought and wet conditions (Figure 2). This includes the lowest PDSI value in over a century recorded during the 2012 drought.

The raw ring with data were analyzed using standard dendrochronology routines to calculate basal area increment (BAI) per year of growth for each tree core.

Conversions from ring width to BAI were done with the statistical software R (RStudio

Team (2015). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA v

3.4.4). 32 RESULTS:

Age distribution-

Across the 51 ponderosa pine trees analyzed, the mean age was 80 years old

(median = 87 years), and the oldest ponderosa pine sampled was estimated to be 166 years old (Fig. 3). Thirty-three of the ages are based on a single core or cross-sectional analysis per tree, while 15 are based on two cores per tree. The largest age class among the sampled trees was 90-100 years old. Of the trees aged, 25% were estimated to be older than 100 years old. These ages are corrected for the estimated time of tree growth to sampling height (1.3 m) (Fraver et al. 2011). The mean age correction for cored trees was

11 years and ranged from 5.6 to 18 years. Ring counts of trees sampled with cross sections (cookies) at ground level were not corrected for height. Note that Fig. 3 is a biased size distribution compared to the total population because trees used for age analyses were not chose randomly.

Count 6

0 20 40 60 80 100 120 140 160 180

Fig. 3- Histogram showing the range of ages of 51 trees measured, including the correction for growth to 1.3 meters height for cored trees.

33 Tree Diameter vs Age-

The relationship between age and diameter was highly significant, R2=0.619 (Fig.

4). Height was not a significant predictor of tree age in a model with diameter and height and was not used in analyses (F = 1.31, p= 0.258). We used a regression model to test the effect of stand density (categorical variable with high and low densities), tree size (cross section diameter or DBH) and their interaction on tree age (Table 1).

175

150

125

100

75 Age (years)

0 0 10 20 30 40 50 Diameter at 1.3m Height (cm)

Fig. 4- Estimated tree age versus tree diameter for 51 ponderosa pines. Ages are based on ring counts of tree cores (10cm) or cross sections (<10cm) and are corrected for estimated time of growth to 1.3 meters height (r2 = 0.619, F = 74.89, P< 0.0001).

For the overall model, R2 equaled 0.666 (p<0.0001), the effect of diameter on tree age was highly significant (Table 1). The significant interaction of diameter and stand density indicates that the dependence of tree age on tree size differed for low and high density stands (Table 1). The main effect of stand density (low vs high) was not significant. The significant stand density by diameter interaction appears to be driven by smaller (<10cm diameter) trees (Fig. 5). The typical small (<10cm) tree in high density 34 plots (i.e. forest) was significantly younger than comparable small trees in low density plots (i.e. open savanna) (Fig. 5, illustrated by representative cross sections in Fig. 6).

Table 1- Two-Way ANOVA model predicting the effect of Diameter, Stand Density (two levels), and the interaction of Density and Diameter on tree age (model r2 = 0.666) SS F Ratio Probability Diameter 40693 84.5 0.0001* Density 13.8 0.029 0.866 Density * Diam. 2953 6.13 0.0172*

Mean(Mean(correct age)) vs. Density 70 Mean

30 Estimated Age (years)

0 Low High Stand Density Fig. 5- Mean age (with standard error) of small (<10 cm diameter) ponderosa pine trees for low and high density plots.

Fig. 6- Representative cross sections of small ponderosa pines (<10cm diameter) collected at ground level in low-density (left) and high-density (right) stands in the Wildcat Hills. Estimated age of the tree on the left is 69 years and the tree on the right is 24 years.

Modeled tree age distribution in low- and high-density stands

In Chapter 1, size distributions for over 2,000 pine and juniper trees measured across 60 plots and spanning a wide range in stand densities were presented. The statistical model predicting tree age using diameter and stand density (R2=0.659, Table 1) allowed us to model the age distribution of all pines in high and low density stands at our study area given their diameters. Although the modeled age distribution has higher variability (e.g. lower confidence) than the original size distribution, the age distributions indicate large differences in both pine age and number, particularly for smaller trees.

In the modeled pine age distribution for high density stands (i.e., the Woodland class from Chapter 1), the average age of pines was 59 years old, the median was 52 years old and the estimated maximum age was 172 years old (Fig. 7). The population is clearly skewed to younger trees, with half the trees estimated to be less than 52 years old

(Fig. 7). 36

20 40 60 80 100 120 140 160

Fig. 7- Histogram showing the modeled age distribution of ponderosa pine trees within high-density woodland plots. Ages are predicted using the model in Table 1.

Within the low density stands (combined Open and Savanna classes from chapter

1), the average age of ponderosa pine trees was 78 years old (median = 76 years, Fig. 8).

The oldest trees in the open and savanna plots were estimated to be 142 years old.

Because the small trees that we aged in the open plots were generally greater than 40 years old (Fig. 5, Fig. 6), the model predicted that all small trees in high density plots are

>40 years old, creating a truncated distribution. The model also suggests that the two largest cohorts for pines in the low-density stands are 50-60 years old. (ca 1960) and 90-

100 years old (ca 1920). Figure 2 indicates that the decade around 1920 had high moisture conditions, while the decade around 1960 had variable conditions with both wet and dry years. 37

20 40 60 80 100 120 140 160

Fig. 8- Histogram showing the modeled age distribution of ponderosa pine trees within low-density plots (open grassland and savanna). Ages are predicted using the model in Table 1.

Tree growth & climatic variability-

The response of tree growth (basal area increment, BAI) to recent climate variability was evaluated by performing simple linear regressions of eight climate variables with annual BAI for the period 2000 to 2016. As in all dendrochronological studies, my tree ring data set is strongest (more samples and less prone to errors) in recent years and grows weaker going back in time. This 17-year period was characterized by three swings from wet to dry years, including the three most severe single-season droughts (based on annual PDSI) in the last 120 years (Fig 2). Growing-season PDSI

(Gpdsi) was the strongest predictor for BAI among the 8 climate variables with an R2 of

0.11 and p<0.0001 (Table 2).

38 Table 2- Significance of linear regressions relating tree growth (BAI) to climatic variables (G and D indicate growing and dormant season, PDSI = Palmer Drought Severity Index). Climatic variable R2 F Ratio Prob > F Gpdsi 0.11 74.9 <0.0001 Dpdsi 0.04 28.7 <0.0001 Gtempmean 0.02 10.11 0.0016 Dtempmean 0.007 4.15 0.0421 Gtempmax 0.02 13.25 0.0003 Dtempmax 0.006 3.41 0.06 Gprecip 0.08 45.42 <0.0001 Dprecip 0.02 14.29 0.0002

Using growing-season PDSI (Gpdsi) to represent climate variability from 2000-

2016, we created a mixed model examining the response of annual tree growth (BAI) to the effects of tree size (categorical: <20cm and >20cm diameter), tree age in the year a ring was formed, Gpdsi, stand density (categorical: high and low), and the stand density by Gpdsi interaction.

The size of a tree was the largest predictor of BAI (Table 3). For ponderosa pine in the Wildcat Hills, BAI increases almost linearly from age 1 to age 60, peaks around age 80 and slowly decreases for ages greater than 90 (Fig. 9). Trees less than 15cm diameter were sampled at ground level with cross sections, while larger trees were sampled with cores at 1.3m height. This inconsistency introduces an irregularity in the

BAI versus age relationship around 15-20 years old. The Tree Size effect is somewhat confounded in our analyses with Tree Age in the Mixed-Model analyses (Table 3), because small trees (cross sections collected from trees <10cm) contributed disproportionately to the data in Figure 9 for younger ages, while larger cored trees contributed to our estimates for older ages. Nevertheless, the two terms together (Tree

Age and Tree Size) account for significant variability in our BAI model, allowing us to 39 more clearly examine the effects of stand density and Gpdsi on tree growth, the focus of this study. On average, low density stands had 32% higher BAI than high density stands

(i.e., a comparison of Least Square Means, which predict the effect of stand density with other variables in the model set to mean values). Pine growth (BAI) responded strongly to Gpdsi, and that response differed for high and low density stands (i.e., a significant

Density x gpdsi interaction).

Table 3- A mixed-model examining the response of BAI to Tree Size (2 levels), Tree Age (increasing annually from 2000-2016), Stand Density (2 levels), Gpdsi (growing season Palmer Drought Severity Index) and the interaction of Density and Gpdsi. N=867 Observations representing the response of 15 trees over 17 years (N=867, r2=0.471, F=153.4, P<0.0001). Effect F Ratio Prob > F Tree Size 484.6 <0.0001 Tree Age 38.7 <0.0001 Stand Density 18.4 <0.0001 Gpdsi 79.9 <0.0001 Density x Gpdsi 12.35 0.0005

1000

800

600

400 Basal Area Increment (cm2) Increment Basal Area 200

0 20 40 60 80 100 120 140 Age

Figure 9- Mean BAI across 50 ponderosa pines versus the age of the tree when a particular tree ring was formed. Because of the range of total ages and missing rings from the center of many large trees, individual values for Mean BAI represent as few as 4 and as many as 50 trees. The splined curve is added for visualization.

Effect of stand density on tree growth from 2000-2016

To examine the significant interaction between stand density and the response of pines to drought (Table 2), we considered the mean growth of small (<10cm) versus large

(>20cm) trees over the 17-year period. Because trees between 10cm and 20cm in diameter were not equally represented for high and low density stands, they were not used in Figures 10 and 11. The growth of both small and large trees tracked climate fluctuations since 2000 (Figure 2), reflecting the significant impact of PDSI on growth

(Table 2). For example, the severe drought of 2012 was followed by moderate drought in

2013 (Figure 2). For both small and large trees, however, the cumulative impact of these droughts was larger in 2013, the second year of drought. By 2014, a wet year, growth had recovered dramatically. Although Figure 10 suggests that large pines may be responding 41 differently to interannual climate variation in low and high density stands, our analyses do not indicate that this difference is statistically significant. For example, the coefficients of variation for BAI in trees from low versus high density stands were similar. A larger sample size of trees or analysis of a longer time series may clarify the patterns in Figure 10.

In contrast, the growth of small trees (<10cm diameter) responded differently to recent climate variability in low and high density stands (Figure 11). This difference was not apparent from 2000 to 2006, which were mostly dry years (Figure 2). However, during recent wet periods following severe drought (2009-2011 and 2014-2016), the growth response of small trees in low density stands was much greater than that observed for small trees in high density stands (Figure 11). Since 2006, the performance of small trees in low and high density stands have diverged, with small pines in the dense woodland stands less able to take advantage of periodic wet years. As discussed above, small pines in low density stands (i.e. open grassland and savannas) tend to be older than small pines in dense stands (Fig. 5, Fig. 6). The BAI analyses (Fig. 11) suggest they are also more resilient to fluctuations between dry and wet conditions, which have become more pronounced in the last decade. 42

1200

1100

1000

900

800 Mean(bai)

700

600

500

2000 2002 2004 2006 2008 2010 2012 2014 2016 year

Fig. 10- Mean basal area increment (BAI) for trees greater than 20 cm DBH for low (red) and high (blue) density stands from 2000-2016.

120

100

60 Mean(bai)

2000 2002 2004 2006 2008 2010 2012 2014 2016 year

Fig. 11- Mean basal area increment (BAI) for trees less than 10 cm DBH for low (red) and high (blue) density stands from 2000-2016.

43 DISCUSSION:

We estimated the age of ~50 trees chosen to represent the ponderosa pine size range found in our study area in the Wildcat Hills (Fig. 3). These ages, together with modeled ages for over 1200 pines surveyed in Chapter 1 (Fig. 7, Fig. 8), indicated that the oldest trees were approximately 170 years old, with large cohorts of trees 90-110 years old and 40-60 years old. There are presumably older trees on the landscape, but they are rare. The 90-110 year-old cohort dominates the basal area of both high- and low-density stands, although high-density stands have large numbers of smaller pines and junipers. Both management and climate may explain the origin of this cohort, which established between 1905 and 1930. Dense ponderosa pine recruitment is documented for this period in the Pine Ridge, the Black Hills, and the Rocky Mountain Front Range as a result of post-settlement logging, grazing practices and fire suppression.

The Wildcat Hill’s climate record (Fig. 2) also indicates 1905-1930 was an usually wet, drought-free period, which would have promoted pine establishment.

Reconstructed drought indices (PDSI) using dendrochronology (Cook and Krusic, 2004) indicate the previous wet period of this duration in this region was 1825-1840, predating the oldest trees we observed, and no comparable wet periods occurred in the 1700’s.

Thus, it is unlikely that the pine woodlands of the Wildcat Hills have been in equilibrium with climate, even before anthropogenic climate change.

The relationship between tree age and tree size (i.e. diameter) differed significantly for ponderosa pines from low-density stands (open grassland and pine savannas) versus high-density stands (pine/juniper woodland). This difference was largely explained by small trees (<10cm diameter), which for a given size, were on 44 average older in low-density stands (Fig. 5 and Fig. 6). Thus, high-density woodland areas had smaller, younger and denser pines on average than open grasslands and savannas in the Wildcat Hills. Our dendrochronological analyses of annual growth (basal area increment, BAI) and recent climate also found that small pines in dense stands responded differently than small pines in open areas to the large climate fluctuations observed during this period 2010 to 2016 (Fig. 11). While small pines in open areas appeared to recover quickly from severe drought events in 2008 and 2012-2013, the small pines in dense stands had less recovery, suggesting that pines in low-density environments were more resilient to drought events.

Small trees generally face more competition from surrounding vegetation than larger trees, but the nature of competition for small trees differs in low- and high-density pine woodlands. In low-density stands, the competition is likely from herbaceous vegetation (mostly grasses) while in high-density stands that competition is from junipers and other pines (Chapter 1). There were both more juniper trees and seedlings (Chapter 1:

Fig.7 and Fig. 10) than small pines and seedlings in dense plots. Herbaceous biomass was also reduced in dense plots. Because junipers are more deeply rooted than grasses and are active year-round, they are able to deplete soil moisture significantly more than grasses

(Eggemeyer et al., 2008). Thus, ponderosa pines with competing junipers in dense plots likely experienced greater soil moisture depletion during droughts (e.g. 2012-2013), and increased competition for moisture following droughts (e.g. 2014-2016). This was clear in our study for small pines (Fig. 11) and may have been the case for large pines (Fig.

10), but further research is needed on the drought response and recovery of large pines in low and high density stands. Long-term soil moisture profiles from the Nebraska Sand 45 Hills show that drought impacts are shorter-term in grasslands than woodlands

(particularly juniper) because of large differences in rooting depth (Adane et al. 2018).

These insights into the age distribution and growth of ponderosa pine, together with predictions of increased drought and climate variability, can assist land managers in targeting specific high-density woodlands in the Wildcat Hills. Thinning of junipers and small pines in higher density woodlands may give higher resilience to remaining ponderosa pine when facing future drought events.

46 CHAPTER 3: MANAGEMENT IMPLICATIONS & FUTURE RESEARCH

From this study, we have evaluated and determined the current status of the

Wildcat Hills in Nebraska as well as discovered existing ponderosa pine management goals from multiple agencies across the United States. We found that understanding the existing landscape is critical for future management practices in specific areas. As the

Wildcat Hills are unique and have not encountered catastrophic disturbance in the last century, management of encroaching juniper species by creating a larger fuel strata gap underneath of the ponderosa pine trees will be critical for the resilience of the dense woodland structures. Implementation or reintroduction of fire in the open and savanna structures will also help prevent juniper encroachment into the grasslands outside of the dense woodlands while rejuvenating the ponderosa pine trees as they thrive in the presence of low intensity fire.

My hope for future research includes a landscape scale extrapolation of the study site with remote sensing and spatial analysis. This small study area in the Wildcat Hills is representative of denser woodland stands, but not of the landscape as a whole. More data collection needs to be done, covering a larger representation of landscape structures (i.e. open, savanna, woodland). To further analyze the impact of juniper encroachment on the individual trees, I would suggest collecting a much larger selection of ponderosa pine and juniper tree cores and tree cross sections. With a larger data set, further predictions could be made instead of only a small glimpse into the existing response.

47 REFERENCES:

Allen, C. D., Savage, M., Falk, D. a, Suckling, K. F., Thomas, W., Schulke, T., … Oct, N. (2002). Ecological Restoration of Southwestern Ponderosa Pine Ecosystems: A Broad Perspective ECOLOGICAL RESTORATION OF SOUTHWESTERN PONDEROSA PINE ECOSYSTEMS: A BROAD PERSPECTIVE. Ecological Applications, 12(5), 1418–1433.

Allen, C. D., Macalady, A. K., Chenchouni, H., Bachelet, D., McDowell, N., Vennetier, M., … Cobb, N. (2010). A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management, 259(4), 660–684.

Allen, C. D., Breshears, D. D., & McDowell, N. G. (2015). On underestimation of global vulnerability to tree mortality and forest die-off from hotter drought in the Anthropocene. Ecosphere, 6(8), art129.

Adane, Z. A., Nasta, P., Zlotnik, V., & Wedin, D. (2018). Impact of grassland conversion to forest on groundwater recharge in the Nebraska Sand Hills. Journal of Hydrology: Regional Studies, 15(December 2017), 171–183.

Altman, J., Fibich, P., Dolezal, J., & Aakala, T. (2014). TRADER: A package for Tree Ring Analysis of Disturbance Events in R. Dendrochronologia, 32(2), 107–112.

Arenas-Castro, S., Fernández-Haeger, J., & Jordano-Barbudo, D. (2015). A Method For Tree-Ring Analysis Using Diva-Gis Freeware On Scanned Core Images. Tree-Ring Research, 71(2), 118–129.

Aughlin, D. A. C. L., & Oore, M. A. M. M. (2011). A century of increasing pine density and associated shifts in understory plant strategies Reports Reports, 92(3), 556–561.

Bailey, J. D., & Covington, W. W. (2002). Evaluating ponderosa pine regeneration rates following ecological restoration treatments in northern Arizona, USA. Forest Ecology and Management, 155(1–3), 271–278.

Biondi, F. (1999). Comparing tree-ring chronologies and repeated timber Inventories as forest monitoring tools. Ecological Applications, 9(1), 216–227.

Biondi, F., & Qeadan, F. (2008). A Theory-Driven Approach to Tree-Ring Standardization: Defining the Biological Trend from Expected Basal Area Increment. Tree-Ring Research, 64(2), 81–96.

Bollenbacher, B. L., Graham, R. T., & Reynolds, K. M. (2014). Regional Forest Landscape Restoration Priorities: Integrating Historical Conditions and an Uncertain Future in the Northern Rocky Mountains. Journal of Forestry, 112(5), 474–483.

Bonan, G. B. (2008). Forests and Climate Change: Forcings, Feedbacks, and the Climate Benefits of Forests. Science, 320(5882), 1444–1449. 48 Brice, B., Lorion, K. K., Griffin, D., Macalady, A. K., Guiterman, C. H., Speer, J. H. J. H., … Wang, H. (2013). Signal Strength In Sub-Annual Tree-Ring Chronologies from Pinus ponderosa In Northern New Mexico. Tree-Ring Research, 69(2), 81–86.

Brown, P. M., & Sieg, C. H. (1999). Historical variability in fire at the ponderosa pine - northern Great Plains prairie ecotone, southeastern Black Hills, South Dakota. Ecoscience, 6(4), 539–547.

Brown, P. M. (2006). Climate Effects on Fire Regimes and Tree Recruitment in Black Hills Ponderosa Pine Forests. Ecology, 87(10), 2500–2510.

Brown, P. M., Battaglia, M. A., Fornwalt, P. J., Gannon, B., Huckaby, L. S., Julian, C., & Cheng, A. S. (2015). Historical (1860) forest structure in ponderosa pine forests of the northern Front Range, Colorado. Canadian Journal of Forest Research, 45(11), 1462–1473.

Burkett, V. R., Wilcox, D. A., Stottlemyer, R., Barrow, W., Fagre, D., Baron, J., … Doyle, T. (2005). Nonlinear dynamics in ecosystem response to climatic change: Case studies and policy implications. Ecological Complexity, 2(4), 357–394.

Chapin, F. S., Matson, P. a, & Mooney, H. a. (2002). Principles of terrestial ecosystem ecology. System (Vol. 21).

Cheek, J. W., Biondi, F., Sibold, J. S., & Tausch, R. (2012). Fuel Analysis and Potential Fire Behavior in Mixed Conifer Woodlands of the Great Basin, Nevada, USA. Physical Geography, 33(3), 205–228.

Cook, E.R., and P.J. Krusic. (2004). The North American Drought Atlas. Lamont- Doherty Earth Observatory and the National Science Foundation.

Cruz, M. G., Alexander, M. E., & Wakimoto, R. H. (2003). Assessing canopy fuel stratum characteristics in crown fire prone fuel types of western North America. International Journal of Wildland Fire, 12(1), 39–50.

Dai, A.. (2017). Dai Global Palmer Drought Severity Index (PDSI). Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory.

Dale, V. H., Joyce, L. A., McNulty, S., Neilson, R. P., Ayres, M. P., Flannigan, M. D., … Wotton, B. M. (2001). Climate Change and Forest Disturbances. BioS, 9, 723–734.

Dannenberg, M. P., & Wise, E. K. (2016). Seasonal climate signals from multiple tree ring metrics: A case study of Pinus ponderosa in the upper Columbia River Basin. Journal of Geophysical Research: Biogeosciences, 121(4), 1178–1189.

De L. Dantas, V., Batalha, M. A., & Pausas, J. G. (2013). Fire drives functional thresholds on the savanna-forest transition. Ecology, 94(11), 2454–2463. 49 Eggemeyer, K. D., Awada, T., Harvey, F. E., Wedin, D. A., Zhou, X., & Zanner, C. W. (2009). Seasonal changes in depth of water uptake for encroaching trees Juniperus virginiana and Pinus ponderosa and two dominant C4 grasses in a semiarid grassland. Tree Physiology, 29(2), 157–169.

Erickson, C. C., & Waring, K. M. (2014). Old Pinus ponderosa growth responses to restoration treatments, climate and drought in a southwestern US landscape. Applied Vegetation Science, 17(1), 97–108.

Evans, M. E. K., Falk, D. A., Arizpe, A., Swetnam, T. L., Babst, F., & Holsinger, K. E. (2017). Fusing tree-ring and forest inventory data to infer influences on tree growth. Ecosphere, 8(7).

Fettig, C. J., Reid, M. L., Bentz, B. J., Sevanto, S., Spittlehouse, D. L., & Wang, T. (2013). Changing Climates, Changing Forests: A Western North American Perspective. Journal of Forestry, 111(3), 214–228.

Fraver, S., Bradford, J. B., & Palik, B. J. (2011). Improving tree age estimates derived from increment cores: A case study of red pine. Forest Science, 57(2), 164–170.

Fulé, P. Z., Crouse, J. E., Paul, J., & Kalies, E. L. (2012). Do thinning and / or burning treatments in western USA ponderosa or Jeffrey pine- dominated forests help restore natural fire behavior? Forest Ecology and Management 269, 68–81.

Fule, P. Z., & Swetnam, T. W. (2013). Unsupported inferences of high-severity fire in historical dry forests of the western United States : response to Williams and Baker cially under changing climate. Reconstructions of dry western US forests in the late 19th century in. Global Ecology and Biogeography, 1–6.

Fule, P. Z., Covington, W. W., & Moore. (1997). Determining Reference Conditions for Ecosystem. Ecological Applications, 7(3), 895–908.

Gartner, M. H., Veblen, T. T., Sherriff, R. L., & Schoennagel, T. L. (2012). Proximity to grasslands influences fire frequency and sensitivity to climate variability in ponderosa pine forests of the Colorado Front Range. International Journal of Wildland Fire, 21(5), 562–571.

Griffin, D. R. (2001). Return to the Magic Well: Echolocation Behavior of Bats and Responses of Insect Prey. BioScience, 51(7), 555–556.

Griffis, K. L., Crawford, J. A., Wagner, M. R., & Moir, W. H. (2001). Understory response to management treatments in northern Arizona ponderosa pine forests. Forest Ecology and Management, 146(1–3), 239–245.

Grote, R., & Erhard, M. (1999). Simulation of tree and stand development under different environmental conditions with a physiologically based model. Forest Ecology and Management, 120(1–3), 59–76. 50 Hart, S. C., & Classen, A. T. (2003). Potential for assessing long-term dynamics in soil nitrogen availability from variations in delta N-15 of tree rings. Isotopes in Environmental and Health Studies, 39(1), 15–28.

Helms, J. A. (1998). The Dictionary of Forestry. Society of American Foresters.

Hessburg, P. F., Spies, T. A., Perry, D. A., Skinner, C. N., Taylor, A. H., Brown, P. M., … Riegel, G. (2016). Tamm Review: Management of mixed-severity fire regime forests in Oregon, Washington, and Northern California. Forest Ecology and Management, 366, 221–250.

Heyerdahl, E. K., Morgan, P., & Riser, J. P. (2008). Multi-season climate synchronized historical fires in dry forests (1650-1900), northern Rockies, USA. Ecology, 89(3), 705–716.

Holling, C. S. (1973). RESILIENCE AND STABILITY OF ECOLOGICAL SYSTEMS. Annual Review of Ecology and Systematics, 4(1973), 1–23.

Homer, C. G., Dewitz, J. A., Yang, L., Jin, S., Danielson, P., Xian, G., … Megown, K. (2015). Completion of the 2011 National Land Cover Database for the conterminous United States-Representing a decade of land cover change information. Photogrammetric Engineering & Remote Sensing, 81(5), 345–354.

Johnson, S. E., & Abrams, M. D. (2009). Age class, longevity and growth rate relationships: Protracted growth increases in old trees in the eastern United States. Tree Physiology, 29(11), 1317–1328.

Johnstone, J. F., Allen, C. D., Franklin, J. F., Frelich, L. E., Harvey, B. J., Higuera, P. E., … Turner, M. G. (2016). Changing disturbance regimes, ecological memory, and forest resilience. Frontiers in Ecology and the Environment, 14(7), 369–378.

Julius, S. H., West, J. M., Nover, D., Hauser, R., Schimel, D. S., Janetos, A. C., … Backlund, P. (2013). Climate change and U.S. natural resources: Advancing the nation’s capability to adapt. Issues in Ecology.

Kaufmann, M. R. (1996). To live fast or not: growth, vigor and longevity of old-growth ponderosa pine and lodgepole pine trees. Tree Physiology, 16(1_2), 139–144.

Kaye, M. W., Woodhouse, C. A., & Jackson, S. T. (2010). Persistence and expansion of ponderosa pine woodlands in the west-central Great Plains during the past two centuries. Journal of Biogeography, 37(9), 1668–1683.

Keane, R. E., Agee, J. A., Fule, P., Keeley, J. E., Key, C., Kitchen, S. G., … Schulte, L. (2008). Ecological Effects of Large Fires on North American Landscapes : Benefit or Catastrophe ? International Journal of Wildland Fire, 17, 696–712. 51 Kerhoulas, L. P., & Kane, J. M. (2012). Sensitivity of ring growth and carbon allocation to climatic variation vary within ponderosa pine trees. Tree Physiology, 32(1), 14– 23.

LANDFIRE, 2008, Existing Vegetation Type Layer, LANDFIRE 1.1.0, U.S. Department of the Interior, Geological Survey. Accessed 28 October 2010 at http://landfire.cr.usgs.gov/viewer/.

Larson, A. J., & Churchill, D. (2012). Tree spatial patterns in fire-frequent forests of western North America, including mechanisms of pattern formation and implications for designing fuel reduction and restoration treatments. Forest Ecology and Management, 267, 74–92.

Liu, Y., Stanturf, J., & Goodrick, S. (2010). Trends in global wildfire potential in a changing climate. Forest Ecology and Management, 259(4), 685–697.

Lloret, F., Keeling, E. G., & Sala, A. (2011). Components of tree resilience: Effects of successive low-growth episodes in old ponderosa pine forests. Oikos, 120(12), 1909–1920.

Long, J. N., & Shaw, J. D. (2012). A Density Management Diagram for Even-Aged Sierra Nevada Mixed-Conifer Stands. Western Journal of Applied Forestry, 27(4), 187–195.

Mast, J. N., & Wolf, J. J. (2004). Ecotonal changes and altered tree spatial patterns in lower mixed-conifer forests, Grand Canyon National Park, Arizona, U.S.A. Landscape Ecology, 19(2), 167–180.

Mast, J. N., & Wolf, J. J. (2006). Spatial patch patterns and altered forest structure in middle elevation versus upper ecotonal mixed-conifer forests, Grand Canyon National Park, Arizona, USA. Forest Ecology and Management, 236(2–3), 241– 250.

McCullough, I. M., Davis, F. W., & Williams, A. P. (2017). A range of possibilities: Assessing geographic variation in climate sensitivity of ponderosa pine using tree rings. Forest Ecology and Management, 402, 223–233.

Meneguzzo, D. M., & Liknes, G. C. (2015). Status and Trends of Eastern Redcedar (Juniperus virginiana) in the Central United States: Analyses and Observations Based on Forest Inventory and Analysis Data. Journal of Forestry, 113(3), 325–334.

Millar, C., & Stephenson, N. (2015). Temperate forest health in an era of emerging megadisturbance. Science, 349(6250), 823–826.

Moore, M. M., Covington, W. W., & Fulé, P. Z. (1999). Reference conditions and ecological restoration: A Southwestern ponderosa pine perspective. Ecological Applications, 9(4), 1266–1277. 52 Negrón, J. F., & Popp, J. B. (2004). Probability of ponderosa pine infestation by mountain pine beetle in the Colorado Front Range. Forest Ecology and Management, 191(1), 17–27.

North, E. A., D’Amato, A. W., Russell, M. B., & Johnson, G. R. (2017). The influence of sidewalk replacement on urban street tree growth. Urban Forestry and Urban Greening, 24(December 2016), 116–124.

Nutto, L., Machado, S. do A., Cavalheiro, R., & da Silva, L. C. (2012). Comparison of different procedures for growth ring measurement of Mimosa scabrella and Pinus taeda. SCIENTIA FORESTALIS, 40(94), 135–144.

Park Williams, A., Allen, C. D., Macalady, A. K., Griffin, D., Woodhouse, C. A., Meko, D. M., … McDowell, N. G. (2012). Temperature as a potent driver of regional forest drought stress and tree mortality. Nature Climate Change, 3, 292. Retrieved from

Phipps, R. L. (1985). Collecting, Preparing, Crossdating, and Measuring Tree Increment Cores. US Geological Survey Water Resources Investigations Report 85-4148, 48. Retrieved from http://library.wur.nl/WebQuery/clc/275330

Pierce, J. L., Meyer, G. a, & Jull, a J. T. (2004). Fire-induced erosion and millennial- scale climate change in northern ponderosa pine forests. Nature, 432(November), 87–90.

Pohl, K. A., Hadley, K. S., & Arabas, K. B. (2006). Decoupling tree-ring signatures of climate variation, fire, and insect outbreaks in central Oregon. Tree-Ring Research, 62(2), 37–50.

PRISM Climate Group, Oregon State University, http://prism.oregonstate.edu, created 4 Feb 2004

Reineke, L. H. (1933). PERFECTING A STAND-DENSITY INDEX FOR EVEN- AGED FORESTS. Journal of Agricultural Research, 46(7), 627–638.

Roccaforte, J. P., Huffman, D. W., Fulé, P. Z., Covington, W. W., Chancellor, W. W., Stoddard, M. T., & Crouse, J. E. (2015). Forest Ecology and Management Forest structure and fuels dynamics following ponderosa pine restoration treatments, White Mountains, Arizona, USA. Forest Ecology and Management, 337, 174–185

Roccaforte, J. P., Huffman, D. W., Fulé, P. Z., Covington, W. W., Chancellor, W. W., Stoddard, M. T., & Crouse, J. E. (2015). Forest structure and fuels dynamics following ponderosa pine restoration treatments, White Mountains, Arizona, USA. Forest Ecology and Management, 337, 174–185.

Roden, J. (2008). Cross-dating of tree ring delta O-18 and delta C-13 time series. Chemical Geology, 252(1–2), 72–79. 53 Romme, W. H., Veblen, T. T., Kaufmann, R., Sherriff, R., & Regan, M. (2003). Ecological Effects of the Hayman Fire. USDA Forest Service, General Technical Report RMRS-GTR-114, 181–195.

Schneider, R., Humpert, M., Stoner, K., & Steinauer, G. (2005). The Nebraska Natural Legacy Project. Nebraska Game and Parks Commision Publications, (25).

Schneider, R., Stoner, K., Steinauer, G., Panella, M., & Humpert, M. (2011). The Nebraska Natural Legacy Project - State Wildlife Action Plan. The Nebraska Game and Parks Commission.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Web Soil Survey. Available online. Accessed 2018.

Soule, P. T., & Knapp, P. A. (2011). Radial Growth and Increased Water-Use Efficiency for Ponderosa Pine Trees in Three Regions in the Western United States. The Professional Geographer, 63(3), 379–391.

Stanton, S. (2007). Effects of dwarf mistletoe on climate response of mature ponderosa pine trees. Tree-Ring Research, 63(2), 69–80.

Stevens, J. T., Safford, H. D., North, M. P., Fried, J. S., Gray, A. N., Brown, P. M., … Taylor, A. H. (2016). Average stand age from forest inventory plots does not describe historical fire regimes in ponderosa pine and mixed-conifer forests of western North America. PLoS ONE, 11(5), 1–21. https://doi.org/10.1371/journal.pone.0147688

Stoddard, M. T., McGlone, C. M., Fulé, P. Z., Laughlin, D. C., & Daniels, M. L. (2011). Native plants dominate understory vegetation following ponderosa pine forest restoration treatments. Western North American Naturalist, 71(2), 206–215.

Swetnam, T. W., Allen, C. D., & Betancourt, J. L. (1999). Applied Historical Ecology: Using the Past To Manage for the Future. Ecological Applications, 9(November), 1189–1206.

Symstad, A. J., & Leis, S. A. (2017). Woody Encroachment in Northern Great Plains Grasslands: Perceptions, Actions, and Needs. Natural Areas Journal, 37(1), 118– 127.

Torbenson, M. C. A., Stahle, D. W., Villanueva Díaz, J., Cook, E. R., & Griffin, D. (2016). The Relationship Between Earlywood and Latewood Ring-Growth Across North America. Tree-Ring Research, 72(2), 53–66.

Turner, M. G., Baker, W. L., Peterson, C. J., & Peet, R. K. (1998). Factors influencing succession: Lessons from large, infrequent natural disturbances. Ecosystems, 1(6), 511–523. 54 Twidwell, D., Fuhlendorf, S. D., Taylor, C. A., & Rogers, W. E. (2013). Refining thresholds in coupled fire–vegetation models to improve management of encroaching woody plants in grasslands. Journal of Applied Ecology, 50(3), 603– 613.

Uzoh, F. C. C., & Oliver, W. W. (2008). Individual tree diameter increment model for managed even-aged stands of ponderosa pine throughout the western United States using a multilevel linear mixed effects model. Forest Ecology and Management, 256(3), 438–445.

Vila-Lameiro, P., & Diaz-Maroto, I. J. (2005). Study of the influence of environmental factors on the width of growth rings in Quercus petraea (Matts.) Liebl. through interpretation of digital images. Wood Research, 50(3), 23–36.

Vila-Lameiro, P., & Diaz-Maroto, I. J. (2007). Use of digital images in computer-assisted dendrochronological analysis of deciduous hardwoods. Wood Research, 52(3), 15– 31.

Watson, E., & Luckman, B. H. (2002). The dendroclimatic signal in Douglas-fir and ponderosa pine tree-ring chronologies from the southern Canadian Cordillera. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere, 32(10), 1858–1874.

Williams, A. P., Allen, C. D., Millar, C. I., Swetnam, T. W., Michaelsen, J., Still, C. J., & Leavitt, S. W. (2010). Forest responses to increasing aridity and warmth in the southwestern United States. Proceedings of the National Academy of Sciences, 107(50), 21289–21294.

Williams, M. A., & Baker, W. L. (2012). Spatially extensive reconstructions show variable-severity fire and heterogeneous structure in historical western United States dry forests. Global Ecology and Biogeography, 21(10), 1042–1052.

Wilson, R. J. S., & Luckman, B. H. (2003). Dendroclimatic reconstruction of maximum summer temperatures from upper tree line sites in Interior British Columbia, Canada. The Holocene, 13, 851–861.

Wood, L. J., & Smith, D. J. (2015). Intra-annual dendroclimatic reconstruction for northern British Columbia, Canada, using wood properties. Trees-Structure and Function, 29(2), 461–474.

Woodall, C. W., Fiedler, C. E., & Milner, K. S. (2003). Stand density index in uneven- aged ponderosa pine stands. Canadian Journal of Forest Research, 33(1), 96–100.

Yocom, L. L., Shive, K. L., Strom, B. A., Sieg, C. H., Hunter, M. E., Stevens-rumann, C. S., & Fulé, P. Z. (2015). Forest Ecology and Management Interactions of fuel treatments, wildfire severity , and carbon dynamics in dry conifer forests. Forest Ecology and Management, 349, 66–72. 55 Ziaco, E., Biondi, F., & Heinrich, I. (2016). Wood Cellular Dendroclimatology: Testing New Proxies in Great Basin Bristlecone Pine. Frontiers in Plant Science, 7(October), 1–13.

Zuur, A. F., Ieno, E. N., & Smith, G. M. (2007). Analysing Ecological Data. Profiles of drug substances, excipients, and related methodology (Vol. 36).

56 Appendix 1– Understory Vegetation Percent Cover and Forest Floor Depth Mean of %Cover F ratio prob > F response Open Savanna Woodland Grass 5.34 0.0075 3.629 4.290 3.680 3.011 Forb 10.60 0.0001 1.042 1.580 0.890 0.730 Bare Soil 0.93 0.4000 2.36 2.750 2.110 2.285 Duff 1.51 0.2300 4.56 4.150 4.670 4.810 Tall Shrub 0.12 0.8908 0.033 0.042 0.036 0.024 Short Shrub 0.04 0.9635 0.183 0.167 0.190 0.190 Small Woody (10 hr fuels) 17.13 0.0001 0.758 0.097 0.860 1.230 Large Woody (>10 hr) 1.21 0.3063 0.092 0.042 0.083 0.143

Appendix Table 1

I analyzed percent cover of grasses, forbs, bare soil, duff, tall shrubs, short shrubs, small woody (10 hr fuels), and large woody (>10 hr fuels) in a oneway ANOVA for each response to vegetation class. The only categories that responded significantly were grass

(p=0.0075), forbs (p=0.0001), and small woody (p=0.0001) (Table 1).

Each of the percent cover variables were also tested in two-way ANOVA analyses

(analyses not presented). The Grazing and Vegetation interaction was only significant for forb percent cover, with an average of 2.2 (~13% cover) of forbs in ungrazed down to 1.4

(~6% cover) in grazed. Interestingly, bare soil did not have a significant response to vegetation, grazing or their interaction; bare soil had approximately ~15% cover across all three vegetation types. This suggest that open sites for seedling establishment are not limiting in the Wildcat Hills.

While analyzing duff depth (cm) vs duff biomass (g/m2) (Fig. 1), we found that the two were positively correlated with an R2 of 0.25 and a prob >F of 0.001. Duff depth 57 was only measured for 41/60 plot points observed. The results suggest that average duff depths in Woodland plots do not limit juniper regeneration.

Fig. 1- duff depth & duff biomass linear relationship.

58 Appendix 2- Fuel strata gap

The fuel strata gap (FSG) was measured at each tree in the 0.09 ha plot area. Fuel stratum gap characteristics are used to estimate potential crown fire behavior within forested systems (Cruz et al., 2003). We took the mean FSG for all trees measured (juniper and pine) in each plot. Figure 2 shows mean FSG versus total number of junipers (trees + regeneration). There were other similar regressions to Fig. 2, but the regressions were not significant. We cannot conclude that FSG increases significantly across the canopy gradient.

AveFSG(m) vs. TotalJuniperDensity(ha) 0.7 Open Savanna

0.6 Woodland

0.5

0.4

AveFSG(m) 0.3

0.2

0.1

0.0 0 500 1000 1500 TotalJuniperDensity(ha)

Fig. 2- Total juniper density (trees + regeneration) by the average fuel strata gap across open, savanna, and woodland plots studied in the Wildcat Hills survey area (Chapter 1). 59 Mean FSG across all plots for which there were measurements (n=40), 75% were less than 0.14 meters, the median was 0.036 meters, and only 10% had a mean FSG > 0.25 meters.

The mean FSG across all junipers is almost zero versus ~15 cm for ponderosa pine trees

(Fig. 3). These two are significantly different but are each very low.

Fig. 3- The mean fuel strata gap (m) for juniper and pine trees across the entire tree database from 2017 data collection (Chapter 1).