Tree size structure of Tectona grandis (Linn f.) stand in Hilltop and Valley-Bottom of Omo Forest Reserve

: Competition for growth resources contributes to size hierarchy in tree populations. Com-7 petition hierarchy of trees is dependent on rate of growth and stages of stand development. How-8 ever, competition hierarchy may not cause size symmetry in tree populations. Size structure of even-9 aged stand can identify mechanisms for growth resources competition among trees. The study in-10 vestigated tree size structure of Teak stand in Valley-Bottom and Hilltop of Omo Forest Reserve. Ten (10) years old Teak plantation was divided into Hilltop and Valley-Bottom stands base on to-12 pography. Five (30m x 30m) sample plots were systematically demarcated in each of Hilltop and Valley-Bottom stands. Tree stems were enumerated and stem densities of both stands were esti-14 mated. Diameter at -breast height and total height were measured using Girth tape and Spiegel Relaskop, respectively. Stem size inequality, diversity and evenness of both stands were evaluated. Data collected were analyzed using descriptive, correlation, regression analysis and t-test at α0.05. Mean diameter and height of Valley-Bottom (11.42±4.83cm dbh and 3.46±1.35m) were not signifi-18 cantly different from Hilltop stands (10.29±4.59 cm dbh and 3.41±1.55m). Stem density of Hilltop (1431.0 stems/ha) was higher than Valley-Bottom stands (1248.0stems/ha). Coefficient of determina-20 tion (R2) of Height-Diameter allometry for Valley-Bottom (0.59) was higher than Hilltop stands 21 (0.45). Diameter distribution of Valley-Bottom and Hilltop expressed bimodality and unimodality, 22 respectively. Height distribution of Valley-Bottom and Hilltop expressed positive skewed uni-23 modality. Inequality was higher in Hilltop than Valley-Bottom for height and diameter. Elevation 24 affected the stem form and size hierarchy of Teak stems in Hilltop habitat than Valley-Bottom hab-25 itat. Different mechanisms were responsible for stand structure of Hilltop and Valley-Bottom Hab-26 itats.


31
There is competition for resources among plant populations. Asymmetric and sym-32 metric models are recognized as two extreme expressions of competition models [1]. 33 There is intrinsic difference between competition symmetry for above-ground and below-34 ground tree growth resources. Asymmetric and symmetric models are considered for 35 light and plant nutrient, respectively. Therefore, tree size symmetry varies with variation 36 in resources availability [2]. Identification of mechanisms that determine size hierarchy in 37 tree populations is critical because of their ecological and management significance [3]. 38 However, understanding the effect of topographic elevation on competition hierarchy is 39 limited [4]. The estimate of size structure of even-aged Tectona grandis plantation in dif-40 ferent elevations is required so as to identify competition mechanisms for tree growth 41 resources at different elevation belts. Tree height and stem diameter are components of 42 tree size. The tree height determines light capturing capacity while stem diameter deter-43 mines mechanical support and water transport efficiency [5]. Allometry and architecture 44 of a tree are regulated by abiotic and biotic factors [5]. Moreover, tree height-diameter 45 relationship reflects the available environmental resources and therefore, can be used to 46 support decisions on silvicultural treatments. However, the effect of elevation on tree 47 height-diameter allometry is yet to be clarified. The hypothesis was to assess the effect of 48 habitat on size inequality within the teak plantation. The aim of the study was to analyse 49 the the spatial difference of the diameter distribution of 10-year-old Tectona grandis plan-50 tation in Omo Forest Reserve. Therefore, this study investigated tree size structure of Teak 51 stands in Hilltop and Valley-Bottom of Omo Forest Reserve. 52 53 54 This study was conducted in 10-year-old Tectona grandis plantation in Area J4 of Omo 55 Forest Reserve. Omo Forest Reserve is located between Latitude 6° 35´ to 7° 05´ N and 56 Longitude 4° 19´ to 4° 40´ E at altitude 150 above sea level (asl) in the Ijebu area of Ogun 57 state in Southwestern Nigeria [6]. Omo Forest Reserve covers 130,500 hectares of land 58 area. It is the largest industrial plantation in Nigeria. The Tectona grandis plantation used 59 for this study was planted in year 2010 using a spacing of 2.0 m x 3.0 m among tree stems 60 and covers 22 hectares of land area. The plantation is located in Fire Blast area of Area J4 61 in Omo Forest Reserve. 62

63
Reconnaissance survey was conducted to access the landscape and stand physiog-64 nomy so as to determine the sampling technique to be adopted. It was observed that the 65 Teak plantation was on steepy landscape. Therefore, Teak plantation was divided into 66 two stands base on natural demarcation of its topography so as to achieve the objective 67 and reduce variation. Therefore, the plantation was subjectively divided into two altitu-68 dinal levels; Hilltop stand is located between 105 and 112 m and Valley-Bottom stand is 69 located between 85 and 104 m above sea level (asl). The sampling method for plot selection 70 was systematic sampling technique. Five sample (30m x 30m) plots were systematically 71 demarcated in each of Hilltop and Valley-Bottom stands. The height and diameter-at 72 (base, breast-height, middle and top) of Teak stems were measured in each plot using 73 Spiegel relaskop and Girth tape and stem density was estimated. 74

75
Stem density was computed for Hilltop and Valley-Bottom stands and converted to 76 hectare. The regression analysis of stem H-D allometry of Hlltop and Valley-Bottom 77 stands were evaluated. Also, diameter-at-breast height (dbh) and height measurements of 78 tree stems were divided into 17 equal interval size classes starting from the smallest to the 79 largest and size-density distribution were represented with histogram of stem diameter 80 and height distributions, respectively. Therefore, diameter-density and height-density 81 distribution of Hilltop and Valley-Bottom stands were characterized by their mean, stand-82 ard deviation and Coefficient of Variation and tested for normality by calculating Skew-83 ness coefficient and Kurtosis. Also, Inequality measures (Gini-coefficient, Coefficient of 84 Variation and Skewness coefficient) were calculated for the diameter and height distribu-85 tions of Hilltop and Valley-Bottom stands. Further analysis was carried out; (i) Significant 86 differences between means were tested using t-test at 0.05 level, (ii) Inequality statistics 87 (Gini-Coefficient, Coefficient of Variation and Skewnes-Coefficient) were correlated with 88 tree size diversity measures (Shannon-Weiner and Simpson-indices) and tree size eveness 89 measures (Eveness and Margalef indices) at 0.05 level. The highly significant correlation 90 values at 0.05 level were extracted from matrices.

The H-D Allometry
The H-D relationship for Hilltop stand was derived from 644 sample tree stems and 94 best described by the equation (Height = 5.73*ln(Dbh) -5.61) which explained 45.5% of 95 variation in tree height while the Valley-Bottom stand was derived from 562 sample tree 96 stems and best described by the equation ( Height = 5.34*ln(Dbh) -5.16) which explained 97 59.1% of variation in tree height. There was significant difference between tree height for 98 a given diameter of stems in Hilltop and Valley-Bottom stands. The diameter-at-breast-99 height increased with exponential increase in height in Hilltop and Valley-Bottom stands 100 of Tectona grandis. Figure 1 and 2 showed that H-D relationship may be site specific. There-101 fore, a single equation can not be used for the prediction of H-D relationship of Gmelina 102 arborea in Omo Forest Reserve.  (Table 1a). Therefore, skewness and kurtosis of stem height in 117 Hilltop stand were higher than Valley-Bottom stand. (Table 1b). Inequality of stem height and diameter was evaluated by Gini-Coefficient (GC), Co-122 efficient of Variation (CV) and Skewness Coefficient (SC). Therefore, inequality of stem 123 diameter and height distribution of Hilltop stand was higher than inequality of Valley-124 Bottom stand (Table 1a and 1b). Also, stand density of Hilltop stand (1431.00 stems/ha) 125 was higher than Valley-Bottom stand (1251.00 stems/ha) (Table 1a and  The result of correlation analysis between inequality measures and diversity indices 143 of stem diameter distribution in Hilltop stand (Table 2a)

Height (m) Figure 3. Height distribution of Teak stand on Valley-Bottom and Hilltop in Omo Forest Reserve
HT freq

BV freq
The result of correlation analysis between inequality measures and diversity indices 158 of stem diameter distribution in Valley-Bottom stand (Table 2b)  The result of correlation analysis between inequality measures and diversity indices 178 of stem height distribution in Valley-Bottom stand (Table 3b)  The relationship between tree height and diameter is an indicator of stem form [7] 203 and therefore was examined in Hilltop and Valley-Bottom stands. The relationship of 204 Height-Diameter allometry is useful to identify competitive effect of tree stems on their 205 morphological feature since relationship between height and diameter depends on site 206 conditions [8]. Therefore, a regression analysis of H-D allometry was used to determine 207 the relationship between tree height and diameter-at-breast of Teak in Hilltop and Valley-208 Bottom habitats. The results showed that variability of H-D allometry in Valley-Bottom 209 stand was higher than variability in Hilltop stand. Stem form of many trees were more 210 than the average (Height-Diameter ratio > 1.0) in Hilltop stand. Therefore, Hilltop stand 211 had trees that allocated more biomass to tree height growth than stem diameter growth. 212 Conversely, stem form of many tree were approximately average in Valley-Bottom stand 213 (Height-Diameter = 1.0). This indicated that relative height growth of most tree stems was 214 almost equal to relative diameter growth. Tree stems in Hilltop stand had increased height 215 growth compared to diameter growth. Hilltop stand displayed higher canopy stature than 216 Valley-Bottom stand. The axial growth is a trait that show strong adaptation where com-217 petition for space is very important. This contrary to the report of [9] that tree growth and 218 competition for light declined with elevation. Therefore, effect of stem density is more 219 significant on tree growth than effect of elevation. The height increased with increase in 220 diameter in Valley-Bottom stand. This suggested that stem form differ among trees of dif-221 ferent sizes [7] and elevations.
[5] stated that allocation of biomass to stem diameter is 222 likely to occur when greater inter tree competition is present or environmental disturb-223 ance. Difference in H-D relationship was found in the two stands. Stem density probably 224 caused difference in the allometric equation of the two sites. Initially Hilltop and Valley-225 Bottom stands were established using 2.0 x 3.0 espacement but a lot of forked stems were 226 observed in the Hilltop stand probably due to water stress at seedling stage. Flooding and 227 water logging were noticed in Valley-Bottom stand. Flooding and water logging during 228 rainy season may reduce rate of plant growth of large tree stems in Valley-Bottom stand. 229 Therefore, size hierarchy is influenced by water availability and duration of water availa-230 bility. Competition may be primarily symmetric when water availability is low and asym-231 metric when water availability is high [10] and [4]. The difference in stem form between 232 Valley-Bottom and Hilltop stands may be caused by water logging as a consequence of 233 difference in elevation.  235 Histogram of frequency distribution allows a visual estimate of the shape of distri-236 bution to be made [11]. Diameter-density distribution of Valley-Bottom stand expressed 237 positively skewed bimodal distribution while Hilltop stand expressed positively skewed 238 reverse J-shaped unimodal distribution. Therefore, Valley-Bottom stand had a second 239 maximum in the middle size class in addition to positive skewness. Histogram of Valley-240 Bottom stand indicated unequal decline in relative growth rates across plant size classes 241 with decreasing stem density. [12] suggested that bimodality distribution was the conse-242 quence of a disjunct distribution of relative growth rates in the population where individ-243 uals share limited resources disproportionately in relation to their relative sizes. Diameter 244 classes of Hilltop stand contained higher stem density (stems/ha) than Valley-Bottom 245 stand except at class 12.06-14.06 and 14.07-16.07 cm dbh. Therefore, forest structure of 246 Hilltop stand was higher than Valley-Bottom stand. Two peaks in diameter distribution 247 of Valley-Bottom stand suggest development of a two tiered canopy of large and small 248 tree stems. Therefore, Valley-Bottom produced bimodal frequency distribution of plant 249 size. This described a segregation of Tectona grandis tree stems into suppressed and dom-250 inant trees. The segregation occurs before the occurrence of substantial mortality in mon-251 oculture stands [13]. The large diameter trees had higher relative growth rates than small 252 diameter trees. Moreover, [13] reported that segregation occurs when large plants inter-253 cept a disproportionately large portion of available light as their canopies overlap those 254 of the smaller trees. The difference between dbh classes of Hilltop and Valley-Bottom 255 stands was the number of stems in the mid-classes of diameter distribution. The major 256 difference between the Hilltop and Valley-Bottom was stem density of the saplings (4.02-257 14.02 cm dbh). This partialy support the report of [14] that vigorous mid-class growth may 258 produce a sigmoid distribution. 259 Positive skewness showed that few large trees suppressed growth of numerous small 260 stems [15]. High coefficient of variation indicates that a higher relative growth rate of stem 261 diameter [16] in Hilltop than Valley Bottom stands. Although there was no significant 262 difference in the stem diameter of both Hilltop and Valley-Bottom stands but tree stems 263 of relatively small diameter occupied Valley-Bottom habitat because selective logging was 264 noticed in the area. The value of skewness of Hilltop stand was higher than Valley-Bot-265 tom. According to [17] skewness indicates interference among tree stems. Therefore, more 266 interference occurred among tree stems of Hilltop stand. 267 Stem diameter inequality of Hilltop stand was higher than Valley-Bottom stand. 268 Therefore, stem diameter inequality was greater at higher tree density [18]. Size asymmet-269 ric competition is more applicable in Hilltop stand than size symmetric competition. It 270 was proposed that skewness can be used as a measure of interference [17]. Also, size 271 asymmetry refers to skewness within the size-frequency distribution while size inequality 272 refers to the uneven allocation of mass among individuals in a population [19]. The dif-273 ference in inequality of diameter distribution of Hilltop and Valley Bottom stands could 274 be a consequence of elevation gradient. The presence of resource depletion increase the 275 skewness and variance of distributions of plant size [20]. Size inequality in plant commu-276 nities arises when a few large individual suppress the growth of the other tree stems [15]. 277 The two stands differ in height-density distribution and size inequality. This suggested 278 that competition intensity of Hilltop was greater than Valley Bottom stands because com-279 petition for resources increases size inequality in tree populations.   286 The Gini coefficient obtained from the stem diameter distribution of Hilltop and Val-287 ley Bottom stands were higher than the Gini obtained from stem height of Hilltop and 288 Valley Bottom stands. Therefore inequality was significantly greater in diameter than in 289 height in the Hilltop stand but not in the Valley. The Gini values were significantly higher 290 in the Hilltop than in the Valley Bottom for both stem diameter and height. This resulted 291 in greater size variability at increased density. Size variability increases with stem density 292 [19]. Size inequality was measured for stem size by Gini-Coefficient, Coefficient of Varia-293 tion and Skewness Coefficient [21] and [22]. Size diversity was measure by Shannon- 294 Weiner and Simpson indices and stem size eveness was measure by Margalef index and 295 Eveness. Size asymmetry refers to skewness within the size frequency distribution while 296 size inequality refers to the uneven relative growth rate among individuals in a popula-297 tion [19]. Environmental constrain may reduce the evenness of plant stems and commu-298 nities [23]. All diversity and evenness indices showed considerable relationship with 299 skewness of stem diameter [24] in Hilltop stand. The value of Gini-Coefficient of height 300 distribution was closely related to Shannon-Weiner diversity index of height and Mar-301 galef index of height. Therefore, inequality measures were closely related to size diversity 302 measures and evenness for stem height distribution. The two main factors taken into con-303 sideration when measuring diversity are richness and evenness. Richness is a measure of 304 the number of different size classes in a population while evenness compares the similar-305 ity of different size classes in a population and is related to other attributes of the popula-306 tion such as competition, structure and stability [24]. 307

308
The variation in height-diameter allometry between Hilltop and Valley-Bottom 309 stands may be caused by difference in elevation. Therefore, stem form differ among trees 310 of different sizes and elevations. The frequency distribution of plant diameter in Valley-311 Bottom was bimodal. High elevation promotes tree size inequality while low elevation 312 promotes homogeneity in tree size classes. Valley-Bottom stand had approximately ho-313 mogeneity tree size classes. The relationship between proportion of different height clas-314 ses (inequality) and height variation (diversity) was greater than that of diameter in Val-315 ley-Bottom stand. Stem height distribution indicated asymmetric competition among tree 316 stem than diameter distribution in Valley-Bottom stand of 10-year-old Tectona grandis 317 plantation. Therefore, height distribution of Valley-Bottom stand was more applicable for 318 description of size asymmetric competition in 10-year-old Tectona grandis plantation.

320
Conflicts of Interest: The authors declare no conflict of interest. 321