KHAN AND ZAKI (2015), FUUAST J. BIOL., 5(2): 307-314 LEAF ARCHITECTURE AND NON- DESTRUCTIVE LAMINA AREA ESTIMATION IN BAUHINIA RACEMOSA LAMK. D. KHAN AND M. JAVED ZAKI Department of Botany, University of Karachi, Karachi-75270, Pakistan. Abstract The architecture and non-destructive lamina area estimation is described in Bauhinia racemosa Lamk. The graphically measured one-sided leaf area (LAM) of 50 individual leaves of Bauhinia racemosa Lamk. varied from 0.55 to 18.53cm 2 (mean = 11.4179 ± 0.5586 cm 2 ; CV = 34.60%). The overall leaf shape, as given by aspect ratios, appeared to be maintained with age. Based on the criterion of LAWG (1999), apex shape was, however, found to be lobed in 62% of the leaves, emarginate in 36 % of the leaves and retuse in 2% leaves (young ones). Apical and basal angles were wide obtuse but apical angle was substantially larger than basal angle. The Leaf length LL was determined as LL = LM + La + Lb where La was the apical leaf extension length and Lb the basal leaf extension length. Allometric methods such as simple linear, power model and mutiple correlation and regression analyses were employed using midrib length (LM), leaf breadth (LB) or leaf length (LL) or their multiplicative parameters (LM x LB or LL x LB) as independent variables against LAM in addition to the arithmatic methods to determine mulplication coefficients (leaf form factors) k and k′ on the basis of k = Leaf area measured / (LM x LB) and k′ = Leaf area measured / (LL x LB) were determined to arrive at simple and useful models to estimate lamina area. The power model was, the good fit model relating LAM with LM x LB or LL x LB. Amongst the two, LAM was obviously better correlated with LM x LB (R = 0.984; F = 1496.64, p < 0.0001) than LL x LB (R = 0.968; F = 712.92, p < 0.0001). k averaged to 1.2727 ± 0.0176 with variability around 9.78% only and k ′ averaged to 0.787014 ± 0.01453 with variability around 13%. Comparison of various models indicated that power based models and arithmetic factors k and k′ were the best fit to estimate leaf area in B. racemosa. The use of k in leaf area estimation appears to be more suitable as k′ involves determination of leaf length, LL = LM + La + Lb, a cumbersome way to work with leaves attached with the plants. Introduction The development of predictive models for leaf area estimation is important and a useful tool in studies related to the plant growth and development. The leaf area is directly related to light interception, photosynthesis, transpiration and carbon gain and storage. It is considered to be the most important single determinant of plant productivity (Linder, 1985; Kathirvelan and Kalaiselvan, 2007). The estimation of leaf area is, however, a time-consuming and laborious task. The applicability of allometric methods in leaf area estimation was shown by Huxley (1924) first time in some grasses. Pearsall (1927) used allometric relationships in carrot and turnip to predict root storage through shoot growth estimation. Leaf area estimation in several species has been investigated by many workers for various reasons (Kemp, 1960; Jain and Misra, 1966; Williams et al., 1973; Aase et .al., 1978; Hatfield et. al., 1976; Elasner and Jubb, 1988; Chinamuthu et. al., 1989; O’Neal et al., 2002; Williams III and Martinson, 2003; Kathirvelan and Kalaiselvan, 2007, Cristofori et al., 2007; Khan, 2008, 2009, Ahmed and Khan, 2011, Khan et al, 2015 a and c). Such simple and accurate methods eliminate the need of expensive leaf area meters (Gamiely et al., 1991). In this paper leaf architecture and lamina area estimation in Bauhinia racemosa Lamk., a useful leguminous arid land tree species, have been undertaken. Materials and Methods Fifty leaves of various sizes from a mature tree of Bauhinia racemosa growing in the campus of University of Karachi were collected and their linear measurements were recorded for the length of midrib (distance from the proximal most to the distal most point of the mid-vein (LM) and lamina breadth (LB) at the broadest points on the margin i.e. perpendicular to LM. To determine true leaf area, the leaf outline was carefully drawn on graph paper and area determined with all possible precision and accuracy. The multiplication factor (k) was calculated by employing the formula, k = Leaf area measured / (LM x LB). Employing average value of the multiplication factor k, leaf areas were calculated as Leaf Area computed = k (LM x LB) for comparison with the measured areas of the leaves. Leaf architectural parameters (apical leaf extension (La) and basal leaf extension (Lb) lengths and apex and base angles) were determined according to LAWG (1999). Since B. racemosa leaves are provided with La and Lb, the leaf length (LL) was determined as LL = LM +La +Lb (LAWG, 1999). Leaf aspect ratios were calculated as LB / LM and also as LL / LB (Lu et al. (2012). The multiplication factor calculated as Leaf area measured / (LL x LB) was designated as k′. The location and dispersion parameters of the data were calculated (Zar, 1994). The skewness and kurtosis (g1 and g2, respectively) were calculated as g 1 = K3 / (K2’) 3/2 and g2 = K4 / (K2’) 2 , respectively - Ks, are moments around mean (see Shaukat and Khan, 1979). The standard errors of skewness and kurtosis (Sg1 and Sg2, respectively) were given as: Sg1=√ 6N (N-1) / (N-2) (N+1) (N+3) and Sg2= √ 24N (N-1) 2 / (N-3) (N-2) (N+3) (N+5). Linear and power law relationships of leaf area with multiplicative parameters of LM x LB and LL x LB were determined. In addition to it, the regression coefficients were also calculated by employing multiple regression method fitting in the allometric model, Y = a + b 1LM + b2LB ± SE and also as Y = a + b1LL + b2LB ± SE (Zar, 1994). The arithmetic and allometric methods were compared for their precision and suitability. The data was analyzed using SPSS version 12.