Functional traits of common New Zealand foredune species at New Brighton, Canterbury

Abstract

The introduction of exotic plant species to the sand dunes of New Zealand has displaced native dune species. To examine our hypotheses that: (1) exotic species have functional traits that allow them to outcompete native species; and (2) that traits of exotic species vary across the dune environment, we measured specific leaf area, leaf dry matter content, leaf area and plant height in five common fore-dune taxa of New Brighton, Christchurch. The native species Spinifex sericeus, Ficinia spiralis and Poa billardierei had traits indicative of a more conservative growth strategy compared with the exotic tussock grass Ammophila arenaria and the exotic succulent Carpobrotus spp., suggesting that natives would be less competitive. In exotics, leaf size was larger at greater distances from the high tide mark, as was plant height in A. arenaria. These differences in traits across the dune environment point to phenotypic plasticity across short, but sharp, environmental gradients.

Keywords:

• Ammophila arenaria
• Carpobrotus chilensis
• Carpobrotus edulis
• Ficinia spiralis
• height
• leaf area
• leaf dry matter content
• Poa billardierei
• specific leaf area
• Spinifex sericeus

Introduction

The sand dune environment of New Zealand has been highly modified since European settlement in the 19th century, resulting in a displacement of native dune species, a loss of vegetation cover and unstable shifting dunes (Hesp 2000). Exotic species were successfully introduced as dune stabilisers, but because they can displace native species, they caused a further decline in dune areas dominated by native species (Hilton et al. 2005). Hence, there is a need to undertake restoration of native sand dune plant communities. To do this successfully, a better understanding of the ecology of native sand dune species and the invasive exotics that threaten them is needed. One approach is to study their plant functional traits, those attributes that impact plant fitness indirectly via their effects on growth, reproduction and survival (Violle et al. 2007). However, few studies have examined the functional traits of sand dune species (Sykes & Wilson 1988; Feagin & Wu 2007; Stanisci et al. 2010; Murphy et al. 2012) and of these, most have focused on qualitative or ordinal traits, rather than quantitative traits (García-Mora et al. 1999; Gallego-Fernandez & Martinez 2011).

Coastal sand dune plants must survive stressful environmental conditions (Esler 1970). Salinity, wind exposure and burial are highest near the ocean and decrease towards the hind-dunes (Miller & Paul 2007; Murphy et al. 2012). Nutrients such as nitrogen (N) and phosphorus (P) are generally lowest in the foredunes (Lane et al. 2008; Syers & Walker 1969). Moisture stress is correlated with the distance to the water table (Gries et al. 2003) and hence likely increases towards dune crests. Plant responses to these gradients are associated with traits such as specific leaf area, leaf dry matter content, leaf area and plant height (Lavorel & Garnier 2002; Garnier & Navas 2012). Hence, these traits can help us understand the resource-use strategies (e.g. nutrient and water use) of sand dune plants. Leaf area is correlated with moisture availability and/or nutrient gradients (Ackerly et al. 2001); this commonly results in smaller leaves in habitats with low nutrient or moisture availability (Ackerly & Reich 1999; Cornelissen et al. 2003). Specific leaf area (SLA) and leaf dry matter content (LDMC) are generally negatively correlated; a low SLA with a high LDMC is related to an efficient conservation of nutrients, and a high SLA with a low LDMC corresponds to a rapid production of biomass (Garnier et al. 2001). Within a growth form, plant height is correlated with the competitive vigour of plants and their ability to reproduce (Gaudet & Keddy 1995; Westoby 1998; Westoby et al. 2002; Cornelissen et al. 2003; Garnier & Navas 2012). Finally, functional traits will vary within species across environmental gradients (Garnier & Navas 2012) and such trait plasticity is important in understanding invasion dynamics (Stanisci et al. 2010; Davidson et al. 2011).

At one dune location, we examined our hypotheses that: (1) the functional plant traits of exotic species enable them to outcompete native species; and (2) traits within exotic species differ across the dune environment, from fore- to hind-dune, in response to the environmental conditions. The aim of this study was to provide information on the resource-use strategy of common sand dune species and the extent of trait plasticity within exotic species, which will aid future restoration efforts by helping to understand invasion processes in this sand dune ecosystem and identifying the management actions that could reduce or reverse these impacts (Funk et al. 2008).

Methods

The sand dunes of New Brighton, Christchurch (43°33′S; 172°47′E) are spread along the southern coast of Pegasus Bay. The mean annual temperature of Christchurch is 12.2 °C and the total annual precipitation is 618.2 mm (NIWA 2010). Dominant plants include Spinifex sericeus R.Br., Ficinia spiralis A.Rich. and Poa billardierei (Spreng.) St.-Yves, the exotic tussock grass Ammophila arenaria L. and the exotic succulent Carpobrotus spp. N.E.Br. When conducting our fieldwork it was not possible to distinguish between the two Carpobrotus spp. that occur in the study area, due to the lack of flowers. It could be either C. edulis (L.) N.E.Brown or C. chilensis (Molina) N.E.Brown. Ten individual adult plants of each exotic taxon were randomly selected within three habitat zones: foredune, crest and hind-dune. The native species occur only in restoration plantings on the foredune (see Bergin & Kimberley 1999; Bergin 2008 for details on plantings), so ten individuals of each species were sampled from this habitat. SLA (mm² mg⁻¹), LDMC (oven dry mass divided by saturated fresh mass; mg g⁻¹), individual leaf area (mm²) and plant height (m) were measured following the standardised methods described by Cornelissen et al. (2003). Only a slice of the epidermis and some parenchyma of the leaves of Carpobrotus spp. were used to measure its SLA and LDMC, as recommended by Cornelissen et al. (2003) and Vendramini et al. (2002).

Three environmental parameters were measured in the field at each sample plant location: distance to hide tide mark (HTM in m), elevation (m above sea level, using differential GPS) and the percentage cover of other species within 1 m² around each individual. The distance to HTM provides a proxy for the salinity gradient in the sand dunes (Miller & Paul 2007). Elevation is associated with the distance to the water table and is therefore a surrogate for moisture stress (Gries et al. 2003). The percentage cover of other species provides a measure of interspecific competition.

SLA, LDMC and leaf area were measured on two leaves of each individual plant and averaged to give one statistical observation (Cornelissen et al. 2003). Analysis of variance (ANOVA) with post hoc Tukey's honestly significant difference (HSD) tests were used to determine if trait values differed among the species. Principal component analysis (PCA) was used to explore relationships among traits. Multiple regression was used to examine the relationship between traits and the environmental parameters separately for each of the two exotic species.

Results

There were significant differences between species for all four traits (Table 1). The native species had a similar height, but showed some differences in SLA, LDMC and leaf area. The exotic tussock grass A. arenaria was the tallest plant and had similar SLA to P. billardierei and F. spiralis. The shortest plant was Carpobrotus spp., which had similar SLA and leaf area to S. sericeus. The first PCA axis explained 60% of variation and was correlated with plant height and leaf area, whereas the second PCA axis was correlated with SLA and accounted for 25% of variation (Fig. 1).

Table 1 Mean (and standard error of the mean) for specific leaf area (SLA), leaf dry matter content (LDMC), leaf area and plant height for each of five sand dune species.

Species	SLA (mm^2 mg^−1)		LDMC (mg g^−1)		Leaf area (mm^2)		Plant height (m)
Poa billardierei	5.0 (0.4)	a	457.0 (9.1)	a	339.0 (175.5)	a	0.52 (0.07)	a
Ficinia spiralis	4.2 (0.5)	a	361.5 (10.6)	b	1800.2 (202.7)	b	0.61 (0.08)	a
Spinifex sericeus	6.8 (0.5)	b	354.3 (10.6)	b	1612.2 (202.7)	bd	0.54 (0.08)	a
Ammophila arenaria	5.3 (0.6)	a	423.1 (12.9)	c	2633.7 (248.2)	c	1.11 (0.1)	b
Carpobrotus spp.	6.6 (0.6)	b	93.8 (12.9)	d	1127.0 (248.2)	d	0.26 (0.1)	c

The same lowercase letter within a column indicates species that did not differ significantly for that trait (Tukey's HSD test, P > 0.05).

Figure 1 Principal components scores for each individual plant (sample) for the first two axes of a PCA on traits and samples for all sand-dune taxa. Lines represent trait vectors: specific leaf area (SLA), leaf dry matter content (LDMC), leaf area and plant height. Native species: Spinifex sericeus (▵), Ficinia spiralis (□) and Poa billardierei (○). Exotic species: Ammophila arenaria (•) and Carpobrotus spp. (▪).

Most plant functional traits varied within exotic species along the environmental gradients of distance to HTM, elevation and per cent cover of other species (Table 2), although SLA did not change across the dune gradient for either exotic species. The LDMC of A. arenaria was positively correlated with elevation. Leaf area and height of A. arenaria increased towards the hind-dune, but height was negatively correlated with the cover of other species. Leaf area of Carpobrotus spp. increased towards the hind-dune and its height increased in the presence of other species.

Table 2 Results of the multiple regression, showing the relationships (regression coefficient values and significance) between traits and environmental parameters within exotic species.

		SLA	LDMC	Leaf area	Plant height
Ammophila arenaria	Distance to HTM (m)	NS	NS	638.9***	0.2*
	Elevation (m)	NS	14.8**	NS	NS
	Cover other species (%)	NS	NS	NS	−0.004*
	Dist*elev	NS	NS	NS	NS
Carpobrotus spp.	Distance to HTM (m)	NS	NS	155.5*	NS
	Elevation (m)	NS	NS	NS	NS
	Cover other species (%)	NS	NS	NS	0.002*
	Dist*elev	NS	NS	NS	NS

HTM, hide tide mark; LDMC, leaf dry matter content; SLA, specific leaf area. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant (P > 0.05).

Discussion

There have been few quantitative assessments of plant functional traits in coastal sand dune communities worldwide (but see Feagin & Wu 2007; Stanisci et al. 2010) and in New Zealand coastal dunes (Murphy et al. 2012). Here, we show how the key functional traits of SLA, LDMC, leaf area and height vary between native and exotic species and within exotic species across the fore- to hind-dune gradient.

The range in SLA of plants in our coastal dune was narrow (4–7 mm² mg⁻¹) and places them at the lower range of SLA values worldwide (<1–300 mm² mg⁻¹; Pérez-Harguindeguy et al. 2013). These low mean SLA values reflect a conservative growth strategy that is characteristic of plants in resource-poor, stressful environments (Wilson et al. 1999). However, these values are in stark contrast to some other published SLA values from coastal sand dunes. Feagin & Wu (2007) recorded SLA values of 36–142 in Texan coastal sand dunes, but did not give units, and Italian coastal dune plants were reported to have SLA values of 81–177 mm² mg⁻¹ (Stanisci et al. 2010). The high SLA values for most species in both these studies places them within the higher range of SLA values globally (Pérez-Harguindeguy et al. 2013), suggesting a different life history strategy to New Zealand foredune species. Murphy et al. (2012) recorded SLA values of 2.8–10.9 mm² mg⁻¹ for shrubs and monocotyledons on a coastal dune at Stewart Island, New Zealand. They recorded an SLA of 2.76 mm² mg⁻¹ for F. spiralis and 4.13 mm² mg⁻¹ for P. billardierei, very similar to the SLA of these species measured in our study, further confirming that native foredune species in New Zealand exhibit conservative growth strategies.

Species with high SLA and low LDMC values are able to rapidly produce biomass compared to low SLA/high LDMC species (Garnier et al. 2001). Carpobrotus spp. best represents the former life history strategy among our study species. This may explain its successful invasion in New Zealand coastal dunes because higher SLA has been linked with successful invasions elsewhere (Grotkopp & Rejmánek 2007; Feng et al. 2008).

Taller plants are generally considered to be better competitors than shorter plants (Falster & Westoby 2003), because height contributes to a competitive advantage (Hacker et al. 2012). Ammophila arenaria is the tallest species on the foredunes at New Brighton and is among the tallest herbaceous species on sand dunes throughout New Zealand (Hesp 2000; Gadgil 2006). This, coupled with its larger leaves, which aid in light capture, gives it a competitive advantage over other foredune species (Violle et al. 2009; Garnier & Navas 2012). The native species P. billardierei and F. spiralis have traits that reflect a slow-growing, conservative resource use strategy that has left them susceptible to faster growing, more competitive exotic species. The higher SLA of S. sericeus suggests a faster growth rate relative to the other natives. This may reflect a need to quickly elongate stems in response to the frequent sand burial that occurs in the foredunes, which is close to the HTM (Maze & Whalley 1992).

Some traits differed within exotic species across the dune environment. Within A. arenaria, leaf area increased with distance to HTM, which seems likely to be due to a decrease in environmental stresses, such as salinity and hence drought stress (Munns 2002), from the foredunes to the hind-dunes (Miller & Paul 2007). Gadgil (2006) noted that A. arenaria is less tolerant to salinity than native foredune species, especially when younger (Sykes & Wilson 1988), but is tolerant to high temperatures, nutrient stress and wind action, and sand deposition stimulates its growth (Hesp 2000). In this study, its height decreased in the occurrence of other species. Generally, a plant will not invest in its vegetative height when the expected decrease in reproductive output, due to this investment, outweighs the potential decrease in shaded conditions (Falster & Westoby 2003). According to Gadgil (2006), A. arenaria does not compete well with S. sericeus on the foredunes, although A. arenaria is able to displace F. spiralis and other native species on active dune sites (Hilton et al. 2005). Carpobrotus spp. seem to respond to competition by growing taller because the height increases with the per cent cover of other species. Carpobrotus spp. are known to compete successfully for surface soil moisture, nutrients and space with other shallowly rooted plants and also affect herbaceous plants and shrub species in sandy areas (D'Antonio 1990). In this study, the leaf area of Carpobrotus spp. increased towards the back of the dunes, likely due to the lower environmental stress (Hesp 2000). The plasticity in leaf area and plant height across the dune environment in A. arenaria and Carpobrotus spp. helps explain their ability to grow in a variety of environments (Bradshaw 1965). Although we did not test for trait variation across the environmental gradient in natives, this was conducted by Murphy et al. (2012), who found a significant increase in SLA, averaged over the seven species observed in their transects, and a decrease in leaf area of P. billardierei, on transects of higher elevation.

Our examination of functional traits indicates that the exotics, Carpobrotus spp. and A. arenaria, are likely to grow faster, and hence are better competitors in the short-term compared with the native foredune plants on New Zealand sand dunes. There is also very little overlap in the traits of natives and exotics studied in this ecosystem (Fig. 1), suggesting that it will not be possible to prevent reinvasion by planting these natives to fill all available niches in the community (Funk et al. 2008). Hence, there will likely need to be ongoing management intervention in the form of intensive weeding of exotic species in areas where there has been planting of native sand dune plants to exclude exotic species and ensure the success of such restoration projects.

Acknowledgements

We thank Hannah Lewis and Ruud van Driel for help with fieldwork and Marleen Pierik for advice. Jill Rapson and an anonymous reviewer provided helpful comments on the manuscript.