Plant Biology ISSN 1435-8603
RESEARCH PAPER
Plasticity in latitudinal patterns of leaf N and P of Oryza rufipogon in China W. Zhou1,2, Z. Wang1,2, W. Xing1 & G. Liu1,3 1 Key Laboratory of Aquatic Botany and Watershed Ecology, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China 2 College of Life Sciences, University of Chinese Academy of Sciences, Beijing, China 3 Hubei Key Laboratory of Wetland Evolution & Ecological Restoration, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
Keywords Common garden transplant; N:P ratio; phenotypic plasticity; population differentiation; soil nutrient availability; temperature; wetland plant. Correspondence G. Liu, Key Laboratory of Aquatic Botany and Watershed Ecology, Wuhan Botanical Garden, the Chinese Academy of Sciences, Wuhan 430074, China. E-mail: [email protected] Editor M. Hawkesford Received: 25 September 2013; Accepted: 15 November 2013 doi:10.1111/plb.12147
ABSTRACT Characterising the adaptability in nature of plant stoichiometric patterns across geographic or environmental gradients is important in advancing our understanding of the organisation of plant–nutrient relationships. We examined correlations between plant nutrient traits, latitude, longitude, climate and soil variables in 34 populations of Oryza rufipogon across its range. We further compared the responses of population transplants at two experimental gardens: one beyond its northern natural range and another near the southern limit, to assess the nature of geographic variation in plant nutrients. The study showed that leaf P of O. rufipogon in the field was negatively correlated with latitude and largely depended on temperature and soil P availability. Leaf N was not related to latitude but was significantly correlated with precipitation and soil N concentration. Leaf N:P ratio was largely determined by absorption efficiency of P. Transplantation revealed that there were no significant associations of leaf nutrients with geographic, climatic or soil variables of origin in either of the experimental gardens, indicating phenotypic plasticity. However, examination of relationships between response ratios of leaf nutrients and change ratio of climate and soil environments, as well as norms of reaction in the transplantation experiment, revealed more complexity, suggesting both substantial genotypic diversity and the existence of genotype 3 environment interactions in these populations of O. rufipogon. These data indicate that adaptive plasticity response of plants to temperature and soil P availability significantly explain the observed shifts in leaf N, P and N:P of O. rufipogon along latitudinal gradients.
INTRODUCTION A fundamental goal of ecological stoichiometry is to explain the pattern of an organism’s C:N:P stoichiometry across ecological gradients. It is widely accepted that plant nutrients may depend primarily on latitude, climate and soil fertility (Reich & Oleksyn 2004; He et al. 2008; Han et al. 2011). Two leading hypotheses, the soil age hypothesis (SAH) and the growth rate hypothesis (GRH), have been proposed to explain the patterns in plant nitrogen (N) and phosphorus (P) with latitude. SAH predicts that leaf N and P increase (and N:P decreases) with latitude, because of geochemical limitations to P availability associated with soil age (Reich & Oleksyn 2004), which makes soil fertility a key factor. In contrast, GRH predicts that leaf N and P decrease (and N:P increases) with latitude because of higher growth rates in warmer environments, thus leading to a high P requirement for RNA and protein synthesis (Elser et al. 2000). A number of broad-scale surveys of leaf stoichiometry generally suggest that leaf P content increases and leaf N:P decreases with an increase in latitude (McGroddy et al. 2004; Reich & Oleksyn 2004; He et al. 2008), supporting SAH but not GRH. Kerkhoff et al. (2005) modified GRH and suggested that higher growth rates are needed in temperate than in tropical climates to complete growth and reproduction within
shorter growing seasons, thus producing a similar pattern expected with SAH. Clearly, the question remains open as to whether the mechanisms underlying the origin and maintenance of observed stoichiometric patterns involve growth adaptation or simple acclimation responses of plants to availability of soil N and P. The challenge to the interpretation of such mechanisms in nature is to understand the relative partitioning of control between climate factors and soil fertility. However, identifying the individual driving factor that underlies the geographical clines is not straightforward because, inevitably, soil nutrients vary collaterally with geographic and climate variables. For example, temperature-related effects on biogeochemistry could lead to leaf N and P decreasing monotonically with latitude across the globe (Aerts & Chapin 1999); on the other hand, soils are geologically older and more weathered, resulting in higher N:P ratios at lower latitudes relative to higher latitudes (Crews et al. 1995). These overlapping variations in soil and climate raise questions about clear separation of these two effects on stoichiometric patterns. A previous study found that pines had higher nutrient absorption efficiency at higher latitudes than at lower latitudes, and interpreted the gradient as reflecting nutrient availability of the area of origin (Oleksyn et al. 2003). However, Lovelock et al. (2007) implied that such
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
917
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
a pattern could also provide some support for the temperature-dependent variation in growth rate with latitude. Because of the extensive geographic range and broad-scale site and environmental heterogeneity, variations in nutrient conservation expressed as local demographic behaviour across a geographic range may arise in fundamentally different ways: through phenotypic plasticity or local genetic differentiation. A history of geographic variation in natural selection might be expected to generate corresponding adaptive evolution along the environmental gradient. Intrinsic differences in physiology among populations may therefore produce differences in nutrient absorption efficiency. Treseder & Vitousek (2001a) reported common garden-grown populations of evergreen tree Metrosideros polymorpha from fertile sites had significantly lower N absorption than plants from N-limited sites, suggesting local genetic differentiation. Alternatively, phenotypic plasticity can be sufficient to compensate functionally for the effects of resource limitation to maintain higher efficiency of absorption (Reich et al. 2001; Bott et al. 2008). Characterising the adaptability in nature of stoichiometric variation to climate and soil across geographic or environmental gradients is important in advancing our understanding of the organisation of plant–nutrient relations and in the face of ongoing environmental changes. Common garden studies of biogeographically distant populations across temperature and soil fertility should go a long way toward discriminating between the contributions of climate and soil, and the potential constraining mechanism (Hedin 2004). Here, studies of populations in situ, combined with transplant experiments at two common gardens, were conducted to investigate the nature of geographic variation in leaf N, P and N:P in common wild rice, Oryza rufipogon throughout its range in China (about 1250 km in latitude and 1400 km in longitude). O. rufipogon is considered the ancestor of Asian cultivated rice (O. sativa), and is the most important germplasm for rice improvement (Oka 1988). A previous study of 47 populations of Chinese O. rufipogon with similar geographic origins to our material revealed that high levels of genetic variability were detected at six polymorphic microsatellite DNA loci, suggesting underlying genotypic differences in this wideranging species (Gao 2004). The objectives of this study were to examine (i) how geographic gradients have potentially impacted leaf N, P and N:P of O. rufipogon; (ii) what are the relative contributions of soil and climate to leaf nutrient patterns; and (iii) to what extent do genetic and environmental factors affect leaf nutrient patterns. MATERIAL AND METHODS Field survey and transplant experiments Material of O. rufipogon was analysed in direct collections from field populations and also after cultivation of plants of the same populations in two common environments (gardens). A detailed account of the field survey of O. rufipogon populations and design of the transplant experiment is provided in Zhou et al. (2013). We collected rhizome fragments from five individuals in each of 34 natural populations, representing the geographic distribution of O. rufipogon in China, between September and November 2010. The rhizome fragments were 918
planted individually in the glasshouse at Wuhan Botanical Garden (CAS) to produce two genetically matched batches of ramets. One batch was transplanted to the northern experimental garden (Huazhong Agricultural University, Wuhan; 30°29′ N, 114°19′ E) and the other to the southern garden (Lingshui County, Hainan Province, Hainan Island; 18°34′ N, 110°02′ E) in a completely randomised block design, with five replicate clones from each population per garden. The flag leaf of rice contributes most to grain yield and hence was chosen for sampling in this study. As flowering of O. rufipogon is in response to shortening days, flag leaves develop from late July (in northern populations) to early September (in southern populations), after a critical day length occurs. Accordingly, flag leaves of at least five individuals in each population were collected from all natural and transplanted populations from August to October 2010, when plants were at a similar development stage. Leaves were dried for 48 h at 65 °C and then ground using a ball mill (NM200; Retsch, Haan, Germany) before determining their total N and P concentrations. Leaf N was determined using the Kjeldahl method (automatic Kjeldahl N determination analyser, JK9870; Nanbei, Henan, China), and leaf P was measured using the Mo–Sb colorimetric method, after H2SO4– HClO4 digestion. The absorbance of each sample was measured at 700 nm, 20 min after addition of the Mo–Sb–ascorbic acid reagent. Three soil samples per site from a depth of ca. 20 cm were collected (reflecting substratum heterogeneity) with a shovel; samples were then immediately passed through a 2-mm sieve and the 0; Fig. 2a) and significantly lower (ratio < 0; Fig. 2b) in the southern transplant garden, whereas AMT showed the reverse trend; soil P showed no significant change in the two transplant gardens relative to the field. In response to these changes, leaf N and P significantly increased in the southern transplant garden or in both transplant gardens, but leaf N:P was not obviously affected (Fig. 2b). The response ratios of leaf N and P were negatively correlated with the change ratios of soil N and positively correlated with soil P in at least one garden or in both gardens (Fig. 3). AMT had a significant positive influence on leaf P but not leaf N in both transplant gardens. Substantial phenotypic variation was detected in leaf N between plants from natural populations and those in transplanted populations. Leaf N content was significantly higher when transplanted southwards than when transplanted northwards (Fig. 4). However, no significant variations were found in leaf P and N:P between the three different environments. Norms of reaction revealed considerable genotypic differences between populations for all leaf nutrients, in concert with environmental effects (Fig. 4). Moreover, two-way ANOVA models detected significant population 9 environment interactions for leaf N, P and N:P (Table 2), providing further evidence for local genetic adaptation. DISCUSSION Variation of natural populations
Leaf nutrient response to transplantation Unlike natural populations, none of the transplanted populations showed significant correlation of leaf nutrients with
Table 1. Correlations (r) between leaf N, P and N:P in natural populations of Oryza rufipogon and geographic, climate and soil variables. Partial correlation coefficients represent the unique contributions of latitude to variations of leaf N, P and N:P after controlling for the effects of a specific climate and soil variables.
latitude latitude (control AMT) latitude (control AMP) latitude (control Log10 soil N) latitude (control Log10 soil P) longitude annual mean temperature (AMT) annual mean precipitation (AMP) Log10 soil N Log10 soil P
Log10 leaf N
Log10 leaf P
leaf N:P
0.19 0.02 0.08 0.06 0.13 0.11 0.21 0.35* 0.43** 0.27
0.67*** 0.10 0.62*** 0.59*** 0.60*** 0.09 0.64*** 0.19 0.31 0.40*
0.52** 0.11 0.63*** 0.58*** 0.50** 0.24 0.52** 0.17 0.09 0.16
*P < 0.05, **P < 0.01, ***P < 0.001.
Oryza rufipogon is a widely distributed tropical or subtropical wetland species that exhibits a northern range limit in southern China and northern Burma that has remained stable; in China it extends some 1250 km northwards from the South China Sea, and also has a range of about 1400 km from east to west. This study demonstrated that populations of O. rufipogon growing in their natural habitat showed clear leaf nutrient trends over these geographic and climatic ranges. Leaf P and N: P tend to be more strongly associated with temperature, whereas foliar N showed no correlation with temperature and was positively related to precipitation. Temperature had no significant effect on leaf N, agreeing with most previous results (Reich & Oleksyn 2004; Ordo~ nez et al. 2009). However, the trends towards decreasing leaf P and increasing leaf N:P at lower temperature are not consistent with previously reported patterns of higher leaf P and lower leaf N:P at high latitudes at global scales (McGroddy et al. 2004; Reich & Oleksyn 2004; Ordo~ nez et al. 2009). Higher leaf P and lower leaf N:P might be expected as a response to variation in plant growth rate: higher temperature at lower latitudes would promote plant growth rates, thus increasing the requirement of P for RNA and protein synthesis (Elser et al. 2000). Hence, the results may support GRH. Similar trends in leaf nutrients and temperature have been found at genus and species levels. For
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
919
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
Fig. 1. Leaf N, P and N:P ratio in relation to latitude, annual mean temperature (AMT), annual mean precipitation (AMP), soil N and P concentrations for northern transplant populations (NEP), natural populations (NP), and southern transplant populations (SEP) of Oryza rufipogon. Significance levels: *P < 0.05; **P < 0.01; ***P < 0.001.
(a)
(b)
Fig. 2. (a) Change ratios of three main environmental factors and (b) response ratio of leaf N, P and N:P in northern transplant populations (NEP) and southern transplant populations (SEP), compared with natural populations (NP). Asterisk labels significant differences from zero in change/ response ratios (Wilcoxon signed rank tests, ***P < 0.001, *P < 0.05).
example, although the general trends of leaf N and P declined and N:P ratios rose with increasing temperatures for all 1280 plant species examined by Reich & Oleksyn (2004), at the 920
genus level the patterns of leaf N, P and N:P varied broadly with temperature. Waite & Sack (2011) detected that, in 12 studied bryophyte species, the dominant Marchantia polymorpha showed decreased P and increased N:P ratios with increasing temperature. The reason for the difference in latitudinal variation in leaf nutrients at species or genus level from that at the vegetative level remains unclear. It is widely accepted that relationships between temperature and plant traits associated with metabolism and growth rates are complicated because of the adaptive adjustments of plant physiology to temperature (Enquist et al. 2003; Kerkhoff et al. 2005). Thus, speciesspecific adaption to temperature is likely a key factor. Leaf N was negatively related to soil N, while leaf P was positively related to soil P and weakly and negatively related to soil N (r = 0.31, P = 0.08). The negative response for leaf N to soil N coincides with the result of Ordo~ nez et al. (2009), where they examined 474 plant species distributed across global sites and found that soil total N had opposite patterns with leaf traits. G€ usewell & Koerselman (2002) analysed the data from field studies, fertilisation experiments and growth experiments in wetlands, and found that plant–soil relationships for N and P could be either positive or negative, and many of them were not correlated. There are several reasons that might help to explain a negative response of leaf N to soil N. First, a nutrient resorption mechanism from senescing leaves makes plants less dependent on soil nutrient availability. In this study, the sampled flag leaf of O. rufipogon is able to acquire substantial nutrients from the older leaves that contribute to grain yield. Evidence from fertilisation experiments indicates that plant N resorption may remain unchanged or decrease with increasing supply of fertiliser in some nutrient-limited ecosystems (Van Heerwaarden et al. 2003; Kozovits et al. 2007; Soudzilovskaia et al. 2007; L€ u & Han 2010; L€ u et al. 2013). Second, plant
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
Zhou, Wang, Xing & Liu
Plasticity in latitudinal patterns of leaf N and P
Fig. 3. Correlations between response ratios for leaf N, P and N:P and change ratios for soil N, P and annual mean temperature (AMT) in northern transplant populations (NEP) and southern transplant populations (SEP). Lines were plotted for relationships with P < 0.05.
Leaf N (mg·g–1)
(a) 25
F2,69 = 4.96, P < 0.01
20 15
ab
b
NP
SEP
a
10 5 0
(b) Leaf P (mg·g–1)
3
F2,69 = 1.70, P = 0.19
2
1
0
(c) 20
Leaf N:P
Fig. 4. Comparison of mean plant nutrients (left-hand panels) and their reaction norms (right-hand panels) for 34 populations of Oryza rufipogon in their natural habitat (NP) and transplanted to northern (NEP) and southern (SEP) gardens: (a) leaf N, (b) leaf P and (c) leaf N:P ratios. Box-plots show median values, 25th and 75th percentiles, and 5th and 95th percentiles. Results from oneway ANOVA and Tukey–Kramer multiple contrast tests are shown. Treatments indicated by different letters are significantly different (P < 0.05).
18 16 14 12 10 8 6 4 2 0
F2,69 = 0.36, P = 0.70
NEP
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
NEP
NP
SEP
Populations
921
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
Table 2. Summaries of two-way ANOVAs of the effects of environment and population of origin on leaf N, P and N:P in 34 populations of Oryza rufipogon grown in three environments (natural environment, northern and southern transplant gardens). environment (E)
leaf N leaf P leaf N:P
E9P
population (P)
df
F
P
df
F
P
df
F
P
2 2 2
5.24 1.87 1.95
RESEARCH PAPER
Plasticity in latitudinal patterns of leaf N and P of Oryza rufipogon in China W. Zhou1,2, Z. Wang1,2, W. Xing1 & G. Liu1,3 1 Key Laboratory of Aquatic Botany and Watershed Ecology, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China 2 College of Life Sciences, University of Chinese Academy of Sciences, Beijing, China 3 Hubei Key Laboratory of Wetland Evolution & Ecological Restoration, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
Keywords Common garden transplant; N:P ratio; phenotypic plasticity; population differentiation; soil nutrient availability; temperature; wetland plant. Correspondence G. Liu, Key Laboratory of Aquatic Botany and Watershed Ecology, Wuhan Botanical Garden, the Chinese Academy of Sciences, Wuhan 430074, China. E-mail: [email protected] Editor M. Hawkesford Received: 25 September 2013; Accepted: 15 November 2013 doi:10.1111/plb.12147
ABSTRACT Characterising the adaptability in nature of plant stoichiometric patterns across geographic or environmental gradients is important in advancing our understanding of the organisation of plant–nutrient relationships. We examined correlations between plant nutrient traits, latitude, longitude, climate and soil variables in 34 populations of Oryza rufipogon across its range. We further compared the responses of population transplants at two experimental gardens: one beyond its northern natural range and another near the southern limit, to assess the nature of geographic variation in plant nutrients. The study showed that leaf P of O. rufipogon in the field was negatively correlated with latitude and largely depended on temperature and soil P availability. Leaf N was not related to latitude but was significantly correlated with precipitation and soil N concentration. Leaf N:P ratio was largely determined by absorption efficiency of P. Transplantation revealed that there were no significant associations of leaf nutrients with geographic, climatic or soil variables of origin in either of the experimental gardens, indicating phenotypic plasticity. However, examination of relationships between response ratios of leaf nutrients and change ratio of climate and soil environments, as well as norms of reaction in the transplantation experiment, revealed more complexity, suggesting both substantial genotypic diversity and the existence of genotype 3 environment interactions in these populations of O. rufipogon. These data indicate that adaptive plasticity response of plants to temperature and soil P availability significantly explain the observed shifts in leaf N, P and N:P of O. rufipogon along latitudinal gradients.
INTRODUCTION A fundamental goal of ecological stoichiometry is to explain the pattern of an organism’s C:N:P stoichiometry across ecological gradients. It is widely accepted that plant nutrients may depend primarily on latitude, climate and soil fertility (Reich & Oleksyn 2004; He et al. 2008; Han et al. 2011). Two leading hypotheses, the soil age hypothesis (SAH) and the growth rate hypothesis (GRH), have been proposed to explain the patterns in plant nitrogen (N) and phosphorus (P) with latitude. SAH predicts that leaf N and P increase (and N:P decreases) with latitude, because of geochemical limitations to P availability associated with soil age (Reich & Oleksyn 2004), which makes soil fertility a key factor. In contrast, GRH predicts that leaf N and P decrease (and N:P increases) with latitude because of higher growth rates in warmer environments, thus leading to a high P requirement for RNA and protein synthesis (Elser et al. 2000). A number of broad-scale surveys of leaf stoichiometry generally suggest that leaf P content increases and leaf N:P decreases with an increase in latitude (McGroddy et al. 2004; Reich & Oleksyn 2004; He et al. 2008), supporting SAH but not GRH. Kerkhoff et al. (2005) modified GRH and suggested that higher growth rates are needed in temperate than in tropical climates to complete growth and reproduction within
shorter growing seasons, thus producing a similar pattern expected with SAH. Clearly, the question remains open as to whether the mechanisms underlying the origin and maintenance of observed stoichiometric patterns involve growth adaptation or simple acclimation responses of plants to availability of soil N and P. The challenge to the interpretation of such mechanisms in nature is to understand the relative partitioning of control between climate factors and soil fertility. However, identifying the individual driving factor that underlies the geographical clines is not straightforward because, inevitably, soil nutrients vary collaterally with geographic and climate variables. For example, temperature-related effects on biogeochemistry could lead to leaf N and P decreasing monotonically with latitude across the globe (Aerts & Chapin 1999); on the other hand, soils are geologically older and more weathered, resulting in higher N:P ratios at lower latitudes relative to higher latitudes (Crews et al. 1995). These overlapping variations in soil and climate raise questions about clear separation of these two effects on stoichiometric patterns. A previous study found that pines had higher nutrient absorption efficiency at higher latitudes than at lower latitudes, and interpreted the gradient as reflecting nutrient availability of the area of origin (Oleksyn et al. 2003). However, Lovelock et al. (2007) implied that such
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
917
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
a pattern could also provide some support for the temperature-dependent variation in growth rate with latitude. Because of the extensive geographic range and broad-scale site and environmental heterogeneity, variations in nutrient conservation expressed as local demographic behaviour across a geographic range may arise in fundamentally different ways: through phenotypic plasticity or local genetic differentiation. A history of geographic variation in natural selection might be expected to generate corresponding adaptive evolution along the environmental gradient. Intrinsic differences in physiology among populations may therefore produce differences in nutrient absorption efficiency. Treseder & Vitousek (2001a) reported common garden-grown populations of evergreen tree Metrosideros polymorpha from fertile sites had significantly lower N absorption than plants from N-limited sites, suggesting local genetic differentiation. Alternatively, phenotypic plasticity can be sufficient to compensate functionally for the effects of resource limitation to maintain higher efficiency of absorption (Reich et al. 2001; Bott et al. 2008). Characterising the adaptability in nature of stoichiometric variation to climate and soil across geographic or environmental gradients is important in advancing our understanding of the organisation of plant–nutrient relations and in the face of ongoing environmental changes. Common garden studies of biogeographically distant populations across temperature and soil fertility should go a long way toward discriminating between the contributions of climate and soil, and the potential constraining mechanism (Hedin 2004). Here, studies of populations in situ, combined with transplant experiments at two common gardens, were conducted to investigate the nature of geographic variation in leaf N, P and N:P in common wild rice, Oryza rufipogon throughout its range in China (about 1250 km in latitude and 1400 km in longitude). O. rufipogon is considered the ancestor of Asian cultivated rice (O. sativa), and is the most important germplasm for rice improvement (Oka 1988). A previous study of 47 populations of Chinese O. rufipogon with similar geographic origins to our material revealed that high levels of genetic variability were detected at six polymorphic microsatellite DNA loci, suggesting underlying genotypic differences in this wideranging species (Gao 2004). The objectives of this study were to examine (i) how geographic gradients have potentially impacted leaf N, P and N:P of O. rufipogon; (ii) what are the relative contributions of soil and climate to leaf nutrient patterns; and (iii) to what extent do genetic and environmental factors affect leaf nutrient patterns. MATERIAL AND METHODS Field survey and transplant experiments Material of O. rufipogon was analysed in direct collections from field populations and also after cultivation of plants of the same populations in two common environments (gardens). A detailed account of the field survey of O. rufipogon populations and design of the transplant experiment is provided in Zhou et al. (2013). We collected rhizome fragments from five individuals in each of 34 natural populations, representing the geographic distribution of O. rufipogon in China, between September and November 2010. The rhizome fragments were 918
planted individually in the glasshouse at Wuhan Botanical Garden (CAS) to produce two genetically matched batches of ramets. One batch was transplanted to the northern experimental garden (Huazhong Agricultural University, Wuhan; 30°29′ N, 114°19′ E) and the other to the southern garden (Lingshui County, Hainan Province, Hainan Island; 18°34′ N, 110°02′ E) in a completely randomised block design, with five replicate clones from each population per garden. The flag leaf of rice contributes most to grain yield and hence was chosen for sampling in this study. As flowering of O. rufipogon is in response to shortening days, flag leaves develop from late July (in northern populations) to early September (in southern populations), after a critical day length occurs. Accordingly, flag leaves of at least five individuals in each population were collected from all natural and transplanted populations from August to October 2010, when plants were at a similar development stage. Leaves were dried for 48 h at 65 °C and then ground using a ball mill (NM200; Retsch, Haan, Germany) before determining their total N and P concentrations. Leaf N was determined using the Kjeldahl method (automatic Kjeldahl N determination analyser, JK9870; Nanbei, Henan, China), and leaf P was measured using the Mo–Sb colorimetric method, after H2SO4– HClO4 digestion. The absorbance of each sample was measured at 700 nm, 20 min after addition of the Mo–Sb–ascorbic acid reagent. Three soil samples per site from a depth of ca. 20 cm were collected (reflecting substratum heterogeneity) with a shovel; samples were then immediately passed through a 2-mm sieve and the 0; Fig. 2a) and significantly lower (ratio < 0; Fig. 2b) in the southern transplant garden, whereas AMT showed the reverse trend; soil P showed no significant change in the two transplant gardens relative to the field. In response to these changes, leaf N and P significantly increased in the southern transplant garden or in both transplant gardens, but leaf N:P was not obviously affected (Fig. 2b). The response ratios of leaf N and P were negatively correlated with the change ratios of soil N and positively correlated with soil P in at least one garden or in both gardens (Fig. 3). AMT had a significant positive influence on leaf P but not leaf N in both transplant gardens. Substantial phenotypic variation was detected in leaf N between plants from natural populations and those in transplanted populations. Leaf N content was significantly higher when transplanted southwards than when transplanted northwards (Fig. 4). However, no significant variations were found in leaf P and N:P between the three different environments. Norms of reaction revealed considerable genotypic differences between populations for all leaf nutrients, in concert with environmental effects (Fig. 4). Moreover, two-way ANOVA models detected significant population 9 environment interactions for leaf N, P and N:P (Table 2), providing further evidence for local genetic adaptation. DISCUSSION Variation of natural populations
Leaf nutrient response to transplantation Unlike natural populations, none of the transplanted populations showed significant correlation of leaf nutrients with
Table 1. Correlations (r) between leaf N, P and N:P in natural populations of Oryza rufipogon and geographic, climate and soil variables. Partial correlation coefficients represent the unique contributions of latitude to variations of leaf N, P and N:P after controlling for the effects of a specific climate and soil variables.
latitude latitude (control AMT) latitude (control AMP) latitude (control Log10 soil N) latitude (control Log10 soil P) longitude annual mean temperature (AMT) annual mean precipitation (AMP) Log10 soil N Log10 soil P
Log10 leaf N
Log10 leaf P
leaf N:P
0.19 0.02 0.08 0.06 0.13 0.11 0.21 0.35* 0.43** 0.27
0.67*** 0.10 0.62*** 0.59*** 0.60*** 0.09 0.64*** 0.19 0.31 0.40*
0.52** 0.11 0.63*** 0.58*** 0.50** 0.24 0.52** 0.17 0.09 0.16
*P < 0.05, **P < 0.01, ***P < 0.001.
Oryza rufipogon is a widely distributed tropical or subtropical wetland species that exhibits a northern range limit in southern China and northern Burma that has remained stable; in China it extends some 1250 km northwards from the South China Sea, and also has a range of about 1400 km from east to west. This study demonstrated that populations of O. rufipogon growing in their natural habitat showed clear leaf nutrient trends over these geographic and climatic ranges. Leaf P and N: P tend to be more strongly associated with temperature, whereas foliar N showed no correlation with temperature and was positively related to precipitation. Temperature had no significant effect on leaf N, agreeing with most previous results (Reich & Oleksyn 2004; Ordo~ nez et al. 2009). However, the trends towards decreasing leaf P and increasing leaf N:P at lower temperature are not consistent with previously reported patterns of higher leaf P and lower leaf N:P at high latitudes at global scales (McGroddy et al. 2004; Reich & Oleksyn 2004; Ordo~ nez et al. 2009). Higher leaf P and lower leaf N:P might be expected as a response to variation in plant growth rate: higher temperature at lower latitudes would promote plant growth rates, thus increasing the requirement of P for RNA and protein synthesis (Elser et al. 2000). Hence, the results may support GRH. Similar trends in leaf nutrients and temperature have been found at genus and species levels. For
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
919
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
Fig. 1. Leaf N, P and N:P ratio in relation to latitude, annual mean temperature (AMT), annual mean precipitation (AMP), soil N and P concentrations for northern transplant populations (NEP), natural populations (NP), and southern transplant populations (SEP) of Oryza rufipogon. Significance levels: *P < 0.05; **P < 0.01; ***P < 0.001.
(a)
(b)
Fig. 2. (a) Change ratios of three main environmental factors and (b) response ratio of leaf N, P and N:P in northern transplant populations (NEP) and southern transplant populations (SEP), compared with natural populations (NP). Asterisk labels significant differences from zero in change/ response ratios (Wilcoxon signed rank tests, ***P < 0.001, *P < 0.05).
example, although the general trends of leaf N and P declined and N:P ratios rose with increasing temperatures for all 1280 plant species examined by Reich & Oleksyn (2004), at the 920
genus level the patterns of leaf N, P and N:P varied broadly with temperature. Waite & Sack (2011) detected that, in 12 studied bryophyte species, the dominant Marchantia polymorpha showed decreased P and increased N:P ratios with increasing temperature. The reason for the difference in latitudinal variation in leaf nutrients at species or genus level from that at the vegetative level remains unclear. It is widely accepted that relationships between temperature and plant traits associated with metabolism and growth rates are complicated because of the adaptive adjustments of plant physiology to temperature (Enquist et al. 2003; Kerkhoff et al. 2005). Thus, speciesspecific adaption to temperature is likely a key factor. Leaf N was negatively related to soil N, while leaf P was positively related to soil P and weakly and negatively related to soil N (r = 0.31, P = 0.08). The negative response for leaf N to soil N coincides with the result of Ordo~ nez et al. (2009), where they examined 474 plant species distributed across global sites and found that soil total N had opposite patterns with leaf traits. G€ usewell & Koerselman (2002) analysed the data from field studies, fertilisation experiments and growth experiments in wetlands, and found that plant–soil relationships for N and P could be either positive or negative, and many of them were not correlated. There are several reasons that might help to explain a negative response of leaf N to soil N. First, a nutrient resorption mechanism from senescing leaves makes plants less dependent on soil nutrient availability. In this study, the sampled flag leaf of O. rufipogon is able to acquire substantial nutrients from the older leaves that contribute to grain yield. Evidence from fertilisation experiments indicates that plant N resorption may remain unchanged or decrease with increasing supply of fertiliser in some nutrient-limited ecosystems (Van Heerwaarden et al. 2003; Kozovits et al. 2007; Soudzilovskaia et al. 2007; L€ u & Han 2010; L€ u et al. 2013). Second, plant
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
Zhou, Wang, Xing & Liu
Plasticity in latitudinal patterns of leaf N and P
Fig. 3. Correlations between response ratios for leaf N, P and N:P and change ratios for soil N, P and annual mean temperature (AMT) in northern transplant populations (NEP) and southern transplant populations (SEP). Lines were plotted for relationships with P < 0.05.
Leaf N (mg·g–1)
(a) 25
F2,69 = 4.96, P < 0.01
20 15
ab
b
NP
SEP
a
10 5 0
(b) Leaf P (mg·g–1)
3
F2,69 = 1.70, P = 0.19
2
1
0
(c) 20
Leaf N:P
Fig. 4. Comparison of mean plant nutrients (left-hand panels) and their reaction norms (right-hand panels) for 34 populations of Oryza rufipogon in their natural habitat (NP) and transplanted to northern (NEP) and southern (SEP) gardens: (a) leaf N, (b) leaf P and (c) leaf N:P ratios. Box-plots show median values, 25th and 75th percentiles, and 5th and 95th percentiles. Results from oneway ANOVA and Tukey–Kramer multiple contrast tests are shown. Treatments indicated by different letters are significantly different (P < 0.05).
18 16 14 12 10 8 6 4 2 0
F2,69 = 0.36, P = 0.70
NEP
Plant Biology 16 (2013) 917–923 © 2014 German Botanical Society and The Royal Botanical Society of the Netherlands
NEP
NP
SEP
Populations
921
Plasticity in latitudinal patterns of leaf N and P
Zhou, Wang, Xing & Liu
Table 2. Summaries of two-way ANOVAs of the effects of environment and population of origin on leaf N, P and N:P in 34 populations of Oryza rufipogon grown in three environments (natural environment, northern and southern transplant gardens). environment (E)
leaf N leaf P leaf N:P
E9P
population (P)
df
F
P
df
F
P
df
F
P
2 2 2
5.24 1.87 1.95
