Effects of soil and microclimatic conditions on the ...

Effects of soil and microclimatic conditions on the community-level plant functional traits across different tropical forest types Yong Jiang, Runguo Zang, Xinghui Lu, Yunfeng Huang, Yi Ding, Wande Liu, Wenxing Long, Junyan Zhang & Zhidong Zhang Plant and Soil An International Journal on Plant-Soil Relationships ISSN 0032-079X Volume 390 Combined 1-2 Plant Soil (2015) 390:351-367 DOI 10.1007/s11104-015-2411-y

1 23

Your article is protected by copyright and all rights are held exclusively by Springer International Publishing Switzerland. This eoffprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Plant Soil (2015) 390:351–367 DOI 10.1007/s11104-015-2411-y

REGULAR ARTICLE

Effects of soil and microclimatic conditions on the community-level plant functional traits across different tropical forest types Yong Jiang & Runguo Zang & Xinghui Lu & Yunfeng Huang & Yi Ding & Wande Liu & Wenxing Long & Junyan Zhang & Zhidong Zhang Received: 9 October 2014 / Accepted: 4 February 2015 / Published online: 13 February 2015 # Springer International Publishing Switzerland 2015

Abstract Background and aims Variations in community-level plant functional traits might indicate the adaptation strategies of vegetations under changing environment. Here, we assess the key factors in controlling functional community compositions across different forest types. Methods We established 12 1-ha permanent forest dynamics sites, two in each of the six old-growth forests. Each site was divided into 25 plots (20 m×20 m). We measured five plant traits and eight abiotic factors in each plot. Correlations between environmental factors and plant traits were assessed by step-wise multiple regression. Results The key abiotic factors controlling communitylevel plant functional traits in TDMRF were CO, SWC and AN. Those in TCF were CO and AN, in TLRF was AP, in TMRF were CO and TP, in TMEF were pH, SOM and TP, and in TMDF were SWC, pH, TN and TP, respectively. The controlling factors for each of the trait Responsible Editor: Harry Olde Venterink . Electronic supplementary material The online version of this article (doi:10.1007/s11104-015-2411-y) contains supplementary material, which is available to authorized users. Y. Jiang : R. Zang (*) : X. Lu : Y. Huang : Y. Ding : W. Liu : W. Long : J. Zhang : Z. Zhang Key Laboratory of Forest Ecology and Environment of State Forestry Administration; Institute of Forest Ecology, Environment and Protection, Chinese Academy of Forestry, Beijing 100091, China e-mail: [email protected] R. Zang e-mail: [email protected]

across the vegetations were: SWC, pH,TN,TP and AN for CWM_SLA; CO and AN for CWM_CC;CO, pH, SOM,TN and AN for CWM_LNC;CO,pH,TN and TP for CWM_LPC;and CO,SWC,TN, AN and AP for CWM_WD. Conclusions Key soil and microclimatic factors controlling plant functional traits change with vegetation types. Each community-level plant functional trait is associated with 3–5 abiotic factors across the vegetations. Keywords Abiotic environment . China . Plant functional traits . Vegetation types . Specific leaf area . Wood density . Tropical forests Abbreviations TDMRF deciduous forest TCF coniferous forest TLRF lowland forest TMRF montane forest TMEF evergreen forest TMDF dwarf forest SLA CC LNC LPC WD

Tropical deciduous monsoon rain forest Tropical coniferous forest Tropical lowland rain forest Tropical montane rain forest Tropical montane evergreen forest Tropical montane dwarf forest Specific leaf area Leaf total chlorophyll content Leaf nitrogen concentration per mass Leaf phosphorus concentration per mass Wood density

Author's personal copy 352

CO SWC SOM TN TP AN AP AT RH

Plant Soil (2015) 390:351–367

Canopy openness Soil water content Soil organic content Soil total nitrogen Soil total phosphorus Soil available nitrogen Soil available phosphorus Air temperature Relative air humidity

Introduction A pervasive way of exploring species distribution is to examine the functions of a plant and the way that these functions are related to the environment (Swenson and Weiser 2010). Plant functional traits are the features (morphological, physiological, phenological) that represent ecological strategies and determine how plants respond to environmental factors, affect other organisms and influence ecosystem properties. Variation in plant functional traits, and trait syndromes, has proven useful for tackling many important ecological questions at a range of scales (Cornelissen et al. 2003). Specific leaf area (SLA), one of the important functional traits, which is positively related to growth rates, leaf turnover rates, foliar nutrient concentrations and photosynthetic capacities (Long et al. 2011b; Roche et al. 2004). Leaf nitrogen concentration per mass (LNC) and leaf phosphorus concentration per mass (LPC) are positively correlated with one another and show similar relationships with photosynthetic rate (Chaturvedi et al. 2011; PérezHarguindeguy et al. 2013). High leaf nitrogen concentration per mass or leaf phosphorus concentration per mass is generally associated with high nutritional quality to the consumers in food webs (Cornelissen et al. 2003). Leaf total chlorophyll content (CC) is linked to leaf nitrogen concentration per mass and hence to photosynthetic rate (Le Maire et al. 2004). Moreover, leaf total chlorophyll content provides information about the physiological state of the plant (Lichtenthaler et al. 1996). It is a direct indicator of photosynthetic capacity and plant productivity (Chaturvedi et al. 2011). Wood density (WD) is a vitally important trait to the ecology of woody plants as it can reflect carbon allocation patterns and growth/survival strategies of plants in different plant communities (Swenson and Enquist 2007). Several studies have demonstrated that many kinds of environmental factors (such as: light, temperature,

soil nitrogen, soil phosphorous, soil organic matter, etc.) and their combination can influence or determine the distribution patterns of vegetations (Baribault et al. 2012; Bodin et al. 2013; Sollins 1998). In some lowland tropical forests, temperature may in case be more important determinant of phenology (Long et al. 2011a) and more N circulates annually through lowland tropical forests and does so at higher concentrations (Martinelli et al. 1999). The previous studies conducted in the tropics have proven that phosphorous was the most likely limited nutrient in low elevation tropical forests owing to its low availabilities (Reich 2014), while nitrogen might become a limiting nutrient in some tropical montane rain forests at higher elevations (Bobbink et al. 2010; Elser et al. 2007). The probable reasons might be:(1) high elevation montane rainforests have lower temperatures;(2) the activities of nitrogen fixation by microbes were low;(3) decomposition of organic matter in soils was slow and the cycling rate of nitrogen was slow;(4) less nitrogen fixation species. Soils in tropical montane cloud forest are usually acidic and, consequently, decomposition of soil organic matter (Hölscher et al. 2004) and soil concentration of exchangeable bases and base saturation are usually low (Ledo et al. 2013), and pH can also be an important factor responsible for the floristic differentiation of vegetations in low temperature regions due to precipitation and moisture availability (Chytrý et al. 2008). Plant strategies result from suites of correlated traits which usually have trade-offs in resources allocation under different environmental conditions, they can therefore be best understood from a whole-plant perspective (Reich 2014). For example, leaves with higher tissue density and toughness, but lower specific leaf area and nutrient concentrations, have lower photosynthetic rates but longer life span (Westoby et al. 2002). Such leaves are more defended against herbivores and pathogens (Agrawal and Fishbein 2006; Hanley et al. 2007), and therefore confer slower growth but higher survival rates (Poorter and Bongers 2006). From a succession perspective, at the high light environment of earlysuccession stages, fast-growing species that were characterized by leaves with short life span and lower tissue density but higher specific leaf area and nutrient concentrations, have higher photosynthetic rates but shorter life spans, which is advantageous for a rapid acquisition of available resources (Ordoñez et al. 2009; Shipley et al. 2006).

Author's personal copy Plant Soil (2015) 390:351–367

The differences in functional community structures are usually linked with the variations in environmental conditions in different vegetations. Long et al. (2011a) showed that air temperature and soil phosphorus availability correlated with trait differences between two types of tropical cloud forests, and across 59 Amazonian plots and four Neotropical forests, respectively wood density was found negatively associated with soil fertility (Baker et al. 2004: Chave et al. 2006), whereas within Guyana the variation in mean wood density was not correlated with either precipitation or soil fertility (ter Steege and Hammond 2001). In a lowland wet forest in Panama, variation in leaf functional traits was driven by microhabitat variation related to plant stature, reflecting the vertical continuum of environmental conditions through the canopy (Santiago and Wright 2007). In temperate forests, tree height of some conifer species was positively related to leaf mass per unit area but negatively related to both in mass-based leaf nitrogen content and in mass and areabased rates of photosynthesis (Rijkers et al. 2000a). Although numerous studies show how plant traits affect the performance of individual species, quantification of the trait responses across different vegetation types or various spatial scales is nevertheless limited (Long et al. 2011a). This is partly because few studies directly linked plant traits to vegetations under differing environmental conditions. Although the occurrence of an individual plant within a local community is controlled by environments operating on functional traits rather than on species identity per se (Messier et al. 2010), a community-level analysis approach can directly connect environmental factors with the dominant functional traits in an ecosystem (Swenson and Weiser 2010; Wright et al. 2005b). Here, we examine the differences in both community-level plant functional traits and environmental factors, and assess which are the key environmental factors in controlling variation of functional plant community structures across the six old-growth forest types. We hypothesize that: (1) The functional community composition for each of the six old-growth forest type was controlled by varied sets of abiotic factors (i.e., when data for each vegetation type were analyzed); and (2) Each of the measured functional trait at community-level was associated with several combinations of abiotic factors across all the six vegetation types (i.e., when data for all the vegetations combined were analyzed).

353

Materials and methods Study site The study site is located in the Bawangling Forest Region (18°53′–19°20′N, 108°58′–109°53′E) on Hainan Island, China (Fig. 1). Hainan Island is located at the northern edge of tropical Asia, with a total area of 33,920 km2. For less than 0.5 % of China’s land area, over 4600 plant species have been recorded, equal to 15 % of the nation’s total, and some 500 species are endemic to the island (Farm et al. 2001). Bawangling National Nature Reserve (BNNR, 18°57′–19°11′N, 109°03′–109°17′E) is a protected area at the southwestern region of Hainan Island (Fig. 1). Summary of site characteristics of the six old-growth forest vegetation types in the tropical nature reserve is listed in Table 1. The low elevation tropical forests (deciduous forest, coniferous forest, and lowland forest) are similar in precipitation; however, the local terrain and soils for each of the forest types are highly variable. Deciduous forest occurs in drought and hot habitats and tree species are characterized by deciduousness and thorny stems. Coniferous forest is located in habitats with low soil nutrient and low water holding-capacity, where environmental conditions are stressful for broadleaved trees, a few species of conifers become dominant. Due to the special geological, environmental and floristic conditions, coniferous forest often has a distinct community structure and contains a unique biodiversity that is different from that of adjacent tropical rainforests. Lowland forest is distributed in better soil conditions, occupies the largest area and is the zonal vegetation type at low elevation (≤800 m) area on Hainan Island. Montane forest is the zonal vegetation type at intermediate elevation which has the best environmental conditions and a high diversity in comparison with other tropical rain forests in the region. Evergreen forest and dwarf forest are usually collectively called the tropical cloud forests due to frequent fog or cloud, low temperature, high humidity and strong winds. The environmental conditions within these two vegetation types are quite different compared to other forests (Long et al. 2011b). Field sampling We established 12 1-ha (100 m×100 m) permanent forest dynamics sites in the old-growth stands of the six forest vegetation types in the Bawangling National


Plant Soil (2015) 390:351–367

Fig. 1 Sketch map of different vegetation types in the Bawangling National Nature Reserve on Hainan Island, China. The dots indicate the locations of the plots. TDMRF represents tropical deciduous monsoon rain forest; TCF represents tropical coniferous

forest; TLRF represents tropical lowland rain forest; TMRF represents tropical montane rain forest; TMEF represents tropical montane evergreen forest; TMDF represents tropical montane dwarf forest

Nature Reserve during 2007–2010. Each permanent site was divided into 25 plots (20 m×20 m). Fielding sampling was carried out in these plots from 2011 to 2013. All of the woody stems with diameter at breast height (dbh)≥1 cm in each plot were identified and measured. The nomenclature of species follows Flora of China (English edition;http://www.efloras.org).

by subtracting the dry weight from the water saturated weight of each sample during the dry season. After taking this moisture value, the three samples were thoroughly mixed before analyses to represent the soil properties of each 20 m×20 m plot. Soil pH, available phosphorus (AP, mg kg−1), total phosphorus (TP, g kg−1), available nitrogen (AN, mg kg−1), total nitrogen (TN, g kg−1), and soil organic matter (SOM, g kg−1) were analyzed in the laboratory according to standard methods (Long et al. 2011a). We estimated the precipitation (mm) using the empirical regression equation: precipitation (precipitation: mm) = 65.4 + 37.1× log (elevation) (Jiang and Liu 1991), based on all of the measured elevations in each of the 20 m×20 m plots

Measurement of environmental factors Soil samples were taken at three points in each 20 m× 20 m plot. A core of the top 20 cm of soil was taken at each point. Soil samples were air-dried, then sieved (2 mm). Soil water content (SWC, %) was calculated

22±0.1

139

Sp1-Sp6

Air temperature(°C)

Number of species

Dominant species

265

150

23±0.1

155±0.8

579

Sp7-Sp12

167

21±0.01

167±0.9 146

21±0.02

170±0.1

668

site 2

Sp13-Sp18

194

21±0.1

168±0.1

577

site1

TLRF

172

21±0.3

164±0.2

594

site 2

Sp19-Sp24

230

19±0.04

175±0.1

904

site1

TMRF

194

19±0.01

178±0.1

923

site 2

Sp25-Sp30

126

17±0.02

179±0.1

1200

site1

TMEF

140

17±0.04

180±0.4

1200

site 2

Sp31-Sp36

105

17±0.2

180±0.4

1305

site1

TMDF

110

16±0.1

181±0.1

1360

site 2

Species data are based on trees with diameter at breast height≥1 cm. Pricipitation and air temperature were expressed as the mean±SD. Dominant species are based on basal areas that rank the top six in each forest type. Sp1: Terminalia hainanensis Exell in Sunyatsenia; Sp2: Croton laevigatus Vahl; Sp3: Lagerstroemia balansae Koehne; Sp4: Streblus ilicifolius (Vidal) Corner; Sp5: Lannea coromandelica (Houtt.) Merr. Sp6: Syzygium cumini (L.) Skeels; Sp7: Pinus latteri Mason; Sp8: Aporusa dioica (Roxb.) Muell. Arg; Sp9: Engelhardtia roxburghiana; Sp10: Schima superba Gardn. et Champ; Sp11: Adinandra hainanensis; Sp12: Memecylon scutellatum (Lour.) Hook. et Arn; Sp13: Cyclobalanopsis patelliformis (Chun) Y. C. Hsu et H. W. Jen; Sp14: Lithocarpus fenzelianus A. Camus; Sp15: Vatica mangachapoi Blanco; Sp16: Cyclobalanopsis bambusaefolia (Hance) Chun ex Y. C. Hsu et H. W. Jen; Sp17: Canarium album (Lour.) Raeusch; Sp18: Winchia calophylla A. DC; Sp19: Lithocarpus fenzelianus A. Camus; Sp20: Castanopsis hystrix Miq; Sp21: Cyclobalanopsis patelliformis (Chun) Y. C. Hsu et H. W. Jen; Sp22: Canarium album (Lour.) Raeusch; Sp23: Xanthophyllum hainanense Hu; Sp24: Mallotus hookerianus (Seem.) Muell. Arg; Sp25: Cyclobalanopsis disciformis (Chun et Tsiang) Y. C. Hsu et H. W. Jen; Sp26: Distylium racemosum; Sp27: Dacrydium pierrei Hickel; Sp28: Syzygium araiocladum Merr. et Perry; Sp29: Syzygium buxifolium Hook. et Arn; Sp30: Ternstroemia gymnanthera (Wight et Arn.) Beddome; Sp31: Distylium racemosum; Sp32: Syzygium buxifolium Hook. et Arn; Sp33: Engelhardtia roxburghiana; Sp34: Cinnamomum tsoi Allen; Sp35: Cyclobalanopsis disciformis (Chun et Tsiang) Y. C. Hsu et H. W. Jen; Sp36: Pentaphylax euryoides Gardn. et Champ

340

160±0.9

Precipitation(mm)

site1

site1

site 2

TCF

TDMRF

Vegetation types

Altitude (m)

Characteristics

Table 1 Major characteristics of the 12 sites across the six old-growth forest vegetation types

Author's personal copy

Plant Soil (2015) 390:351–367 355


(300 plots in total). To determine understory irradiance, each 20 m×20 m plot was subdivided into four 10 m× 10 m quadrates, at each 10 m×10 m quadrate centre, hemispherical canopy photographs were taken at 1.5 m above ground level using a fish-eye lens (HMV1v8, Delta-T Devices Ltd, Cambridge, UK) mounted on a tripod. Canopy cover was calculated from each photograph as the percentage of closed-canopy pixel using the Gap Light Analyzer software (Frazer et al. 1999). Canopy openness (CO, %) was then obtained from the formula Canopy Openness =1- Canopy Cover. Air temperature and relative air humidity for each vegetation type were obtained from each HOBO data logger; totally 12 HOBOs (Onset, U23-001, USA) were placed 1.3 m above ground level in the center of each forest dynamics site repectively. Since we had only 12 HOBOs mounted to measure air temperature (AT, °C) and relative air humidity (RH; %), we could not obtain data on AT and RH at the plot level, so we did not consider these two factors in step-wise multiple regression. However, we just put the annual average values measured over 1 year (2012 to 2013) at the 12 site centers by 12 HOBOs on the appendix. 1 for describing the general microclimatic differences among the six vegetation types. Trait measurement All woody species (dbh≥1 cm) were sampled to measure functional traits data (supporting information appendix 5). For species with more than five welldeveloped individuals, the five functional traits were measured. For species with less than five individuals, we assigned these species the mean value measured for the same species in other plots. Totally, in the 300 plots, there are 610 species but 497 species were measured. For each individual, five to ten leaves were collected and made measurements of width, length and area. Specific leaf area (SLA, cm2 g−1) for broadleaved species was calculated as leaf area divided by leaf dry mass (after drying for 72 h at 70 °C). SLA for needle leaves was calculated by the method of Gower (Gower et al. 1999). Leaf total chlorophyll content (CC) was estimated using three values per lamina from a SPAD 502Plus meter (Konica Minolta, Osaka, Japan). Leaf nutrition characteristics, including leaf nitrogen concentration per mass (LNC, mg g−1) and leaf phosphorus concentration per mass (LPC, mg g−1), were measured by standardized protocols for plant functional trait measurements (Pérez-

Plant Soil (2015) 390:351–367

Harguindeguy et al. 2013). To characterize species’ wood density (WD, g cm−2), we sampled branches (1 cm≤diameter≤2 cm) from 1 to 10 individuals across the species in each plot. We removed the pith, phloem, and bark, measured fresh volume by water displacement and determined dry mass after drying for 72 h at 70 °C (Cornwell et al. 2006). Meanwhile, we chose ten species of hardwood and corkwood respectively, ten individuals per species were sampled nearby the study sites, and cores of stem were taken with an increment borer (Haglof Increment Borer, Sweden) after sampling of branches. We found that the core density linearly related to the branch density for hardwood (ρcore = 1.054ρbranch, R2 =0.95, P