Relative competitive abilities and productivity in Ginkgo and broad bean and wheat mixtures in southern China
- First Online:
- Cite this article as:
- Cao, F., Kimmins, J.P., Jolliffe, P.A. et al. Agroforest Syst (2010) 79: 369. doi:10.1007/s10457-009-9268-0
- 160 Views
Ginkgo (Ginkgo biloba L.) is a multi-value deciduous tree species grown for the production of nuts, timber and foliage for medicinal products. Understanding the ecological and biological basis for Ginkgo agroforestry systems is essential for the design of optimum Ginkgo-crop species combinations. We established two greenhouse replacement series to examine interactions between Ginkgo and wheat (Triticum aestivum L.); and Ginkgo and broad bean (Vicia faba L.). The results showed that crop species were more competitive than Ginkgo at low Ginkgo density, but less competitive than Ginkgo at high Ginkgo density. Ginkgo: wheat ratio 5:1 and Ginkgo: broad bean ratio 5:1 had relative yield total (RYT) and relative land output (RLO) values of more than one and the largest total land output (TLO) values in respective mixtures. Therefore, these two ratios might be considered optimum Ginkgo: crop ratio for enhancing the combined biomass of the Ginkgo and crop in respective mixtures. Broad bean and wheat were more competitive than Ginkgo, which was less affected by wheat than by broad bean. However, there were compensatory interactions between Ginkgo and wheat, and Ginkgo and broad bean. There was significant belowground competition for soil N between Ginkgo and the two crop species in the Ginkgo/crop mixtures. The two mixtures outperformed monocultures of the individual species when comparing the mixtures with the crop monoculture system.
KeywordsIntercropRelative yieldReplacement seriesVector analysisNitrogen use efficiencyTotal land output
In agroforestry, numerous studies on intercropping have been performed to investigate potential agronomic benefits (Rodrigues et al. 2009; Andersen et al. 2007). One of the potential benefits is that differences in the way crop species and tree species utilize resources can improve intercropping system yields and increase sustainability (Vandermeer 1989).
In a mixed species plant community, competition between species may be alleviated through different phenological, morphological and physiological characteristics (i.e., niche separation). Although niche is defined in a number of different ways by ecologists, competition and the concept of niche differentiation are key themes of research on crop mixtures (Putnam et al.1985; Pilbeam et al.1994, 1995; Siame et al. 1997). Spatial and/or temporal offsets in the utilization of environmental resources may reduce competition in mixed species associations compared to pure stands (Veresoglou and Fitter 1984; Ong et al.1996), which may allow for greater productivity (Spitters 1983; Chui and Shibles 1984; Ofori and Stern 1987). This concept has been stated as the central biophysical hypothesis for research in agroforestry (Cannell et al. 1996), and the advancement of agroforestry and other intercropping systems requires a better understanding of plant competition and other aspects of the complexity of mixtures (Sanchez 1995; Nair 1998).
Species mixtures are widely used in agricultural, forestry and agroforestry systems based on the assumption of multiple benefits of mixed plant species communities (Vandermeer 1989). Species mixtures possess greater structural and functional diversity than single species populations, which may enable mixtures to exploit environmental resources more completely (Jolliffe 1997). Sheltering of one species by another may facilitate growth of the former by reducing physiological stresses (Callaway et al. 2003). Against these benefits, there can be antagonistic interactions between plant species in mixtures. This may result from allelochemical inhibition and/or competition for resources, and productivity is not necessarily enhanced when species are mixed.
Ginkgo has been used in traditional Chinese medicine for thousands of years. Today Ginkgo is one of the leading healthy remedy in China, France, Germany, and the United States though Ginkgo product is not FDA approved prescription medicine in US. There is increasing interest among local farmers to establish Ginkgo/crop intercropping plantations for leaf and nut production. Knowledge of possible interactions, competitive abilities, and growth strategies of Ginkgo and companion annual crop species in intercropping systems is fundamental to provide advice on optimum Ginkgo and crop combinations and silvicultural management measures for farmers. Large areas of Ginkgo agroforestry have been established in southern China in the past 20 years, but there have been few studies of Ginkgo/crop competition and mixtures. Comprehending the nature and extent of the interactions at the Ginkgo/annual crop interface in various circumstances will help us understand more about the way agroforestry systems function and will be essential for the development of agroforestry systems to manage Ginkgo agroforestry systems.
The objectives for this study were to: (1), estimate the effect of Ginkgo/crop mixtures on biomass of Ginkgo, wheat, and broad bean and their combined biomass; (2), assess the relative competitive abilities of different crop species in the Ginkgo/crop mixtures; (3), detect whether Ginkgo/crop mixtures are more productive than monocultures; and (4), search for the most productive Ginkgo/crop combinations. We chose a three-component mixture of wheat, broad bean and Ginkgo. These three species have widely different habits and canopy development patterns.
Materials and methods
Study site and plant materials
The research site was located in Nanjing, Jiangsu, China. Based on 20-year climate records, mean annual temperature of the region is 15°C, with the extreme of 38.8°C in summer and −12.5°C in winter. Mean daily temperature is 1.9°C in January and 27.7°C in July. Mean annual precipitation is 1,000 mm and the frost-free period is 229 days. The species chosen for this study were the deciduous broadleaf Ginkgo (Ginkgo biloba L.), broad bean (Vicia faba L.), and wheat (Triticum aestivum L. cv. “Feng Shou No. 2”).
There are two kinds of Ginkgo plantation management systems employed by farmers in Jiangsu. First, one- or two-year-old Ginkgo seedlings are planted with close spacing (e.g., 0.3 m × 0.3 m or 0.5 m × 0.5 m) to establish plantations for leaf production. When the plantation reaches five years old, Ginkgo seedlings are cut 20 cm above the ground level every 3 years. Second, three-year-old grafted seedlings are used to establish plantations for nut production with wide spacing (e. g., 5 m × 5 m or 6 m × 5 m). At early stage of the plantation, farmers can harvest leaves, but when Ginkgo starts to produce nuts (at age of approximately 10 years old), they will harvest nuts instead of harvesting leaves. Because of limited land area, farmers in this area will do their best to intercrop annual crops, vegetables before Ginkgo canopy closes.
The replacement series experiments of Ginkgo: wheat and Ginkgo: broad bean
Experiment 1 (replacement series 1)
Ginkgoes per pot
Wheat per pot
Experiment 2 (replacement series 2)
Ginkgoes per pot
Broad beans per pot
The experiment was installed on November 8, 1999 and lasted for 214 days (one growing season for both annual species wheat and bean) and the crop seedlings were harvested on June 10, 2000. Two-year-old Ginkgo seedlings and one-month-old crop (wheat and broad bean) seedlings were obtained from Taixing Forest Station, Jiangsu. Before planting, similar initial sizes of seedlings were sought to ensure that any differences found in competitive ability were not due to initial transplant size differences. In this study, mean Ginkgo height and root collar diameter were 37.6 cm and 1.10 cm, respectively. During the growing season, weeds were regularly hand removed to control weed competition in pots.
After harvesting, all crop plants were separated into root, seed, and shoot (stem and leaf) to determine the components of crop biomass, and Ginkgo seedlings were separated into leaf, stem, and fine and coarse root to determine the components of seedling biomass. Roots were carefully hand-washed in clean water in order to remove adhering soil particles and preserve broken fine roots. Each of the seedling components was oven-dried at 70°C for 48 h and then weighed to the nearest 0.01 g.
Soil and Ginkgo nutrient analysis
Soil samples were collected from each pot after harvesting, air-dried in the greenhouse and taken to the laboratory for N, P and K analysis. Hydrolyzable N, hereafter referred to as “available N”, in soil was determined by alkali hydrolysis micro-diffusion method in which the soil samples were hydrolyzed with 1.8 M sodium hydroxide (NaOH) at 40°C for 24 h, and titrated with 0.01 M hydrochloric acid (HCl). Available (Olsen) P in soil was determined colorimetrically by the method of Murphy and Riley (1962) after extracted using 0.5 M sodium bicarbonate (pH 8.5). Available (exchangeable) K was measured by the atomic absorption spectrometry (Varian-95) extracted with 1 M neutral ammonium acetate. Seedling samples were oven-dried at 70°C (Allen 1974) and then ground in a Wiley mill to pass a 1 mm sieve. Total N was determined by the Kjeldahl method, after block digestion using a sulfuric acid-hydrogen peroxide solution (Lowther 1980), phosphorus by the molybdenum blue method (Allen 1974) with HCl and NH4F as extractants, and K, Ca, and Mg by atomic absorption.
Data on plant biomass components was used to calculate relative yield total (RYT), relative land output (RLO), and relative yield per plant (RY), and total land output (TLO) (Reyes et al. 2009; Walck et al. 1999; Fowler 1982; Jolliffe 1997, 2000). One-way ANOVA was conducted on some indices of competition and mixture advantages to test the effects of plant combination (or Ginkgo: crop ratio) using the GLM procedure in the SYSTAT statistical package (Wilkinson 1996). The probability for statistical significance was 0.05. If significant differences were detected among the levels of density, then a multiple comparison procedure (Tukey test) was used.
Biomass of individual species and their combined biomass
Relative yield total (RYT)
Relative yield total (RYT) represents the sum of the proportional changes in yield that occurred in the mixtures and it measures the degree to which the two components of a two-species mixture make demands on the same resources. RYT values were larger than one for all treatments in this study except Ginkgo: broad bean ratios 1:5 and 2:4, indicating that better production from the mixture than from the monocultures (Fig. 2).
Relative yield per plant
Relative yield per Ginkgo (RYPG) and per crop plant (RYPC), relative yield total (RYT), relative land output (RLO), and total land output (TLO, g pot−1) for the Ginkgo/broad bean and Ginkgo/wheat mixtures
(Treatments) Ginkgo: crop ratio
Ginkgo/broad bean mixture
1G + 5B
2G + 4B
3G + 3B
4G + 2B
5G + 1B
1G + 5W
2G + 4W
3G + 3W
4G + 2W
5G + 1W
Relative land output
Relative land output (RLO) is the ratio of combined biomass of species components in mixture and pure stands. RLO is used to detect whether mixtures overyield monocultures. If mixtures are more productive than monocultures, RLO will exceed one. Table 2 shows that Ginkgo: broad bean ratio 5:1, and Ginkgo: wheat ratios 5:1 and 4:2 had RLO values of greater than one, indicating that these three species combinations in mixtures were, to some extent, more productive than monocultures. This may imply that interspecific competition for resource factors (light, water, and nutrients) is low.
Total land output
Total land output (TLO) is simply the sum of all component species yield per unit land area. Like RLO, TLO is also intended to quantify the productivity of mixed plant associations. Ginkgo: crop ratio had significant (p = 0.001) effect on TLO (equivalent to combined biomass per pot in this study). Changing Ginkgo: crop ratio from 1:5 to 5:1 resulted in a linear increase in TLO. Ginkgo: crop ratio 5:1 in both mixtures had the highest TLO (Table 2). There was no significant (p = 0.065) difference in TLO between the two mixtures, but TLO in the Ginkgo/broad bean mixture was slightly higher than those in the Ginkgo/wheat mixture.
Soil N, P, and K concentration
Ginkgo: crop ratio had significant (p < 0.01) effects on soil exchangeable N concentration, but no significant impact on soil available P and K. Soil exchangeable N decreased as the Ginkgo: crop ratio varied from 1:5 to 5:1 in the two mixtures (Fig. 4a). Soil N availability in the Ginkgo/broad bean was slightly higher than that in the Ginkgo/wheat mixtures. There were no significant differences in available P concentration between the two mixtures although the Ginkgo/broad bean had a slightly higher soil available P concentration than the Ginkgo/wheat mixture (Fig. 4b). Soil available P declined significantly (p < 0.05) with increasing Ginkgo density from 1 to 5 plants per pot in the absence of crop species; however, there was no significant Ginkgo: crop ratio treatment influence on soil available P in the mixtures. Soil in Ginkgo monoculture had higher levels of available K than soil in Ginkgo crop mixtures, but there was no significant difference between the two mixtures. Number of seedlings of Ginkgoes per pot and Ginkgo: crop ratio had no apparent influence in soil K levels (Fig. 4c). While no significant differences in soil available P and K concentration between the mixtures were found, soil available P and K concentrations in the Ginkgo/broad bean mixture were higher than those in the Ginkgo/wheat treatment.
Leaf biomass-based NUE and per-plant-biomass-based NUE
Ginkgo leaf biomass-based nitrogen use efficiency (LNUE) and per-plant biomass-based nitrogen use efficiency (PNUE) in the two Ginkgo/crop mixtures
LNUE (g g−1 N)
PNUE (g g−1 N)
Ginkgo/broad bean mixture
1G + 5B
2G + 4B
3G + 3B
4G + 2B
5G + 1B
1G + 5W
2G + 4W
3G + 3W
4G + 2W
5G + 1W
The difference in RYT between the two mixture systems is predominantly results from the difference in Ginkgo biomass production. Values of RYT of Ginkgo and broad bean mixtures were larger than 1.0 only when the ratio of broad bean was ≥50% in the mixture (Table 2). In comparison, values of RYT were larger than 1.0 in any of the proportions for Ginkgo and wheat mixture. This demonstrates that Ginkgo and wheat used different resources when grown together and higher yield was achieved when Ginkgo mixed with wheat (Table 2). The concave RY curves of Ginkgo indicate the effect of interspecific competition on plants of Ginkgo when grown with broad bean or wheat was greater than that of intraspecific competition (Vanclay 2006). On the other hand, the convex curves of broad bean and wheat indicate the effect of interspecific competition on plants of either of these two species when grown with Ginkgo was less than that of intraspecific competition (Fig. 1) (Harper 1979). Thus, the competitive hierarchy suggested by the de Wit diagrams in this study is: broad bean > wheat > Ginkgo.
RYT represents the sum of the proportional changes in yield that occurred in the mixtures (Fowler 1982) and measures the degree to which the two components of a mixture make demands on the same resources (Taylor and Aarssen 1990). The early use of RYT was to assess ecological niche differentiation, but it was later used to evaluate the relative productive performance of mixtures (Trenbath 1974). If RYT values are larger than one, then mixtures have an overyielding advantage compared to monocultures. Mixed cropping may be of benefit only when RYT is larger than one. In this study RYT for all ratios of Ginkgo and wheat mixtures had RYT large than 1. It indicates that Ginkgo and wheat mixture demonstrated the most advantage of intercropping and niche differentiation.
The main problems associated with the replacement series have been extensively debated (Jolliffe et al.1984; Law and Watkinson 1987; Snaydon 1991; Jolliffe 2000) primarily because the relative yields of the replacement series and competitive relationships are strongly dependent on the initial planting density, and because differences in initial size of the two species may bias results. The problems with this approach may be minimized if the density chosen is high enough or the duration of the experiment is long enough to reach the range of constant final yield. Ideally replacement series should be carried out at a number of densities, but due to severe space restrictions we were not able to compare different total densities. However, provided that the density used allows competition to occur through much of the life cycle of the two crop species, but not for Ginkgo, the replacement design is a quick and amendable method to analyze the quantitative effects of competition between two components (Taylor and Aarssen 1989).
In this study, a total density of six plants per container was used to try to overcome some difficulties in replacement series design. It was assumed that a density of six Ginkgoes per pot (equivalent to 96 Ginkgoes per m2) is high enough to reach the range of constant final yield in order to make some useful interpretation from the Ginkgo/crop replacement series. Also we used two year-old Ginkgo seedlings and one-month old crop seedlings in order to represent the operation in field for Ginkgo and crop system. The initial size differences of Ginkgo and crop seedlings might have affected the level of competition. In practice it is impossible to sow Ginkgo and crop seeds at the same time because crops are annual plants and Ginkgo will be harvested for leaf for many years to come once established. The life spent of the two crop species does not allow the longer period of the experiment.
There is considerable ecological and agronomic interest in whether a mixture yields more (overyield) than a monoculture (Trenbath 1974). It has long been supposed that mixtures of terrestrial plant species tend to be more productive than pure stands, because mixtures possess greater structural and functional diversity than monoculture, which enables mixtures to exploit environmental resources more efficiently. Some researchers have suggested that the higher yield in mixture stand was caused by niche differentiation in rooting depth (Berendse 1981). After analyzing data from 54 published experiments involving binary species associations, Jolliffe (1997) concluded that mixtures were significantly more productive than pure stands in 38 experiments, and significantly less productive in only eight experiments.
The mechanisms resulting in RYT larger than one were summarized by Walck et al. (1999) and Trenbath (1976). Weigelt et al. (2007) found that high RYT values, high relative yield of mixture (RYM) values, and overyielding all tend to occur when soil depth and/or root stratification give species the opportunity to root at different depths, suggesting that differences in rooting depth is a major facet of niche differentiation. Similarly, the overyielding observed in the two mixtures in this study maybe attributed in part to the fact that there was a significant ecological niche differentiation in the Ginkgo/wheat (or broad bean) mixture. There are several possible mechanisms to explain this. First, Ginkgo and crop species used in this study differ in their growth phenology. Ginkgo has a period of leafless dormancy between early December and early April. During this period, winter crops grow with less competition from Ginkgo especially for light. Ginkgo may not suffer too much from competition with annual winter crop species. Therefore, competitive relations in mixture may be manipulated by choice of variety of either species (Andrews 1974) and by the time of planting (Spitters 1979). Second, Ginkgo and annual crop species differ in their rooting depths. Wheat and broad bean are shallow-rooting annual crop species, whereas Ginkgo has a well-developed taproot system. Mixtures of Ginkgo and the crop species can effectively exploit resources from different soil layers. Third, broad beans can fix nitrogen to supple its own N and, hence, improve soil fertility. From this analysis, it seems that although it is difficult to determine whether competition for resources is the sole cause of interference, RYT can be used, to some extent, to evaluate niche differentiation existing in mixtures.
To contribute to a better understanding of the mechanisms involved in tree/crop interactions, Imo and Timmer (1998) developed a new diagnostic method (vector competition analysis) for screening alternative strategies for vegetation control in integrated crop management. In this technique, tree and weed interactions are evaluated in a bivariate model depicting vectors of changing biomass and nutrient uptake relative to competition-free status. Mead and Mansur (1993) used a similar method to examine the competition for nutrients and moisture between Pinus radiata trees and pasture in an agroforestry system in New Zealand. Imo and Timmer (2000) employed this approach to analyze the competitive ability of each plant component in a Leucaena-Maize alley cropping system in western Kenya. The advantage of this graphical diagnostic approach is that vector patterns simplify the interaction of treatment responses, facilitate multiple site and treatment comparisons, and identify possible mechanisms associated with observed crop responses. In this study, we used this approach to examine competition responses of Ginkgo and crop species to combined Ginkgo: crop ratios in two mixtures. The results showed that all Ginkgo: crop (wheat and broad bean) ratios exhibited a compensatory interaction type, which indicates some degree of niche differentiation. Thus, Ginkgo and crop intercropping management is beneficial. This study indicated that vector competition analysis is a useful and simple way to interpret treatment response of plant components in Ginkgo/crop intercropping system.
In this study, relative yield total (RYT), relative land output (RLO) and total land output (TLO) were used to evaluate Ginkgo/wheat (or broad bean) mixture productivity. For these three indices value larger than 1 indicates that production from mixture is higher than that from monoculture. As reported by many studies (e.g., Mead and Willey 1980; Willey 1985), RYT has been extensively employed to evaluate whether overyielding is achieved when species are mixed, compared to growing them in monoculture systems. Based on Jolliffe and Wanjau (1999), RYT compares mixtures and monocultures that have different land area and numbers of plants. Therefore, comparisons between the mixtures and monocultures based on RYT are inappropriate since RYT involves different inputs of land and plant material.
Total land output (TLO) is simply the sum of all component species yields per unit land area in mixtures. Compared to RYT, RLO was not developed to interpret interference, but simply to quantitatively assess productivity of mixtures (Jolliffe 2000; Weigelt and Jolliffe 2003). Jolliffe and Wanjau (1999) concluded that RLO is a generalization of an index developed by Wilson (1988) to cope with the problem. RLO is the ratio of combined yields in the mixture to those in monoculture (Jolliffe 1997, 2000). For RLO, equivalent inputs of land and number of plants are included in the mixtures and monocultures. Despite this argument, RLO is positively correlated with RYT; and values of RYT and RLO usually differ by a small percentage. Contrary to the two relative indices, RYT and RLO, total land output (TLO) simply measures total absolute production by a mixture, independent of densities and species combinations (Perry et al. 2009; Jolliffe and Wanjau 1999). Weigelt and Jolliffe (2003) further pointed out that TLO is not a direct measure to evaluate competition processes although it may reflect the outcome of competition somewhat. Results from this study are in accordance with these findings. For example, highest RYT and RLO matched highest TLO (i.e., total biomass per pot) at Ginkgo: broad bean ratio 5:1 in the Ginkgo/broad bean mixture; however the Ginkgo/wheat mixture with highest RYT and RLO at Ginkgo: wheat ratio 4:2 did not have highest TLO (Table 2). Plants compete with each other for the available growth factors like light, water, nitrogen and minerals; and biomass production is approximately linearly related to the uptake of the resource that limits growth. Biomass production reflects the production of dry matter over long periods of time and integrates the effects of varying environmental factors (Grime 1979; Wilson and Keddy 1986). Based on this knowledge, it seems that the Ginkgo/crop mixture production can be better measured by TLO that is equivalent to total combined biomass per pot measured in this study.
The soil nutrient results after one growing season indicate that Ginkgo pure stand did not use lots of nutrients as the two crop species (Fig. 3), which will make this intercropping system possible. It is necessary to point out that the soil under pure Ginkgo plantation had higher nutrient concentration. It indicates that Ginkgo does not need lots of nutrient as much as the two crop species.
Vector analysis showed that there were compensatory interactions between Ginkgo and crop species in the mixtures. RYT value larger than one indicated that Ginkgo/crop mixtures were more productive than monocultures.
Results from analysis of relative yield (RY) of each plant species and relative yield per Ginkgo (RYPG) and per crop plant (RYPC) indicated that the intra- and inter-specific components of interference were not in balance, and that wheat and broad bean were more competitive than Ginkgo.
Leaf biomass-based nitrogen use efficiency (LNUE) and per-Ginkgo-biomass-based nitrogen use efficiency (PNUE) increased as Ginkgo: crop ratio changed from 1:5 to 5:1. Ginkgoes in the Ginkgo/broad bean mixture had lower LNUE and PNUE than those in the Ginkgo/wheat mixture. High LNUE and PNUE matched high total land output (RLO).
Relative yield total (RYT) and relative land output (RLO) values for the Ginkgo/wheat mixture were significantly higher than those for the Ginkgo/broad bean mixture, suggesting that Ginkgo/wheat intercropping systems were more productive than monoculture compared to the Ginkgo/broad bean mixture. Ginkgo: wheat ratio 5:1 and Ginkgo: broad bean ratio 5:1 had RYT and RLO values of more than one and the largest TLO values in respective mixtures. Therefore, these two ratios might be considered optimum Ginkgo: crop ratio for enhancing the combined biomass of the Ginkgo and crop in respective mixtures.
Compared with Ginkgo monoculture, mixing resulted in a significant reduction in soil N, P and K concentrations, in the two mixtures. Soil N concentration decreased more significantly than soil P and K concentration as Ginkgo: crop ratio changed from 1:5 to 5:1. These data suggest that Ginkgo and crop in the mixtures mainly compete for N resource.
The authors would like to thank the anonymous reviewers for their critical review and thoughtful comments.