Root functional diversity of native and non-native C3 and C4 grass species in Hawai‘i

By Courtney L. Angelo* and Stephanie Pau

AbstractC3 and C4 plants are often reported to differ in functional traits and resource-use strategies, while non-native plants may differ from native plants in functional traits, leading to different resource-use strategies that facilitate their invasion. In this study, we compared root functional traits of native and non-native C3 and C4 grasses with the prediction that different resource acquisition strategies would be observed among these groups. We examined five root functional traits (mean diameter, specific root length, root tissue density, % fine roots (diameter <0.2 mm), and root length density) among natural communities of C3 and C4 grass species along an elevation gradient in Hawai‘i to classify resource-use strategies. We additionally examined how root functional traits were related to environmental characteristics (mean annual, seasonal, and July soil moisture; mean annual precipitation and temperature; and soil % nitrogen) along this elevation gradient. Root traits broadly corresponded to differences in photosynthetic pathway (C3

vs. C4; P < 0.002). However, significant variation occurred within the C3 functional group (P < 0.05), whereas all C4 species had similar root functional traits. Grass assemblages were associated with differences in seasonal soil moisture, but root traits did not sort consistently along any environmental gradient. Principal component and cluster analyses of root functional traits showed that non-native C3 species tended to be resource acquisitive whereas the native C3 species was resource conservative, similar to native and non-native C4 species. Evaluating functional traits of native versus non-native species will provide a better understanding of invasion dynamics and suggest possible restoration and conservation strategies in grassland communities where native species still persist.

Corresponding Author E-mail: courtneyangelo@gmail.com

Pacific Science, vol. 71, no. 2December 9, 2016 (Early view)

Introduction

Grasses are commonly classified into C3 or C4 functional groups based on differences in

photosynthetic pathway. C3 and C4 photosynthetic pathways indicate a key difference in plant

physiology (Sage and Monson 1999). The C4 photosynthetic pathway has greater efficiency than

the C3 photosynthetic pathway under conditions of high temperature, high light, and increased

aridity. C4 plants gain this advantage by an alternative biochemical mechanism compared with C3

plants, where CO2 is internally concentrated at high levels and photorespiration is minimized

(Ehleringer et al. 1997). This translates into lower stomatal conductance, higher water-use

efficiency, and higher photosynthetic rates for C4 plants compared with C3 plants in warm-xeric

conditions (Sage and Kubien 2003). Functional classifications such as C3 and C4 have been

shown to provide insight into ecosystem processes such as rates of resource-use, productivity,

and biogeochemical cycling (Sage and Monson 1999). However, the usefulness of placing

species within these groups has been increasing called into question (Reich et al. 2001, Craine et

al. 2002a, b, Reich et al. 2003, Wright et al. 2006).

Recent research has shown that there is not a clear distinction in plant functional traits

grouped by commonly used functional groups (Reich et al. 2001, Craine et al. 2002a, Reich et al.

2003, Wright et al. 2006). For example, Wright et al. (2006) found that typical functional

classifications schemes (C3, C4, forbs, woody, and legumes) have low explanatory power and in

some cases are no better than organizing species into random groups. Additionally, Reich et al.

(2001) found that although functional groups differed in their response to CO2 and nitrogen [N]

supply, there was also substantial variation in species’ responses within functional groups,

suggesting that simple generalizations about plant and ecosystem processes may not be made

based on functional classifications alone.

2

In order for classification schemes to be useful, most members of one functional group

need to regularly differ in their functionality compared to members of another group (Reich et al.

2003). C4 species have been shown to exhibit a resource strategy based on resource conservation.

C4 grass species have been found to have lower root tissue [N] concentrations, less enriched in

δ15N, larger diameters (DM) and higher root tissue density (RTD) than C3 species in tropical and

subtropical grasslands (Craine et al. 2005). In contrast, some work suggests that C3 grass species

have a diversity of root traits corresponding to variation in resource availability in different

habitats (Craine and Lee 2003, Fort et al. 2012). For example, in a temperate grassland, C3 grass

species in a resource-poor environment were found to have root functional traits such as low

specific root length (SRL) and large root diameters (DM), whereas C3 species in a resource-rich

environment were found to have root traits such as high SRL, fine DM, and high tissue [N]

concentrations (Fort et al. 2012). Thus, the C3-C4 functional grouping alone may not allow

prediction or understanding of a species' contribution to ecosystem functions.

The distinction between resource acquisition and conservation has been recognized in the

leaf economic spectrum (LES), which describes the coordinated variation in functional leaf traits

across an extensive range of species and climates (Chapin 1980, Wright et al. 2004, Reich 2014,

Funk 2013). A continuous root economic spectrum (RES), which explores a transition in these

relationships among root traits, has followed from the widely recognized LES (Wright et al.

2004, Reich 2014). The LES and RES have been extended to characterize traits of native

compared to non-native species (Chapin 1980, Reich et al. 1997, Craine 2009, Reich 2014), with

researchers suggesting that functional traits of native and non-native species differ along these

spectrums (Leishman et al. 2007, Ordonez et al. 2010, Penuelas et al. 2010, Ordonez and Olff

2013, Larson and Funk 2016). Recent research has found native species usually are found at the

slow growth end of the spectrum (Leishman et al. 2007, Ordonez et al. 2010, Penuelas et al.

3

2010, Ordonez and Olff 2013, Larson and Funk 2016) and generally possess traits corresponding

with resource conservation (e.g. a long leaf and/or root life span, low specific leaf area and/or

specific root length, low tissue nutrient content, and thick leaf and/or root tissue density), which

typically maximize resource-use efficiency (RUE) (Chapin 1980, Reich et al. 1997, Craine 2009,

Reich 2014). At the other end of the spectrum, non-native invasive species possess traits that

promote resource acquisition and fast overall growth (e.g. a short leaf and/or root life span, high

specific leaf area and/or specific root length, high tissue nutrient content, and thin leaf and/or

root tissue density). For example, native C3 in contrast to non-native C3 grasses have been shown

to exhibit root functional traits such as high DM and low SRL consistent with a resource strategy

based on conservation (Larson and Funk 2016).

High-resource environments typically experience more invasion than low-resource

environments due to non-native species having functional traits that exploit the acquisition of

resources, accompanied by their high growth potential (Huenneke et al.1990, Daehler 2003). In

contrast, native species are thought to have a competitive advantage in low-resource

environments because of adaptations that allow them to tolerate resource-poor conditions

(Daehler 2003). However, many plant invasions have been shown to occur in low-resource

environments (Vitousek and Walker 1989, James and Richards 2006, Funk and Vitousek 2007,

Funk 2008, Han et al. 2012, see review in Funk 2013). Recent research has found that non-native

species can invade in both high and low-resource environments (Funk and Vitousek 2007,

Leishman et al. 2007, Ordonez et al. 2010, Penuelas et al. 2010, Ordonez and Olff 2013, Funk

2013). Two alternative mechanisms have been recognized for the invasion success of non-native

species relative to native species in shared environments. First, the mechanism of ‘phenotypic

divergence’ has been suggested (i.e., limiting similarity; Hutchinson 1959, MacArthur and

Levins 1967), which proposes that non-native species are phenotypically distinct from native

4

species, conferring greater invasion success in a community (Ordonez et al. 2010, Penuelas et al.

2010, see review in Funk 2013, Ordonez and Olff 2013). In contrast, ‘phenotypic convergence’

or ‘habitat filtering’ (Keddy 1992, Weiher, Clarke, and Keddy 1998), has also been proposed as a

mechanism for non-native species’ invasion success (Smith and Knapp 2001, Daehler 2003, see

review in Funk 2013). Thus, species that possess traits similar to established species may have

greater invasion success. Specifically in low-resource environments, higher RUE of non-natives

compared to natives has been proposed as a mechanism for invasion success in these systems

(Funk and Vitousek 2007, Funk 2008).

In the Hawaiian Islands, part of a biodiversity hotspot (Myers et al. 2000), anthropogenic

influences have altered the native vegetation (Taylor 1992, Loope and Giambelluca 1998, Pau et

al. 2012) since the arrival of the Polynesians circa 800-1290 A.D. (Kirch and McCoy 2007,

Wilmshurst et al. 2011). In recent decades, one of the most dramatic changes to the native

vegetation is the invasion of non-native grasses, which increase fire frequency and size (Angelo

and Daehler 2013). In Hawai‘i, most non-native grasses that have altered the grass-fire cycle and

threaten native plant communities are C4 grasses (D’Antonio and Vitousek 1992). Due to

anthropogenic pressures put on low elevation landscapes in Hawai‘i, most native species are

found in remnant populations at high elevations and in protected areas such as national parks.

However, fire-adapted grass species have shifted their distributions upward in elevation in

response to increases in fire frequency and climate warming (Angelo and Daehler 2013), putting

these remnant native plant populations in jeopardy. Therefore, evaluating native versus non-

native root functional trait diversity may aid in our understanding of invasion dynamics (Craine

and Lee 2003), and assist in conservation and restoration efforts of native ecosystems (Funk

2013).

5

Much of the work on tradeoffs between resource acquisition and conservation has

focused on leaf traits (Reich et al. 1997, Diaz et al. 2004), yet a number of root traits are useful

indicators of ecosystem functions (Craine et al. 2002b, Craine and Lee 2003, Picon-Cochard et

al. 2012, Leuschner et al. 2013); thus, our understanding of root trait variation is limited

compared to foliar trait diversity (Mommer and Weemstra 2012, McCormack et al. 2014). Here

we examine root functional traits (mean diameter, specific root length, root tissue density, % fine

roots (diameter <0.2 mm), and root length density) among C3-C4 native and non-native grasses

found along a well-studied elevational gradient of C3-C4 turnover in Hawai‘i (Rundel 1980,

Angelo and Daehler 2013). We address the following questions: Do community composition and

root traits turnover with site resources (soil moisture and % nitrogen [N]) along the elevational

gradient? Do root traits group into traditional C3-C4 classifications? How do root traits of C3 and

C4 native and non-native species differ? We hypothesize that root traits will sort into traditional

C3-C4 classifications with wider variability within the C3 functional group depending on site

resources. Secondly, we hypothesize that root traits of native and non-native species will differ

with native species having a set of traits reflecting resource conservation (low SRL, high DM,

low FRP, low RLD, and high RTD), and non-native species having a suite of traits reflecting

resource-acquisition (high SRL, low DM, high FRP, high RLD, and low RTD). Since the current

grass flora in Hawai‘i is primarily non-native, grass assemblages are expected to be shaped by

ecological sorting and competitive ability, as opposed to insular evolutionary processes (Edwards

and Still 2008, Angelo and Daehler 2013).

Materials and Methods

Field Sites and Field Methods

Roots of nine grass species, in total 66 plants, were collected in nine field sites during late

July to early August 2013, along an elevation gradient on Mauna Loa, Hawai‘i (see supplemental

6

Figure S1 and supplemental Table S1, available online). Field sites were within Hawai‘i

Volcanoes National Park (HAVO) along the Chain of Craters and Mauna Loa Strip Roads. Plots

50 m x 2 m in size were established in natural habitats (i.e., ~ 30 m away from the road) with

open canopies every 100 -150 m change in elevation along the Mauna Loa (ML) gradient. Our

transect extended from our C4-dominated wettest site at 540 m with 2043 mm mean annual

precipitation (MAP) to our C3-dominated driest site at 2050 m with 1600 mm MAP. The study

sites included C3 and C4 grasses with C4 grasses dominating low to mid elevation sites and C3

grasses dominating mid to high elevation sites (Angelo and Pau 2015) (see supplemental Figure

S1 supplemental Table S1; available online).

In each plot, the most dominant grasses (at least 25 % cover) were selected for root

extraction. Species richness was low, thus at almost all sites all grasses in the community were

sampled and those that were not sampled totaled < 5% cover. Three 5-cm-diameter and 15-cm-

length root samples were collected using an AMS sliding hammer corer (American Falls, ID,

USA) for all dominant grass species at each site as point measurements directly under the plant.

Our soils were of volcanic nature and very shallow with volcanic rock found at depths circa 15

cm at most sites, thus, the majority of roots in our study were captured by sampling to this depth.

Mature individual plants were chosen based on a species’ average size along the transect and the

presence of flowering stalks. In each plot, six soil samples were also collected using the AMS

sliding hammer corer (American Falls, ID, USA) to determine soil % [N] content. Phosphorus

content was not measured since plant growth in young soils in the Hawaiian Islands, such as in

this study, is primarily limited by soil [N] (Harrington et al. 2001, Vitousek et al. 2010). All root

and soil samples were stored in a cooler on ice during transport to the Department of Natural

Resource and Environmental Management and the Stable Isotope Biogeochemistry Lab at the

University of Hawai‘i at Manoa (UHM).

7

Mean annual (MAM), seasonal (SSM) (November to March), and July soil moisture

(month of root extractions), MAP, and mean air temperature (MAT) were extracted using ArcGIS

10.2.2 (ESRI, Redlands, CA) for each plot using raster grids produced by Giambelluca et al.

(2013, 2014). Site elevations were recorded using a Geographic Positioning System (GPS).

MAM ranged from 0.53-0.57 (evapotranspiration (mm)/ available soil moisture (mm)), SSM

ranged from 0.56-0.63 and MAP ranged from 1661- 2043 mm, which increased with decreasing

elevation. MAT ranged from 19.30 ºC to 10.78 ºC at the lowest and highest elevations,

respectively.

Lab Analysis

Soils were taken to the Stable Isotope Biogeochemistry Lab at the UHM for nitrogen

content analysis. Soils were then analyzed on a Costech ECS 4010 Elemental Combustion

System (Costech Analytical Technologies Inc., Valencia, CA) coupled with a ThermoFinnigan

DeltaXP isotope ratio mass spectrometer (Thermo Electron Corporation, Bremen Germany).

For root samples, soil was removed from live root tissue through a series of sieves (0.5

mm mesh and smaller for fine roots) and then washed to remove dirt particulates. Roots were

carefully sieved in the lab to make sure that all roots below a focal plant belonged to that plant.

For all species, roots were determined to be alive based on root color and turgidity, along with

the presence of root hairs. Furthermore, grass roots had characteristics that made them easy to

distinguish from woody or forb roots. C3 and C4 grass roots also tended to have specific

morphology that made them easy to tell apart. For example, C3 grasses had very fibrous roots

versus thicker, less dense roots systems for C4 grasses. To determine root functional trait

measurements, all clean roots from the soil core were placed in water in a clear acrylic tray (21 x

29.7 cm2) and scanned at 600 dots per inch with a scanner. The entire root mass was scanned

with root sample diameter <2 mm, the upper diameter limit that root functional traits can be

8

analyzed on (Pérez-Harguindeguy et al. 2013). WinRHIZO Pro 2013 software (Regent

Instruments, Quebec, Canada) was used to analyze scanned root images. This program analyzes

roots present on an image and quantifies total root length and the diameter of each segment of

length. We used twenty root diameter classes (0.1 mm each) and the total amount of root length

and root volume present in a sample was determined. Specific root length (SRL) was calculated

by dividing total root length (m) of a sample by dry root mass (g). The average diameter (DM)

was determined by taking the median value of each root diameter class (mm) and multiplying it

by the root length segment (cm) for that class and dividing it by the total root length for the

entire sample (cm) (Fort et al. 2012). Root tissue density (RTD) was calculated as the ratio of a

sample’s mass (g) and the total root volume (cm3) (Pérez-Harguindeguy et al. 2013). The

percentage of fine roots (FRP) was determined as the ratio of root length (cm) with a diameter <

0.2 mm to the total root length (cm) (Roumet et al. 2006, Fort et al. 2012). Root length density

(RLD) was calculated as the total root length (cm) of a sample divided by the volume of the soil

core (cm3) for that sample (Fort et al. 2012). Root samples were oven dried (60 ºC) to a constant

mass for each sample; dry mass was used to calculate functional trait values.

Statistical Analyses

First, a permutational multivariate analysis of variance (MANOVA) was performed using

the ‘vegan’ package in R to examine changes in community composition (% cover of each

species and abundance weighted means of trait values by site) with differences in MAM, SSM,

MAP, and soil % [N], each tested separately (based on a Bray-Curtis dissimilarity matrix and

9999 permutations).

Next, we used F-tests in analysis of variance using functional groups (C3 and C4) and

species nested within functional groups to examine species variation in mean root trait values

within groups. Differences among species in each functional group were tested using a post hoc

9

Tukey’s HSD test. We additionally used multiple ANOVAs to determine how much variation the

native and non-native groups explained species’ mean root trait values. Data that did not meet the

assumption of normality were log - or square root-transformed.

Root functional traits were also examined in a principal component analysis (PCA) to

determine how root traits and species differ and if variation in species and traits correspond to a

gradient of resource acquisitive vs. conservative strategies (three samples per species/site, n =

65). PCA is an orthogonal transformation of correlated variables intended to reduce the number

of variables to uncorrelated components. All root functional traits were normalized by one

standard deviation before PCA analysis. A cluster analysis based on K-means was then used to

cluster species from the PCA ordination into similar groups based on root traits. K-means is a

partitioning method that can determine the number of clusters a data set fit into by plotting the

within group sums of squares. Additionally, functions pam and pamk in the R package, ‘fpc’,

were used to determine the number of clusters estimated by the optimum average silhouette

width. We also used the function cascadeKM in the R package, ‘vegan’, which uses the Calinski

criterion to determine the optimal number of clusters. Pearson’s correlations were used to

determine if C3 or C4 root traits, PCA scores of root traits, or root traits grouped by cluster

analysis had relationships with MAM, SSM or soil % [N] content. All statistical analyses were

done in R version 3.0.3 (http://www.r-project.org).

Results

Species and Site Diversity

Along the ML gradient there were four abundant C3 grass species (Anthoxanthum

odoratum, Deschampsia nubigena, Ehrharta stipoides, and Holcus lanatus) and five abundant C4

grass species (Hyparrhenia rufa, Panicum tenuifolium, Paspalum dilatatum, Pennisetum

clandestinum, and Schizachyrium condensatum) (Table 1). Of these, three C3 species were non-

10

native, whereas one species was native. Four C4 species were non-native whereas one was native.

Non-native C4 grasses, H. rufa, P. clandestinum, and S. condensatum, were found in wet and

warm, low to mid elevation sites whereas the non-native grass, E. stipoides (C3), was found at

mid elevations where air temperature and soil moisture were intermediate along our gradient.

The nonnative species, A. odoratum (C3), H. lanatus (C3), and P. dilatatum (C4) spanned mesic

and dry sites and covered the largest gradient of air temperature. Lastly, the two native species in

our study, D. nubigena (C3) and P. tenuifolium (C4) coexisted with non-native C3 and C4 species

at high elevations where water resources are low. The nine grass species are distributed in four

grass tribes; two C3 tribes (Aveneae and Ehrharteae) and two C4 tribes (Andropogoneae and

Paniceae) (Table 1). Soil % [N] content spanned 0.08 - 0.81%. Soil % [N] did not significantly

correlate with any soil moisture index (P > 0.05), nor was soil % [N] significantly correlated

with elevation (P > 0.05). Therefore, as elevation increased and soil moisture decreased, soil [N]

did not change.

The permutational MANOVA showed that only differences in SSM were significantly

associated with community composition based on % cover (F = 1.93; P < 0.05) but not trait

composition. Sites with greater seasonality in soil moisture tended to be comprised of non-native

C4 grasses at lower elevations and mixed non-native C3 and C4 grasses at mid-elevations. Sites

with low seasonality in soil moisture tended to be comprised of the native C3 and C4 grasses,

along with non-native C3 grasses.

Comparison of Root Traits between C3 and C4 Native and Non-Native Species

For each of the root traits, photosynthetic pathway generally explained more variation

than native/non-native status (photosynthetic pathway: R2 = 0.36 (DM), R2 = 0.47 (SRL), R2 =

0.13 (RTD), R2 = 0.24 (FRP), and R2 = 0.33 (RLD); native/non-native status: R2 values are as

follows: R2 = 0.23 (DM), R2 = 0.08 (SRL), R2 = -0.01 (RTD), R2 = 0.37 (FRP), and R2 = 0.04

11

(RLD). C4 species had larger mean root diameters (x̅ = 0.39 mm) than C3 species (x̅ = 0.27 mm)

(F = 5.85, df =1, 7, P < 0.05) (Fig. 1). C4 species had significantly higher mean RTD compared

to C3 species (x̅ = 0.24, 0.19 g cm-3, respectively) (F=8.72, df = 1, 7, P = 0.01) (Fig. 2). In

contrast, C3 species had higher SRL (x̅ = 42.42 m g-1) than C4 species (x̅ = 16.60 m g-1; F = 9.46,

df = 1, 7, P = 0.02) (Fig. 3). C3 species had higher mean FRP values than C4 species (x̅ = 55 %

versus 37%, respectively), but were not significantly different (F= 4.40, df =1, 7, P = 0.07) (Fig.

4). C3 species had significantly higher mean RLD compared to C4 species as a group (x̅ = 8.71

cm cm-3 versus 3.65 cm cm-3, respectively; F= 17.83, df =1, 7, P = 0.003) (Fig. 5).

However there was considerable variation within photosynthetic groups. C3 species had

more within functional group variation than C4 species for the variables DM, SRL, and RLD (F

= 13.30, df = 7, 56, P < 0.001; F = 9.52, df = 7, 56, P < 0.001; F = 2.44, df = 7, 56, P < 0.03;

respectively) (Figs. 1, 3, 5). Both C3 and C4 functional groups had significant among species

variation within groups for FRP (F= 13.28, df = 7, 56, P < 0.02) (Fig. 4); there was no

significant variation among species within groups for RTD (F= 1.06, df = 7, 56, P = 0.40; Fig.

2).

The root functional traits of the native C3 species, Deschampsia nubigena differed from

the C3 non-native species, and its trait values typically resembled C4 species values. Using a post

hoc Tukey’s HSD test, the native species, D. nubigena (C3) had a significantly larger diameter

than non-native C3 species (P < 0.05). On average D. nubigena had a root diameter similar to

native and non-native C4 species (x̅ = 0.38 mm and x̅ = 0.39 mm, respectively; Fig. 1).

Non-native C3 species, Anthoxanthum odoratum and Holcus lanatus, had significantly

higher mean SRL values than the native, D. nubigena (P < 0.001), while non-native Ehrharta

stipoides (C3) and D. nubigena mean SRL values were not significantly different (P = 0.99). In

12

contrast, E. stipoides had mean SRL values significantly different than A. odoratum (P = 0.006),

but not H. lanatus (P = 0.07) (Fig. 3).

FRP values for the native, D. nubigena and non-native C3 species were significantly

different (P < 0.001). Within the C4 functional group, Panicum tenuifolium (C4) and Paspalum

dilatatum (C4) mean FRP values were significantly different (P = 0.02). Both native species in

our study, D. nubigena and P. tenuifolium had the lowest mean FRP values across functional

groups (x̅ = 34% and 28%, respectively; Fig. 4).

D. nubigena had the lowest RLD out of the C3 functional group. However, considerable

variation occurred within the C3 functional group for this variable. The non-native, A. odoratum

and the native, D. nubigena had significantly different mean RLD values (P = 0.009), while non-

native, E. stipoides (x̅ = 10.53 cm cm-3) and non-native, H. lanatus (x̅ = 7.80 cm cm-3) had

intermediate mean RLD values that were not significantly different from A. odoratum (x̅ = 12.99

cm cm-3) or D. nubigena (x̅ = 5.66 cm cm-3; P > 0.05) (Fig. 5).

Root functional traits of C3 and C4 grasses were not found to correlate with MAM, SSM,

or soil % [N] (all P > 0.05).

Identification of Resource Conservative or Acquisitive Strategies

The first two axes of the principal component analysis explained 88% of the variability

(Component 1: 72%, Component 2: 16%). Component 1 separates species by resource

acquisition and conservation strategies (e.g. high values indicate a more acquisitive strategy

versus low values indicate a conservative strategy). Three root functional traits had a positive

correlation with component 1: SRL, FRP, and RLD, while DM had a negative correlation with

component 1 (correlation coefficients were > 0.43; see Table 2; Fig. 6). RTD was found to

comprise a large portion of component 2 (correlation coefficient = 0.89; see Table 2; Fig. 7). C4

species always had low values for component 1, whereas C3 species’ values for component 1

13

spanned a gradient of values. For component 1, non-native species, Holcus lanatus and Ehrharta

stipoides, were in the transition zone between resource groups, while the native species,

Deschampsia nubigena, was more similar to C4 species. For component 2, species’ scores did not

clearly differentiate functional groups (Fig. 6). PCA component 1 scores of all species did not

correlate with any of the environmental variables evaluated in this study (MAM, SSM, % [N]; all

P > 0.05). PCA component 1 scores of species in each photosynthetic pathway (C3 and C4) were

not found to correlate with MAM, SSM, or soil % [N] (all P > 0.05).

Cluster analysis revealed that there is a gradient of resource-use strategies for C3 species

from resource acquisition to conservation. Our data best fit into two clusters separating C3 into a

resource acquisitive group and C4 species into a resource conservative group. However, the

native C3 species, D. nubigena, was consistently placed into the C4 cluster irrespective of site,

whereas two other non-native C3 species, H. lanatus and E. stipoides, were placed in the C4

cluster depending on site (Fig. 7). Grass species’ PCA scores within the resource acquisitive and

resource conservative groups did not correlate with annual and seasonal soil moisture or soil

nitrogen content (all P > 0.05).

Discussion

Our primary goal was to determine how root traits of grass species sort into traditional

functional classifications (C3 and C4). Even though a classification based on C3 and C4 functional

groups broadly explained variation in root traits, we found considerable diversity in root

functional traits among C3 species, with the exception of root tissue density. C4 species’ trait

variation was only seen in FRP due to the native C4 grass exhibiting significantly lower FRP than

one non-native C4 grass. Other studies have also found more variation within the C3 functional

group than between C3 and C4 functional groups when evaluating resource-use traits (Craine et

al. 2002b, Craine and Lee 2003, Fort et al. 2012, Larson and Funk 2016). Here we show that the

14

functional diversity of C3 species was more closely aligned with their native/non-native status

and habitat preference than their photosynthetic pathway.

We found that C3 species formed a gradient from Anthoxanthum (non-native) >Holcus

(non-native) >Ehrharta (non-native) > Deschampsia (native) across resource-use strategies –

acquisitive to conservative. Non-native C3 species having traits associated with a resource

acquisitive strategy had high mean values for SRL, FRP, and RLD. These traits have been shown

to allow for rapid nutrient uptake and assimilation (Craine et al. 2002b, Tjoelker et al. 2005) and

quick root turnover (Ryser 1998). In previous studies, SRL and FRP have also been positively

associated with root nitrogen concentration (Tjoelker et al. 2005, Fort et al. 2012). In our study,

plants with high SRL and FRP were coupled with high RLD, a trait that has been related to a

species’ ability for nitrogen-acquisition efficiency under competition (Craine 2006, Fargione and

Tilman 2006) and high affinity for soil exploration and acquisition of resources (Fort et al. 2014).

The non-native C3 species, Holcus lanatus and Ehrharta stipoides, appeared intermediate in

resource-use strategies because they displayed high plasticity among sites. In contrast, the native

C3 species, Deschampsia nubigena, displayed a resource conservative strategy with large DM,

low SRL, FRP, and RLD, similar to coexisting native and non-native C4 species. These traits

have been linked with more investment in root structure and low root turnover allowing high

water transport capacity and prolonged retention of resources (Ryser 1998, Eissenstat et al. 2000,

Tjoelker et al. 2005, Roumet et al. 2006, Hernandez et al. 2010).

Our study aimed to determine differences in root functional traits of native and non-

native C3 and C4 grasses. We found that both native species, D. nubigena (C3) and Panicum

tenuifolium (C4), had root functional traits reflecting resource conservation in a low-resource

environment. In contrast, non-native C3 and C4 species had functional traits that did not

consistently sort along the environmental gradient. Non-native C4 species in our study displayed

15

conservative root functional traits across dry and wet sites, whereas, non-native C3 root

functional traits varied in low-resource sites. Craine and Lee (2003) found that native C3 grasses

along an elevation gradient in New Zealand had functional traits associated with conservative

resource-use in a resource-poor environment. They also showed that non-native C3 grasses were

found in productive sites with a more acquisitive resource-use strategy; however we found non-

native C3 grasses spanned across different environments with varied functional traits. Similar to

our study, Craine et al. (2002a) found that some native C3 grass species were more similar to C4

grass species than to non-native C3 species. For example, a native C3 species, Koeleria cristata

(Ledeb.) Schult., was found to have more similar root functional traits such as lower SRL, large

DM, and a similar C:N ratio to a C4 grass, Bouteloua gracilis (Willd. ex Kunth) Lag. ex Griffiths,

than to other grass species in the study. Thus C3 functional diversity appears more closely

associated with their native/non-native status and habitat preference than their photosynthetic

pathway.

Our results suggest that root functional traits of non-native C3 species may be more

phenotypically plastic than native and non-native C4 or native C3 species. We found that two of

our three non-native C3 species, Holcus lanatus and Ehrharta stipoides had conservative

functional traits in resource-poor sites and acquisitive traits in resource-rich sites. Root traits are

considered to be highly plastic and strongly influenced by resource supply (Eissenstat et al. 2000,

Ryser and Eek 2000, Larson and Funk 2016). C3 species in particular have been shown to have

root phenotypic plasticity in response to changes in resource availability (Ryser and Lambers

1995, Ryser 1998, Ryser and Eek 2000) and water availability compared to C4 species (Nippert

and Knapp 2007a, b). Larson and Funk (2016) found that root traits shifted towards resource-

conserving attributes in response to drought conditions, but the trend varied with species as we

found in this study. Furthermore, it has been shown that non-native species have the ability to be

16

phenotypically plastic in resource-rich and poor sites enabling them to exploit a variety of habitat

types (Funk 2008). Thus, the relationship between a species ability to adapt to its’ environment

and complex biotic interactions may alter the resource-use strategy of non-native C3 grass

species.

Many native Hawaiian grass species are now rare or extinct and the most commonly

encountered species are of non-native origin (Daehler and Carino 1998). Angelo and Daehler

(2013) found that only one native C3 species, Deschampsia nubigena, and two C4 species,

Panicum tenuifolium and Heteropogon contortus, typically coexist with non-native species along

the Mauna Loa (ML) gradient. Due to the volcanic substrate found along the ML gradient and

the inability to extract roots in young lava at low elevations, the low elevation C4 species,

Heteropogon contortus, was not examined for root functional traits in this study. Thus, although

we examined root traits of only two native species, our data appear to be representative of the

extant grass system found along the ML gradient in Hawai‘i. Native Deschampsia nubigena and

Panicum tenuifolium have been known to achieve higher cover after fire and compete with non-

native grasses (Smith and Tunison 1992); whereas, other native C3 and C4 grasses of smaller

stature have not been able to survive these anthropogenic alterations to the ecosystem (Angelo

and Daehler 2013).

In conclusion, we showed that commonly used C3-C4 functional classifications did not

always separate clearly according to root functional traits. There was significant variation among

C3 species in all root traits with the exception of root tissue density. Although C4 species

consistently exhibited traits associated with a conservative resource strategy, root traits of C3

species were more closely aligned with their native/non-native status and habitat preference.

Both native species in this study displayed resource-conservative root traits in low-resource

environments, whereas non-native species exhibited a diversity of root traits along our gradient.

17

Our results suggest that successful invasive species need not be functionally distinct from the

dominant native species to exploit vacant niches (Hutchinson 1959, MacArthur and Levins 1967,

Hamilton et al. 2005, Ordonez et al. 2010, van Kleunen, Weber, and Fischer 2010) in low-

resource environments. Instead, abiotic tolerances may exert stronger controls on community

composition in low-resource environments (i.e., environmental filtering), limiting the range of

possible traits to resource conservative traits in resource-poor environments (Funk 2013). C4 and

some C3 species had a conservative resource strategy in low-resource environments; whereas,

non-native C3 species displayed an acquisitive resource strategy in productive sites and non-

native C4 species had a resource strategy that remained conservative. Although there is growing

evidence that invasive species have traits aligned with resource acquisition (Leishman et al.

2007, Ordonez et al. 2010, Penuelas et al. 2010, Ordonez and Olff 2013, Larson and Funk 2016),

our results show that root traits of invasive C4 species are aligned with a conservative resource

strategy. Understanding which functional traits may confer success among native and invasive

species is important for identifying strategies for native plant conservation and predicting areas

of invasive plant dominance. The conservative resource-use strategy inferred by root traits of the

native grasses in this study suggests that the native grasses may be more successful at resource-

poor sites.

18

Acknowledgments

This work was supported by the Department of Geography at Florida State University

(FSU) and the FSU Committee on Faculty Research Support (COFRS). This work could not

have been completed without access to Hawai‘i Volcanoes National Park and the assistance and

cooperation of Rhonda Loh in the permitting process and HAVO park rangers during survey

periods. We thank Brian Popp’s lab for soil preparation and analysis and Travis Idol’s lab for

space and use of equipment at the University of Hawai‘i at Manoa (UHM). We additionally

thank Susan Crow and David Beilman for assistance at UHM, Mike Patterson (FSU) for help in

field collections and Marian Chau (UHM) for help with processing root samples. Finally, we

thank Jennifer Bufford (Bio-Protection Research Centre, Lincoln University) for help with R-

coding and statistical support, along with Kevin McCarthy and Hoang Tran from the Statistics

Department at FSU.

19

Literature Cited

Angelo, C. L., and C. C. Daehler. 2013. Upward expansion of fire‐adapted grasses along a

warming tropical elevation gradient. Ecography 36: 551-559.

Angelo, C. L., and S. Pau. 2015. Root biomass and soil δ13C in C3 and C4 grasslands along a

precipitation gradient. Plant Ecol. 216: 615-627.

Chapin III, F. S. 1980. The mineral nutrition of wild plants. Annu. Rev. Ecol. Syst. 11:233-260.

Craine, J. M. 2006. Competition for nutrients and optimal root allocation. Plant Soil 285: 171-

185.

Craine, J. M. 2009. Resource strategies of wild plants. Princeton University Press, Princeton.

Craine, J.M., and W. G. Lee. 2003. Covariation in leaf and root traits for native and non-native

grasses along an altitudinal gradient in New Zealand. Oecologia 134: 471-478.

Craine, J.M., D. Wedin, F. S. Chapin, and P. Reich. 2002a. Relationship between the structure of

root systems and resource use for 11 North American grassland plants. Plant Ecol. 165: 85-100.

Craine, J.M., D. Tilman, D. Wedin, P. Reich, M. Tjoelker, and J. Knops. 2002b. Functional

traits, productivity and effects on nitrogen cycling of 33 grassland species. Funct. Ecol. 16: 563-

574.

20

Craine, J. M., W. G. Lee, W. J. Bond, R. J. Williams, and L. C. Johnson. 2005. Environmental

constraints on a global relationship among leaf and root traits of grasses. Ecology 86: 12-19.

D'Antonio, C. M., and P. M. Vitousek. 1992. Biological invasions by exotic grasses, the

grass/fire cycle, and global change. Annu. Rev. Ecol. Syst. 23: 63-87.

Daehler, C. C. 2003. Performance comparisons of co-occurring native and alien invasive plants:

implications for conservation and restoration. Annu. Rev. Ecol. Evo. Syst. 34: 183-211.

Daehler, C. C., and D. A. Carino. 1998. Recent replacement of native pili grass (Heteropogon

contortus) by invasive African grasses in the Hawaiian Islands. Pac. Sci. 52: 220-227.

Diaz, S., J. Hodgson, K. Thompson, M. Cabido, J. Cornelissen, A. Jalili, G. Montserrat‐Martí, J.

Grime, F. Zarrinkamar, and Y. Asri. 2004. The plant traits that drive ecosystems: evidence from

three continents. J. Veg. Sci. 15: 295-304.

Edwards, E. J., and C. J. Still. 2008. Climate, phylogeny and the ecological distribution of C4

grasses. Ecol. Lett. 11: 266-276.

Ehleringer, J. R., T. E. Cerling, and B. R. Helliker. 1997. C4 photosynthesis, atmospheric CO2,

and climate. Oecologia 112: 285-299.

Eissenstat, D., C. Wells, R. Yanai, and J. Whitbeck. 2000. Building roots in a changing

environment: implications for root longevity. New Phyt. 147: 33-42.

21

Fargione, J., and D. Tilman. 2006. Plant species traits and capacity for resource reduction predict

yield and abundance under competition in nitrogen‐limited grassland. Funct. Ecol. 20: 533-540.

Fort, F., C. Jouany, and P. Cruz. 2012. Root and leaf functional trait relations in Poaceae

species: implications of differing resource-acquisition strategies. J. Plant Ecol. rts034.

Fort, F., P. Cruz, and C. Jouany. 2014. Hierarchy of root functional trait values and plasticity

drive early‐stage competition for water and phosphorus among grasses. Funct. Ecol. 28: 1030

-1040.

Funk, J. L. 2008. Differences in plasticity between invasive and native plants from a low

resource environment. J. Ecol. 96: 1162-1173.

---. 2013. The physiology of invasive plants in low-resource environments. Conserv.

Physiol. 1: cot026.

Funk, J. L., and P. M. Vitousek. 2007. Resource-use efficiency and plant invasion in low

resource systems. Nature 446: 1079-1081.

Giambelluca, T. W., Q. Chen, A. G. Frazier, J. P. Price, Y.-L. Chen, P.-S. Chu, J. K. Eischeid,

and D. M. Delparte. 2013. Online rainfall atlas of Hawai‘i. B. Am. Meteorol. Soc. 94: 313-316.

Giambelluca, T.W., X. Shuai, M. Barnes, R. Alliss, R. Longman, T. Miura, Q. Chen, A. Frazier,

22

R. Mudd, and L. Cuo. 2014. Evapotranspiration of Hawai‘i. Final report submitted to the US

Army Corps of Engineers—Honolulu District, and the Commission on Water Resource

Management, State of Hawai‘i.

Hamilton, M. A., B. R. Murray, M. W. Cadotte, G. C. Hose, A. C. Baker, C. J. Harris, and D.

Licari. 2005. Life‐history correlates of plant invasiveness at regional and continental scales.

Ecol Lett 8: 1066-1074.

Han, Y., Y. M. Buckley, and J. Firn. 2012. An invasive grass shows colonization advantage

over native grasses under conditions of low resource availability. Plant Ecol. 213: 1117-1130.

Harrington, R. A., J. H. Fownes, and P. M. Vitousek. 2001. Production and resource use

efficiencies in N-and P-limited tropical forests: a comparison of responses to long-term

fertilization. Ecosystems 4: 646-657.

Hernández, E., A. Vilagrosa, J. Pausas, and J. Bellot. 2010. Morphological traits and water use

strategies in seedlings of Mediterranean coexisting species. Plant Ecol. 207: 233-244.

Huenneke, L. F., S. P. Hamburg, R. Koide, H. A. Mooney, and P. M. Vitousek. 1990. Effects of

soil resources on plant invasion and community structure in Californian serpentine grassland.

Ecology 71: 478-491.

Hutchinson, G. E. 1959. Homage to Santa Rosalia or why are there so many kinds of animals?

American Naturalist. 93: 145-159.

23

James, J., and J. Richards 2006. Plant nitrogen capture in pulse‐driven systems: interactions

between root responses and soil processes. J. Ecol. 94: 765-777.

Keddy, P. A. 1992. Assembly and response rules: two goals for predictive community ecology.

J. Veg. Sci. 3: 157-164.

Kirch, P. V., and M. D. McCoy. 2007. Reconfiguring the Hawaiian cultural sequence: results

of re-dating the Hālawa dune site (Mo-A1-3), Moloka'i Island. J. Polynesian

Soc. 116: 385-406.

Larson, J. E., and J. L. Funk. 2016. Seedling root responses to soil moisture and the identification

of a belowground trait spectrum across three growth forms. New Phyt. 210: 827-838.

Leishman, M. R., T. Haslehurst, A. Ares, and Z. Baruch. 2007. Leaf trait relationships of native

and invasive plants: community‐and global‐scale comparisons. New Phyt. 176: 635-643.

Leuschner, C., S. Gebel, and L. Rose. 2013. Root trait responses of six temperate grassland

species to intensive mowing and NPK fertilisation: a field study in a temperate grassland. Plant

Soil 373: 687-698.

Loope, L. L., and T. W. Giambelluca. 1998. Vulnerability of island tropical montane cloud

forests to climate change, with special reference to East Maui, Hawai‘i. Climatic Change 39:

503-517.

MacArthur, R., and R. Levins. 1967. The limiting similarity, convergence, and divergence of

24

coexisting species. Am. Nat. 101:377-385.

McCormack, M. L., E. Lavely and Z. Ma. 2014. Fine‐root and mycorrhizal traits help explain

ecosystem processes and responses to global change. New Phyt. 204: 455-458.

Mommer, L., and M. Weemstra. 2012. The role of roots in the resource economics spectrum.

New Phyt. 195: 725-727.

Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. Da Fonseca, and J. Kent. 2000.

Biodiversity hotspots for conservation priorities. Nature. 403: 853-858.

Nippert, J. B., and A. K. Knapp. 2007a. Linking water uptake with rooting patterns in grassland

species. Oecologia 153: 261-272.

Nippert, J. B., and A. K. Knapp. 2007b. Soil water partitioning contributes to species coexistence

in tallgrass prairie. Oikos 116: 1017-1029.

Ordonez, A., and H. Olff . 2013. Do alien plant species profit more from high resource supply

than natives? A trait‐based analysis. Global Ecol. Biogeogr. 22: 648-658.

Ordonez, A., I. J. Wright, and H. Olff. 2010. Functional differences between native and alien

species: a global‐scale comparison. Funct. Ecol. 24: 1353-1361.

Pau, S., G. M. MacDonald, and T. W. Gillespie. 2012. A dynamic history of climate change and

human impact on the environment from Keālia Pond, Maui, Hawaiian Islands. Ann. Assoc. Am.

25

Geogr. 102: 748-762.

Penuelas, J., J. Sardans, J. Llusia, S. M. Owen, J. Carnicer, T. W. Giambelluca, E. L. Rezende,

M. Waite, and U. Niinemets. 2010. Faster returns on ‘leaf economics’ and different

biogeochemical niche in invasive compared with native plant species. Global Change Biol.

16: 2171-2185.

Pérez-Harguindeguy, N., S. Díaz, E. Garnier, S. Lavorel, H. Poorter, P. Jaureguiberry, M. Bret-

Harte, W. Cornwell, J. Craine, and D. Gurvich. 2013. New handbook for standardised

measurement of plant functional traits worldwide. Aust. J. Bot. 61: 167-234.

Picon-Cochard, C., R. Pilon, E. Tarroux, L. Pagès, J. Robertson, and L. Dawson. 2012. Effect of

species, root branching order and season on the root traits of 13 perennial grass species. Plant

Soil 353: 47-57.

Reich, P. B. 2014. The world‐wide ‘fast–slow’plant economics spectrum: a traits manifesto. J.

Ecol. 102: 275-301.

Reich, P. B., M. B. Walters, and D. S. Ellsworth. 1997. From tropics to tundra: global

convergence in plant functioning. Proc. Natl. Acad. Sci. 94: 13730-13734.

Reich, P. B., D. Tilman, J. Craine, D. Ellsworth, M. G. Tjoelker, J. Knops, D. Wedin, S. Naeem,

D. Bahauddin, and J. Goth. 2001. Do species and functional groups differ in acquisition and use

of C, N and water under varying atmospheric CO2 and N availability regimes? A field test with

26

16 grassland species. New Phyt. 150: 435-448.

Reich, P. B., C. Buschena, M. Tjoelker, K. Wrage, J. Knops, D. Tilman, and J.-L. Machado.

2003. Variation in growth rate and ecophysiology among 34 grassland and savanna species under

contrasting N supply: a test of functional group differences. New Phyt. 157: 617-631.

Roumet, C., C. Urcelay, and S. Díaz. 2006. Suites of root traits differ between annual and

perennial species growing in the field. New Phyt. 170: 357-368.

Rundel, P. W. 1980. The ecological distribution of C4 and C3 grasses in the Hawaiian Islands.

Oecologia 45: 354-359.

Ryser, P. 1998. Intra-and interspecific variation in root length, root turnover and the underlying

parameters. Pages 441-465 in H. Lambers, H. Poorter, and M. M. I. vanVuuren, eds. Inherent

variation in plant growth. Physiological mechanisms and ecological consequences. Backhuys

Publishers, Leiden.

Ryser, P., and H. Lambers. 1995. Root and leaf attributes accounting for the performance of fast-

and slow-growing grasses at different nutrient supply. Plant Soil 170: 251-265.

Ryser, P., and L. Eek. 2000. Consequences of phenotypic plasticity vs. interspecific differences

in leaf and root traits for acquisition of aboveground and belowground resources. Am. J. Bot. 87:

402-411.

27

Sage, R. F., and R. K. Monson. 1999. C4 plant biology. Academic Press, San Diego.

Sage, R. F., and D. S. Kubien. 2003. Quo vadis C4? An ecophysiological perspective on global

change and the future of C4 plants. Photosynthesis Res.77: 209-225.

Smith, C. W., and J. T. Tunison. 1992. Fire and alien plants in Hawai‘i: research and

management implications for native ecosystems. Pages 394-408 in C.P. Stone, C.W. Smith, and

J.T. Tunison, eds. Alien plant invasions in native ecosystems of Hawai‘i: management and

research. University of Hawai‘i Press, Honolulu.

Smith, M. D. and A. K. Knapp. 2001. Physiological and morphological traits of exotic,

invasive exotic, and native plant species in tallgrass prairie. Inter. J. Plant Sci. 162:

785-792.

Taylor, D. 1992. Controlling weeds in natural areas in Hawai‘i: A manager's perspective. Pages

752-756 in C.P. Stone, C.W. Smith, and J.T. Tunison, eds. Alien plant invasions in native

ecosystems of Hawai‘i: management and research. University of Hawai‘i Press, Honolulu.

Tjoelker, M., J. Craine, D. Wedin, P. Reich, and D. Tilman. 2005. Linking leaf and root trait

syndromes among 39 grassland and savannah species. New Phyt. 167: 493-508.

van Kleunen, M., E. Weber, and M. Fischer. 2010. A meta‐analysis of trait differences between

invasive and non‐invasive plant species. Ecol. Lett. 13: 235-245.

28

Vitousek, P. M., and L. R. Walker. 1989. Biological invasion by Myrica faya in Hawai‘i: plant

demography, nitrogen fixation, ecosystem effects. Ecol. Monogr. 59: 247-265.

Vitousek, P. M., S. Porder, B. Z. Houlton, and O. A. Chadwick. 2010. Terrestrial phosphorus

limitation: mechanisms, implications, and nitrogen-phosphorus interactions. Ecol. Appl. 20: 5-

15.

Weiher, E., G. P. Clarke, and P. A. Keddy. 1998. Community assembly rules, morphological

dispersion, and the coexistence of plant species. Oikos 81: 309-322.

Wilmshurst, J. M., T. L. Hunt, C. P. Lipo, and A. J. Anderson. 2011. High-precision radiocarbon

dating shows recent and rapid initial human colonization of East Polynesia. Proc.

Nat. Acad. Sci. 108: 1815-1820.

Wright, I. J., P. B. Reich, M. Westoby, D. D. Ackerly, Z. Baruch, F. Bongers, J. Cavender-Bares,

T. Chapin, J. H. Cornelissen, and M. Diemer. 2004. The worldwide leaf economics spectrum.

Nature. 428: 821-827.

Wright, J. P., S. Naeem, A. Hector, C. Lehman, P. B. Reich, B. Schmid, and D. Tilman. 2006.

Conventional functional classification schemes underestimate the relationship with ecosystem

functioning. Ecol. Lett. 9: 111-120.

29

Table 1. Species studied and their phylogenetic, photosynthetic pathway, native/non-native affiliations

Species Tribe Subfamily Clade PP Native/Non-native

Anthoxanthum odoratum L. Aveneae Pooideae BEP C3 Non-nativeDeschampsia nubigena Hillebr. Aveneae Pooideae BEP C3 NativeHolcus lanatus L. Aveneae Pooideae BEP C3 Non-nativeEhrharta stipoides (Labill.) R. Br. Ehrharteae Ehrhartoideae BEP C3 Non-nativeHyparrhenia rufa (Nees) Stapf Andropogoneae Panicoideae PACMA

D

C4 Non-native

Schizachyrium condensatum (Kunth)

Nees

Andropogoneae Panicoideae PACMA

D

C4 Non-native

Panicum tenuifolium Hook. and Arn. Paniceae Panicoideae PACMA

D

C4 Native

Paspalum dilatatum Poir. Paniceae Panicoideae PACMA

D

C4 Non-native

Pennisetum clandestinum Hochst. ex Chiov.

Paniceae Panicoideae PACMAD

C4 Non-native

*Species are ordered alphabetically by tribe

30

Table 2. Component loadings for the first two principal components. Also included is the percentage of total variation explained by all

principal component analysis (PCA) axes that is associated with each axis

Correlation coefficientsComponent 1 Component 2

Roots traitsMean diameter (DM, mm) -0.49 0.36Specific root length (SRL, m g-1) 0.50 -0.04Root tissue density (RTD, g cm-3) -0.29 -0.89Fine root percentage (FRP, % root length with diameter <0.2 mm) 0.48 -0.26Root length density (RLD, cm cm-3) 0.44 0.13% Variation explained 72% 16%

31

Figure 1. Box plots for root diameter (mm) separated by C3 species (black line) and C4 species (gray line) and C3 tribes (Aveneae,

black dotted line; Ehrharteae, black dashed line) and C4 tribes (Andropogoneae, gray dotted line; Paniceae, gray dashed line). Results

for analysis of variance and significant differences determined by Tukey’s post-hoc test for species within functional groups (C3 or C4).

Root diameter for C3 and C4 functional groups (P < 0.05) and for species nested within functional group (P < 0.001). Letters a-d

represent significant differences within functional groups.

32

Figure 2. Box plots for root tissue density (g cm-3) separated by C3 species (black line) and C4 species (gray line) and C3 tribes

(Aveneae, black dotted line; Ehrharteae, black dashed line) and C4 tribes (Andropogoneae, gray dotted line; Paniceae, gray dashed

line). Results for analysis of variance and significant differences determined by Tukey’s post-hoc test for species within functional

groups (C3 or C4). Root tissue density for C3 and C4 functional groups (P = 0.01) and for C3 and C4 species nested within functional

group (P = 0.40). Letters a-d represent significant differences within functional groups.

33

Figure 3. Box plots for specific root length (m g-1) separated by C3 species (black line) and C4 species (gray line) and C3 tribes

(Aveneae, black dotted line; Ehrharteae, black dashed line) and C4 tribes (Andropogoneae, gray dotted line; Paniceae, gray dashed

line). Results for analysis of variance and significant differences determined by Tukey’s post-hoc test for species within functional

groups (C3 or C4). Specific root length for C3 and C4 functional groups (P = 0.02) and for species nested within functional group (P <

0.001). Letters a-d represent significant differences within functional groups.

34

Figure 4. Box plots for fine root percentage (% root length < 0.2mm) separated by C3 species (black line) and C4 species (gray line)

and C3 tribes (Aveneae, black dotted line; Ehrharteae, black dashed line) and C4 tribes (Andropogoneae, gray dotted line; Paniceae,

gray dashed line). Results for analysis of variance and significant differences determined by Tukey’s post-hoc test for species within

functional groups (C3 or C4). Fine root percentage for C3 and C4 functional groups (P = 0.07) and for species nested within functional

group (P < 0.02). Letters a-d represent significant differences within functional groups.

35

Figure 5. Box plots for root length density (cm cm-3) separated by C3 species (black line) and C4 species (gray line) and C3 tribes

(Aveneae, black dotted line; Ehrharteae, black dashed line) and C4 tribes (Andropogoneae, gray dotted line; Paniceae, gray dashed

line). Results for analysis of variance and significant differences determined by Tukey’s post-hoc test for species within functional

groups (C3 or C4). Root length density for C3 and C4 functional groups (P = 0.003) and species nested within functional group (P <

0.03). Letters a-d represent significant differences within functional groups.

36

Figure 6. Principal component analyses of functional trait syndromes of grass species along the

Mauna Loa elevation gradient (those given in Table 2). Trait abbreviations: DM, mean diameter;

SRL, specific root length; RTD, root tissue density; FRP, fine root percentage; RLD, root length

density. Black letters are C3 grasses (A = Anthoxanthum, D = Deschampsia, E = Ehrharta, H =

Holcus) and gray letters are C4 grasses (Y = Hyparrhenia, P = Panicum, U = Paspalum, N =

Pennisetum, S = Schizachyrium)

37

Figure 7. Cluster analyses for 5 variables (those given in Table 2). Two clusters were determined, but C3 species were included in both

clusters. Components 1 and 2 correspond to the multivariate components of the PCA. The native C3 grass, Deschampsia nubigena,

always clustered with native and non-native C4 species. Black letters are C3 grasses (A = Anthoxanthum, D = Deschampsia, E =

Ehrharta, H = Holcus) and gray letters are C4 grasses (Y = Hyparrhenia, P = Panicum, U = Paspalum, N = Pennisetum, S =

Schizachyrium)

38

Figure S1. Site locations (black dots) on the island of Hawai‘i, all within Hawai‘i Volcanoes National Park. White letters represent C3

grasses (A = Anthoxanthum, D = Deschampsia, E = Ehrharta, H = Holcus) and C4 grasses (Y = Hyparrhenia, P = Panicum, U =

Paspalum, N = Pennisetum, S = Schizachyrium) at each site along our transect.

39

Table S1. Photosynthetic pathways, native/non-native pairs, and environmental data at each site

Photosynthetic Types

Native/Non-Native Pairs

Elevation (m) Annual Temperature (˚ C)

Annual Precipitation (mm)

Site (%) Nitrogen

Annual Soil Moisture

Seasonal Soil Moisture

C4 Non-native 540 19.3 2043 0.56 0.57 0.63C4 Non-native 1233 15.86 1814 0.08 0.54 0.60C3 and C4 Non-native 1298 15.46 1862 0.61 0.54 0.61C4 Non-native 1364 15.22 1876 0.79 0.55 0.61C3 and C4 Non-native 1419 14.73 1900 0.59 0.55 0.61C3 and C4 Native and

non-native

1514 14.08 1907 0.72 0.55 0.60

C3 Native and

non-native

1669 13.03 1756 0.81 0.54 0.59

C3 and C4 Native and

non-native

1871 11.61 1753 0.71 0.53 0.58

C3 and C4 Native and

non-native

2050 10.78 1661 0.57 0.53 0.56

40