Improving the sample plan for the assessment index of site productivity based on height in uneven-aged mixed beech and fir forests (with spruce) Admir Avdagić 1) , Besim Balić 1) , Ahmet Lojo 1) , Azra Čabaravdić 1) , Aida Ibrahimspahić 1) 1) Department for Forest Management and Urban Greenery, Faculty of Forestry, University of Sarajevo, 71000 Sarajevo, Bosnia and Herzegovina Corresponding author: Admir Avdagić Faculty of Forestry, University of Sarajevo, 71000 Sarajevo, Bosnia and Herzegovina Email: a.avdagic@sfsa.unsa.ba Website: www.sfsa.unsa.ba Phone: +387 33 812 490/491 Contact 1. BALIĆ, B. (2011): Proizvodnost, strukturna izgrađenost i modeli rasta i prirasta izdanačkih šuma bukve na području Kantona Sarajevo. Disertacija. Šumarski fakultet Univerziteta u Sarajevu. 2. BALIĆ, B., LOJO, A., IBRAHIMSPAHIĆ, A. (2007): Utvrđivanje bonitetne dispozicije staništa izdanačkih šuma bukve na području Kantona Sarajevo. Radovi Šumarskog fakulteta u Sarajevu. Radovi Šumarskog fakulteta u Sarajevu. No.1, –str 67-79. 3. KRAMER, H., AKÇA, A. (1995): Letfaden zur Waldmesslehre. J.D.Sauerländer's Verlag. 3. erweiterte Auflage. Frankfurt am Main. 4. KUŠAR G. (2007): Zanesljivost ugotavljanja volumna dreves in lesne zaloge sastojev z enoparametrskim funkcijami in stratifikacijo. Disertacija. Univerza v Ljubljani. Biotehniška fakulteta. str. 32-38 5. LOJO, A. (2011): Formiranje gazdinskih klasa unutar šuma bukve i jele i šuma bukve i jele sa smrčom na krečnjacima i dolomitima. Disertacija. Šumarski fakultet Univerziteta u Sarajevu 6. LOJO, A., BALIĆ, B. (2008): Preliminarne zapreminske tablice bukve izdanačkog porijekla u BiH. Šesti simpozij poljoprivrede, veterinarstva, šumarstva i biotehnologije “Strategija razvoja domaće proizvodnje”, Goražde, Knjiga kratkih sadržaja, 2008.-str. 99-100. 7. MATIĆ, V. (1977): Metodika izrade šumskoprivredni osnova za šume u društvenoj svojini na području SRBIH. Radovi šumarskog fakulteta u sarajevu. Posebno izdanje. Šumarski fakultet u Sarajevu. 8. SKOVSGAARD J.P., VANCLAY J.K. (2012): Forest site productivity: review of the evolution of dendrometric concepts for even_aged stands References Beech and fir forests (with spruce) are the most common forests in Bosnia and Herzegovina. According to the data from the first forest inventory in large areas of B&H (MATIĆ et al., 1971), they occupy the area of 562,237 ha or 49.7 % of high forests in Bosnia and Herzegovina (1,130,183 ha) and 25% of the total forest land area in Bosnia and Herzegovina (2,501,465 ha). Index of site productivity of a particular stand has been evaluated by constructing a height curve for the given tree species. After overlaying the determined height curve with the appropriate index of site productivity disposition curves for a given tree species and analysing the curve height position, difficulties in assessing the index of site productivity have been encountered. Height of the curve of determined stands often intersects two or more curves. In such cases, assessment of the index of site productivity is quite unreliable. Introduction • Prodan model was chosen as the best dendrometric model for all investigated species of trees, although the differences between the function is negligible. • Tree heights below 30 cm in diameter showed greater variability than trees over 30 cm in diameter. The 7.5 cm class indicates the highest variability. • Partial site class/bonity calculated using the weighted arithmetic mean should be used for calculating site class/bonity and volume, or basal area should be used as weights to avoid subjectivity in the assessment. • The minimum number of trees required for a reliable assessment of the site class/bonity is at least 25 trees (which are a „carrier” of stand volume) • There is no significant difference for the calculated site classes/bonity between all 12 replications at equal distance between the sample centres • In regard to the complete survey, there is a significant difference between the site class for sample plots and total survey for diameter class lower than 30 cm. • Three experimental plots of 9.0 hectares each with different amounts of tree species have been placed on the field (Figure 2). Total experimental area is 27 hectares. The plots are selected using a IRS 5m satellite image. • Within the three separate plots, 30,110 trees in total have been included in the survey. From this total, 16,898 trees of fir were measured (56%), 8,379 trees of beech (28%), 4,753 of other broadleaves (16%) and 80 trees of spruce (0.3%). For every tree above the threshold of 5 cm the diameter at breast height (1.3 m) and height is measured. • For this purpose, the following methods were used: the method of correlation analysis (analytical method equalization curve by applying an appropriate height dendrometrics function), analysis of variance, multiple post hoc tests (Fisher's LSD test and Dunnett's test), the method of comparison, the GIS spatial analysis. • Systematic grid sampling that has been replicated 12 times, with five different plot grids Methods and Materials We propose a new plan of measuring the height of trees between 30 and 70 cm DBH on every second sample plot. This way the costs would remain the same but the assessment results would undoubtedly be more reliable. Furthermore, this model will ensure that there is a sufficient number of trees for a reliable assessment of site class/bonity, avoiding the trees that show a high variability in height. Conclusions • Determine the tree height variability for each diameter class within the same stand per tree species; • Assess the minimum number of trees for a reliable assessment of the habitat site class/bonity in the forests of beech, fir and spruce in limestone; • Create the plan for assessing the site class/bonity of the habitats on the basis of tree height in the forests of beech, fir with spruce in limestone, and the reliability assessment. Results Chart 1. Prodan model for fir. Objectives Figure 1. Map of research area, north- west Bosnia and Herzegovina Figure 2. Positions of the plots on satellite images IRS 5m with the example of sample plot grid 100x100 m               − ⋅ − − = ) ( 4 , 0 3 . . 3 5 . 0 3 H H H H b p i 0 5 10 15 20 25 30 35 40 45 0 10 20 30 40 50 60 70 80 90 100 110 120 Height (m) DBH (cm) I III V H_1 H_3 H_5 0 5 10 15 20 25 30 35 40 45 0 10 20 30 40 50 60 70 80 90 100 110 120 Height (m) DBH (cm) I III V H_1 H_3 H_5 0 5 10 15 20 25 30 35 40 45 0 10 20 30 40 50 60 70 80 90 100 110 120 Height (m) DBH (cm) I III V H_1 H_3 H_5  = ∑   ∙   ∑   b = ∑   ∙  ∑    = 1,3 +  1,3 2  +  1,3 +  1,3 2 +  1,3 3 Formula 2. Model for calculating the relative site class/bonity for any pair of measured data (D 1,30 i H) (BALIĆ, 2011) Formula 1. Analytical expression for data which has been already equalized and presented the results in “Tables for High and Coppice Forests in B&H (DRINĆ, et al, 1990) (BALIĆ, 2011) Chart 2. Prodan model for spruce Chart 3. Prodan model for beech. Chart 4. Comparison of graphical functions with the existing bonity disposition for fir Chart 5. Comparison of graphical functions with the existing bonity disposition for spruce Chart 6. Comparison of graphical functions with the existing bonity disposition for beech Formula 3. Calculating a site class/bonity trough weighted arithmetic mean where volume is used as weights. (LOJO, 2011) Formula 4. Calculating a site class/bonity trough weighted arithmetic mean where basal area is used as weights. (BALIĆ, 2011) Faculty of Forestry, University of Sarajevo, Bosnia and Herzegovina Scan for pdf.