- Email: [email protected]
Effect of Canopy Structure on Degree of Asymmetry of Competition in Two Forest Stands in Northern Japan
Annals of Botany 77 : 565–571, 1996
Effect of Canopy Structure on Degree of Asymmetry of Competition in Two Forest Stands in Northern Japan K I H A C H I R O K I K U Z A WA* and K I Y O S H I U M E K I Hokkaido Forestry Research Institute, Koshunai, Bibai, Hokkaido 079-01, Japan Received : 26 June 1995
Accepted : 6 November 1995
The canopy structure in terms of the vertical distribution of leaf mass and the degree of asymmetry of competition between individual trees was studied in two types of forest stand in Hokkaido, northern Japan : a naturally regenerated stand of Betula spp. and an artificial plantation of Picea abies. The canopy structure in the Betula stand was more hierarchical ; larger individuals were not heavily shaded even in the lowest part of their crowns and smaller individuals were heavily shaded by their larger neighbours. The canopy structure in the Picea stand was less hierarchical ; even larger individuals were shaded in the lowest part of their crowns and smaller individuals were not heavily shaded by their neighbours. Application of the general formula of sizedependent mean growth rate revealed that competition in the Betula stand was more one-sided than that in the Picea stand. This result was consistent with the trends in the change over time in size equality in both stands. Even if competition is mediated by light, which often makes competition one-sided, the degree of one-sidedness in competition can be variable depending on canopy structure. # 1996 Annals of Botany Company Key words : Betula ermanii, canopy structure, crown shape, Picea abies, one-sided competition, two-sided competition.
INTRODUCTION The degree of asymmetry, or one-sidedness, in competition affects many aspects of population dynamics and}or microevolution in plants (Weiner, 1990). For example, one-sided competition in a population leads to strong hierarchy in its size distribution, makes the size structure stable against perturbations, and inhibits coexistence of species (Hara, 1992). Within a species, asymmetry accelerates the selection of a single genotype (Weiner, 1990). The degree of one-sidedness in competition is believed to be determined, to a large extent, by the nature of the resource for which plants compete with their neighbours (Weiner, 1985, 1986 ; Hara, 1986 ; Weiner and Thomas, 1986 ; Weiner, 1990). Competition for light is often onesided, because light comes directionally from above so that taller plants can easily shade shorter ones but not ice ersa. On the other hand, competition for underground resources, such as water or nutrients, is often two-sided. In the real world, competition cannot be absolutely onesided but often asymmetrically two-sided even if it is only for light (Yokozawa and Hara, 1992), because larger plants can be shaded by shorter neighbours in lower parts of their crowns. Thus we can expect that the crown shape of individuals and the canopy structure of the populations are important in determining the degree of asymmetry of competition. This expectation has been tested in herbaceous populations (Ellison and Rabinowitz, 1989 ; Geber, 1989). However, the relationship between plant morphology and * Present address : Center for Ecological Research, Kyoto University, Kyoto 606-01, Japan.
0305-7364}96}06056507 $18.00}0
the degree of asymmetry of competition remains unclear (Geber, 1989). Moreover, it has not been tested in populations of trees which have a short exponential growth phase, so that competition may be more influenced by plant form (Geber, 1989). In this paper, we estimate the canopy structure and the degree of asymmetry of competition in two forest stands between which crown shape is different—a naturally regenerated stand of Betula ermanii Cham. and an artificial plantation of Picea abies Karsten—and discuss a possible effect of the canopy structure on the degree of asymmetry of competition.
MATERIALS AND METHODS Data collection The Betula stand is located on a gentle mountain slope at about 400 m altitude in the Tobetsu area of Hokkaido State Forest, western Hokkaido, northern Japan. Annual precipitation and mean annual temperature in this area are 1500 mm and 3±5 °C, respectively. The soil type is a brown forest soil II (Land Bureau, National Land Agency, 1975). In 1971, herbicide was applied to eliminate dense Sasa kurilensis Makino et Shibata vegetation—dwarf bamboo which inhibits the regeneration of tree seedlings. Dead Sasa culms and rhizomes were removed by a bulldozer, together with litter and humus. After this scarification, the site was naturally invaded by hardwood seedlings. Betula ermanii was the dominant species in the regeneration. A few individuals of Betula maximowicziana Regel and Phellodendron amurense Ruprecht were also found. In 1981, two # 1996 Annals of Botany Company
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry
plots (each 0±01 ha in area) were established in the regenerated stand. In one of the plots, extremely heavy thinning was carried out in 1984 ; the stem density of this plot was 1900 stems ha−". The other plot was left intact ; the stem density was 26 800 stems ha−". All individuals in the two plots were tagged, and their diameter at breast height (dbh), tree height, and the height of the base of the lowest living branch were measured every 2 years from 1984 to 1992. For this stand, data in 1984 and 1992 were used as the initial and final measurement data, respectively. The Picea abies stand was planted in 1961 on a gentle mountain slope at about 220 m altitude in the experimental forest of Hokkaido Forestry Research Institute, central Hokkaido, northern Japan. Annual precipitation and mean annual temperature in this area are 1300 mm and 7±2 °C, respectively. The soil type is a brown forest soil II (Land Bureau, National Land Agency, 1977). In 1982, six plots (each 0±099 ha in area) were established in this stand, and thinned to varying densities : 495, 1060, 1505, 2000, 2384, and 2687 stems ha−". The same measurements as those for the Betula stand were carried out. For the Picea stand, data taken in 1982 and 1986 were used as the initial and final measurement data, respectively.
Estimation of canopy structure The vertical distribution of leaf mass for an individual tree was estimated by two methods : a simple one and a realistic one. In the simple method, the leaf mass of an individual was assumed to be distributed homogeneously between the crown top and the base of the lowest living branch. In the realistic method, it was assumed that the cumulative leaf mass from the top of the crown and crown depth have an allometric relationship with a variable allometric ratio (Hashimoto, 1990, 1991, 1992). This model is given by : Fz ¯ a zb ecz, where Fz denotes the cumulative leaf mass from the crown apex to the point z distant from the crown apex ; a, b and c are parameters. The derivative of this function [F z! ¯ a zb−" ecz(bcz)] represents the density distribution of leaf mass. The parameter b determines the shape of the allometric curve. The value of ®b}c gives the distance from the crown apex at which Fz (cumulative leaf mass) is maximum. At this point (z ¯®b}c), F z! is zero indicating that this point is crown base ; ®b}c gives crown length (the distance from the crown apex to the crown base). To determine species–specific values for parameters b and c, the log-transformed model was applied to data of stratified clipping of individual trees of Betula ( Y. Takahashi, unpubl. res.) and Picea (T. Asai, unpubl. res.). Because of the scarcity of data for Betula, data of other deciduous hardwood species (Komiyama, 1992) were also analysed. The restriction that ®b}c equals observed crown length was placed on the regressions to make the estimated density of leaf mass take zero value at the observed crown base. With this restriction, only parameter b should be determined and is relevant to describe the species-specific
0
0 A
B 1
Distance from crown apex (z: m)
566
2 2 4 3 4
6
5 8 6 0.0
1.0
2.0 3.0 4.0 0.0 0.4 0.8 Distribution density (F': kg m–1)
1.2
1.6
F. 1. Vertical distribution density of foliage of an individual tree of Betula platyphylla (A) and Picea glehnii (B). The fitted line is F !z ¯ a zb−" ecz (bcz) ; a ¯ 3±54, b ¯ 1±45, c ¯®0±16 for Betula, and a ¯ 0±14, b ¯ 4±02, c ¯®0±61 for Picea.
shape of the vertical leaf mass distribution, because parameter a is the scaling factor that is mainly dependent on the size of individual trees. Figure 1 shows the examples of leaf mass distributions and fitted density distribution functions (F z!) for individual trees of Betula and Picea. The leaf mass in the Betula tree was more concentrated in the upper part of the crown than that in the Picea tree. Results for other hardwoods [not shown ; data in Komiyama (1992)] and those of Hashimoto (1990) for Cryptomeria japonica D. Don also indicated this tendency ; estimated parameter b ranged from 2 to 3 for hardwoods and from 3 to 4 for conifers with a few exceptions. We therefore set a fixed b value for each species (b ¯ 2±5 for Betula and b ¯ 3±5 for Picea) for simplicity, though vertical leaf mass distribution varies with age and}or the degree of suppression by larger competitors (Hashimoto, 1990, 1991). In both methods, it was assumed that the total leaf mass of an individual tree (wL, kg) had allometric relationship with dbh (d, cm) and tree height (h, m) : wL ¯ 0±00036 d #h for Betula and wL ¯ 0±0029 d #±)&'( for Picea. These allometric relationships were those for Betula platyphylla var. japonica Hara (Takahashi, Asai and Kikuzawa, 1974) and for Picea abies (Yoshimura, 1967). The vertical leaf mass distribution for a stand was obtained by summing those for constituting individual trees using data in the initial measurement.
Degree of asymmetry of competition The change in size inequality over time from the initial to the final measurement and the relationship between the size of an individual and its growth were analysed to estimate the degree of asymmetry of competition. When mortality is negligible, size inequality is believed to increase over time
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry
¯ 0±2004 (d #h)!±)!'$, for hardwood trees (Kikuzawa, 1988) and ¯ 0±0000892 (100d )!±)!'$ 0±9935("!!d), for Picea abies (Nakajima, 1943). The relationship between the size of individual and its growth was analysed by fitting a modified general formula of size-dependent mean growth rate proposed by Yokozawa and Hara (1992). Yokozawa and Hara’s formula is given by : G(t, w) ¯ w[a ®a wm®c C(t, w)®c C(t, w )], " # ! ! " where G(t, w) denotes the mean growth rate of individuals of plant mass w at time t ; C(t, w) is proportional to cumulative basal area of individuals larger than w at time t (defined later) ; w is the minimum individual mass in a population ; ! a , a , m, c and c are constants. C(t, w) is given by : ! " " # C(t, w) ¯
&
wmax
x#f(t, x) dx,
w
where f(t, x) is the distribution density function of dbh x at time t, and wmax is the plant mass of individual of the maximum dbh. C(t, w) represents the effect of one-sided competition on the mean growth rate, and C(t, w ) that of ! two-sided competition. For simplicity, the term wm was replaced by lnw. Thus, a Gompertz growth curve for a single isolated tree was assumed. The model used was the following : G(t, ) ¯ [a ®a ln ®c C(t, )®c C(t, )] (1) ! " " # ! where denotes stem volume. Annual growth rate in terms of stem volume was calculated by dividing the volume increment by the duration of growth (8 years for the Betula stand, and 4 years for the Picea stand). Individuals were divided into ten size classes of the same ranges in each plot, and mean growth rate, G(t, ), was calculated in each size class. The model was fitted by multiple regression using the midpoint of size classes as variable , and the degree of asymmetry of competition was estimated by the significance of the regression coefficients c " and c . In the Betula stand, data of individuals of Betula spp. # were used as the criterion variable ; data of other species were involved only in the explanatory variables [C(t, ) and C(t, )]. ! RESULTS Density and change in mean size Mortality was very low in all plots except for the dense plot in the Betula stand in which 60 % of the individuals died
0.4
0.1 Mean volume (m3)
under one-sided competition, but not to do so under twosided competition (Ellison and Rabinowitz, 1989). The size inequality was expressed by the coefficient of variation of individual size. In this paper, stem volume (, m$) was used to express the size of individuals. Stem volume was estimated from dbh (d, m) and tree height (h, m) using allometric functions :
567
0.01
0.001 100
1000 Density (ha–1)
10 000 20 000
F. 2. Changes over time in stem density and mean volume of the studied plots. (_) Betula plots in 1984, (U) Betula plots in 1992, (+) Picea plots in 1982, (E) Picea plots in 1986.
during the measurement period of 8 years (Fig. 2). Mortality in this plot was strongly size-dependent. If we divide the population into two halves based on size, 85 % of the smaller individuals died, but only 37 % of the larger individuals. Because the mortality was concentrated in smaller individuals and there was no evidence for sizeindependent mortality (e.g. attack by insect or pathogen), the mortality was attributable mostly to competition. The increase of mean stem volume in a plot was dependent on stem density in the plot in both stands (Fig. 3). Canopy structure Figure 4 shows the estimated vertical distribution of leaf mass, and the distribution of tree height, and the height of the base of the lowest living branch in two representative plots in the Betula and Picea stand at the initial measurement. Only trees that survived until the final measurement were used in the distributions of tree height and the height of the lowest living branch base. The canopy structures in other plots in each stand were similar to those shown in Fig. 4. For the Betula stand, the vertical distributions of leaf mass estimated by the two different methods were very similar to each other. On the other hand, for the Picea stand, the leaf mass distribution estimated with the assumption that leaf mass is distributed homogeneously between tree height and the base of the lowest living branch (the simple method) had a flatter peak than its counterpart (the realistic method). However, the relative position of the leaf mass distribution to the distribution of tree height and the height of the base of the lowest living branch did not depend on the estimating method. Therefore, only the leaf mass distribution estimated by the realistic method is discussed hereafter.
568
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry 0.025 A 0.02
Degree of asymmetry of competition
0.015
Change in mean volume (m3)
0.01
0.005
0
0
4000
8000
12 000
16 000
20 000
0.25 B 0.2
0.15
0.1
0.05
0
than short trees in the Picea stand so that even tall individuals were shaded by short individuals to some extent.
0
500
1000 1500 2000 Density (ha–1)
2500
3000
F. 3. Relationship between the change in mean volume and the initial stem density. A, Betula ; B, Picea.
The vertical distributions of leaf mass estimated by the realistic method were similar to a normal distribution for both Betula and Picea stand, which agreed with other researchers’ results (Whitehead, 1978 ; Beadle, Talbot and Jarvis, 1982 ; Hashimoto, 1990, 1991, 1992). The relative height of the peak of leaf mass distribution to the canopy height for the Betula stand was higher than that for the Picea stand. The relative range of tree height distribution to canopy height was also different between these stands. There were a considerable number of individuals lower than the peak of leaf mass distribution in the Betula stand, whereas there were few such individuals in the Picea stand. The distribution of height of the base of the lowest living branch of the Betula stand was also wider than that of the Picea stand. The distributions of tree height and the base of the lowest living branch crossed in the Betula stand, but did not in the Picea stand. This means that the lowest portion of crowns of some tall individuals was higher than tree height of short trees in the Betula stand so that tall individuals were not shaded by short individuals, and that the lowest portion of crowns of tall individuals was lower
Although the overall positive relationship between density and coefficient of variation is apparent in both stands (Fig. 5), this tendency was artificially made due to the thinning practice in which smaller trees were cut more heavily than larger ones in low-density plots, and does not necessarily mean that the competition is one-sided by itself. In the Betula stand, coefficients of variation did not change in the dense plot through time, and increased little in the sparse plot (Fig. 5 A). In the Picea stand, coefficients of variation decreased or remained at the same level through time (Fig. 5 B). Figure 6 shows the annual mean growth rate of size classes in representative plots that had contrasting densities. In the Betula stand, mean growth rate of the largest size class in the dense plot did not differ very much from that in the sparse plot, indicating that the growth of the largest trees was hardly reduced by competition. On the other hand, mean growth rates of the largest size class in the dense plots in the Picea stand were smaller than that in the sparse plot, indicating that the growth of the largest trees was reduced to some extent by competition. Table 1 shows the estimated parameters of the general formula of size-dependent mean growth rate for both stands. For the Betula stand, the term lnw in eqn (1) was neglected, because the relationship between growth rate and size in the sparse stand, in which competition was very weak, did not show curvilinearity, and because the sample size was not very large to determine the significance of many parameters. For the Betula stand, parameter c for one" sided competition was significant at the 1 % level of probability, and c for two-sided competition at the 5 % # level. For the Picea stand, parameter c did not differ " significantly from zero, whereas c was highly significant # (P ! 0±1 %). DISCUSSION The mean growth rate in a plot decreased with density (Fig. 3), and considerable mortality was found in the dense plot in the Betula stand (Fig. 2). The existence of competition in both stands is obvious. The results of applying the general size-dependent growth model [eqn (1)] indicated that the competition in the Betula stand was more one-sided than that in the Picea stand. In general, size inequality increases over time in populations in which one-sided competition prevails (Ellison and Rabinowitz, 1989). The coefficient of variation decreased or remained at the same level over time in the Picea stand, indicating that the competition in this stand was not onesided. The coefficient of variation increased in the sparse plot in the Betula stand. However, it remained unchanged in the dense plot. It is probably because the size-dependent mortality in this plot, which was concentrated in smaller individuals, countered the increase of size inequality by onesided competition. Therefore the tendency in the changes in coefficient of variation was consistent with the results of the
569
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry –1 Number of individuals (ha )
12
0
600
1200
1800
14
0
100
200
300
400
500
600
700
800
12
10
10 Height (m)
8 8 6 6 4 4 2
2
0
2
4
6
8
10
12
14
0 16 18 –1 Leaf mass (kg ha )
50
100
150
200
250
300
F. 4. Vertical distribution of tree heights (± ± ± ±), bases of the lowest living branches (– – – –), and estimated leaf density by the realistic method (––––) and by the simple method (— — —) in two representative plots in the Betula (A) and Picea stands (B).
130
70
Coefficient of variation (%)
A
B
110
60
90
50
70
40
50
0
4000
8000
12 000
16 000
30 20 000 0 Density (ha–1)
500
1000
1500
2000
2500
3000
F. 5. Change in coefficients of variation over time from the initial measurement (+) to the final (E) related with stem density in the Betula (A) and the Picea stands (B).
growth rate analysis ; the competition in the Betula stand was more one-sided than that in the Picea stand. Because the degree of asymmetry of competition plays important roles in population dynamics (Weiner, 1990), the mechanism by which competition becomes one-sided or two-sided should be clarified. Many researchers have pointed out that if plants compete for light, competition is one-sided, whereas if plants compete for underground resources, competition is two-sided (Weiner, 1985, 1986,
1990 ; Hara, 1986 ; Weiner and Thomas, 1986). For clonal plants, the degree of physiological integration in a genet also affects the degree of asymmetry of competition between ramets (Suzuki, 1994 a, b). Another factor that affects the degree of asymmetry of competition is plant form and}or canopy structure (Turner and Rabinowitz, 1983 ; Yokozawa and Hara, 1992). If larger plants are shaded by smaller neighbours in the lower part of their crown, the competition cannot be absolutely one-sided but asymmetrically two-
570
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry 0.009
0.09
G (m3 year–1)
A
B
0.008
0.08
0.007
0.07
0.006
0.06
0.005
0.05
0.004
0.04
0.003
0.03
0.002
0.02
0.001
0.01
0.01
0.005
0
0.015
0.02
0.025
0.03 0 Volume (m3)
0.05
0.1
0.15
0.2
0.25
0.3
0.35
F. 6. Relationship between mean growth rate in size classes and initial size in dense (E) and sparse plots (+) in the Betula (A) and Picea stands (B).
T 1. Coefficients of determination (r#) and estimated parameters of the general size-dependent mean growth formula for the Betula and Picea stands Species
r#
a
!
a
"
c
"
c
Betula Picea
0±98*** 0±95***
0±34*** 0±32***
— ®0±0022
0±0230** ®0±0008
0±0042* 0±0055***
#
***, **, and * show that the coefficient of determination or the estimated parameter is significant at 0±1, 1, and 5 % level, respectively.
sided (Yokzawa and Hara, 1992). In this situation, the degree to which smaller plants shade the lower part of the crown of larger plants, which determines the degree of onesidedness in competition, depends on crown shape and}or canopy structure (Weiner and Thomas, 1986 ; Yokozawa and Hara, 1992). The results presented in this paper suggest that the difference in the degree of asymmetry of competition between the Betula stand and the Picea stand resulted from the difference in canopy structure and}or plant form in these stands. The more hierarchical canopy structure and relatively shallow crown shape in the Betula stand made the competition relatively one-sided. On the other hand, the canopy structure and crown shape in the Picea stand, in which even larger individuals can be partially shaded by smaller neighbours, made the competition more two-sided. Although no data are available for underground resources in the analysed stands, there are several lines of circumstantial evidence that this sort of resource was not relevant in determining the degree of asymmetry of competition for these stands. First, in forests in Hokkaido, it is believed that water is not a very strong limiting factor. In the two stands analysed, annual precipitation is greater than 1300 mm, which is enough for normal growth of trees,
and there is not a considerable difference in precipitation between the two stands. Moreover, these stands are similar in soil type and topography. Second, in the Betula stand, Sasa (dwarf bamboo) vegetation, litter, and humus were removed to enhance the regeneration of tree seedlings. This silvicultural practice lessened the total amount of nutrient, and might affect the degree of asymmetry of competition to make competition more two-sided. However, the growth rate analysis revealed that competition in the Betula stand was more one-sided than that in the Picea stand. Thus there is little possibility that competition for nutrients is stronger in the Picea stand than in the Betula stand, and that this made the competition in the Picea stand more two-sided than that in the Betula stand. In this study, the pronounced factor that made the canopy structure of the two stands different from each other was species-specific crown shape : wide and shallow crown of Betula s. narrow and deep crown of Picea. However, canopy structure can be affected by factors other than characteristics of species. For example, the vertical distribution of leaves in a plant population or a community is affected by age structure of the stand (Hashimoto, 1991), successional status (Aber, 1979), site quality (Aber, Pastor and Melillo, 1982), density of individual trees (Whitehead, 1978 ; Weiner, Berntson and Thomas, 1990), and species composition (Harcombe and Marks, 1977). The degree of asymmetry of competition can change with these factors. The effect of these factors on the degree of asymmetry of competition should be clarified.
A C K N O W L E D G E M E N TS The authors are grateful to Y. Takahashi and T. Asai for their unpublished data and R. Milne and an anonymous referee for reviewing the manuscript.
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry LITERATURE CITED Aber JD. 1979. Foliage-height profiles and succession in northern hardwood forests. Ecology 60 : 18–23. Aber JD, Pastor J, Melillo JM. 1982. Changes in forest canopy structure along a site quality gradient in southern Wisconsin. The American Midland Naturalist 108 : 256–265. Beadle CL, Talbot H, Jarvis PG. 1982. Canopy structure and leaf area index in a mature Scots Pine forest. Forestry 55 : 105–123. Ellison AM, Rabinowitz D. 1989. Effects of plant morphology and emergence time on size hierarchy formation in experimental populations of two varieties of cultivated peas (Pisum satium). American Journal of Botany 76 : 427–436. Geber MA. 1989. Interplay of morphology and development on size inequality : a Polygonum greenhouse study. Ecological Monographs 59 : 267–288. Hara T. 1986. Effects of density and extinction coefficient on size variability in plant population. Annals of Botany 57 : 885–892. Hara T. 1992. Effects of the mode of competition on stationary size distribution in plant populations. Annals of Botany 69 : 509–513. Harcombe PA, Marks PL. 1977. Understory structure of a mesic forest in southeast Texas. Ecology 58 : 1144–1151. Hashimoto R. 1990. Analysis of the morphology and structure of crowns in a young sugi (Cryptomeria japonica) stand. Tree Physiology 6 : 119–134. Hashimoto R. 1991. Canopy development in young sugi (Cryptomeria japonica) stands in relation to changes with age in crown morphology and structure. Tree Physiology 8 : 129–143. Hashimoto R. 1992. Distribution structure of foliage and the light interception in a young Cryptomeria japonica stand. Journal of the Japanese Forestry Society 74 : 466–474. Kikuzawa K. 1988. Intraspecific competition in a natural stand of Betula ermanii. Annals of Botany 61 : 727–734. Komiyama A. 1992. A study on the methods for resource estimation of deciduous hardwoods. Gifu : Gifu University. (In Japanese). Land Bureau, National Land Agency. 1975. Land classification map (Hokkaido I : Ishikari district). Tokyo : Land Bureau, National Land Agency.
571
Land Bureau, National Land Agency. 1977. Land classification map (Hokkaido III : Sorachi district). Tokyo : Land Bureau, National Land Agency. Nakajima K. 1943. Tables of stem olume of trees in Hokkaido. Sapporo : Hokkaido branch of Korin-Kai. (In Japanese). Suzuki J. 1994 a. Growth dynamics of shoot height and foliage structure of a rhizomatous perennial herb, Polygonum cuspidatum. Annals of Botany 73 : 629–638. Suzuki J. 1994 b. Shoot growth dynamics and the mode of competition of two rhizomatous Polygonum species in the alpine meadow of Mt. Fuji. Folia Geobotanica et Phytotaxonomica, Praha 29 : 203–216. Takahashi Y, Asai T, Kikuzawa K. 1974. On biomass estimation of Betula platyphylla var. japonica forest stand in Nayoro. Bulletin of the Hokkaido Forest Experiment Station 12 : 29–37. Turner MD, Rabinowits D. 1983. Factors affecting frequency distributions of plant mass : the absence of dominance and suppression in competing monocultures of Festuca paradoxa. Ecology 64 : 469–475. Yokozawa M, Hara T. 1992. A canopy photosynthesis model for the dynamics of size structure and self-thinning in plant populations. Annals of Botany 70 : 305–316. Yoshimura K. 1967. Growth and biomass of Norway spruce forest in Ashu experimental forest. Bulletin of the Kyoto Uniersity Forests 39 : 27–34. Weiner J. 1985. Size hierarchies in experimental populations of annual plants. Ecology 66 : 743–752. Weiner J. 1986. How competition for light and nutrients affects size variability in Ipomoea tricolor populations. Ecology 67 : 1425–1427. Weiner J. 1990. Asymmetric competition in plant populations. Trends in Ecology and Eolution 5 : 360–364. Weiner J, Bernston GM, Thomas SC. 1990. Competition and growth form in a woodland annual. Journal of Ecology 78 : 459–469. Weiner J, Thomas SC. 1986. Size variability and competition in plant monocultures. Oikos 47 : 211–222. Whitehead D. 1978. The estimation of foliage area from sapwood basal area in Scots Pine. Forestry 51 : 137–149.
Effect of Canopy Structure on Degree of Asymmetry of Competition in Two Forest Stands in Northern Japan K I H A C H I R O K I K U Z A WA* and K I Y O S H I U M E K I Hokkaido Forestry Research Institute, Koshunai, Bibai, Hokkaido 079-01, Japan Received : 26 June 1995
Accepted : 6 November 1995
The canopy structure in terms of the vertical distribution of leaf mass and the degree of asymmetry of competition between individual trees was studied in two types of forest stand in Hokkaido, northern Japan : a naturally regenerated stand of Betula spp. and an artificial plantation of Picea abies. The canopy structure in the Betula stand was more hierarchical ; larger individuals were not heavily shaded even in the lowest part of their crowns and smaller individuals were heavily shaded by their larger neighbours. The canopy structure in the Picea stand was less hierarchical ; even larger individuals were shaded in the lowest part of their crowns and smaller individuals were not heavily shaded by their neighbours. Application of the general formula of sizedependent mean growth rate revealed that competition in the Betula stand was more one-sided than that in the Picea stand. This result was consistent with the trends in the change over time in size equality in both stands. Even if competition is mediated by light, which often makes competition one-sided, the degree of one-sidedness in competition can be variable depending on canopy structure. # 1996 Annals of Botany Company Key words : Betula ermanii, canopy structure, crown shape, Picea abies, one-sided competition, two-sided competition.
INTRODUCTION The degree of asymmetry, or one-sidedness, in competition affects many aspects of population dynamics and}or microevolution in plants (Weiner, 1990). For example, one-sided competition in a population leads to strong hierarchy in its size distribution, makes the size structure stable against perturbations, and inhibits coexistence of species (Hara, 1992). Within a species, asymmetry accelerates the selection of a single genotype (Weiner, 1990). The degree of one-sidedness in competition is believed to be determined, to a large extent, by the nature of the resource for which plants compete with their neighbours (Weiner, 1985, 1986 ; Hara, 1986 ; Weiner and Thomas, 1986 ; Weiner, 1990). Competition for light is often onesided, because light comes directionally from above so that taller plants can easily shade shorter ones but not ice ersa. On the other hand, competition for underground resources, such as water or nutrients, is often two-sided. In the real world, competition cannot be absolutely onesided but often asymmetrically two-sided even if it is only for light (Yokozawa and Hara, 1992), because larger plants can be shaded by shorter neighbours in lower parts of their crowns. Thus we can expect that the crown shape of individuals and the canopy structure of the populations are important in determining the degree of asymmetry of competition. This expectation has been tested in herbaceous populations (Ellison and Rabinowitz, 1989 ; Geber, 1989). However, the relationship between plant morphology and * Present address : Center for Ecological Research, Kyoto University, Kyoto 606-01, Japan.
0305-7364}96}06056507 $18.00}0
the degree of asymmetry of competition remains unclear (Geber, 1989). Moreover, it has not been tested in populations of trees which have a short exponential growth phase, so that competition may be more influenced by plant form (Geber, 1989). In this paper, we estimate the canopy structure and the degree of asymmetry of competition in two forest stands between which crown shape is different—a naturally regenerated stand of Betula ermanii Cham. and an artificial plantation of Picea abies Karsten—and discuss a possible effect of the canopy structure on the degree of asymmetry of competition.
MATERIALS AND METHODS Data collection The Betula stand is located on a gentle mountain slope at about 400 m altitude in the Tobetsu area of Hokkaido State Forest, western Hokkaido, northern Japan. Annual precipitation and mean annual temperature in this area are 1500 mm and 3±5 °C, respectively. The soil type is a brown forest soil II (Land Bureau, National Land Agency, 1975). In 1971, herbicide was applied to eliminate dense Sasa kurilensis Makino et Shibata vegetation—dwarf bamboo which inhibits the regeneration of tree seedlings. Dead Sasa culms and rhizomes were removed by a bulldozer, together with litter and humus. After this scarification, the site was naturally invaded by hardwood seedlings. Betula ermanii was the dominant species in the regeneration. A few individuals of Betula maximowicziana Regel and Phellodendron amurense Ruprecht were also found. In 1981, two # 1996 Annals of Botany Company
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry
plots (each 0±01 ha in area) were established in the regenerated stand. In one of the plots, extremely heavy thinning was carried out in 1984 ; the stem density of this plot was 1900 stems ha−". The other plot was left intact ; the stem density was 26 800 stems ha−". All individuals in the two plots were tagged, and their diameter at breast height (dbh), tree height, and the height of the base of the lowest living branch were measured every 2 years from 1984 to 1992. For this stand, data in 1984 and 1992 were used as the initial and final measurement data, respectively. The Picea abies stand was planted in 1961 on a gentle mountain slope at about 220 m altitude in the experimental forest of Hokkaido Forestry Research Institute, central Hokkaido, northern Japan. Annual precipitation and mean annual temperature in this area are 1300 mm and 7±2 °C, respectively. The soil type is a brown forest soil II (Land Bureau, National Land Agency, 1977). In 1982, six plots (each 0±099 ha in area) were established in this stand, and thinned to varying densities : 495, 1060, 1505, 2000, 2384, and 2687 stems ha−". The same measurements as those for the Betula stand were carried out. For the Picea stand, data taken in 1982 and 1986 were used as the initial and final measurement data, respectively.
Estimation of canopy structure The vertical distribution of leaf mass for an individual tree was estimated by two methods : a simple one and a realistic one. In the simple method, the leaf mass of an individual was assumed to be distributed homogeneously between the crown top and the base of the lowest living branch. In the realistic method, it was assumed that the cumulative leaf mass from the top of the crown and crown depth have an allometric relationship with a variable allometric ratio (Hashimoto, 1990, 1991, 1992). This model is given by : Fz ¯ a zb ecz, where Fz denotes the cumulative leaf mass from the crown apex to the point z distant from the crown apex ; a, b and c are parameters. The derivative of this function [F z! ¯ a zb−" ecz(bcz)] represents the density distribution of leaf mass. The parameter b determines the shape of the allometric curve. The value of ®b}c gives the distance from the crown apex at which Fz (cumulative leaf mass) is maximum. At this point (z ¯®b}c), F z! is zero indicating that this point is crown base ; ®b}c gives crown length (the distance from the crown apex to the crown base). To determine species–specific values for parameters b and c, the log-transformed model was applied to data of stratified clipping of individual trees of Betula ( Y. Takahashi, unpubl. res.) and Picea (T. Asai, unpubl. res.). Because of the scarcity of data for Betula, data of other deciduous hardwood species (Komiyama, 1992) were also analysed. The restriction that ®b}c equals observed crown length was placed on the regressions to make the estimated density of leaf mass take zero value at the observed crown base. With this restriction, only parameter b should be determined and is relevant to describe the species-specific
0
0 A
B 1
Distance from crown apex (z: m)
566
2 2 4 3 4
6
5 8 6 0.0
1.0
2.0 3.0 4.0 0.0 0.4 0.8 Distribution density (F': kg m–1)
1.2
1.6
F. 1. Vertical distribution density of foliage of an individual tree of Betula platyphylla (A) and Picea glehnii (B). The fitted line is F !z ¯ a zb−" ecz (bcz) ; a ¯ 3±54, b ¯ 1±45, c ¯®0±16 for Betula, and a ¯ 0±14, b ¯ 4±02, c ¯®0±61 for Picea.
shape of the vertical leaf mass distribution, because parameter a is the scaling factor that is mainly dependent on the size of individual trees. Figure 1 shows the examples of leaf mass distributions and fitted density distribution functions (F z!) for individual trees of Betula and Picea. The leaf mass in the Betula tree was more concentrated in the upper part of the crown than that in the Picea tree. Results for other hardwoods [not shown ; data in Komiyama (1992)] and those of Hashimoto (1990) for Cryptomeria japonica D. Don also indicated this tendency ; estimated parameter b ranged from 2 to 3 for hardwoods and from 3 to 4 for conifers with a few exceptions. We therefore set a fixed b value for each species (b ¯ 2±5 for Betula and b ¯ 3±5 for Picea) for simplicity, though vertical leaf mass distribution varies with age and}or the degree of suppression by larger competitors (Hashimoto, 1990, 1991). In both methods, it was assumed that the total leaf mass of an individual tree (wL, kg) had allometric relationship with dbh (d, cm) and tree height (h, m) : wL ¯ 0±00036 d #h for Betula and wL ¯ 0±0029 d #±)&'( for Picea. These allometric relationships were those for Betula platyphylla var. japonica Hara (Takahashi, Asai and Kikuzawa, 1974) and for Picea abies (Yoshimura, 1967). The vertical leaf mass distribution for a stand was obtained by summing those for constituting individual trees using data in the initial measurement.
Degree of asymmetry of competition The change in size inequality over time from the initial to the final measurement and the relationship between the size of an individual and its growth were analysed to estimate the degree of asymmetry of competition. When mortality is negligible, size inequality is believed to increase over time
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry
¯ 0±2004 (d #h)!±)!'$, for hardwood trees (Kikuzawa, 1988) and ¯ 0±0000892 (100d )!±)!'$ 0±9935("!!d), for Picea abies (Nakajima, 1943). The relationship between the size of individual and its growth was analysed by fitting a modified general formula of size-dependent mean growth rate proposed by Yokozawa and Hara (1992). Yokozawa and Hara’s formula is given by : G(t, w) ¯ w[a ®a wm®c C(t, w)®c C(t, w )], " # ! ! " where G(t, w) denotes the mean growth rate of individuals of plant mass w at time t ; C(t, w) is proportional to cumulative basal area of individuals larger than w at time t (defined later) ; w is the minimum individual mass in a population ; ! a , a , m, c and c are constants. C(t, w) is given by : ! " " # C(t, w) ¯
&
wmax
x#f(t, x) dx,
w
where f(t, x) is the distribution density function of dbh x at time t, and wmax is the plant mass of individual of the maximum dbh. C(t, w) represents the effect of one-sided competition on the mean growth rate, and C(t, w ) that of ! two-sided competition. For simplicity, the term wm was replaced by lnw. Thus, a Gompertz growth curve for a single isolated tree was assumed. The model used was the following : G(t, ) ¯ [a ®a ln ®c C(t, )®c C(t, )] (1) ! " " # ! where denotes stem volume. Annual growth rate in terms of stem volume was calculated by dividing the volume increment by the duration of growth (8 years for the Betula stand, and 4 years for the Picea stand). Individuals were divided into ten size classes of the same ranges in each plot, and mean growth rate, G(t, ), was calculated in each size class. The model was fitted by multiple regression using the midpoint of size classes as variable , and the degree of asymmetry of competition was estimated by the significance of the regression coefficients c " and c . In the Betula stand, data of individuals of Betula spp. # were used as the criterion variable ; data of other species were involved only in the explanatory variables [C(t, ) and C(t, )]. ! RESULTS Density and change in mean size Mortality was very low in all plots except for the dense plot in the Betula stand in which 60 % of the individuals died
0.4
0.1 Mean volume (m3)
under one-sided competition, but not to do so under twosided competition (Ellison and Rabinowitz, 1989). The size inequality was expressed by the coefficient of variation of individual size. In this paper, stem volume (, m$) was used to express the size of individuals. Stem volume was estimated from dbh (d, m) and tree height (h, m) using allometric functions :
567
0.01
0.001 100
1000 Density (ha–1)
10 000 20 000
F. 2. Changes over time in stem density and mean volume of the studied plots. (_) Betula plots in 1984, (U) Betula plots in 1992, (+) Picea plots in 1982, (E) Picea plots in 1986.
during the measurement period of 8 years (Fig. 2). Mortality in this plot was strongly size-dependent. If we divide the population into two halves based on size, 85 % of the smaller individuals died, but only 37 % of the larger individuals. Because the mortality was concentrated in smaller individuals and there was no evidence for sizeindependent mortality (e.g. attack by insect or pathogen), the mortality was attributable mostly to competition. The increase of mean stem volume in a plot was dependent on stem density in the plot in both stands (Fig. 3). Canopy structure Figure 4 shows the estimated vertical distribution of leaf mass, and the distribution of tree height, and the height of the base of the lowest living branch in two representative plots in the Betula and Picea stand at the initial measurement. Only trees that survived until the final measurement were used in the distributions of tree height and the height of the lowest living branch base. The canopy structures in other plots in each stand were similar to those shown in Fig. 4. For the Betula stand, the vertical distributions of leaf mass estimated by the two different methods were very similar to each other. On the other hand, for the Picea stand, the leaf mass distribution estimated with the assumption that leaf mass is distributed homogeneously between tree height and the base of the lowest living branch (the simple method) had a flatter peak than its counterpart (the realistic method). However, the relative position of the leaf mass distribution to the distribution of tree height and the height of the base of the lowest living branch did not depend on the estimating method. Therefore, only the leaf mass distribution estimated by the realistic method is discussed hereafter.
568
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry 0.025 A 0.02
Degree of asymmetry of competition
0.015
Change in mean volume (m3)
0.01
0.005
0
0
4000
8000
12 000
16 000
20 000
0.25 B 0.2
0.15
0.1
0.05
0
than short trees in the Picea stand so that even tall individuals were shaded by short individuals to some extent.
0
500
1000 1500 2000 Density (ha–1)
2500
3000
F. 3. Relationship between the change in mean volume and the initial stem density. A, Betula ; B, Picea.
The vertical distributions of leaf mass estimated by the realistic method were similar to a normal distribution for both Betula and Picea stand, which agreed with other researchers’ results (Whitehead, 1978 ; Beadle, Talbot and Jarvis, 1982 ; Hashimoto, 1990, 1991, 1992). The relative height of the peak of leaf mass distribution to the canopy height for the Betula stand was higher than that for the Picea stand. The relative range of tree height distribution to canopy height was also different between these stands. There were a considerable number of individuals lower than the peak of leaf mass distribution in the Betula stand, whereas there were few such individuals in the Picea stand. The distribution of height of the base of the lowest living branch of the Betula stand was also wider than that of the Picea stand. The distributions of tree height and the base of the lowest living branch crossed in the Betula stand, but did not in the Picea stand. This means that the lowest portion of crowns of some tall individuals was higher than tree height of short trees in the Betula stand so that tall individuals were not shaded by short individuals, and that the lowest portion of crowns of tall individuals was lower
Although the overall positive relationship between density and coefficient of variation is apparent in both stands (Fig. 5), this tendency was artificially made due to the thinning practice in which smaller trees were cut more heavily than larger ones in low-density plots, and does not necessarily mean that the competition is one-sided by itself. In the Betula stand, coefficients of variation did not change in the dense plot through time, and increased little in the sparse plot (Fig. 5 A). In the Picea stand, coefficients of variation decreased or remained at the same level through time (Fig. 5 B). Figure 6 shows the annual mean growth rate of size classes in representative plots that had contrasting densities. In the Betula stand, mean growth rate of the largest size class in the dense plot did not differ very much from that in the sparse plot, indicating that the growth of the largest trees was hardly reduced by competition. On the other hand, mean growth rates of the largest size class in the dense plots in the Picea stand were smaller than that in the sparse plot, indicating that the growth of the largest trees was reduced to some extent by competition. Table 1 shows the estimated parameters of the general formula of size-dependent mean growth rate for both stands. For the Betula stand, the term lnw in eqn (1) was neglected, because the relationship between growth rate and size in the sparse stand, in which competition was very weak, did not show curvilinearity, and because the sample size was not very large to determine the significance of many parameters. For the Betula stand, parameter c for one" sided competition was significant at the 1 % level of probability, and c for two-sided competition at the 5 % # level. For the Picea stand, parameter c did not differ " significantly from zero, whereas c was highly significant # (P ! 0±1 %). DISCUSSION The mean growth rate in a plot decreased with density (Fig. 3), and considerable mortality was found in the dense plot in the Betula stand (Fig. 2). The existence of competition in both stands is obvious. The results of applying the general size-dependent growth model [eqn (1)] indicated that the competition in the Betula stand was more one-sided than that in the Picea stand. In general, size inequality increases over time in populations in which one-sided competition prevails (Ellison and Rabinowitz, 1989). The coefficient of variation decreased or remained at the same level over time in the Picea stand, indicating that the competition in this stand was not onesided. The coefficient of variation increased in the sparse plot in the Betula stand. However, it remained unchanged in the dense plot. It is probably because the size-dependent mortality in this plot, which was concentrated in smaller individuals, countered the increase of size inequality by onesided competition. Therefore the tendency in the changes in coefficient of variation was consistent with the results of the
569
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry –1 Number of individuals (ha )
12
0
600
1200
1800
14
0
100
200
300
400
500
600
700
800
12
10
10 Height (m)
8 8 6 6 4 4 2
2
0
2
4
6
8
10
12
14
0 16 18 –1 Leaf mass (kg ha )
50
100
150
200
250
300
F. 4. Vertical distribution of tree heights (± ± ± ±), bases of the lowest living branches (– – – –), and estimated leaf density by the realistic method (––––) and by the simple method (— — —) in two representative plots in the Betula (A) and Picea stands (B).
130
70
Coefficient of variation (%)
A
B
110
60
90
50
70
40
50
0
4000
8000
12 000
16 000
30 20 000 0 Density (ha–1)
500
1000
1500
2000
2500
3000
F. 5. Change in coefficients of variation over time from the initial measurement (+) to the final (E) related with stem density in the Betula (A) and the Picea stands (B).
growth rate analysis ; the competition in the Betula stand was more one-sided than that in the Picea stand. Because the degree of asymmetry of competition plays important roles in population dynamics (Weiner, 1990), the mechanism by which competition becomes one-sided or two-sided should be clarified. Many researchers have pointed out that if plants compete for light, competition is one-sided, whereas if plants compete for underground resources, competition is two-sided (Weiner, 1985, 1986,
1990 ; Hara, 1986 ; Weiner and Thomas, 1986). For clonal plants, the degree of physiological integration in a genet also affects the degree of asymmetry of competition between ramets (Suzuki, 1994 a, b). Another factor that affects the degree of asymmetry of competition is plant form and}or canopy structure (Turner and Rabinowitz, 1983 ; Yokozawa and Hara, 1992). If larger plants are shaded by smaller neighbours in the lower part of their crown, the competition cannot be absolutely one-sided but asymmetrically two-
570
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry 0.009
0.09
G (m3 year–1)
A
B
0.008
0.08
0.007
0.07
0.006
0.06
0.005
0.05
0.004
0.04
0.003
0.03
0.002
0.02
0.001
0.01
0.01
0.005
0
0.015
0.02
0.025
0.03 0 Volume (m3)
0.05
0.1
0.15
0.2
0.25
0.3
0.35
F. 6. Relationship between mean growth rate in size classes and initial size in dense (E) and sparse plots (+) in the Betula (A) and Picea stands (B).
T 1. Coefficients of determination (r#) and estimated parameters of the general size-dependent mean growth formula for the Betula and Picea stands Species
r#
a
!
a
"
c
"
c
Betula Picea
0±98*** 0±95***
0±34*** 0±32***
— ®0±0022
0±0230** ®0±0008
0±0042* 0±0055***
#
***, **, and * show that the coefficient of determination or the estimated parameter is significant at 0±1, 1, and 5 % level, respectively.
sided (Yokzawa and Hara, 1992). In this situation, the degree to which smaller plants shade the lower part of the crown of larger plants, which determines the degree of onesidedness in competition, depends on crown shape and}or canopy structure (Weiner and Thomas, 1986 ; Yokozawa and Hara, 1992). The results presented in this paper suggest that the difference in the degree of asymmetry of competition between the Betula stand and the Picea stand resulted from the difference in canopy structure and}or plant form in these stands. The more hierarchical canopy structure and relatively shallow crown shape in the Betula stand made the competition relatively one-sided. On the other hand, the canopy structure and crown shape in the Picea stand, in which even larger individuals can be partially shaded by smaller neighbours, made the competition more two-sided. Although no data are available for underground resources in the analysed stands, there are several lines of circumstantial evidence that this sort of resource was not relevant in determining the degree of asymmetry of competition for these stands. First, in forests in Hokkaido, it is believed that water is not a very strong limiting factor. In the two stands analysed, annual precipitation is greater than 1300 mm, which is enough for normal growth of trees,
and there is not a considerable difference in precipitation between the two stands. Moreover, these stands are similar in soil type and topography. Second, in the Betula stand, Sasa (dwarf bamboo) vegetation, litter, and humus were removed to enhance the regeneration of tree seedlings. This silvicultural practice lessened the total amount of nutrient, and might affect the degree of asymmetry of competition to make competition more two-sided. However, the growth rate analysis revealed that competition in the Betula stand was more one-sided than that in the Picea stand. Thus there is little possibility that competition for nutrients is stronger in the Picea stand than in the Betula stand, and that this made the competition in the Picea stand more two-sided than that in the Betula stand. In this study, the pronounced factor that made the canopy structure of the two stands different from each other was species-specific crown shape : wide and shallow crown of Betula s. narrow and deep crown of Picea. However, canopy structure can be affected by factors other than characteristics of species. For example, the vertical distribution of leaves in a plant population or a community is affected by age structure of the stand (Hashimoto, 1991), successional status (Aber, 1979), site quality (Aber, Pastor and Melillo, 1982), density of individual trees (Whitehead, 1978 ; Weiner, Berntson and Thomas, 1990), and species composition (Harcombe and Marks, 1977). The degree of asymmetry of competition can change with these factors. The effect of these factors on the degree of asymmetry of competition should be clarified.
A C K N O W L E D G E M E N TS The authors are grateful to Y. Takahashi and T. Asai for their unpublished data and R. Milne and an anonymous referee for reviewing the manuscript.
Kikuzawa and Umeki—Effect of Canopy Structure on Competitie Asymmetry LITERATURE CITED Aber JD. 1979. Foliage-height profiles and succession in northern hardwood forests. Ecology 60 : 18–23. Aber JD, Pastor J, Melillo JM. 1982. Changes in forest canopy structure along a site quality gradient in southern Wisconsin. The American Midland Naturalist 108 : 256–265. Beadle CL, Talbot H, Jarvis PG. 1982. Canopy structure and leaf area index in a mature Scots Pine forest. Forestry 55 : 105–123. Ellison AM, Rabinowitz D. 1989. Effects of plant morphology and emergence time on size hierarchy formation in experimental populations of two varieties of cultivated peas (Pisum satium). American Journal of Botany 76 : 427–436. Geber MA. 1989. Interplay of morphology and development on size inequality : a Polygonum greenhouse study. Ecological Monographs 59 : 267–288. Hara T. 1986. Effects of density and extinction coefficient on size variability in plant population. Annals of Botany 57 : 885–892. Hara T. 1992. Effects of the mode of competition on stationary size distribution in plant populations. Annals of Botany 69 : 509–513. Harcombe PA, Marks PL. 1977. Understory structure of a mesic forest in southeast Texas. Ecology 58 : 1144–1151. Hashimoto R. 1990. Analysis of the morphology and structure of crowns in a young sugi (Cryptomeria japonica) stand. Tree Physiology 6 : 119–134. Hashimoto R. 1991. Canopy development in young sugi (Cryptomeria japonica) stands in relation to changes with age in crown morphology and structure. Tree Physiology 8 : 129–143. Hashimoto R. 1992. Distribution structure of foliage and the light interception in a young Cryptomeria japonica stand. Journal of the Japanese Forestry Society 74 : 466–474. Kikuzawa K. 1988. Intraspecific competition in a natural stand of Betula ermanii. Annals of Botany 61 : 727–734. Komiyama A. 1992. A study on the methods for resource estimation of deciduous hardwoods. Gifu : Gifu University. (In Japanese). Land Bureau, National Land Agency. 1975. Land classification map (Hokkaido I : Ishikari district). Tokyo : Land Bureau, National Land Agency.
571
Land Bureau, National Land Agency. 1977. Land classification map (Hokkaido III : Sorachi district). Tokyo : Land Bureau, National Land Agency. Nakajima K. 1943. Tables of stem olume of trees in Hokkaido. Sapporo : Hokkaido branch of Korin-Kai. (In Japanese). Suzuki J. 1994 a. Growth dynamics of shoot height and foliage structure of a rhizomatous perennial herb, Polygonum cuspidatum. Annals of Botany 73 : 629–638. Suzuki J. 1994 b. Shoot growth dynamics and the mode of competition of two rhizomatous Polygonum species in the alpine meadow of Mt. Fuji. Folia Geobotanica et Phytotaxonomica, Praha 29 : 203–216. Takahashi Y, Asai T, Kikuzawa K. 1974. On biomass estimation of Betula platyphylla var. japonica forest stand in Nayoro. Bulletin of the Hokkaido Forest Experiment Station 12 : 29–37. Turner MD, Rabinowits D. 1983. Factors affecting frequency distributions of plant mass : the absence of dominance and suppression in competing monocultures of Festuca paradoxa. Ecology 64 : 469–475. Yokozawa M, Hara T. 1992. A canopy photosynthesis model for the dynamics of size structure and self-thinning in plant populations. Annals of Botany 70 : 305–316. Yoshimura K. 1967. Growth and biomass of Norway spruce forest in Ashu experimental forest. Bulletin of the Kyoto Uniersity Forests 39 : 27–34. Weiner J. 1985. Size hierarchies in experimental populations of annual plants. Ecology 66 : 743–752. Weiner J. 1986. How competition for light and nutrients affects size variability in Ipomoea tricolor populations. Ecology 67 : 1425–1427. Weiner J. 1990. Asymmetric competition in plant populations. Trends in Ecology and Eolution 5 : 360–364. Weiner J, Bernston GM, Thomas SC. 1990. Competition and growth form in a woodland annual. Journal of Ecology 78 : 459–469. Weiner J, Thomas SC. 1986. Size variability and competition in plant monocultures. Oikos 47 : 211–222. Whitehead D. 1978. The estimation of foliage area from sapwood basal area in Scots Pine. Forestry 51 : 137–149.