INVESTIGATING THE GROWTH DYNAMICS OF MĀMANE (SOPHORA

CHRYSOPHYLLA) ON MAUNAKEA, HAWAIʻI USING RADIOCARBON DATING AND CLASSICAL DENDROCHRONOLOGY METHODS

A THESIS SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAIʻI AT HILO IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF

MASTER OF SCIENCE

IN

TROPICAL CONSERVATION BIOLOGY AND ENVIRONMENTAL SCIENCE

NOVEMBER 2012

By

Kainana S. Francisco

Thesis Committee:

Patrick J. Hart, Chairperson Paul C. Banko James O. Juvik

Keywords: tree-rings, annual rings, crossdating, tree age, tree growth rate, tropical trees

! ii!

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Patrick Hart (Biology, UH Hilo), for providing all the

project resources and helping to conceive and design this project, and for his active role in

supporting me to further my education and opening the door to the many opportunities that have

led me to where I am today. Thank you to my committee members Dr. Paul Banko (USGS

Pacific Island Ecosystems Research Center) and Dr. Jim Juvik (Geography, UH Hilo) for their

guidance and support of my research. Thank you to Dr. Axel Timmerman (International Pacific

Research Center, UH Mānoa) for connecting us with tree-ring experts Dr. Edward R. Cook

(Tree-Ring Lab, Lamont-Doherty Earth Observatory, Columbia University), and Dr. Jinbao Li

(International Pacific Research Center, UH Mānoa), both of whom have been very instrumental

in the concept and design of this project and in the analysis of the data; thanks for being such

awesome mentors to me. Additional thanks to Jinbao for always answering the million questions

I had, graciously spending hours with me on the phone and computer helping me to wrap-up

analysis, and editing and commenting on my thesis writing. Thank you to the many student

assistants and interns for all their hard work in the collection and preparation of samples - Joshua

Pang-Ching and Iwikauikaua Joaquin for doing the chain sawing of my samples, and Rhea

Thompson and Leinaʻala Hall for helping with sanding. Thank you to Susan Cordell (U.S.

Forest Service, Institute of Pacific Islands Forestry) for helping us get access to our Pōhakuloa

site, and to Sam Brooks (U.S. Forest Service, Institute of Pacific Islands Forestry) for escorting

us and chain sawing my samples on several occasions at the Pōhakuloa site. And a very special

thank you to our laboratory and field technician, Tishanna Bailey Ben, who has helped

immensely with all my field and lab work, and probably dreams of tree-rings as much as I do.

Thank you to the faculty and staff at the Kīpuka Native Hawaiian Student Center, who have not

! iii!

only provided unlimited encouragement and support throughout my academic career at UH Hilo,

but also graciously allowed me to use their outdoor areas to process my samples. Māmane

samples were collected with the collaboration of the Hawaiʻi State Division of Forestry and

Wildlife (Permit - DOFAWHA2010-01). This project was funded by the National Science

Foundation’s Centers for Research Excellence in Science and Technology (CREST) (Award

#0833211).

! iv!

ABSTRACT

When trees form annual growth rings, the patterns can be used to provide information on tree age

and growth dynamics that is essential to developing appropriate forest management strategies.

Annual rings are not commonly produced in tropical trees because they grow in a relatively

aseasonal environment, however, in the subalpine zones of Hawaiʻi’s highest volcanoes, there is

often strong seasonal variability in temperature and rainfall. Using a combination of classical

dendrochronology methods and radiocarbon dating I determined that annual growth rings occur

in māmane (Sophora chrysophylla), a native hardwood tree species, on Maunakea, Hawaiʻi. I

developed a baseline chronology for māmane, and used this to explore the relationship between

local precipitation and the growth dynamics of this native tree. I also used tree-rings to estimate

the age and growth rates of these trees. This study is the first in the eastern tropical Pacific

region to utilize dendrochronology methods to gain a better understanding of the stand history,

growth dynamics, and life history strategies of any of a native forest tree. The results from this

research will greatly benefit efforts to protect, conserve, and effectively manage the subalpine

forest ecosystem on Maunakea.

! v!

TABLE&OF&CONTENTS!

ACKNOWLEDGEMENTS!.........................................................................................................................................!ii!

ABSTRACT!..................................................................................................................................................................!iv!

LIST!OF!TABLES!.......................................................................................................................................................!vi!

LIST!OF!FIGURES!....................................................................................................................................................!vii!

CHAPTER!1.!!INTRODUCTION!.............................................................................................................................!1!

CHAPTER!2.!!METHODS!..........................................................................................................................................!6!

Study!Sites!......................................................................................................................................................!6!

Study!Species!.................................................................................................................................................!6!

Field-Methods!................................................................................................................................................!7!

Sample-Preparation!....................................................................................................................................!9!

Radiocarbon-Dating!................................................................................................................................!10!

Crossdating-Methods!...............................................................................................................................!11!

Analysis-Methods!......................................................................................................................................!12!

Estimating-Growth-Rates-and-Age!.....................................................................................................!13!

CHAPTER!3.!!RESULTS!.........................................................................................................................................!15!

Radiocarbon-Dating!................................................................................................................................!15!

Crossdating-Summary-Statistics!.........................................................................................................!15!

Quality-of-the-Chronology!......................................................................................................................!16!

The-Relationship-Between-Ring-Patterns-and-Rainfall!..............................................................!17!

Growth-Rate-and-Age-Estimates-of-Māmane!..................................................................................!17!

CHAPTER!4.!!DISCUSSION!..................................................................................................................................!18!

APPENDICES!............................................................................................................................................................!24!

LITERATURE!CITED!..............................................................................................................................................!34&

! vi!

LIST OF TABLES Table 1. Estimates of ring numbers and median calibrated 14C age (BP represents before

2010), calibrated age interval, and 95.4% confidence interval probability from OXCAL 4.0 for eight māmane (Sophora chrysophylla) cross-sections from Puʻulāʻau, Hawaiʻi.

Table 2. General summary statistics from Cofecha of the pine/cedar (Pinus radiata, Pinus

jeffreyi, Cedrus deodara) and māmane (Sophora chrysophylla) tree-ring chronologies on Maunakea, Hawaiʻi.

Table 3. Physical characteristics, estimated age, and estimated growth rates of māmane

(Sophora chrysophylla) samples at three sites on Maunakea, Hawaiʻi.

! vii!

LIST OF FIGURES Figure 1. Map of Study Sites on Maunakea, Hawaiʻi. Figure 2. Average Monthly Rainfall (mm) at Pōhakuloa, Hawaiʻi from 1938-1976. Figure 3. Raw Tree-Ring Series and Mean of Māmane (Sophora chrysophylla) from

Pōhakuloa, Hawaiʻi Figure 4. Standard Chronology of Māmane (Sophora chrysophylla) from Pōhakuloa,

Hawaiʻi Figure 5. Detrended RBar and EPS of Māmane (Sophora chrysophylla) from Pōhakuloa,

Hawaiʻi Figure 6. Correlation of Māmane (Sophora chrysophylla) Ring Width to Monthly

Precipitation of Current Year and a 1-Year Lag at Pōhakuloa, Hawaiʻi. Figure 7. Age as a Function of Size of Māmane (Sophora chrysophylla) on Maunakea,

Hawaiʻi.

! 1!

CHAPTER 1. INTRODUCTION

Hahai no ka ua i ka ululāʻau. Rains always follow the forest.

(405 - Pukui, 1983)

In the Hawaiian worldview our existence is directly linked to the health and continuity of

our forest environments. In today’s changing climate the need to protect and sustain our natural

resources is important to our island communities. Public understanding of the role functioning

ecosystems play in our lives is growing, and therefore conservation and protection of endemic

species, biodiversity, and native forests are spreading and becoming mainstream ideas.

Understanding past, present, and future growth dynamics of our forest is key to understanding

how forest structure is affected by climate change, and will allow us to make better-informed

decisions about the management of these resources.

Dendrochronology is the study of the chronological sequence of annual growth rings in

trees (Stokes & Smiley 1968). It is often referred to as tree-ring dating, and is perhaps the most

common method for understanding past growth dynamics of forests (Stokes & Smiley 1968).

When trees form annual growth rings, the ring patterns can be used to gain information that can

help to better understand the stand history, growth rates, and life history strategies of a particular

forest ecosystem (Chambers et al. 1998; Worbes et al. 2003). This information is essential to

developing sustainable management systems of these valuable, finite resources, regardless of

whether it is for the conservation of biodiversity or for agricultural forestry (Jacoby 1989;

Kurokawa et al. 2003; Steenkamp et al. 2008).

Tree growth occurs in two ways. Primary growth involves an increase in tree height and

the lengthening of stems and branches through activity of the apical meristem, while secondary

growth involves an increase in tree diameter and the thickening of stems and branches through

! 2!

activity of the lateral meristem (Stokes & Smiley 1968; Schweingruber et al. 2006). The lateral

meristem is also responsible for the formation of the vascular cambium; cells formed inside of

the cambium develop into the xylem, which creates the woody part of the tree, while cells

formed outside of the cambium develop into the phloem, which is responsible for transporting

nutrients throughout the entire tree (Stokes & Smiley 1968). During each growing season, xylem

growth is laid down outside of the previous season’s growth, and appears as rings in a cross-

section of a tree (Stokes & Smiley 1968). The patterns of tree-rings in woody plants are created

by seasonal climatic growth factors, such as temperature or precipitation (Worbes 1995). The

amount of cambial growth depends on the level of stress caused by these climatic growth factors

(Worbes 1995).

Dendrochronology has been widely developed and practiced in the seasonal temperate

and semi-arid zones, where the existence of annual tree-rings was established centuries ago

(Worbes 1995). Until recently, tropical forests were viewed as places that were very difficult, if

not impossible, to practice tree-ring research (Jacoby 1989). Tropical trees have long been

described as having relatively fast and continuous growth due to the presumed uniform tropical

climate with a general lack of seasonality. This has led many researchers to accept the concept

of a lack of annual rings in the trees of tropical regions (Détienne 1989; Jacoby 1989). As a

result, the tropics were not a primary focus for dendrochronology, and knowledge of tropical

forests under natural conditions was weak (Worbes et al. 2003).

Surprisingly, tree-ring research conducted in the early twentieth century, such as Coster’s

investigations on the timbers of Java (1927 and 1928) or Berlage’s study on Tectona grandis

(1931), successfully demonstrated the existence of annual rings in the tropics (Jacoby 1989;

Worbes 1989). The lack of awareness of such successful work has misled many to accept the

! 3!

concept of a lack of annual rings in the tropics. Contrary to popular belief of a uniform tropical

climate, many tropical regions have been documented to have a distinct seasonal climate

(Worbes 1995). Studies have shown that species of trees growing in tropical forests with a

distinct and predictable dry season, pronounced seasonal rainfall, or seasonal flooding produce

annual growth rings (Détienne 1989; Worbes 1989; Worbes and Junk 1989; Stahle et al. 1999;

Worbes 1999; Enquist & Leffler 2001; Trouet et al. 2001; Dezzeo et al. 2003; Worbes et al.

2003; Fichtler et al. 2004; Schöngart et al. 2004). A study by Fichtler et al. (2003) on several

non-pioneer tree species in an old-growth, lowland wet forest in Costa Rica has shown that

annual rings can even be present in an everwet (non-seasonally wet) tropical rain forest. The

evidence supporting the existence of annual tree-rings in the tropics is substantial; annual tree-

rings have been demonstrated for many tree species in more than twenty tropical countries

(Worbes 2002).

Several studies have been conducted in Hawaiʻi examining tree growth rates through

repeated diameter measurements (Walker & Vitousek 1991; Gerrish & Mueller-Dombois 1999;

Pearson & Vitousek 2001; Hart 2010). In addition, Hart (2010) estimated the age of trees in an

old-growth montane Hawaiian wet forest using a combination of radiocarbon dating and the

crown-class model, a method based on diameter at breast height measurements and crown class

to estimate the age of individual trees. However, there are currently no published studies on

forest dynamics and life history strategies of trees in Hawaiian forest ecosystems based on

dendrochronology methods. Although the periodicity of the growth rings in Hawaiian forest

trees have not yet been documented, the evidence supporting the existence of annual growth

rings in other tropical areas raises the possibility that they may occur in some trees in Hawaiʻi, in

! 4!

places such as Hawaiʻi’s highest volcanoes, where strong seasonal variability in temperature and

rainfall occur.

In January 2010, I began exploring the slopes of Maunaloa and Maunakea on the island

of Hawaiʻi searching for possible candidate tree species that could be studied using tree-ring

dating methods. The most promising of the tree species examined was the māmane (Sophora

chrysophylla), a native hardwood tree that has long been known to produce rings, and is

ecologically significant to this ecosystem.

Maunakea is home to the largest remaining stand of māmane subalpine woodland,

approximately 22,000 ha, situated between an elevation of about 1,800 and 2,900 m (Scowcroft

1983; Scowcroft & Giffin 1983; Scowcroft & Conrad 1992; Hess et al. 1999). Open woodland

of almost pure māmane can be found on the eastern, northern, and western slopes of Maunakea,

while the south and southwestern slopes of Maunakea are comprised of a dense, mixed woodland

of māmane and naio (Myoporum sandwicense) (Scowcroft & Giffin 1983; Scott et al. 1984).

The māmane subalpine woodland on Maunakea provides habitat for the endemic

Hawaiian honeycreeper, palila (Loxioides bailleui), a U.S. federally listed endangered species

(Banko et al. 2002a,b). The developing seeds of the māmane are a primary food source for

palila, as well as for the larvae of Cydia spp., a specialist moth (Banko et al. 2002a,b). The

production of māmane flowers and seed pods occurs seasonally along the elevational gradient

with about four months separating peaks in reproduction, while production within an elevation

stratum varies greatly; as a result, this food source is available to the palila and the Cydia larvae

throughout the year (Banko et al. 2002b). The seeds of the māmane contain levels of

quinolizidine alkaloids that are toxic and potentially lethal to vertebrates, but the palila and

Cydia larvae are adapted to utilizing this food resource (Banko et al. 2002a). The specialized

! 5!

relationship that the Cydia larvae and the palila have with the māmane makes their survival

dependent on the health of the māmane subalpine woodlands.

No dendrochronology studies have been conducted in Hawaiʻi to understand the stand

history, growth dynamics, and life history strategies of any of our native forest trees.

Furthermore, there are no existing tree-ring chronologies for the eastern tropical Pacific region.

My thesis research focuses on determining if annual growth rings occur in the native hardwood

māmane tree on Maunakea, Hawaiʻi through the use of classical dendrochronology methods in

combination with radiocarbon dating. I developed a baseline chronology for māmane, and used

this to explore the relationship between local precipitation and the growth dynamics of this

native tree. I also used tree-rings to estimate the age and growth rates of these trees.

! 6!

CHAPTER 2. METHODS Study Sites

Located above the tradewind inversion, the climate of the subalpine woodlands on the

slopes of Maunakea is highly variable, but can typically be described as being cool and dry

(Scowcroft 1983; Juvik et al. 1993). Temperatures in the area range from nighttime lows near

freezing to daytime highs above 68°F (20°C) (van Riper 1980; Scowcroft & Conrad 1992).

Areas within the rainshadow of the mountain at all elevations receive a very low amount of

rainfall, with annual averages of 35-75 cm (van Riper 1980; Scott et al. 1984; Scowcroft &

Conrad 1992). The wet season generally occurs from November to April while the dry season

occurs from May to October (van Riper 1980; Scowcroft 1983). The soils in the area are very

fine sandy loams formed from volcanic ash, sand, and cinders (Scowcroft 1983). Furthermore,

the soils are poorly developed, unstable, and well drained, resulting in low water-retention

capabilities (Scowcroft 1983; Scowcroft & Conrad 1992). Precipitation is the chief climatic

factor affecting the vegetation in the subalpine zone of Maunakea (Hartt & Neal 1940).

Three sites were chosen for collecting māmane on various facing slopes of Maunakea

(Figure 1). The Pōhakuloa site is located on the southern slope (UTM 237964E, 2184036N)

with an elevational range of 2,045-2,100 m, and is a recently burned area from a fire in August

2009. The Puʻulāʻau site is located on the western slope (UTM 231041E, 2195195N) with an

elevational range of 2,200-2,785 m. The Puʻumali site is located on the northern slope (UTM

244771E, 2202276N) with an elevational range of 2,170-2,370 m.

Study Species

Sophora is one of many genera in the Fabaceae family (Rock 1920). This genus, of

nearly fifty species, is widely distributed throughout many temperate, tropical, and sub-tropical

! 7!

regions (Cumbie & Mertz 1962; Wagner et al. 1990). It is usually described as being woody,

and includes forms ranging from tall trees, to small, herbaceous plants, to sometimes perennial

herbs (Cumbie & Mertz 1962; Wagner et al. 1990). Sophora is represented in Hawaiʻi by one

endemic species, Sophora chrysophylla, or māmane or mamani (Rock 1920; Wagner et al.

1990). Māmane can be found from near sea level up to an elevation of 10,000 ft (near 3,000 m)

(Rock 1920; Lamb 1981). It is scattered throughout the dry shrublands and forests, the mesic

forests, and the rarely wet forests of the main islands, with the exception of Niʻihau and

Kahoʻolawe (Wagner et al. 1990). Māmane also holds a dominant role in the composition of

subalpine vegetation on East Maui and Hawaiʻi (Wagner et al. 1990). Māmane trees typically

average less than 8 m in height, but can reach heights of up to 15 m (Wagner et al. 1990;

Scowcroft & Conrad 1992). The trunk is sometimes as wide as two to three feet (0.6-0.9 m) in

diameter, and is covered in about a half-inch (1.3 cm) thick layer of a light-brown, deeply

corrugated bark (Rock 1913; Lamb 1981). The flowers are golden-yellow and pea-like in shape

(Lamb 1981). The seed pods are green when young, and brown as they mature and dry; they are

about 2-16 cm long and 0.5-2 cm wide (Wagner et al. 1990). The seeds are small, about 4-8 mm

in length, and range in color from yellow to orange when mature (Lamb 1981; Wagner et al.

1990). Māmane wood is hard, dense, heavy, smooth, and very durable (Abbott 1992; Handy &

Pukui 1998). The heart wood is a deep brown, while the sapwood is light brown to straw-

colored (Lamb 1981).

Field Methods

The use of increment borers to obtain radial cores from live trees is only possible in tree

species with very distinct rings and concentric ring formation (Worbes 1995). Hawaiian trees

often grow asymmetrically, and the oldest portion of the trunk is rarely located at the center

! 8!

(Hart 2010). During the early exploratory phase of this study, I inspected several increment

cores of māmane from the Puʻulāʻau site, and the rings were not clearly defined under the

microscope. In addition, the non-concentric growth habit of māmane made it necessary to

inspect entire cross-sections of a tree in order to visually identify complete ring formations.

Dead māmane trees at each study area were haphazardly located and a 3-6 cm thick

cross-sectional disc was collected from each individual with the use of a chainsaw. Samples

from Puʻulāʻau (n = 29) were collected in March-May 2010 and in April-May 2011. Samples

from Puʻumali (n = 15) were collected in May-June 2011. Samples from Pōhakuloa (n = 15)

were collected in September and November 2011. Standard sampling protocol in

dendrochronology involves collecting samples well below the first branches, but above any

buttresses of a tree trunk (Stokes & Smiley 1968, Speer 2010). However, samples in this study

were collected with attention to avoiding distorted growth patterns. Most samples were obtained

from tree trunks that were uniformly round and where the majority of the wood was intact, which

sometimes occurred several meters above the base of the tree trunk. All samples were labeled

with an identification number and notes on site name, GPS coordinates, elevation, diameter of

tree at point of collection, and date of collection were recorded in a logbook. Samples were

transported to the laboratory and stored in an air-conditioned room.

With no known mortality dates of these trees, I explored the possibility of assigning

calendar years to māmane rings by crossdating the ring width series to that of nearby pine and

cedar populations, which included Monterey Pine (Pinus radiata), Jeffrey Pine (Pinus jeffreyi),

and Deodar Cedar (Cedrus deodara). An increment borer was used to benignly extract a

continuous core of wood from the bark to the pith of these trees (Stokes & Smiley 1968, Speer

2010). 14 trees were sampled at 1,775 m (UTM 224759E, 2192586N) and 2,275 m (UTM

! 9!

228510E, 2195095N) on the western slope of Maunakea, and 10 trees were sampled at 2,470 m

on the northeastern slope of Maunakea (UTM 249561E, 2201240N). Two cores were collected

from each tree. Cores were stored in plastic straws, and labeled with an identification number.

Notes on site name, GPS coordinates of the grove, species, elevation, and date of collection were

recorded in a logbook. Samples were then transported to the laboratory.

Sample Preparation

My methods of sample preparation for both cross-sections and increment cores are

adapted from the standard guidelines for surfacing wood specimens used for tree-ring studies

(Stokes & Smiley 1968, Speer 2010). Since the cross-sections of māmane are from dead trees

from a very dry climate, the wood is already dried, and there is no need to dry them further. The

surfaces of the cross-sections are first leveled with an electric hand-held planer perpendicular to

the growth of the trunk. The surfaces are then sanded with an electric hand-held belt sander

using coarse grit P50, to P80, to P120 sandpaper. Finally, the surfaces are polished using an

electric hand-held orbital sander using fine grit P320, to P400, to P600 sandpaper.

Cores from the pine and cedar trees were removed from the plastic straws, and fixed to

wooden core mounts using white multi-purpose glue, and bound with masking tape to prevent

the sample from warping as the glue dried. Samples were allowed to dry for at least 48 hours in

an air-conditioned room. Once dry, the tape was removed and the cores were sanded with an

electric handheld belt sander using coarse grit P120 sandpaper, and then polished with an electric

handheld orbital sander using fine grit P320, to P400, to P600 sandpaper.

! 10!

Radiocarbon Dating

As trees fix atmospheric carbon dioxide into organic material during photosynthesis they

incorporate a quantity of radioisotope 14C of natural origin that approximates the level in the

atmosphere (Hua et al. 1999). When the center-most wood of a tree dies, the 14C in the plant

decays at a fixed exponential rate. Natural radiocarbon dating can directly determine the age of a

tree by using an equation based on the remaining amount of 14C in the wood and the fixed

exponential rate of radioactive decay. Because atmospheric radiocarbon is highly variable across

time, the result from a radiocarbon dating usually provides two or more possible ages (Worbes &

Junk 1999). Accuracy for the radiocarbon dating of organic material between 50 and 350 years

old is not great due to the fluctuations in atmospheric 14C during this period, however, the

accuracy of radiocarbon dating improves with age for trees older than 350 years (Stuiver et al.

1998). It is, therefore, essential to collect samples from the center of the tree, which is the oldest

part of the tree trunk.

The center can be easily identified by a single ring approximately 1 mm in diameter (Hart

2010). As with many large tropical trees, the centers often decompose. Using a 25 mm drill bit

a minimum of 5 mg of wood was taken from, or as close, to the approximate center of the 10

largest māmane cross-sections from the Puʻulāʻau site; larger samples were used for 14C analysis

because smaller trees were less likely to be old enough to obtain accurate age estimates (i.e.

greater than 350 y; Hart 2010). Wood shavings were stored in small, plastic vials, labeled with

the cross-section’s identification number, and sent to the NSF-Arizona Accelerator Mass

Spectrometry Laboratory at the University of Arizona to be radiocarbon dated.

The resulting radiocarbon values were calibrated using the program OXCAL 4.0 and the

calibration curve INTCAL04 for the northern hemisphere (Bronk Ramsey 1995, Bronk Ramsey

! 11!

2001, Reimer et al. 2004). OXCAL 4.0 identified all possible locations where the 14C ages

intersected the calibration curve, and calculated relative probabilities (summing to 95.4%) for

each date range.

The number of rings for each sample was estimated based on a sample’s diameter and the

average growth rate estimate, and when not available, the site’s average growth rate estimate.

Estimated ring counts may have been slightly underestimated since samples were collected from

dead trees that did not always have bark, which may have led to weathering or other loss of

several rings. Ring number estimates for each sample were later compared to the 14C calibrated

age estimates.

Crossdating Methods

Crossdating is the most fundamental tool in dendrochronology, and allows for the

determination of the exact year of growth for every ring (Stokes & Smiley 1968, Speer 2010).

This practice of matching ring patterns of wide and narrow rings in a tree helps to determine true

ring boundaries, checks that all rings are represented in a sample, and shows where rings are

missing or where a tree might have formed more than one ring in a single year (Stokes & Smiley

1968, Speer 2010). A tree-ring series produced without crossdating will most likely be

erroneous due to locally absent or false rings (Stokes & Smiley 1968, Speer 2010).

The ring boundaries of the māmane cross-sections were visually identified, checked, and

marked across three different radii that were drawn through the least disturbed sectors of the

wood starting from the same innermost growth ring to the outermost distinguishable and visually

cross-dated ring. The ring boundaries of the pine and cedar cores were also visually identified,

checked, and marked across the two subsamples of each sampled tree. These methods were

adapted from general crossdating protocols (Stokes & Smiley 1968, Speer 2010). Ring widths

! 12!

were then digitally measured at a precision of 0.001mm using a Velmex Measuring System

(Velmex, Inc.) and the MeasureJ2X computer software program (Speer 2010). Since the death

dates of the māmane samples were unknown, ring widths were measured arbitrarily with no

connection to calendar dates, and the produced tree-ring series were considered undated. Ring

widths from the pine and cedar cores were measured as dated rings since they were extracted

from live trees, and the growth ring closest to the bark used as a reference for the most recent

growing season.

The raw data were then imported into Cofecha, a statistical software program that

numerically verifies the accuracy of the crossdating (Grissino-Mayer 2001). Crossdating was

first conducted within individual pine and cedar cores and māmane cross-sections. After

crossdating within individual trees was verified as being accurate, Cofecha was then used to

check the crossdating within each site between the undated māmane samples and the dated

pine/cedar samples. An additional unpublished dataset of eight dated māmane ring width series

from the Pōhakuloa area was also used in the crossdating for the Pōhakuloa site (Note: This

unpublished dataset was provided by Patrick Baker, but no information on the locality or

number of samples/sub-samples was available. Attempts to contact him and find out about the

data were unsuccessful). Based upon results from the Cofecha output on within-site crossdating,

the necessary adjustments were made in calendar dates to the undated māmane samples using the

editing software program Edrm (Grissino-Mayer 2001). Updated raw ring-width data were then

compiled for each site and re-checked in Cofecha for dating accuracy.

Analysis Methods

Raw ring-width data was analyzed and a site chronology was developed for the

Pōhakuloa site using ARSTAN, a software program that removes age-dependent and non-climate

! 13!

trends in the data by a process known as standardization (Cook 1985). I used a Friedman

variable span super smoother growth curve with a variable span tweeter sensitivity level set at

nine (least sensitive) to detrend the tree-ring series; this option is useful in modeling the growth

curves of disturbed tree-ring series that do not evolve through time in a homogeneous way. To

calculate the mean chronology, I used a robust (biweight) mean with bootstrap confidence limits.

I set the running RBar with a 20-year window and a 19-year overlap. All other options were set

to default.

The resulting standard chronology index was then used to investigate the relationship

between local climate and annual growth patterns of the māmane from the Pōhakuloa site using

the software program PcReg. Local precipitation data was gathered from a nearby rain gauge

station at 1,986 m operated by the Hawaiʻi State Division of Forestry and Wildlife (SKN #107,

UTM 235330.14E, 2185636.34N), and spans from 1938-1976 (Figure 2) (Giambelluca et al.

2011). I conducted a Pearson correlation analysis to examine the relationship between the

current year’s tree-ring width and the monthly precipitation within the present year, as well as

the previous year (1-year lag). This will help determine the season most critical for tree growth.

Estimating Growth Rates and Age

Samples that were internally crossdated were evaluated for growth rate and age estimates.

Raw tree-ring measurements were used to calculate the mean ring width of each sample, which

represents the average yearly increase in the radius of the sample. An estimated average growth

rate for each site was also calculated. Majority of the māmane cross-sections were not

crossdatable on the outer edges near the bark and the inner sections near the heartwood. To age

the entire tree estimates were made based on the diameter of a cross-section and its calculated

mean ring width. An average of age and diameter size for each site was also calculated. The age

! 14!

of a sample may have been slightly underestimated since samples were collected from dead trees

that did not always have bark, which may have led to weathering or other loss of several rings.

! 15!

CHAPTER 3. RESULTS Radiocarbon Dating

Of the ten samples sent in for radiocarbon dating two were deemed too young to get

accurate dates. Median calibrated 14C ages of the other eight samples ranged from 106-180 years

(Table 1). 14C intersected the calibration curve between two and four times. Age intervals are

listed in the table in order of decreasing probability. Probabilities for Age Interval 1 (highest

probability interval) ranged from 46.6-69.2%, with variability in ages ranging from 89-142 years.

Of the eight samples included in the ring number estimate and 14C age comparison, two were

aged based on the sample’s average growth rate estimate (M4 and M5), and the other six were

aged based on a site average growth rate estimate. Five of the eight estimated ring counts fell

within the first age interval of highest probability (M5, M11, M13, M15, and M16).

Crossdating Summary Statistics

An 86-year long chronology spanning 1927-2012 was established from 6 of the 24 pine

and cedar trees sampled at the three different groves on the western and northeastern slopes of

Maunakea. This chronology included 591 rings with a 0.788 series intercorrelation, 0.328 auto-

correlation, and 0.402 mean sensitivity.

Due to the lack of clear ring boundaries in some māmane cross-sections, it was only

possible to crossdate a portion of the samples from two of the three study sites. Of the 29

māmane cross-sections collected from Puʻulāʻau, 12 were internally crossdated, of which 9 were

crossdated to the dated pine/cedar chronology. Of the 15 cross-sections collected from Puʻumali,

12 were internally crossdated, of which 7 were crossdated to the dated pine/cedar chronology.

Crossdating of māmane to pine/cedar yielded low correlation values. Further analysis of these

! 16!

sites have been postponed until there are more māmane samples that have been internally

crossdated that can be used to establish a reliable chronology with higher correlation values.

All 15 māmane cross-sections collected from Pōhakuloa were internally crossdated. In

addition, seven of the eight māmane series from the Baker collection crossdated well with the

pine/cedar chronology prior to the 1980s. Low correlation values resulted when I attempted to

crossdate the undated māmane series to the combined Baker and pine/cedar chronology.

Assuming that the māmane series from the Baker collection had been crossdated correctly with

the pine/cedar, and excluding the low correlation values after 1980, I then crossdated a total of

45 radii from the 15 undated māmane series to the 7 Baker series, and established an 86-year

long chronology spanning from 1926-2011. This chronology included 2003 rings with a 0.501

series intercorrelation, 0.194 auto-correlation, and 0.510 mean sensitivity. General summary

statistics of the pine/cedar and māmane chronologies are outlined in Table 2.

Quality of the Chronology

Individual raw tree-ring series (n = 52) and the mean of all the series from the 15 cross-

sections and seven previously dated series from the Baker collection from the Pōhakuloa site are

presented in Figure 3. Figure 4 presents the mean of all the raw tree-ring series and the

detrended curve of the mean, which is the standardized site chronology.

RBar is a measure of the strength of the common growth signal within a chronology

(Wigley et al. 1984). The RBar graph in Figure 5 shows a relatively strong common signal for

the chronology, with the least amount of variation in signal strength, expressed by smaller error

bars, from the 1970s to the late 1990s indicating a stronger common signal amongst the samples

for that time period. Expressed Population Signal (EPS) is a measure of the common variability

within a chronology in relation to sample depth (Wigley et al. 1984). A chronology with an EPS

! 17!

value greater than or equal to 0.85 can be reliably used for climate analysis (Wigley et al. 1984).

As seen in Figure 5, the EPS values for the Pōhakuloa māmane chronology are above the 0.85

level (red line).

The Relationship Between Ring Patterns and Rainfall

Results of the comparison between local precipitation data and the growth of māmane

from the Pōhakuloa site show that August of the previous year is a significant predictor of

growth (r = 0.377, p = 0.019) (Figure 6).

Growth Rate and Age Estimates of Māmane

On average the diameter of māmane increased at an annual rate of 3.78 mm at Puʻulāʻau,

3.71 mm at Puʻumali, and 4.65 mm at Pōhakuloa (Table 3). Ages of māmane trees were a

function of size (r = 0.767; p ≤ 0.001), and averaged 78 yrs for a 30 cm diameter at Puʻulāʻau, 68

yrs for a 25 cm diameter at Puʻumali, and 36 yrs for a 16 cm diameter at Pōhakuloa (Figure 7).

! 18!

CHAPTER 4. DISCUSSION

This study is the first in the eastern tropical Pacific region to utilize dendrochronology

methods to gain a better understanding of the stand history, growth dynamics, and life history

strategies of any of our native forest trees. The results from the research presented in this report

will greatly benefit efforts to protect, conserve, and effectively manage the subalpine forest

ecosystem on Maunakea.

Results from the radiocarbon analysis of this study show that the samples of māmane

analyzed are quite young. The samples submitted for radiocarbon dating were the largest in

diameter and therefore most likely the oldest; yet, 20% of the samples submitted for radiocarbon

dating were too young to be accurately dated by this method. Furthermore, although there were

high probability values associated with particular radiocarbon ages, the range of ages was

sometimes quite broad. Although radiocarbon analysis seemed to be a useful method of early

exploration for this study, it is neither the most cost-efficient nor the most accurate method for an

in-depth investigation into the stand dynamics of a species such as māmane.

Although difficult and time consuming, crossdating of māmane is possible, and it

provided a relatively cost-effective way to examine age structure and growth dynamics of this

native tree. Through this study I demonstrated that chronologies can be developed from māmane

cross-sections using standard dendrochronology methods, that māmane produces annual growth

rings, and that these chronologies can be reliably used in ecological and climate change studies.

In this discussion, I also address key points that can further strengthen the reliability of these

māmane chronologies and the results of the overall study.

Several statistics can be used to assess the quality of a chronology. The series

intercorrelation refers to the average of correlation coefficients based on the comparison of an

! 19!

individual sample to a master chronology developed from the remaining samples. The mean

sensitivity refers to the relative difference in ring width from one year to the next. The auto-

correlation refers to the correlation of a given ring width to the ring width in the previous year.

Tree-ring chronologies with low auto-correlation values are usually associated with high values

of mean series intercorrelation and mean sensitivity, and tend to contain high frequency variance

in the series and strong climate signals, which is useful for climate reconstruction (Fritts 1976).

The Pōhakuloa site statistics include high mean series intercorrelation and mean sensitivity, and

low auto-correlation values. The Pōhakuloa chronology has a high RBar value, as well as EPS

values greater than 0.85. These statistics show high inter-annual variations and strong common

signals, which indicate the chronology’s reliability as a proxy for climate change studies, as well

as its usefulness in studying the ecology of māmane.

Although there was high variability around the calibrated age estimates by radiocarbon

dating for māmane, estimated ring counts fell within or reasonably close to within the highest

probability age interval, providing support for the possibility of annual ring formation in

māmane. Crossdating is based on the interannual variation in the periodical growth patterns,

induced by seasonality in the climate, of many trees from a shared region. Trees that do not

show any synchronization in the variation of growth patterns will not crossdate (Stahle 1999).

My ability to crossdate multiple time-series within and amongst individual samples of māmane

provides strong evidence that the growth rings in māmane are indeed annual. Stronger evidence

for annual growth rings is provided when time-series of at least 50 to 100 years are crossdated

(Stahle 1999). Although the Pōhakuloa chronology has several time-series that are at least 50

years in length, the argument for annual growth rings in māmane on Maunakea can be further

strengthened once the chronologies for the Puʻulāʻau and Puʻumali sites are established since

! 20!

there are many more samples from these sites that are much longer than 50 years. Further

support that growth rings are indeed annual and not just growth bands can be provided when

trees of the same species at different locations in a given climate region are crossdated (Stahle

1999). Region-wide crossdating for māmane can be attempted once the chronologies for the

Puʻulāʻau and Puʻumali sites are established, however there may be a possibility that the climate

of these sites are too different for crossdating to occur.

Additional evidence supporting annual growth rings is provided if the site chronology is

strongly correlated with regional climatic data (Stahle 1999). Māmane growth at the Pōhakuloa

site was found to be significantly correlated to the rainfall that occurred in the previous August.

This information not only provides additional evidence of annual growth rings in māmane, but it

also has climatic and ecological implications. Māmane growth is greatest in a year following a

summer that ends with a heavier rainfall (wetter August). When considering the phenological

timeline of māmane, tree growth occurring in the late summer seems like a reasonable idea. In

most years leaf flush occurs by August, and there are few, if any, flowers and seed pods being

produced (P.C. Banko, 2012, personal communication). In August, at the end of summer and

into the fall, it seems likely that energy is being concentrated towards tree growth in the form of

wood production. Along with the age and growth rate estimates of this study, not only can this

information allow for further investigation into past climate and forest composition of this area,

but it can also help to understand how forest stand development is being affected by a changing

climate, and how the potential of restoration efforts on the mountain can be maximized, all of

which can help guide the management decisions of the mountain.

As with other tropical hardwood tree species, the development of centuries-long tree-ring

chronologies is quite challenging (Stahle 1999). Although many tropical tree species have

! 21!

annual growth rings, studying these trees is often complicated by complex wood anatomy, and

many times annual rings are clear in young trees with relatively large rings, but are often much

more difficult to identify in older trees with suppressed growth (Stahle 1999). Since the trees

were younger at Pōhakuloa compared to those of the other two sites, the growth rings in the

cross-sections were relatively concentric, evenly formed, and well-defined, which may have

resulted in the high correlation values. This may help to explain the success of crossdating all

samples to the bark at Pōhakuloa. In the case of the Baker tree-ring series used in the

development of the Pōhakuloa site chronology, are the correlations after 1980 weak because of

measuring inaccuracies due to anatomical complications as described above? This would need

to be verified by finding out more information on the Baker samples, but from my experience

working with māmane cross-sections it seems this could likely be a possibility. Although the

cross-sections can be rather old and large, the clear portions of the wood with the distinguishable

ring boundaries were usually found on smaller sections of the sample closer to the heartwood,

which led to shorter time-series being measured. If this somehow affected the ability of the

measured time-series from the other two sites to have overlapping timelines then it could explain

why I was not able to complete crossdating of all samples at the other two sites. This would

highlight the importance of collecting more samples that would fill the gaps in the timeline

where samples I have now do not overlap. This could lead to chronologies that go much farther

back in time at the other two sites than what I have now for the Pōhakuloa site.

As shown in my methods to anchor the rings of māmane to calendar dates, developing

centuries-long māmane chronologies, and chronologies of other native tree species based on

dead wood, seems linked to establishing chronologies of species, such as pine and cedar trees,

where current records are available through increment cores. Summarized below are several key

! 22!

strategies outlined by Stahle (1999) for expanding dendrochronology into the tropics, and would

be very useful in identifying species that could provide current chronological records. (1)

Identify species of known dendrochronological value that have a natural distribution that extends

into the tropics (e.g. Pseudotsuga menziesii in North America, Cedrela angustifolia and Juglans

australis in South America). (2) Build upon previous successes in a particular botanical family

(e.g. Tectona grandis, Vitex keniensis, Vitex payous, and Premna maxima from the Verbenaceae

family, and Taxodium distichum and Taxodium mucronatum from the Taxodiaceae family) or

genus (Pinus includes many tropical species, some of which are useful for dendrochronology).

(3) Look for species that have a clear seasonal segregation of leaf flush and leaf fall (e.g. Tectona

grandis and Pterocarpus angolensis), and with long periods between the fall and flush cycle,

because this appears to be an important factor in the formation of annual growth bands. (4)

Species that flush after the onset of the wet season rather than during the dry season seem to be

more useful. Choose species from the seasonally dry deciduous forests of the tropics,

particularly forests, which experience a single prolonged dry season. (5) Using anatomical

descriptions of tropical trees species can help to set sampling priorities.

To strengthen the Pōhakuloa chronology I suggest to not only find out about the Baker

samples, but also search for larger māmane samples that would provide longer time-series for

cross-dating, which would also extend the chronology further back in time. To continue

developing the māmane chronologies at Puʻulāʻau and Puʻumali I suggest that the primary focus

should be on developing a better anchoring chronology that goes as far back in time as possible,

and searching the entire mountain for trees that can add to this chronology, so that the māmane

samples already collected can be crossdated. Monterey Pine sampled on the north side of the

mountain was difficult to work with and produced very short series lengths. Jeffrey Pine

! 23!

sampled at the lower elevation site on the west side of the mountain has very unclear ring

boundaries, and could not be used. Jeffrey Pine sampled at the higher elevation site on the west

side of the mountain has clearer ring boundaries compared to that of the lower elevation site.

Growth rings are much clearer in the early part of tree growth, but tend to get pinched and more

difficult to work with towards the bark. Deodar Cedar from the higher elevation site on the west

side of the mountain has very clear and distinct ring boundaries, and contributed the majority of

the samples used to establish the anchoring chronology in this study.

Dendrochronology has not yet been developed in Hawaiʻi; this project represents an

introductory exploration of dendrochronology into native Hawaiian forest ecosystems, beginning

with a focus on a subalpine population of māmane on Maunakea, Hawaiʻi. Overall,

dendrochronology provides another way for natural resource managers to expand their

knowledge on the functioning of our native forest ecosystems and to gather the much needed

information that will enable them to develop more effective management systems. Non-

destructive methods of tree-ring dating will allow for the estimation of age and growth rates of

other native Hawaiian tree species alongside māmane, and would allow for the development of

growth models that can be used to assess the current state and project the future structure and

composition of native Hawaiian forest ecosystems. This type of research holds a very promising

future in Hawaiʻi in that it has the possibility of providing a historical, current, and future

portrayal of the tree population dynamics within a Hawaiian forest.

! 24!

Table 1. Estimates of ring numbers and median calibrated 14C age (BP represents before 2010), calibrated age interval, and 95.4% confidence interval probability from OXCAL 4.0 for eight māmane (Sophora chrysophylla) cross-sections from Puʻulāʻau, Hawaiʻi. Numbers of rings may be slightly underestimated due to weathered out surfaces. 14C ages intersected the calibration curve between two and four times. Age intervals are listed in order of decreasing probability. Five of the eight estimated ring counts fell within the first age interval of highest probability.

Sample ID

Ring Number Estimate (yrs)

14C Calibrated Age (median yrs BP)

Age Interval 1 (yrs BP),

Probability (%)

Age Interval 2 (yrs BP),

Probability (%)

Age Interval 3 (yrs BP),

Probability (%)

Age Interval 4 (yrs BP),

Probability (%)

M4 118 180 136 - 225, 48.0 253 - 310, 26.7 (-4) - 34, 17.2 73 - 115, 3.6

M5 112 126 9 - 151, 59.1 173 - 275, 36.3 M10 95 173 124 - 231, 47.0 243 - 296, 18.6 (-3) - 37, 17.6 65 - 119, 12.2

M11 89 114 11 - 150, 65.2 186 - 270, 30.2 M13 105 114 11 - 150, 65.6 186 - 270, 29.8 M15 122 106 20 - 145, 69.2 215 - 268, 26.2 M16 102 114 11 - 150, 65.3 187 - 270, 30.1

M17 110 152 166 - 285, 46.6 58 - 155, 31.8 (-2) - 43, 16.9

! 25!

Table 2. General summary statistics from Cofecha of the pine/cedar (Pinus radiata, Pinus jeffreyi, Cedrus deodara) and māmane (Sophora chrysophylla) tree-ring chronologies on Maunakea, Hawaiʻi.

Statistic Pine & Cedar Māmane

Time span 1927-2012 (86 yrs) 1926-2011 (86 yrs)

Number of cores/radii (trees) 12 (6) 52 (22)

Number of rings in series 591 2003

Mean length of series (yrs) 49.2 38.5

Series intercorrelation 0.788 0.501

Standard deviation 1.909 1.246

Auto-correlation 0.328 0.194

Mean sensitivity 0.402 0.510

! 26!

Table 3. Physical characteristics, estimated age, and estimated growth rates of māmane (Sophora chrysophylla) samples at three sites on Maunakea, Hawaiʻi.

Site Sample ID

Mean Ring Width (mm)

Diameter of Sample (cm)

Estimated Age (yrs)

Estimated Site Growth Rate

(mm)

Puʻu

lāʻa

u

M2 1.34 21.3 79

3.78

M4 2.31 54.8 118 M5 2.49 55.7 112 M9 1.48 26.3 89

M14 2.27 28.4 63 M20 1.97 23.6 60 M21 2.03 25.4 62 M27 1.95 24.3 62 M34 1.82 20.0 55 M35 1.66 27.7 83 M37 1.58 26.1 83 M38 1.76 26.1 74

Puʻu

mal

i

M40 2.04 26.7 65

3.71

M41 2.03 26.0 64 M42 2.15 28.5 66 M43 2.09 30.4 73 M44 1.83 20.1 55 M46 1.67 25.3 76 M49 1.56 26.0 83 M50 1.85 25.7 69 M51 1.72 21.4 62 M52 2.06 22.5 55 M54 1.68 25.9 77 M55 1.58 21.5 68

Pōha

kulo

a

PM1 2.28 13.8 30

4.65

PM2 2.91 16.2 28 PM3 2.04 12.2 30 PM4 2.23 21.1 47 PM5 1.90 13.6 36 PM6 2.33 22.6 48 PM7 2.14 13.6 32 PM8 2.84 15.7 28 PM9 1.69 13.2 39

PM10 2.73 19.5 36 PM11 2.43 18.9 39 PM12 2.70 16.9 31 PM13 1.97 17.7 45 PM14 2.36 14.2 30 PM15 2.34 16.1 34

! 27!

Figure 1. Study site map of Maunakea, Hawaiʻi. Cross-sections were obtained from dead māmane (Sophora chrysophylla) trees at

three sites on various slopes of Maunakea - Puʻulāʻau (n = 29), Puʻumali (n = 15), and Pōhakuloa (n = 15). Increment cores were obtained from nearby pine and cedar populations of Monterey Pine (Pinus radiata), Jeffrey Pine (Pinus jeffreyi), and Deodar Cedar (Cedrus deodara), from the western (n = 14) and northeastern (n = 10) slopes of Maunakea. Local precipitation data was gathered from a rain gauge station at 1,986 m in the Pōhakuloa site operated by the Hawaiʻi State Division of Forestry and Wildlife (SKN #107).

#*

¯0 1 20.5 Miles

Māmane  Tree  Ring  Project  

Study  Sites  on  Maunakea,  Hawai'i

5000

7000

9000

11000

13000

Legend

Pōhakuloa

Pu'ulā'au

Pu'umali

Pine/Cedar  Collection  Sites

#* Pōhakuloa  Rain  Gauge

Elevation  (ft)

! 28!

Figure 2. Average monthly rainfall (mm) from a 1,986 m rain gauge station at Pōhakuloa, Hawaiʻi (SKN #107, UTM 235330.14E,

2185636.34N) during 1938-1976 (Giambelluca et al. 2011).

0"

10"

20"

30"

40"

50"

60"

JAN" FEB" MAR" APR" MAY" JUN" JUL" AUG" SEP" OCT" NOV" DEC"

Average'Mon

thly'Rainfall'(mm)'

Month'

! 29!

Figure 3. Raw tree-ring series (black, n = 52) and the mean of all series (red) from 15 cross-

sections of māmane (Sophora chrysophylla) and seven previously dated māmane series (unpublished data from P. Baker) from Pōhakuloa, Hawaiʻi.

Year%

Raw%ring%width%(m

m)%

! 30!

Figure 4. (Top) Raw tree-ring series mean (black) and detrended curve (red) of an 86-year long

chronology (1926-2011) of māmane (Sophora chrysophylla) from Pōhakuloa, Hawaiʻi based on 15 cross-sections and seven previously dated māmane series (unpublished data from P. Baker). (Bottom) Number of samples used to produce the chronology.

Year%

Inde

x%Num

ber%o

f%sam

ples%

! 31!

Figure 5. (Top) Detrended RBar, with a 20-year window and a 19-year overlap, and Expressed

Population Signal (EPS) values of an 86-year long chronology (1926-2011) of māmane (Sophora chrysophylla) from Pōhakuloa, Hawaiʻi based on 15 cross-sections and seven previously dated māmane series (unpublished data from P. Baker). RBar is a measure of the strength of the common growth signal within a chronology (Wigley et al. 1984). This graph shows a relatively strong common signal for the chronology, with the least amount of variation in signal strength, expressed by smaller error bars, from the 1970s to the late 1990s indicating a stronger common signal amongst the samples for that time period. (Bottom) Expressed Population Signal (EPS) is a measure of the common variability within a chronology in relation to sample depth (Wigley et al. 1984). A chronology with an EPS value greater than or equal to 0.85 can be reliably used for climate analysis (Wigley et al. 1984). The EPS values for the Pōhakuloa māmane chronology are above the 0.85 level (red line). These statistics indicate that this chronology is reliable, and can be used in climate change or ecological studies.

Year%

RBar%

EPS%

! 32!

Figure 6. A comparison between local precipitation data (1938-1976) and an 86-year long chronology, spanning from 1926-2011, of

māmane (Sophora chrysophylla) from Pōhakuloa, Hawaiʻi. A Pearson correlation analysis was conducted on the current year’s tree-ring width and the monthly precipitation within the present year, as well as the previous year (1-year lag). Results indicate that rainfall from the previous August is a significant predictor of growth, highlighted in the graph by the red bar (r = 0.377, p = 0.019).

!0.400%

!0.300%

!0.200%

!0.100%

0.000%

0.100%

0.200%

0.300%

0.400%

0.500%

JAN%(1)%

FEB%(2)%

MAR%(3)%

APR%(4)%

MAY%(5)%

JUN%(6)%

JUL%(7)%

AUG%(8)%

SEP%(9)%

OCT%(10)%

NOV%(11)%

DEC%(12)%

JAN%(13)%

FEB%(14)%

MAR%(15)%

APR%(16)%

MAY%(17)%

JUN%(18)%

JUL%(19)%

AUG%(20)%

SEP%(21)%

OCT%(22)%

NOV%(23)%

DEC%(24)%

Pearson(Co

rrela+

on((r)(

Monthly(Precipita+on(

! 33!

Figure 7. Age as a function of size (diameter of cross-sections) of māmane (Sophora chrysophylla) on Maunakea, Hawaiʻi. Māmane

from Puʻulāʻau averaged 78 yrs for a 30 cm diameter, 68 yrs for a 25 cm diameter at Puʻumali, and 36 yrs for a 16 cm diameter at Pōhakuloa.

y"="2.1982x"+"7.6585"R²"="0.76736"p"≤""0.001"

20"

40"

60"

80"

100"

120"

5.0" 15.0" 25.0" 35.0" 45.0" 55.0"

Es#m

ated

)Age)(yrs))

Diameter)of)Cross5sec#on)(cm))

! 34!

LITERATURE CITED

Abbott, I.A. 1992. Lāʻau Hawaiʻi Traditional Hawaiian Uses of Plants. Honolulu, Hawaiʻi: Bishop Museum Press.

Banko, P.C., M.L. Cipollini, G.W. Breton, E. Paulk, M. Wink, and I. Izhaki. 2002a. Seed chemistry of Sophora chrysophylla (Māmane) in relation to diet of specialist avian seed predator Loxioides bailleui (Palila) in Hawaiʻi. Journal of chemical Ecology 28: 1393-1410.

Banko, P.C., P.T. Oboyski, J.W. Slotterback, S.J. Dougill, D.M. Goltz, L. Johnson, M.E. Laut, and T.C. Murray. 2002b. Availability of food resources, distribution of invasive species, and conservation of a Hawaiian bird along a gradient of elevation. Journal of Biogeography 29: 789-808.

Bronk Ramsey, C. 1995. Radiocarbon calibration and analysis of stratigraphy: the OxCal program. Radiocarbon 37: 425-430.

Bronk Ramsey, C. 2001. Development of the radiocarbon calibration program OxCal. Radiocarbon 43: 355-363.

Chambers, J.Q., N. Higuchi, and J.P. Schimel. 1998. Ancient trees in Amazonia. Nature 391: 135-136.

Cook, E.R. 1985. A time-series analysis approach to tree-ring standardization. Ph.D. Dissertation, The University of Arizona. Tucson, Arizona, USA.

Cumbie, B.G., and D. Mertz. 1962. Xylem anatomy of Sophora (Leguminosae) in relation to habit. American Journal of Botany 49: 33-40.

Détienne, P. 1989. Appearance and periodicity of growth rings in some tropical woods. IAWA Bulletin n.s. 10: 123-132.

Dezzeo, N., M. Worbes, I. Ishii, and R. Herrera. 2003. Annual tree rings revealed by radiocarbon dating in seasonally flooded forest of the Mapire River, a tributary of the lower Orinoco River, Venezuela. Plant Ecology 168: 165-175.

Enquist, B.J., and A.J. Leffler. 2001. Long-term tree ring chronologies from sympatric tropical dry-forest trees: individualistic responses to climatic variation. Journal of Tropical Ecology 17: 41-60.

Fichtler, E., D.A. Clark, and M. Worbes. 2003. Age and long-term growth of trees in an old-growth tropical rain forest, based on analyses of tree rings and 14C. Biotropica 35: 306-317.

Fichtler, E., V. Trouet, H. Beeckman, P. Coppin, and M. Worbes. 2004. Climatic signals in tree rings of Burkea africana and Pterocarpus angolensis from semiarid forests in Namibia. Trees 18: 442-451.

Fritts, H.C. 1976. Tree Rings and Climate. New York, New York: Academic Press.

! 35!

Gerrish, G., and D. Mueller-Dombois. 1999. Measuring stem growth rates for determining age and cohort analysis of a tropical evergreen tree. Pacific Science 53: 418-429.

Giambelluca, T.W., Q. Chen, A.G. Frazier, J.P. Price, Y-L. Chen, P-S. Chu, J. Eischeid, and D. Delparte. 2011. The Rainfall Atlas of Hawai‘i. http://rainfall.geography.hawaii.edu.

Grissino-Mayer, H.D. 2001. Evaluating crossdating accuracy: a manual and tutorial for the computer program Cofecha. Tree-Ring Research 57(2): 205-221.

Handy, E.S.C. and M.K. Pukui. 1998. The Polynesian Family System in Kaʻū, Hawaiʻi. Honolulu, Hawaiʻi: Mutual Publishing.

Hart, P.J. 2010. Tree growth and age in an ancient Hawaiian wet forest: vegetation dynamics at two spatial scales. Journal of Tropical Ecology 25: 1-11.

Hartt, C.E., and M.C. Neal. 1940. The plant ecology of Maunakea, Hawaiʻi. Ecology 21: 237-266.

Hess, S.C., P.C. Banko, G.J. Brenner, and J.D. Jacobi. 1999. Factors related to the recovery of subalpine woodland on Maunakea, Hawaiʻi. Biotropica 31: 212-219.

Hua, Q., M. Barbetti, M. Worbes, J. Head, and V.A. Levchenko. 1999. Review of radiocarbon data from atmospheric and tree ring samples for the period 1945-1977 A.D. IAWA Journal 20: 261-283.

Jacoby, G.C. 1989. Overview of tree-ring analysis in tropical regions. IAWA Bulletin n.s. 10: 99-108.

Juvik, J.O., D. Nullet, P. Banko, and K. Hughes. 1993. Forest climatology near the tree line in Hawaiʻi. Agricultural and Forest Meteorology 66: 159-172.

Kurokawa, H., T. Yoshida, T. Nakamura, J. Lai, and R. Nakashizuka. 2003. The age of tropical rain-forest canopy species, Borneo ironwood (Eusideroxylon zwageri), determined by 14C dating. Journal of Tropical Ecology 19: 1-7.

Lamb, S.H. 1981. Native Trees & Shrubs of the Hawaiian Islands. Santa Fe, New Mexico: The Sunstone Press.

Pearson, H.L., and P.M. Vitousek. 2001. Stand dynamics, nitrogen accumulation, and symbiotic nitrogen fixation in regenerating stands of Acacia koa. Ecological Applications 11: 1381-1394.

Pukui, M.K. 1983. ʻŌlelo Noʻeau: Hawaiian Proverbs and Poetical Sayings. Honolulu, Hawaiʻi: Bishop Museum Press.

Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, C.J.H. Bertrand, P.G. Blackwell, C.E. Buck, G.S. Burr, K.B. Cutler, P.E. Damon, R.L. Edwards, R.G. Fairbanks, M. Friedrich, T.P. Guilderson, A.G. Hogg, K.A. Hughen, B. Kromer, G. McCormac, S. Manning, C. Bronk

! 36!

Ramsey, R.W. Reimer, S. Remmele, J.R. Southon, M. Stuiver, S. Talamo, F.W. Taylor, J. van der Plicht, and C.E. Weyhenmeyer. 2004. IntCal04 terrestrial radiocarbon age calibration, 0-26 cal kyr BP. Radiocarbon 46: 1029-1058.

Rock, J.F. 1913. The Indigenous Trees of the Hawaiian Islands. Honolulu, Hawaiʻi: Pacific Tropical Botanical Garden.

Rock, J.F. 1920. The Leguminous Plants of Hawaiʻi: Being an Account of the Native, Introduced and Naturalized Trees, Shrubs, Vines and Herbs, Belonging to the Family Leguminosae. Honolulu, Hawaiʻi: Hawaiian Sugar Planters’ Association.

Schöngart, J., W.J. Junk, M.T.F. Piedade, J.M. Ayress, A. Hüttermann, and M. Worbes. 2004. Teleconnection between tree growth in the Amazonian floodplains and the El Niño-Southern Oscillation effect. Global Change Biology 10: 683-692.

Schweingruber, F.H., A. Borner, and E.D. Schulze. 2006. Atlas of Woody Plant Stems: Evolution, Structure, and Environmental Modifications. Berlin, Germany: Springer.

Scott, J.M., S. Mountainspring, C. van Riper III, C.B. Kepler, J.D. Jacobi, T.A. Burr, and J.G. Giffin. 1984. Annual variation in the distribution, abundance, and habitat response of the palila (Loxioides bailleui). Auk 101: 647-664.

Scowcroft, P.G. 1983. Tree cover changes in Māmane (Sophora chrysophylla) forests grazed by sheep and cattle. Pacific Science 37: 109-119.

Scowcroft, P.G., and C.E. Conrad. 1992. Alien and native plant response to release from feral sheep browsing on Maunakea. In: Stone, C.P., C.W. Smith, and J.T. Tunison. Alien plant invasions in native ecosystems in Hawaiʻi. Management and Research Cooperative National Park Resource Studies Unit, University of Hawaiʻi: 625-665.

Scowcroft, P.G., and J.G. Giffin. 1983. Feral herbivores suppress Māmane and other browse species on Maunakea, Hawaiʻi. Journal of Range Management 36: 638-645.

Speer, J.H. 2010. Fundamentals of Tree-ring Research. Tucson, Arizona: The University of Arizona Press.

Stahle, D.W. 1999. Useful strategies for the development of tropical tree-ring chronologies. IAWA Journal 20: 249-253.

Stahle, D.W., P.T. Mushove, M.K. Cleaveland, F. Roig, and G.A. Haynes. 1999. Management implications of annual growth rings in Pterocarpus angolensis from Zimbabwe. Forest Ecology and Management 124: 217-229.

Steenkamp, C.J., J.C. Vogel, A. Fuls, N. van Rooyen, and M.W. van Rooyen. 2008. Age determination of Acacia erioloba trees in the Kalahari. Journal of Arid Environments 72: 302-313.

! 37!

Stokes, M.A., and T.L. Smiley. 1968. An Introduction to Tree-ring Dating. Tucson, Arizona: The University of Arizona Press.

Stuiver, M., P.J. Reimer, E. Bard, J.W. Beck, G.S. Burr, K.A. Hughen, B. Kromer, G. McCormac, J. Van Der Plicht, and M. Spurk. 1998. IntCal98 radiocarbon age calibration, 24,000-0 cal BP. Radiocarbon 40: 1041-1084.

Trouet, V., K. Haneca, P. Coppin, and H. Beeckman. 2001. Tree ring analysis of Brachystegia spiciformis and Isoberlinia tomentosa: evaluation of the ENSO-signal, in the miombo woodland of eastern Africa. IAWA Journal 22: 385-399.

van Riper, C. III. 1980. The phenology of the dryland forest of Maunakea, Hawaiʻi, and the impact of recent environmental perturbations. Biotropica 12: 282-291.

Wagner, W.L., D.R. Herbst, and S.H. Sohmer. 1990. Manual of the Flowering Plants of Hawaiʻi. Honolulu, Hawaiʻi: University of Hawaiʻi Press & Bishop Museum Press.

Walker, L.R., and P.M. Vitousek. 1991. An invader alters germination and growth of a native dominant tree in Hawaiʻi. Ecology 72: 1449-1455.

Wigley, T.M.L., K.R. Briffa, and P.D. Jones. 1984. On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. Journal of Climate and Applied Meteorology 23: 201-213.

Worbes, M. 1989. Growth rings, increment and age of trees in inundation forests, savannas and a mountain forest in the Neotropics. IAWA n.s. 10: 109-122.

Worbes, M. 1995. How to measure growth dynamics in tropical trees – a review. IAWA Journal 16: 337-351.

Worbes, M. 1999. Annual growth rings, rainfall-dependent growth and long-term growth patterns of tropical trees from the Caparo Forest Reserve in Venezuela. Journal of Ecology 87: 391-403.

Worbes, M. 2002. One hundred years of tree-ring research in the tropics – a brief history and an outlook to future challenges. Dendrochronologia 20: 217-231.

Worbes, M., and W.J. Junk. 1989. Dating tropical trees by means of 14C from bomb tests. Ecology 70: 503-507.

Worbes, M., and W.J. Junk. 1999. How old are tropical trees? The persistence of a myth. IAWA Journal 20: 255-260.

Worbes, M., R. Staschel, A. Roloff, and W.J. Junk. 2003. Tree ring analysis reveals age structure, dynamics and wood production of a natural forest stand in Cameroon. Forest Ecology and Management 173: 105-123.