Uprooting resistance of two tropical tree species for sand dune stabilization

Coastal windbreak restoration is important in Taiwan for agroforestry and sand dune stabilization. Australian pine (Casuarina equisetifolia Forst.) is the main species in windbreaks. It often suffers from serious uprooting and waterlogging damages, whereas sea hibiscus (Hibiscus tiliaceus L.) is more resistant to wind and tolerant to waterlogging. It is suggested that sea hibiscus can be substituted for Australian pine in coastal windbreak restoration. However, the adaptive mechanism of its root system to wind is not well understood. In this study, a field experiment was conducted to investigate the anchorage capabilities and root morphology of 10-year-old Australian pine and sea hibiscus plants. The results showed that root system morphologies of Australian pine and sea hibiscus plants belonged to taproot system and heart system, respectively. Root systems of both species were distributed towards northeast and southwest, which coincided with the monsoon directions. Sea hibiscus plants had significantly larger root collar diameter, longer taproot length, larger root biomass and shoot biomass than that of Australian pine plants. Additionally, sea hibiscus plants had significantly larger root volume than Australian pine plants. Moreover, sea hibiscus developed significantly stronger root functional traits, that is, root density (245%), root tissue density (300%) and the root to shoot ratio (138%) than Australian pine plants. Consistently, the root maximum uprooting resistance of sea hibiscus plants was significantly higher than that of Australian pine plants. These results demonstrate that sea hibiscus plants have stronger anchorage capability and they are more suitable for windbreak restoration and sand dune stabilization.


INTRODUCTION
Australian pine (Casuarina equisetifolia Forst.), belonging to the Casuarinaceae family, is a nitrogen-fixing tree species (Diem and Dommergues, 1983;Mailly et al., 1994;Subbarao and Rodríguez-Barrueco, 1995;Tani et al., 2003) introduced to Taiwan in 1910 for windbreak reforestation (Liu and Liao, 1980).Sea hibiscus (Hibiscus tiliaceus L.), affiliated with the family Malvaceae, is also an introduced tropical tree species for windbreak and *Corresponding author.E-mail: jtlee@mail.Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License sand dune stabilization in coastal agricultural lands of Taiwan (Chang, 1993;Liu and Liao, 1980).Both species have high potential for agroforestry, lumber production and coastal dune stabilization.There are 10,000 ha of windbreak forests along the coast of Taiwan.Australian pine is the main planting species used for reforestation.However, Australian pine has suffered from short life span of 20 years, serious insect pest of Lymantria xylina and brown root rot disease caused by Phellinus noxius (Chang, 1995;Ann et al., 2002).In the southwestern coast of Taiwan, Australian pine often suffers from serious uprooting and waterlogging caused by typhoons in summer and autumn (Chen et al., 2008).However, sea hibiscus, with a shorter trunk and many crooked, sprawling, intertwined branches, exhibits higher wind resistance.It is suggested that sea hibiscus can be substituted for Australian pine in coastal windbreak restoration.However, its ability to withstand uprooting by wind has not been investigated.Therefore, a comparison of anchorage capability between Australian pine and sea hibiscus plants is important for the implementation of windbreak restoration projects.
Typhoon, that is, tropical cyclone, comes with windstorm and torrential rain in summer and autumn, is one of the major causes of destruction of windbreak forests in Taiwan.Wind plays an important role in the growth and development of coastal tree species.Several studies have demonstrated that wind affects root morphology and mechanical properties of trees (Telewski and Jaffe, 1986;Stokes et al., 1995;Wu et al., 2016).Stokes and Mattheck (1996) classified tree root systems into heart system, plate system and taproot system.Lee et al. (2016) indicated that root distribution pattern of edge trees of 8-year-old Australian pine windbreak forest coincides with the monsoon direction, and its average maximum uprooting force amounts to 1.04 ± 0.29 kN.However, the effect of wind on root system morphology and mechanical properties of sea hibiscus have not been investigated.This study was conducted to (1) investigate the effect of wind on root system morphology and anchorage capability of 10-year-old Australian pine and sea hibiscus plants grown in windbreak plantation and (2) compare the differences in root morphology and anchorage capability between these two species in order to apply in ecotechnological projects for sand dune stabilization.

Sample selection and site characteristics
In July 2016, fourteen plants each of 10-year-old Australian pine and sea hibiscus were randomly selected from the windward side of a windbreak plantation located at the coastal area of Taisi Township, Yunlin County,Taiwan (120°10'22.3" E,23°43'19.7" N).The climate conditions of the experimental site are as follows: average temperature 22.7°C, relative humidity 85%, annual precipitation 1043.9 mm, northeast monsoon from October to March with average wind speed of 8.6 to 10.8 m/s, and southwest monsoon from April to September with average wind speed of 4.9 to 7.3 m/s.Seven plants each of the Australian pine and sea hibiscus were randomly sampled for growth and root system morphology investigation.Another 7 trees of each species were for uprooting test.Bulk soil samples were collected from the 0 to 30 cm layer.The samples were transported to the laboratory, air-dried, crushed to break the clods and passed through a 2-mm sieve to separate the fine particles from coarse fragments.The soil particle size analysis was performed by hydrometer method (Sheldrick and Wang, 1993).The soil texture was classified as loamy sand soil, which contained 75.1% sand, 15.3% clay and 9.6% silt.The average soil dry weight was 12.3 kN m -3 and the average soil moisture content was 23.5%.

Growth and root system morphology
Tree height and root collar diameter of 7 sampled trees of each species were measured with measuring tape and digital caliper.Tree root systems were completely exposed by the hand excavation method (Böhm, 1979).The rooting depth, root distribution, root number and length were measured.Photos of plant root systems were taken for further measurements of morphological traits.For each plant, root morphological traits, that is, taproot length, total root length, external root surface root area and root tip number were measured with the WinRHIZOPro Image Analysis System (Regent Instruments, Quebec, Canada) (Bouma et al., 2000).Water displacement method was used for measuring total root volume (Pang et al., 2011).Then, roots and stems were dried in an oven at 75°C till constant weight for biomass measurement.Root functional traits, that is, root mass density (RD in kg m -3 , root dry mass per unit volume of soil), root length density (RLD in km m -3 , root length per unit volume of soil), root tissue density (RTD in g cm -3 , root dry mass per unit volume of root), specific root length (SRL in m g -1 , root length per unit root dry mass), and root to shoot ratio (R/S, dry root mass divided by dry leaf and shoot mass) were calculated (Stokes et al., 2009;Burylo et al., 2009Burylo et al., , 2012;;Poorter et al., 2012;Gould et al., 2016).All root traits were pooled together and analyzed.

Uprooting test
The rest of the seven plants each of Australian pine and sea hibiscus were used for vertical uprooting tests.Before vertical uprooting test, tree height and root collar diameter were measured, and the stem was cut off at 20 cm above the base, and the bark was removed to prevent slippage of the pulling device.The vertical uprooting test was carried out with an in situ pullout instrument (U-Soft USPA-003, U-Soft Tech, Taipei, Taiwan) equipped with a load cell (Kyowa 5T, LUK-5TBS, Tokyo, Japan) connected to a loading recorder and control unit, and fixed on a triangular steel frame.The raw data of uprooting resistance and displacement were obtained through a portable computer system.Subsequently, the instrument was connected to the pulling device and a constant vertical pulling force was applied automatically with a speed of 2 mm min -1 , recording the change in resistance and displacement along the way, until the resisting force dropped sharply and the plant uprooted (Lee et al., 2017).

Data analysis
Variations in data of growth and root morphological traits between Australian pine and sea hibiscus plants were analyzed by SPSS 22.0 software (SPSS, Chicago, IL., USA) using t-tests, one-way analysis of variance (ANOVA) and Scheffé's method.The relationships between uprooting resistance and morphological traits

Plant growth
Statistical analysis showed that there were significant differences in growth and biomass between Australian pine and sea hibiscus plants.Sea hibiscus plants had significantly higher average root collar diameter, taproot length, root biomass and shoot biomass than that of Australian pine plants (Table 1).However, mean height of Australian pine plants was significantly higher than sea hibiscus plants.Australian pine is a deciduous tree, which can grow straight to over 30 m in height in Australia (Subbarao and Rodríguez-Barrueco, 1995), whereas sea hibiscus is a medium sized tree 5 m in height, and has numerous sprawling branches.The effect of windbreak for Australian pine is noticeable on reducing wind-blown soil erosion on the lee side due to its tall trunk and canopy.Usually, the effect can cover an area up to 10 times the height of the Australian pine windbreak.However, the effect of windbreak for sea hibiscus is for sheltering agricultural areas and providing sand dune stabilization due to its short trunk, sprawling branches and wide canopy.In general, the vulnerability of trees to wind damage increases with tree height (Peltola et al., 1999).Zubizarreta-Gerendiain et al. ( 2012) also indicated that the susceptibility of trees to wind damage is largely affected by tree species-specific, tree height and rooting characteristics.Thus, Australian pine is more vulnerable to wind damage, which coincides well with wind damage frequency in windbreak plantations.Additionally, Australian pine-infested beaches are more prone to sand loss and erosion due to the allelochemical leachates, which reduce germination and establishment of native vegetation (Wheeler et al., 2011).The results demonstrated that sea hibiscus plants develop longer taproot and larger biomass than Australian pine plants in coastal sand dune areas.Thus, sea hibiscus plants play a more important role in sand dune stabilization than Australian pine plants.

Root system morphology
Sea hibiscus plants developed bigger and deeper root systems than Australian pine plants (Figure 1).The pattern of root growth in 10-year-old Australian pine plants showed that the taproot developed to a depth of 26 cm, with many smaller lateral roots grew horizontally (Table 1 and Figure 1a).In contrast, for sea hibiscus plants, its taproot extended to a depth of 41 cm, with many bigger lateral roots which grew horizontally and vertically (Table 1 and Figure 1b).The types of root system architecture of 10-year-old Australian pine and sea hibiscus plants were classified into taproot system and heart root system, respectively.The effects of wind loading on tree root system development have been well documented (Crook and Ennos, 1996;Nicol and Ray, 1996).Dupuy et al. (2005) indicated that overturning resistance was greatest in tap-and heart-root systems whatever the soil type.Saifuddin and Normanzia (2016) reported that Leucaena (Leucaena leucocephala (Lam.) de Wit), with taproot system, is suitable for slope protection.Stokes et al. (2007) found that Silver fir (Abies alba Mill.) with taproot system and European beech (Fagus sylvatica L.) with heart root system were more resistant to overturning than Norway spruce (Picea abies L.) with plate-like root system.The results revealed that sea hibiscus plants developed larger and deeper root system than that of Australian pine plants.The analysis of morphological parameters by WinRHIZOPro revealed that sea hibiscus plants had significantly larger total root length (70%), external root surface area (60%), and root volume (126%) than Australian pine plants (Table 2).However, there was no significant difference in total root tip number between Australian pine and sea hibiscus plants.Analysis of root functional traits showed that the root density (RD), root tissue density (RTD) and root to shoot biomass ratio (R/S) of sea hibiscus plants were significantly higher than that of Australian pine plants.On average, sea hibiscus plants had significantly higher RD (245%), RTD (300%) and R/S (138%) than Australian pine plants.However, Australian pine plants had significantly higher specific root length (SRL, 400%) than sea hibiscus plants (Table 3).Our data clearly demonstrates that sea hibiscus plants allocate more biomass to roots than Australian pine plants.Coutard et al. (2008) indicated that in young Sweet cherry (Prunus avium), mechanical wind stimuli increase the allocation of biomass towards roots.Nwoke et al. (2016b) also suggested that plant species with the desirable traits of high values of relative root dry weight, high values of root density and low values of specific root length can be used   for soil and water conservation projects.This investigation showed that root systems of Australian pine and sea hibiscus plants were distributed towards northeast and southwest (Table 4).The root distribution patterns of these two species coincide with the monsoon directions.Nicoll and Ray (1996) showed that tree root systems allocated more root biomass to leeward side than the windward side in response to the prevailing wind action.Tamasi et al. (2005) also found that under wind loading, trees developed an asymmetric root distribution between the windward and leeward sides of the root system.Liu et al. (2008) reported that root system of 70year-old Tarim salt cedar (Tamarix taklamakanensis M. T. Liu) had more structural root mass and length on the leeward side than the windward side relative to the prevailing wind direction.Wheeler et al. (2011) also indicated that with shallow roots and tall canopy, Australian pines are among the first trees to fall in high winds.

Root anchorage capability
In west coast of Taiwan, strong wind in monsoon seasons affects the growth and development of windbreak trees and is one of the important stresses regarding uprooting resistance.In this study, seven replicated uprooting tests were carried out to investigate the uprooting resistance of the Australian pine and sea hibiscus plants.The average maximum uprooting resistance of sea hibiscus plants was  2.04±0.91 kN, which was significantly higher than that of Australian pine plants (1.03±0.36kN) (Table 5).The uprooting resistance-displacement curves showed a steep slope up to the maximum uprooting force (Figure 2).Stokes et al. (2005) showed that European beech (F.sylvatica L.) with heart root system was twice as resistant to uprooting as Silver fir (A.alba Mill) with taproot system and three times more resistant than Norway spruce (P.abies L.) with plate root system.Regression analysis showed the linear positive relationship between uprooting resistance and tree height, root collar diameter, root biomass and taproot length.The maximum uprooting resistance increased with increasing tree height, root collar diameter, root biomass, shoot biomass and taproot length, with the taproot length showing the strongest relation (Table 6).Dupuy et al. (2005) indicated that rooting depth was a determinant factor of tree anchorage in sandy soils.Burylo et al. (2009) showed that the Several studies have demonstrated that uprooting resistance significantly influence the soil reinforcement capability of plants (Normaniza et al., 2008;Ali and Osman, 2008;Ali, 2010;Osman et al., 2011;Saifuddin et al., 2015).Our data showed that the maximum uprooting resistance of sea hibiscus plants is significantly higher than Australian pine plants, suggesting a superior root anchorage capability by sea hibiscus plants.Taken together, these results demonstrate that sea hibiscus develop larger root system and higher uprooting resistance than Australian pine plants.These findings are important in the application of sea hibiscus in ecotechnological projects for sand dune stabilization.This is the first report to show the growth, root morphology and anchorage capability of sea hibiscus in coastal windbreak forest.Furthermore, future researches on mixed species forest system and root-soil mechanics of coastal tree species, such as Australian pine, sea hibiscus, Alexandrian laurel (Calophyllum inophyllum L.) and sea mango (Cerbera manghas L.) are needed for windbreak restoration and sand dune stabilization projects.

Conclusion
In this study, root excavation study revealed that the root system morphologies of Australian pine and sea hibiscus were classified as taproot system and heart root system, respectively.Root systems of both species were distributed towards northeast and southwest, which coincided with the directions of winter and summer monsoons.However, sea hibiscus developed larger root collar diameter, longer taproot length, higher root and shoot biomass, and larger root volume than Australian pine.On average, sea hibiscus plants had significantly larger total root length, external root surface area and root volume than that of Australian pine plants.Root functional traits, such as RD, RTD and R/S of sea hibiscus plants were significantly higher than that of Australian pine plants.The average maximum uprooting resistance of sea hibiscus plants was twice higher than that of Australian pine plants.Regression analysis indicated the linear positive relationship between uprooting resistance and tree height, root collar diameter, root biomass, shoot biomass and taproot length.Taken together, these results clearly demonstrate that sea hibiscus has superior root system morphology and anchorage capability than Australian pine.Sea hibiscus is highly recommended for windbreak restoration and sand dune stabilization practices.

Table 1 .
Growth and biomass of 10-year-old C. equisetifolia and H. tiliaceus plants.Shoot biomass.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level.

Table 2 .
Root morphological traits for 10-year-old C. equisetifolia and H. tiliaceus plants.b 2788.1±448.1 b 826.5±62.1 b 3237.0±292.9 a H. teliaceus 2667.1±135.8 a 4506.6±687.6 a 1871.6±32.2 a 3601.8±145.1 a TRL: Total root length; ERSA: External root surface area; RV: Root volume; RT: Root tip number.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level.
a RD: Root density; RLD: Root length density; RTD: Root tissue density; SRL: Specific root length; R/S: Root to shoot biomass ratio.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level.

Table 4 .
Root distribution of 10-year-old C. equisetifolia and H. tiliaceus plants.Southwest.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level.

Table 5 .
Average maximum uprooting resistance of 10-year-old C. equisetifolia and H. tiliaceus plants.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level. a

Table 6 .
Relation between morphological traits and uprooting resistance for 10-year-old C. equisetifolia and H. tiliaceus plants.Uprooting resistance.All values are the means ± standard error of 7 replicates.Values in the same column with different superscript letters significantly differ at 5% significant level.
Nwoke et al. (2016a)Root collar diameter; Rb: Root biomass; Sb: Shoot biomass; Ltr: Taproot length; Ur: maximum uprooting force is most positively correlated with stem basal diameter and taproot length is the best predictor of anchorage strength.Liu et al. (2013)showed that uprooting resistance of Chinese willow (Salix matsudana Koidz) was significantly positively correlated with plant height, total number of roots and root biomass.Nwoke et al. (2016a)reported that maximum uprooting force has a linear relationship with stem basal diameter.