Most plants are clonal in nature. Clonal ramets can share water, nutrients and photosynthate, especially when they experience patchy resources. Patch contrast (i.e. a difference in resources among patches) and patch direction (i.e. source–sink relations) are among the basic attributes of spatial patchiness. Here, I hypothesize that young established ramets in nutrient-rich patches support old ramets in nutrient-poor patches when ramets are subjected to different patch contrasts and patch directions.
In a greenhouse experiment, old and young ramets of Glechoma longituba were grown in four combinations consisting of patch contrast and patch direction. Minus patch direction refers to a patch combination in which parent ramets grow in nutrient-rich patches while connected daughter ramets grow in nutrient-poor ones and plus patch direction is the opposite direction. I measured photosynthesis and fluorescence traits, harvested all ramets, took morphological measures, weighed their dry mass and determined their nutrient uptake and use.
For parental ramets of G. longituba, patch contrast and patch direction and their interactions had no significant effects on net photosynthetic rate, maximal fluorescence yield, photochemical quenching (quenching refers to any process which decreases the fluorescence intensity of a given substance), non-photochemical quenching, nutrient uptake, biomass and stolon weight ratio. Patch direction alone significantly affected root weight ratio. Large patch contrast enhanced N use efficiency (NUE) and P use efficiency (PUE); plus patch direction decreased NUE, but increased PUE; the patch contrast by patch direction interaction affected PUE and K use efficiency (KUE). There were significant interactions between patch direction and patch contrast on PUE and KUE. It is concluded that soil nutrient patchiness may influence nutrient use strategies, but not nutrient uptake, photosynthesis and growth of parent ramets of G. longituba connected to daughter ramets, and that patch contrast and patch direction jointly affect PUE and KUE.
Most plants are clonal in nature. Resource sharing provides clonal plants with advantages in colonizing space, sampling the environment, tolerating stresses, overcoming establishment risk and responding to competing neighbors (Hartnett and Bazzaz 1983, 1984; Hutchings and Wijesinghe 1997, 2008; Slade and Hutchings 1987; Stuefer 1996) and, therefore, has received increased attention during the past decades (Janečekš et al. 2008). In the field, soil nutrients are patchily distributed, thereby resulting in favorable or unfavorable nutrient patches (Hutchings and de Kroon 1994; Hutchings and Wijesinghe 1997; Jackson and Caldwell 1993; Poor et al. 2005; Robinson 1994; Stuefer 1996; Stuefer and Hutchings 1994). Patch contrast (i.e. a difference in resources among patches (Kotliar and Wiens 1990)) and patch direction (i.e. source–sink relations (Marshall 1990)) are among the basic attributes of spatial heterogeneity of resources. Thus, it is necessary to study plant performance in response to different patch combinations.
Since nutrients are associated with physiological processes of plants (Larcher 2003) and soil nutrients are patchily distributed, it is required to consider physiological activities (e.g. photosynthesis and fluorescence) under the context of heterogeneity. To my knowledge, no studies explicitly address effects of patchy nutrients on nutrient use and photosynthesis of plants. Here, I focused on photosynthetic rate, fluorescence traits, nutrient uptake and use because these are linked with nutrient status in plants (Larcher 2003). Patch direction refers to patch combinations and is associated with nutrient levels in patches. For parent ramets, their daughter ramets probably face diverse new microhabitats. When parent ramets position their daughter ramets in resource-richer patches, the cost of parental care may be relatively low and vice versa. Thus, I concentrated on the performance of parent ramets connected to daughter ramets in four habitat types. If nutrient levels are greater in parent ramets’ patches than in daughter ramets’ patches, patch direction is minus, and plus patch direction is the opposite direction. Glechoma longituba L. positions its ramets in a variety of microhabitats, thereby allowing them to experience diverse nutrient patchiness (Chu et al. 2006; Liao et al. 2003; Zhang and He 2009a, 2009b).
Resource sharing can involve acropetal (from old ramets to young ramets) and basipetal (from young ramets to old ramets) movement within clones (Alpert 1991; Alpert and Mooney 1986; Dietz and Steinlein 2001; Friedman and Alpert 1991). A major question in this context is concerned with the conditions under which the two processes will occur (Dietz and Steinlein 2001). For example, for Glechoma hederacea, the movement of resources in the xylem is acropetal (Price and Hutchings 1992), and the movement of N in Fragaria chiloensis is also mainly acropetal (Alpert and Mooney 1986). Stuefer and Hutchings (1994) pointed out that a difference in water potential is needed to transport nutrients to mother ramets through the xylem. Previous studies demonstrated that old established ramets under favorable conditions support young developing ramets, especially when these were under stressful environments (de Kroon et al. 2005; Dietz and Steinlein 2001). However, few studies tested whether young established ramets can support old ramets. I hypothesize: if physiological integration is basipetal, then old ramets in nutrient-poor patches may benefit from nutrient-rich patches due to the transport of nutrients (Marshall 1990; Pitelka and Ashmun 1985). I predicted that, if young established ramets support old ramets, then this support allows old ramets to have higher biomass production when they are connected to young ramets in favorable conditions than when they are connected to young ramets under unfavorable conditions.
Nutrient patchiness can shift source–sink relations and induce the translocation of nutrients within a clone (Farley and Fitter 1999; Huber-Sannwald and Jackson 2001; Hutchings and Wijesinghe 1997;,Stuefer 1996). I hypothesize that the potential of nutrient translocation increases with patch contrast intensity and also predicted that nutrient patchiness modifies the nutrient uptake and use efficiency of parent ramets.
MATERIALS AND METHODS
Glechoma longituba (Lamiaceae) is distributed in forests and along roadsides and creeks. Ramets of G. longituba have two zygomorphic leaves originating from a stolon node, and every leaf axil bears one bud that may grow into a secondary stolon (Wu and Chen 1974). Previous studies with G. longituba revealed a high capacity for clonal integration (Chu et al. 2006; Liao et al. 2003).
Two factors (nutrient-patch contrast and nutrient-patch direction), each with two levels, were crossed in a full factorial design, creating four habitat types. Nutrient-patch contrast consisted of small contrast and large contrast, and nutrient-patch direction included plus direction and minus direction. Patch contrast refers to the degree of difference between patches (Kotliar and Wiens 1990). Plus direction refers to the patch combination in which parent ramets grew in nutrient-poor patches while connected daughter ramets grew in nutrient-rich ones and minus direction is the opposite direction.
Glechoma longituba materials were derived from 10 clones within a small population in a temperate deciduous forest on the Dongling Mountain, located 120 km northwest of Beijing. Therefore, the plants were from no more than 10 genotypes. These clones were vegetatively propagated in a greenhouse at the Institute of Botany, Chinese Academy of Sciences (IBCAS). Here, I used different genotypes to get general responses of ramets. Similar-sized parent–daughter ramet pairs, each ramet with two leaves (∼3–4 cm in height), were grown in 60-cm-long × 20-cm-wide × 25-cm-high plastic trays filled with river sand. This sand was chosen because its grain sizes were homogeneous and its contents of N and P were less than 0.003%. The experiment was conducted in a plastic greenhouse at the IBCAS. The parent ramets were given 0.3% of nutrient solution; the daughter ramets were given 0.1, 0.2, 0.4 and 0.5% of nutrient solution, respectively. I used a water-soluble fertilizer (Peters Professional, Scotts Company) with an elemental ratio of 20N: 20P2O5: 20 K2O to achieve different nutrient supplies. A previous study has showed that the biomass production of G. longituba plants significantly increased along the gradient of these five levels of nutrients under uniform nutrient conditions (Zhang and He, 2009a). When nutrient levels were 0.4 and 0.5% for daughter ramets, the nutrient-patch direction was plus; when nutrient levels were 0.1 and 0.2% for daughter ramets, the nutrient-patch direction was minus. The differences between the concentration of nutrient solution supplied to the parent and daughter ramets were 0.1 and 0.2% for small contrast and large contrast, respectively. Nutrient solution was supplied once a week to maintain the nutrient levels. During the experiment, each ramet group (i.e. all ramets in a growth container) was given 200 ml of water every 2–4 days, depending on the local weather and development stages; the soil water content was monitored with a Soil Moisture Meter MP-160 (Meridian, Australia) and soil water content was ∼10%. To minimize the effect of location, all the trays were randomly distributed in the central area of the greenhouse at the IBCAS. It is important to note that I did not rotate these trays because the trays were too heavy and the ramets per tray were so many. There were 10 replicates for each combination of patch contrast and patch direction. The experiment ran from August to October 2007, and the total amount of nutrient solution for each ramet group was 1000 ml.
In mid-September, fully developed, similar-sized leaves were chosen for physiological measurements. Net photosynthetic rate (Pn) was determined with an open flow gas exchange system (LI-COR 6400) under growing conditions (photosynthetically active radiation: ∼1500 μmol m−2 s−1). Fluorescence induction kinetics is normally studied after a dark adaptation of the leaf, and the plant does not receive any light during this period (Reigosa and Weiss 2001). After a dark adaptation of at least 2.0 h sufficient for the photosystem II (PSII) reaction centers to open, maximal chlorophyll fluorescence yield, photochemical quenching and non-photochemical quenching were measured with a PAM-101/102/103 Chlorophyll Fluorescence System (Heinz Walz GmbH, Germany). Pn indicates the apparent net photosynthetic capacity. The three fluorescence traits are among the basic characteristics of PSII and reflect internal mechanisms of photosynthesis. Maximal chlorophyll fluorescence yield indicates the PSII activity and decreases after exposure to high but not injurious temperatures, photochemical quenching is related to the energy transformation in the reaction centers of the PSII and non-photochemical quenching represents, after the chlorophyll excitement, the different energy return ways (mainly as heat and energy transfer to PSI) (Reigosa and Weiss 2001). Six and eight different plants were chosen for the determinations of Pn and fluorescence, respectively. Specifically, only one fully developed, similar-sized leaf per ramet was chosen.
At the end of the experiment, all parent ramets were harvested and then separated into leaves, stolons and roots. Three samples of stolons and roots from each parent ramet were randomly collected to determine their specific stolon length (ratio of stolon length to its biomass, m g−1) and specific root length (ratio of root length to its biomass, m g−1). All the materials were dried at 85°C for 72 h and weighed. Stolon weight ratio (ratio of stolon biomass to total biomass, g g−1) and root weight ratio (ratio of root biomass to total biomass, g g−1) were calculated.
All the oven-dried materials were ground into a fine powder for nutrient analyses (i.e. nitrogen (N), phosphorus (P), and potassium (K)). The contents of N and P were determined after standard acid digestion with an autoanalyser, and the content of K was determined by extracting 10 g of plants with 50 ml of 2 M NaCl and through analyzing with an atomic absorption spectrophotometer. The nutrient (i.e. N, P and K) uptake was estimated through calculating the total amount of nutrients in leaves, stolons and roots. Nutrient use efficiency was calculated as the ratio of ramets’ biomass to their nutrient uptake (g g−1).
I analysed the impacts of nutrient-patch contrast and nutrient-patch direction as fixed factors, as well as their interactions on 13 traits of parent ramets, using two-way ANOVA. All statistical analyses were carried out using SPSS 13.0 (SPSS Inc., Chicago).
For parent ramets of G. longituba, net photosynthetic rate, maximal chlorophyll fluorescence yield, photochemical quenching and non-photochemical quenching were not significantly affected by nutrient-patch contrast and direction and their interactions (Fig. 1A–D and Table 1). In other words, the changes in nutrient-patch contrast or direction did not result in dramatic influences on photosynthesis (all P > 0.05; Table 1). In contrast, this nutrient patchiness did influence nutrient (N, P and K) use efficiencies, but not nutrient uptake (Fig. 2 and Table 1). For example, large contrast significantly enhanced N use efficiency (NUE) (P = 0.000) and P use efficiency (PUE) (P = 0.000) (Fig. 2A and B); plus direction significantly decreased NUE (P = 0.007) (Fig. 2A), but increased PUE (P = 0.000) (Fig. 2B); Also, the patch contrast × patch direction interaction conferred significant impacts on PUE (P = 0.000) and K use efficiency (KUE) (P = 0.001) (Fig. 2B and C and Table 1).
|PD||PC||PD × PC|
|PD||PC||PD × PC|
PD, patch direction; PC, patch contrast. Values of P < 0.05 are in bold. Pn, net photosynthetic rate; MFY, maximal fluorescence yield; qP, photochemical quenching; qN, non-photochemical quenching; SWR, stolon weight ratio; RWR, root weight ratio. The degrees of freedom for the errors: F1,20 for Pn; F1,28 for MFY, qP and qN; F1,36 for N uptake, P uptake, K uptake, NUE, PUE, KUE, biomass, SWR, and RWR.
Two-way analysis of variance (ANOVA) showed that nutrient-patch contrast and direction and their interactions had no significant effects on total biomass and stolon weight ratio in parent ramets (Fig. 3A and B; Table 1). However, patch direction conferred a dramatic effect on root weight ratio in parent ramets (P = 0.001), but patch contrast (P = 0.396) and its interaction with patch direction (P = 0.201) had no influence on root weight ratio (Fig. 3C).
In this study, the performance of parental ramets of G. longituba was addressed only, thereby not extrapolating the overall outcome for the whole ramets. Although only four out of 13 traits were significantly affected by nutrient patchiness (Table 1), levels of significance of these effects were very high. This indicates that these effects are not an artifact due to multiple testing. Nutrient use efficiency has been intensively studied in the past decades (Aerts and Chapin 2000), but these studies are based on the assumption that habitats of plants are homogeneous. As far as I know, this study is the first to provide evidence that nutrient-patch contrast and direction have profound effects on nutrient use efficiency. These findings suggest that nutrient translocation from young to old ramets of G. longituba does occur. The differences in nutrient solution concentrations in my experiment were relatively small compared to the field. Thus, effects of nutrient-patch contrast on nutrient use efficiency are likely to be common in nature.
For old fragments of G. longituba, NUE and PUE tended to be greater in large contrast than in small contrast. This phenomenon has been observed in a previous experiment (Zhang and He 2009a). More interestingly, nutrient-patch direction conferred opposite influences on NUE and PUE, implying that there exists a trade-off between NUE and PUE. This trade-off has been observed in previous studies (Aerts and Chapin 2000). Environment strongly affects plant nutrient concentration by changing allocation among organs and the major environmental effect on tissue nutrient composition is to alter the concentration of nutrients associated with metabolism (Lambers et al. 1998). High nutrient use efficiency is apparently a result of low nutrient requirement for photosynthesis (Aerts and Chapin 2000). Soil N and P limit plant growth in most terrestrial ecosystems (Elser et al. 2007; Güsewell 2004; Vitousek 1982). Previous studies demonstrated that soil N and P are very limited in warm-temperate forests on the Dongling Mountain (Chen 1997). In my study, soil nutrient patchiness conferred much stronger effects on NUE and PUE than on KUE, suggesting that G. longituba fragments exhibit more sensitive responses to soil N and P than soil K.
Previous other studies have shown that the rate of photosynthesis increases in the ramets remaining well-lit when plants are partially subjected to shading (Hartnett and Bazzaz 1983; Lambers et al. 1998), but this phenomenon was not observed in my study. Slade and Hutchings (1987) found that there was independence between ramets of G. hederacea. Additionally, physiological integration or independence between ramets depends on different structural levels (Price and Hutchings 1992). Stuefer and Hutchings (1994) found no evidence for assimilate translocation between interconnected fragments of G. hederacea, and the absence of this transport was most probably caused by an irreversible acropetal flow of photosynthates. In some clonal species, assimilates are predominantly moved acropetally and to a much lesser extent in a basipetal direction, and such asymmetry in intraclonal resource translocation may strongly constrain the viability of integration responses to environmental patchiness (Stuefer 1996).
Patch direction refers to patch combinations and is associated with nutrient levels in patches. In my study, parent ramets were supplied with constant nutrient level, and daughter ramets were provided with higher or lower nutrient levels compared to the nutrient level experienced by parent ramets. When daughter ramets were grown in nutrient-richer patches, patch direction was plus and vice versa. The direction of resource translocation between ramets is determined by relative strengths of resource gradients (Marshall 1990; Pitelka and Ashmun 1985). Birch and Hutchings (1994, 1999) found no evidence for basipetal transport of nutrients within G. hederacea clones. Stuefer and Hutchings (1994) proposed that a water potential difference is required for nutrients to be transported back through the xylem to reach mother ramets. Watson and Casper (1984) and Watson (1986) proposed that ‘integrated physiological unit’ affects the transport of resources, i.e. the transport of resources is sectorially restricted and plant sectoriality is determined by its vascular structure (Stuefer 1996; Vuorisalo and Hutchings 1996). Some ramets appear to maintain their independent functioning as a normal feature of the carbon allocation within the clone (Marshall 1996). In G. hederacea, the vascular architecture of shoots strongly affects patterns of resource integration, mainly because stolon branches originating from the same node lack direct vascular contacts and are hence unable to share resources (Price and Hutchings 1992; Price et al. 1996). Aegopodium podagraria from forests are more dependent on resource sharing than those from gardens, indicating that origin matters for levels of resource sharing (Nilsson and D'Hertefeldt 2008).
Spatial combinations of resource patches confer profound effects on the growth of clonal plants (Day et al. 2003; Hutchings and Wijesinghe 2008; Janečekš et al. 2008; Wijesinghe and Hutchings 1997, 1999; Zhang and He 2009b). If the basipetal movement of resources occurs between G. longituba fragments, then the old fragments will have a higher photosynthetic capacity and biomass production when they are connected to young fragments under nutrient-richer patches than they are connected to young fragments under nutrient-poorer patches. However, the dry mass of the old fragments of G. longituba was independent of the nutrient levels experienced by their connected young fragments. In recent studies, biomass accumulation of G. longituba fragments remained unchanged along a gradient of nutrient-patch contrasts (Zhang and He 2009a). Physiological changes of plants can provide mechanisms for their growth responses (Englund and Cooper 2003). Net photosynthetic rate reflects the apparent light use efficiency while chlorophyll fluorescence reflects the intrinsic mechanisms of PSII (Larcher 2003; Reigosa and Weiss 2001; Taiz and Zeiger 2002). Net photosynthetic rate, maximal chlorophyll fluorescence yield, photochemical quenching and non-photochemical quenching did not vary with nutrient levels experienced by interconnected fragments. Thus, old fragments have a similar capacity for photosynthesis, thereby enabling them to produce equal amounts of biomass.
In summary, plant materials used in my experiment were from 10 clones in temperate deciduous forests. Thus, my findings can reflect the responses of parental ramets of G. longituba in temperate forests. It is important to note that G. longituba is widely distributed in diverse habitats in subtropic and temperate areas (Wu and Chen 1974) and it is difficult to extrapolate the findings in temperate forests to other habitats or subtropic areas because there are differences in climate and other factors among them. It is concluded: (i) clonal integration, which is affected by patch contrast and patch direction, may influence nutrient use efficiency, but not nutrient uptake, photosynthesis and growth of parent ramets of G. longituba connected to daughter ramets, and (ii) patch contrast by patch direction interactions confer effects on both PUE and KUE, but not NUE.
National Natural Science Foundation of China (40435014).
I am grateful to the associate editor Markus Fischer and two anonymous reviewers for their valuable suggestions that greatly improve the quality of my article.