Pedunculate Oak (Quercus robur L.) Crown Defoliation as an Indicator of Timber Value

Abstract

Pedunculate oak (Quercus robur L.), an ecologically and economically important tree species has been significantly affected by oak dieback in recent years. Since one of the symptoms of oak dieback is crown defoliation, this research aimed to determine the quantity, quality, average tree value, and wood defects that influence grading in different stages of oak dieback indicated by tree crown defoliation degree. The research was conducted in a 62- and 116-year-old stand of the lowland Croatian forest. In total, 115 pedunculate oak trees were sampled and processed in 983 logs that were analyzed. The prescribed single-entry volume tables underestimate harvesting volume by 5.45% on site A and 6.16% on site B, while the calculation of net harvesting volume underestimates net volume by 0.26% on site A and overestimates net volume on site B by 4.59%. The analysis of wood defect presence showed that insect holes, rot, and covered knots were the main reasons for the degradation of quality class. Dead trees showed a decreased average tree value in DBH classes 32.5–42.5 cm compared to the healthy trees. Based on the findings of this research, tree crown defoliation degree could be used as a timber quality and average tree value indicator.

1. Introduction

Forest and wooded land cover more than 43.5% of the land area in the European Union [1]. European forests are negatively affected by climate change, particularly but not only in mono-specific and even-aged forest stands [1]. The reduced productivity of certain forest ecosystems in Croatia such as oak stands, was recognized as a possibility and stated in the Climate Change Adaptation Strategy of the Republic of Croatia [2]. Pedunculate oak (Quercus robur L.) is an important ecological and economical tree species in European forests [3] and an important renewable resource respected because of its beauty, good mechanical properties, and durability, and it is used in many ways, e.g., construction, veneer, flooring, furniture, fuelwood, etc. [4]. It is the second most common tree species in the Croatian state-owned forests, spreading across 205,726.62 ha (24.22%) with 29.93% (61,909,035 m³) of total state-owned growing stock [5]. Pedunculate oak dieback is a known and well-described problem in the European and Croatian forests. In Croatian forests, pedunculate oak dieback started at the beginning of the 20th century, with a significant increase in recent times [6]. Pedunculate oak salvage fellings in the period from 1996 to 2005 amounted to 2,696,062 m³, of which 92% was in thinning stands of Croatian forests [7]. The symptoms of oak dieback are crown defoliation, thinning of the canopy, bleeding bark, bark cracks, etc. [8]. Many studies emphasize pedunculate oak as an ecologically and economically important tree species [9,10,11]. Therefore, pedunculate oak dieback presents a serious problem affecting the quality and consequently the value of oak assortments [12].

For timber trade transparency, harvested timber requires efficient and accurate measurements [13]. To create an accurate estimate of the merchantable timber volume, two important types of data are crucial: correctly estimated harvesting volume and accurate assortment tables. While models for determining tree and stand volume have been very well researched, verified models for timber quality estimation are still lacking [14,15]. Estimating accurate timber quality distribution has strong benefits for owners and forest managers who can adjust silviculture measures [14] but also offers valuable information in harvesting efficiency calculation [16]. The methods for assessing timber quality can be divided into destructive methods, when trees are felled and processed, after which wood quality is determined, and non-destructive methods, when assessment is performed on standing trees [17].

Wood quality is defined as a combination of intrinsic characteristics, key attributes, and factors influencing its suitability for a specific use [14,18], and it depends on the numerous influencing factors. The further use of wood and the final price of each log is heavily influenced by the complete set of wood characteristics [17]. Generally, small-sized and low-quality oak timber is less valuable compared to the branch-free wood of straight, large-dimension oak timber [19]. Wood quality influencing factors can be divided into stand-level factors, e.g., site condition, competition intensity, light availability, genetics and disturbance [20,21], and tree-level factors, e.g., crown development, branch size and location, knot size, type and placement, stem shape, and taper [22]. Forest management plans are essential to implement appropriate silvicultural practices resulting in the production of high-quality timber assortments. Many studies emphasize the influence of silviculture measurements on the wood quality [23,24,25], especially thinning, as one of the main silvicultural measures [19,26,27]. In contrast to the known results of the controlled silvicultural measures, numerous natural disturbances can appear during the stand rotation that could negatively change conditions in which the stands are developing and consequently influence timber quality.

The negative effect of pedunculate oak dieback is recognized through reduced wood density, cellulose content, moisture content, and the questionable possibility of using dead trees’ wood [28]. Further, the influence of wood defects, like frost cracks [29], knots [30,31], red heartwood [32,33,34,35], etc., have been studied in detail, while the combined influence of all wood characteristics on classification and wood quality has rarely been studied [17,36,37,38].

Tree crown defoliation degree is mostly used as a rapid indicator of tree health and vigor assessment [39]. Moreover, researchers were focused on determining the correlation between tree crown defoliation degree and different types of stressors on tree health [40], to improve growth rate models [41] and to improve tree mortality models [42]. On the other hand, the influence of tree crown defoliation degree on timber quality and value has rarely been researched [43].

Following the outlined lack of knowledge about the influence of tree crown defoliation degree and the combined impact of present wood defects on timber quality, this study aimed to determine the following: (a) the tree crown defoliation degree influence on the quantity and quality of produced assortments and average tree value; (b) the intensity of wood defect presence in different oak dieback stages; (c) the level of discrepancies in the harvesting plans and harvested wood quality and quantity.

2. Materials and Methods

2.1. Sample Preparation

Investigation of the influence of tree crown defoliation degree class (TCDD) on the pedunculate oak timber quantity and quality was performed in the Croatian floodplain forests on two sites (

Table 1. Location, phytocenosis, and age of researched sites.

Site	Management Unit; Subcompartment	Coordinates	Phytocenosis	Age
A	Kusare; 6a	φ 45°07′10″ N; λ 18°36′17″ E	Genisto elatae-Quercetum roboris Horvat 1938	62
B	Česma; 101c, 96c, 97c	φ 45°49′52″ N; λ 16°37′22″ E	Carpino betuli-Quercetum roboris typicum Rauš 1975	116

) where, via a management and harvesting plan, a combination of salvage felling and thinning was performed. On the chosen sites, tree marking was performed by »Croatian Forests« Ltd., Zagreb, Croatia, a state company that manages 73% of the Croatian forest area. Sample trees were randomly chosen between trees marked for felling based on the diameter at breast height (DBH) and TCDD class. The tree crown defoliation degree was estimated according to the ICP guidelines for visual assessment of the crown condition and damaging agents [44]. The visual assessment was conducted using a standard ocular method, which requires finding a reference tree (an undefoliated and healthy tree) that serves as a base to assess the amount of missing leaves on the defoliated trees. The chosen trees were classified into the four following classes: healthy tree (H) if crown defoliation was less than 25%, 3A tree (3A) if crown defoliation was 61%–80%, 3B tree (3B) if crown defoliation was 81%–99%, and dead tree (D) if crown was 100% defoliated.

2.2. Quantity and Quality Determination

To determine the quantity and quality of the produced log assortments, a destructive method was used. Tree bucking was performed based on the wood’s potential for its end use in the wood industry following HRN standard [45,46,47], which is in use by »Croatian Forests« Ltd. Following minimal dimensions and allowed wood defects prescribed by the used standard, produced logs were classified as veneer logs of the first quality class (VLI), and second quality class (VLII) [45], saw logs of the first quality class (SLI), second quality class (SLII), and third quality class (SLIII) [46], and fuel wood (FW) [47]. Additionally, quality classes were assigned following HRN EN 1316-1 standard grades [48], which classify produced round wood based on its quality into following classes: A, B, C, D, and FW (where A is the best quality class and D is the worst quality class).

To determine the volume of sampled trees, HRN D.B0.022 [49] and HRN EN 1309-2 [50] standards were used. Both standards have specific rules for log volume determination (

Table 2. Prescribed quantity determination rules used by standards.

	HRN D.B0.022 [49]	HRN EN 1309-2 [50]
Length	- The shortest length is measured with an accuracy of at least one centimeter; - Timber with a notch cut or butt trimming should be measured from the end of the notch cut or butt trimming surface; - Length is rounded to the nearest 0.1 m; - The length allowance for the saw logs is 10 cm.	- The shortest length is measured with an accuracy of at least one centimeter; - Timber with a notch cut or butt trimming should be measured from the middle of the notch cut or butt trimming surface; - The length is expressed in meters to one place of decimal rounded down; - The length allowance is not given.
Diameter	- Measured two mid-length under bark diameters to an accuracy of at least one centimeter; - In case of an over-bark measure, conversion to an under-bark diameter is necessary; - The arithmetic mean of two measurements is rounded down to the nearest centimeter.	- Measured two mid-length under bark diameters to an accuracy of at least one centimeter; - In case of an over-bark measure, conversion to an under-bark diameter is necessary; - The arithmetic mean of two measurements is rounded down to the nearest centimeter.
Volume calculation	$V={\displaystyle \frac{d^2\ast 3.14}{40000}}\ast l$	$V={\displaystyle \frac{d^2\ast 3.1416}{40000}}\ast l$

V—log volume; l—log length; d—log diameter.

) and grading rules in quality determination. For subsequent data processing and gross volume calculation, the actual log dimensions were measured. Log length was measured with an accuracy of 0.01 m, and two mid-length diameters were measured with an accuracy of 0.1 cm. Additionally, the length of the notch cut was measured for the butt log and unprocessed crown length for each tree.

Data was recorded using UMTPlus Max 20.1.12 software installed on a Huawei P30 (Huawei Technologies Co., Ltd. Shenzhen, China) smartphone. The original interface of the mentioned software was customized following Figure 1, to allow us to record data of the log’s dimensions, present wood defects, and quality classes on the tree level. To assign quality class, the entrance criteria were measured dimensions of the produced log, followed by the presence of wood defects, which was visually determined. Since the wood grading was performed on the site, only wood defects that were present on the log’s cross section and bark were determined and measured. If wood defects were not present, the highest possible quality class was assigned based on the log dimensions. If wood defects were present, their presence was recorded and additional measurements were made to determine their dimensions (e.g., knot diameter) and frequency. Following the used standard and wood defect presence, the highest possible quality class was assigned.

2.3. Data Processing

2.3.1. Processing of Collected Data

Recorded data processing was started by creating a database in MS Excel 2412 software. Firstly, the database was checked to determine the correctness of the data, after which the distribution of sampled trees in five-centimeter-wide DBH classes and TCDD classes was made for each site. Single log gross volume was calculated based on the actual length and over-bark diameter using Huber’s formula. Tree gross volume was calculated by adding up the gross volume of each measured roundwood tree part (over-bark diameter > 7.0 cm) of a tree. To calculate the log’s net volume, measured dimensions were rounded, and calculation was performed following rules prescribed by HRN and HRN EN standards (Table 2), after which a single log’s net volume was summed on the tree level. Following the »Croatian Forests« Ltd. price list (fco. forest road) [51], the total value of produced logs was calculated, after which the average tree value (EUR/m³) was calculated by dividing the total tree value by net tree volume determined by HRN standard.

2.3.2. Calculation of Expected Harvesting Volume and Value

The distribution of the number of sampled trees regarding DBH classes and TCDD classes was used to create harvesting plans based on the single-entry volume tables and assortment tables that are in operational use by »Croatian Forests« Ltd. The single-entry volume tables provide data on tree gross volume (over-bark diameter > 7.0 cm) across DBH classes, while the assortment tables provide the share of quality classes across DBH classes. To calculate the expected gross volume, the number of trees in each DBH class was multiplied by the volume predicted by single-entry volume tables. Assortment tables used in Croatian forestry are divided into three types: (a) assortment tables for thinning; (b) assortment tables for regeneration felling; and (c) assortment tables for dead trees. According to the Regulation on tree marking, classification of forest products, consignment note, and forest order [52], trees with a defoliated crown over 60% should be marked for salvage felling, so 3A and 3B trees were treated as D trees in calculations. Following that, assortment tables for thinning were used to calculate the expected assortment structures of H trees, and assortment tables for dead trees were used to calculate the expected assortment structures of 3A, 3B, and D trees. Based on the calculated net volume and »Croatian Forests« Ltd. price list (fco. forest road) [51], the expected average tree value was calculated.

2.3.3. Data Analysis

All the statistical analyses were performed with TIBCO Statistica 14.0.0.15 software. To determine the influence of tree height and tree diameter on the determined gross and net tree volume across the sites and TCDD classes multiple linear regression analyses were performed. To determine the influence of the TCDD class on the average tree value, the percentage deviation (Δ_EUR/m³, %) between the average tree value of defoliated trees and the average tree value of H trees was calculated in DBH classes 17.5–27.5 for site A and in DBH classes 32.5–42.5 for site B. In the same DBH classes, percentage deviation was calculated between the determined average tree value (of each TCDD class) and the expected average tree value. Additionally, a t-test was used to compare differences between expected and determined gross and net volume as well as between expected and determined average tree values. Descriptive statistics were used to present data of measured logs.

To determine the influence of wood defect presence on the timber grading, MS Excel 2412 software was used. Firstly, the dimensions of produced logs and minimum dimensions prescribed by the HRN standard for each quality class were used as criteria to count how many of the produced logs could potentially be classified, based on the dimensions, but were not classified into the highest quality class because of present wood defects. The percentage of volume with particular defects was analyzed, as well as the number of occurrences of wood defects in each TCDD class, and tested with analysis of variance, as it was tested in the previous study [17]. If statistically significant differences had occurred, additionally, Tukey’s honest significant differences test was performed. For all statistical tests, statistical significance was accepted at α < 0.05.

3. Results

On the two sites, 115 pedunculate oak trees, in different TCDD and DBH classes, were measured (

Table 3. Number of sampled trees regarding site and DBH and TCDD classes.

Site	A				B
TCDD	H	3A	3B	D	H	3A	3B	D
DBH, cm
17.5	4	4	4	4
22.5	4	4	4	4
27.5	4	4	4	4
32.5		1	1	2	1	2	2	2
37.5					3	5	1	3
42.5					4	5	1	5
47.5					4		3	8
52.5						3	1
57.5						2	2	6
Total	12	13	13	14	12	17	10	24

TCDD—tree crown defoliation degree; DBH—diameter class; H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

).

3.1. Gross and Net Volume

According to the multiple regression parameters, gross and both HRN and HRN EN net tree volume were influenced by tree diameter and tree height (

Table A1. Multiple regression parameters of determined tree gross volume across sites.

Site	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	Intercept			−0.895993	0.077101	−11.6211	<0.000001	0.9255
	DBH	0.877355	0.055110	0.054270	0.003409	15.9200	<0.000001
	TH	0.114946	0.055110	0.009371	0.004493	2.0857	0.042231
B	Intercept			−3.25230	0.348282	−9.33811	<0.000001	0.9219
	DBH	0.943940	0.036367	0.11528	0.004441	25.95616	<0.000001
	TH	0.092623	0.036367	0.02576	0.010116	2.54692	0.013451

SE—standard error; M—multiple; DBH—diameter at breast height (cm); TH—tree height (m).

,

Table A2. Multiple regression parameters of determined net volume (HRN standard) across sites.

Site	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	Intercept			−0.789219	0.064765	−12.1858	<0.000001	0.9176
	DBH	0.756087	0.057965	0.037352	0.002864	13.0439	<0.000001
	TH	0.260303	0.057965	0.016948	0.003774	4.4907	0.000043
B	Intercept			−2.36296	0.357909	−6.60214	<0.000001	0.8303
	DBH	0.873155	0.053623	0.07432	0.004564	16.28309	<0.000001
	TH	0.171874	0.053623	0.03332	0.010395	3.20521	0.002164

SE—standard error; M—multiple; DBH—diameter at breast height (cm); TH—tree height (m).

and

Table A3. Multiple regression parameters of determined net volume (HRN EN standard) across sites.

Site	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	Intercept			−0.703978	0.065842	−10.6919	<0.000001	0.9029
	DBH	0.767185	0.062901	0.035506	0.002911	12.1966	<0.000001
	TH	0.237788	0.062901	0.014504	0.003837	3.7803	0.000426
B	Intercept			−2.39716	0.365996	−6.54970	<0.000001	0.8277
	DBH	0.871981	0.054023	0.07533	0.004667	16.14091	<0.000001
	TH	0.171102	0.054023	0.03367	0.010630	3.16721	0.002420

SE—standard error; M—multiple; DBH—diameter at breast height (cm); TH—tree height (m).

). The detailed analysis showed that the determined tree gross and both HRN and HRN EN net volume depend on the tree diameter, regardless of the TCDD class (

Table A4. Multiple regression parameters of determined gross volume across different tree crown defoliation degrees.

Site	TCDD	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	H	Intercept			−0.944892	0.130021	−7.26720	0.000047	0.9497
		DBH	0.669850	0.141008	0.040324	0.008488	4.75044	0.001044
		TH	0.339549	0.141008	0.023256	0.009658	2.40801	0.039377
	3A	Intercept			−0.782452	0.214127	−3.65415	0.004432	0.9401
		DBH	1.000256	0.153995	0.063739	0.009813	6.49539	0.000069
		TH	−0.035699	0.153995	−0.003468	0.014960	−0.23182	0.821351
	3B	Intercept			−0.765915	0.137236	−5.58101	0.000234	0.9169
		DBH	0.753857	0.134787	0.040095	0.007169	5.59293	0.000230
		TH	0.255229	0.134787	0.016167	0.008538	1.89357	0.087546
	D	Intercept			−1.21219	0.116266	−10.4259	<0.000001	0.9816
		DBH	0.840599	0.059081	0.05547	0.003898	14.2279	<0.000001
		TH	0.195145	0.059081	0.02387	0.007227	3.3030	0.007040
B	H	Intercept			−2.01548	1.012736	−1.99014	0.077779	0.8877
		DBH	0.916219	0.183091	0.10037	0.020057	5.00418	0.000735
		TH	0.032522	0.183091	0.00825	0.046433	0.17763	0.862949
	3A	Intercept			−4.80159	0.741306	−6.47720	0.000015	0.9546
		DBH	0.896395	0.063435	0.12383	0.008763	14.13096	<0.000001
		TH	0.159142	0.063435	0.06325	0.025214	2.50875	0.025040
	3B	Intercept			−2.13664	0.623551	−3.42657	0.011036	0.9629
		DBH	0.957435	0.079469	0.09600	0.007968	12.04794	0.000006
		TH	0.056150	0.079469	0.01512	0.021394	0.70657	0.502670
	D	Intercept			−3.42132	0.623452	−5.48770	0.000019	0.9357
		DBH	0.962968	0.055930	0.12836	0.007455	17.21738	<0.000001
		TH	0.027300	0.055930	0.00977	0.020011	0.48811	0.630530

TCDD—tree crown defoliation degree; SE—standard error; M—multiple; DBH—diameter at breast height (cm); TH—tree height (m); H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

,

Table A5. Multiple regression parameters of determined net (HRN standard) volume across different tree crown defoliation degrees.

Site	TCDD	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	H	Intercept			−0.817610	0.159722	−5.11897	0.000629	0.9017
		DBH	0.603633	0.197155	0.031926	0.010427	3.06171	0.013534
		TH	0.382197	0.197155	0.022998	0.011864	1.93856	0.084504
	3A	Intercept			−0.604190	0.125079	−4.83047	0.000691	0.9672
		DBH	1.004185	0.113878	0.050546	0.005732	8.81808	0.000005
		TH	−0.024047	0.113878	−0.001845	0.008739	−0.21117	0.836997
	3B	Intercept			−0.657669	0.082275	−7.99355	0.000012	0.9531
		DBH	0.693333	0.101268	0.029425	0.004298	6.84648	0.000045
		TH	0.345560	0.101268	0.017466	0.005118	3.41232	0.006632
	D	Intercept			−1.00378	0.164221	−6.11240	0.000076	0.9378
		DBH	0.738492	0.108729	0.03740	0.005506	6.79206	0.000030
		TH	0.289376	0.108729	0.02717	0.010208	2.66145	0.022129
B	H	Intercept			−1.39031	0.903489	−1.53883	0.158230	0.8638
		DBH	0.945106	0.201665	0.08386	0.017893	4.68651	0.001142
		TH	−0.019907	0.201665	−0.00409	0.041424	−0.09871	0.923531
	3A	Intercept			−4.17589	0.821709	−5.08196	0.000167	0.8971
		DBH	0.790596	0.095502	0.08041	0.009714	8.27833	0.000001
		TH	0.278827	0.095502	0.08160	0.027948	2.91959	0.011200
	3B	Intercept			−0.318653	0.672547	−0.473800	0.650064	0.8955
		DBH	0.985283	0.133309	0.063521	0.008594	7.390951	0.000151
		TH	−0.111002	0.133309	−0.019213	0.023075	−0.832663	0.432522
	D	Intercept			−2.30243	0.585152	−3.93476	0.000759	0.8811
		DBH	0.931052	0.076054	0.08566	0.006997	12.24200	<0.000001
		TH	0.045096	0.076054	0.01114	0.018782	0.59295	0.559549

TCDD—tree crown defoliation degree; SE—standard error; M—multiple; DBH—diameter at breast height (cm); TH—tree height (m); H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

and

Table A6. Multiple regression parameters of determined net (HRN EN standard) volume across different tree crown defoliation degrees.

Site	TCDD	Parameter	b*	SE of b*	b	SE of b	t	p	M R2
A	H	Intercept			−0.715533	0.158848	−4.50450	0.001479	0.8835
		DBH	0.587337	0.214635	0.028378	0.010370	2.73645	0.022981
		TH	0.388856	0.214635	0.021376	0.011799	1.81171	0.103455
	3A	Intercept			−0.510605	0.125370	−4.07280	0.002240	0.9619
		DBH	1.031622	0.122840	0.048251	0.005745	8.39810	0.000008
		TH	−0.059387	0.122840	−0.004235	0.008759	−0.48345	0.639188
	3B	Intercept			−0.605704	0.086370	−7.01289	0.000037	0.9419
		DBH	0.685286	0.112691	0.027437	0.004512	6.08113	0.000119
		TH	0.347916	0.112691	0.016589	0.005373	3.08735	0.011496
	D	Intercept			−0.898389	0.175483	−5.11951	0.000334	0.9209
		DBH	0.753278	0.122603	0.036151	0.005884	6.14406	0.000073
		TH	0.261848	0.122603	0.023296	0.010908	2.13575	0.056018
B	H	Intercept			−1.32460	0.943279	−1.40425	0.193800	0.8582
		DBH	0.963338	0.205809	0.08744	0.018681	4.68074	0.001151
		TH	−0.047226	0.205809	−0.00992	0.043249	−0.22947	0.823636
	3A	Intercept			−4.30969	0.820168	−5.25464	0.000122	0.8991
		DBH	0.781018	0.094577	0.08007	0.009695	8.25805	0.000001
		TH	0.294198	0.094577	0.08678	0.027896	3.11069	0.007668
	3B	Intercept			−0.271563	0.652458	−0.41622	0.689717	0.9054
		DBH	0.995608	0.126851	0.065440	0.008338	7.84864	0.000103
		TH	−0.128318	0.126851	−0.022644	0.022385	−1.01156	0.345427
	D	Intercept			−2.33548	0.611021	−3.82226	0.000993	0.8747
		DBH	0.927839	0.078059	0.08685	0.007307	11.88635	<0.000001
		TH	0.044156	0.078059	0.01109	0.019612	0.56567	0.577612

TCDD—tree crown defoliation degree; SE—standard error; M–multiple; DBH—diameter at breast height (cm); TH—tree height (m); H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

). Additionally, the tree gross volume of the H and D trees on site A, as well as the tree gross volume of the 3A trees on site B, was influenced by tree height (Table A4).

The total determined gross volume according to the HRN standard on site A was 30.76 m³, which is a 5.45% higher gross volume overall compared to the expected gross volume of 29.17 m³. On site B, determined gross volume (164.99 m³) according to the HRN standard overall was 6.16% higher than the expected gross volume (155.41 m³). The parameters of the t-test did not show a statistically significant difference between expected and determined gross volume (

Table 4. t-test parameters comparing determined and expected gross volume (m³).

Site	Mean D	Mean E	t-Value	df	p	N D	N T	SD D	SD T	F-Ratio Variances	p Variances
A	0.5915	0.5610	0.5562	102	0.5793	52	52	0.2889	0.2703	1.1425	0.6362
B	2.6189	2.4669	0.8828	124	0.3791	63	63	0.9510	0.9823	1.0668	0.7998

D—determined gross volume according to the HRN standard; E—expected gross volume by tariff series; SD—standard deviation.

).

On site A, the total measured net HRN volume was 25.29 m³, which is 0.26% higher than the expected net volume. On site B, the measured net HRN volume was 4.59% lower than the expected net volume of 127.96 m³. Despite the mentioned differences in expected and determined net HRN volume, the parameters of the t-test did not show a statistically significant difference between expected and determined net volume (

Table 5. t-test parameters comparing determined and expected net volume (m³).

Site	Mean D	Mean E	t-Value	df	p	N D	N T	SD D	SD T	F-Ratio Variances	p Variances
A	0.4863	0.4285	1.3681	102	0.1743	52	52	0.2307	0.1991	1.3427	0.2959
B	1.9378	2.0312	−0.7153	124	0.4757	63	63	0.6628	0.7962	1.4429	0.1516

D—determined net volume according to the HRN standard; E—expected net volume; SD—standard deviation.

).

3.2. Round Timber Features

Descriptive statistics of measured logs regarding the quality class and TCDD class are shown in Appendix A 

Table A7. Descriptive statistics of measured logs on site A.

TCDD		H		3A		3B			D
QC		SLII	FW	SLII	FW	SLII	SLIII	FW	SLI	SLII	SLIII	FW
N		1	56	4	55	2	1	37	1	5	1	36
Diameter, cm	Average	26.0	15.6	25.5	16.8	26.0	26.0	18.4	30.0	26.8	25.0	18.7
	SD	0.0	5.8	1.0	5.6	1.4	0.0	3.8	0.0	0.4	0.0	4.3
	Min.	26.0	7.0	25.0	7.0	25.0	26.0	12.0	30.0	26.0	25.0	13.0
	Max.	26.0	26.0	27.0	28.0	27.0	26.0	26.0	30.0	27.0	25.0	26.0
Length, m	Average	5.1	4.5	3.1	4.7	3.6	4.7	4.9	2.0	5.0	3.4	5.0
	SD	0.0	1.3	0.4	1.1	0.6	0.0	1.3	0.0	1.2	0.0	1.0
	Min.	5.1	2.0	2.5	2.1	3.1	4.7	2.3	2.0	3.3	3.4	3.0
	Max.	5.1	7.6	3.3	7.1	4.0	4.7	7.2	2.0	6.7	3.4	6.6
Gross volume, m3	Average	0.352	0.109	0.212	0.123	0.268	0.307	0.142	0.195	0.352	0.206	0.159
	SD	0.000	0.074	0.038	0.073	0.080	0.000	0.054	0.000	0.086	0.000	0.086
	Min.	0.352	0.010	0.175	0.014	0.212	0.307	0.035	0.195	0.233	0.206	0.043
	Max.	0.352	0.257	0.264	0.275	0.325	0.307	0.248	0.195	0.470	0.206	0.345
Net HRN volume,	Average	0.271	0.100	0.158	0.114	0.191	0.249	0.130	0.141	0.283	0.167	0.148
	SD	0.000	0.069	0.027	0.070	0.054	0.000	0.050	0.000	0.076	0.000	0.083
	Min.	0.271	0.009	0.123	0.012	0.152	0.249	0.033	0.141	0.175	0.167	0.040
	Max.	0.271	0.232	0.189	0.265	0.229	0.249	0.233	0.141	0.383	0.167	0.324
Value, EUR	Average	29.92	3.42	17.43	3.89	21.07	19.03	4.42	23.29	31.33	12.73	5.05
	SD	0.00	2.35	3.00	2.38	6.00	0.00	1.71	0.00	8.38	0.00	2.82
	Min.	29.92	0.29	13.57	0.41	16.82	19.03	1.13	23.29	19.36	12.73	1.36
	Max.	29.92	7.89	20.87	9.02	25.31	19.03	7.92	23.29	42.39	12.73	11.03

TCDD—tree crown defoliation degree; H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown; SL—saw log; FW—fuelwood; N—number of measured logs.

for site A and in

Table A8. Descriptive statistics of measured logs on site B.

TCDD		H				3A
QC		SLI	SLII	SLIII	FW	VLI	SLI	SLII	SLIII		FW
N		9	10	19	167	4	5	15	28		203
Diameter, cm	Average	37.4	31.1	32.2	13.3	46.5	35.6	31.3	32.6		13.7
	SD	2.9	3.4	4.9	7.5	4.1	1.5	2.6	6.2		7.8
	Min.	33.0	27.0	26.0	7.0	43.0	33.0	28.0	25.0		7.0
	Max.	41.0	37.0	46.0	41.0	51.0	37.0	37.0	45.0		42.0
Length, m	Average	3.2	3.8	4.4	2.1	4.5	4.5	3.7	4.4		1.9
	SD	1.0	1.4	1.7	1.5	1.6	1.3	0.8	1.3		1.5
	Min.	2.0	2.0	2.0	1.0	3.0	3.0	2.8	2.0		1.0
	Max.	5.2	5.6	7.0	6.3	6.6	5.6	5.8	6.8		7.3
Gross volume, m3	Average	0.470	0.362	0.443	0.064	0.945	0.577	0.378	0.497		0.066
	SD	0.168	0.078	0.131	0.107	0.205	0.113	0.079	0.253		0.117
	Min.	0.260	0.234	0.211	0.004	0.720	0.437	0.279	0.174		0.004
	Max.	0.847	0.480	0.655	0.584	1.218	0.692	0.558	1.064		0.684
Net HRN volume, m3	Average	0.358	0.277	0.342	0.060	0.737	0.441	0.284	0.383		0.062
	SD	0.130	0.059	0.105	0.102	0.163	0.110	0.061	0.202		0.112
	Min.	0.197	0.182	0.161	0.004	0.565	0.305	0.211	0.133		0.004
	Max.	0.653	0.370	0.508	0.567	0.958	0.549	0.438	0.858		0.679
Value, EUR	Average	68.82	30.58	26.38	2.08	285.31	72.73	31.43	30.47		2.16
	SD	32.35	6.54	7.92	3.45	77.08	18.20	6.72	17.63		3.79
	Min.	32.41	20.07	12.27	0.16	196.61	50.31	23.35	10.11		0.16
	Max.	141.46	40.87	38.79	19.32	364.20	90.56	48.37	72.80		23.10
TCDD		3B					D
QC		VLII	SLI	SLII	SLIII	FW	VLII	SLI	SLII	SLIII	FW
N		1	5	6	12	98	1	7	12	30	152
Diameter, cm	Average	43.0	40.2	38.0	35.1	14.6	53.0	39.9	39.5	36.4	20.3
	SD	0.0	7.5	3.3	7.1	8.2	0.0	4.6	6.6	6.4	11.2
	Min.	43.0	31.0	33.0	27.0	7.0	53.0	33.0	29.0	26.0	7.0
	Max.	43.0	51.0	42.0	47.0	43.0	53.0	45.0	49.0	53.0	49.0
Length, m	Average	3.4	4.9	4.2	4.6	2.1	3.3	4.0	4.2	3.8	2.6
	SD	0.0	0.7	1.2	1.4	1.7	0.0	1.1	1.1	1.4	1.8
	Min.	3.4	4.1	2.8	2.0	1.0	3.3	3.0	3.2	2.0	1.0
	Max.	3.4	5.8	5.5	7.0	8.3	3.3	6.2	6.9	8.0	7.6
Gross volume, m3	Average	0.662	0.804	0.597	0.573	0.084	0.910	0.641	0.673	0.534	0.170
	SD	0.000	0.249	0.125	0.271	0.136	0.000	0.139	0.234	0.291	0.197
	Min.	0.662	0.495	0.452	0.325	0.004	0.910	0.467	0.343	0.220	0.004
	Max.	0.662	1.065	0.757	1.246	0.598	0.910	0.888	1.064	1.326	0.852
Net HRN volume, m3	Average	0.494	0.627	0.464	0.445	0.079	0.728	0.486	0.512	0.417	0.161
	SD	0.000	0.205	0.098	0.223	0.131	0.000	0.104	0.184	0.232	0.188
	Min.	0.494	0.377	0.334	0.240	0.004	0.728	0.368	0.258	0.169	0.004
	Max.	0.494	0.842	0.591	1.016	0.581	0.728	0.668	0.829	1.058	0.842
Value, EUR	Average	147.18	138.81	56.15	35.33	2.74	286.66	98.37	66.68	33.76	5.52
	SD	0.00	73.70	13.18	19.21	4.43	0.00	25.62	29.32	22.02	6.36
	Min.	147.18	62.18	36.96	18.32	0.16	286.66	60.60	28.47	12.88	0.16
	Max.	147.18	243.65	75.36	86.17	19.77	286.66	144.60	118.76	109.29	28.66

TCDD—tree crown defoliation degree; H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown; QC—quality class; VL—veneer log; SL—saw log; FW—fuelwood; N—number of measured logs.

for site B. On site A, 199 logs were measured in the following quality classes: SLI—1 log; SLII—12 logs; SLIII—2 logs; and FW—184 logs. The high share of FW results from the small tree diameter, because only 36% of measured trees could have an assortment of the higher-quality class, e.g., SLIII or SLII. On site B, 4 logs of VLI, 2 logs of VLII, 26 logs of SLI, 43 logs of SLII, and 89 logs of SLIII were produced. On site B, except for long fuelwood (length above 2.0 m), one-meter-long fuelwood was also produced, and in total, 620 pieces of fuelwood were produced.

For the total number of present wood defects, the highest share had rot (30.1%) followed by covered knots (24.5%), insect holes (20.9%), unsound knots (9.4%), sweep (7.6%), sound knots (2.8%), ring shake (2.3%), eccentric pith (1.0%), sapwood rot (0.8%), double pith (0.3%), and cracks (0.3%). The influence of wood defects on the log declassification is shown in

Table 6. Influence of wood defect presence (%) on the assortment declassification.

QC	N	Rot	Sapwood Rot	Insect Holes > 3.00 mm	Insect Holes < 3.00 mm	Covered Knots	Unsound Knots	Sound Knots	Sweep	Ring Shake	Eccentric Pith	Cracks	Double Pith
Declassified VL logs
SLI	12	7.1	100.0	15.4	33.3	35.7	-	50.0	20.0	-	100.0	-	-
SLII	8	14.3	-	19.2	33.3	7.1	14.3	-	40.0	-	-	-	-
SLIII	17	42.9	-	34.6	33.3	50.0	42.9	50.0	20.0	100.0	-	100.0	-
FW	12	35.7	-	30.8	-	7.1	42.9	-	20.0	-	-	-	-
Total *	49	16.7	1.2	31.0	10.7	16.7	8.3	2.4	6.0	4.8	1.2	1.2	-
Declassified SLI logs
SLII	27	4.5	50.0	6.7	66.7	32.7	5.3	71.4	50.0	40.0	100	-	-
SLIII	44	27.3	50.0	43.3	33.3	50.0	26.3	28.6	20.0	40.0	-	-	-
FW	53	68.2	-	50.0	-	17.3	68.4	-	30.0	20.0	-	-	-
Total *	124	25.1	1.1	17.1	1.7	29.7	10.9	4.0	5.7	2.9	1.7	-	-
Declassified SLII and SLIII logs
SLIII	30	13.3	-	15.4	-	66.7	45.5	-	40.0	-	-	-	100.0
FW	74	86.7	-	84.6	100.0	33.3	54.5	100.0	60.0	-	-	-	-
Total *	104	45.1	-	9.8	0.8	22.6	8.3	1.5	11.3	-	-	-	0.8

QC—quality class; N—number of declassified logs; VL—veneer log; SL—saw log; FW—fuelwood; * share of present wood defects in total number of present wood defects.

. The row »total« shows the percentage of wood defect presence in the total number of the wood defects that exclude specific quality class, while rows with quality classes show the percentage of wood defect presence out of a total number of each wood defect. In total, 49 VL, 124 SLI, and 104 SLII and SLIII were declassified because of the wood defect presence. VL were mostly declassified because of insect holes (41.7%), and knots (27.4%), while SLI were mostly declassified because of knots (44.6%) and rot (25.1%) presence. The analysis of SLII and SLIII declassification reasons was conducted together since these two classes have the same minimum dimensions. SLII and SLIII logs were mostly declassified because of rot (45.1%) and covered knots (22.6%).

The percentage of wood defect presence along TCDD classes is shown in

Table 7. Wood defect presence (%) along tree crown defoliation classes.

TCDD	Rot	Sapwood Rot	Insect Holes > 3.00 mm	Insect Holes < 3.00 mm	Covered Knots	Unsound Knots	Sound Knots	Sweep	Ring Shake	Eccentric Pith	Cracks	Double Pith
H	24.0	-	-	-	30.7	13.3	8.0	16.0	5.3	2.7	-	-
3A	22.8	1.1	5.4	-	40.2	8.7	4.3	9.8	4.3	2.2	-	1.1
3B	33.3	1.7	16.7	3.3	23.3	11.7	-	6.7	1.7	-	1.7	-
D	35.8	0.6	32.7	6.7	13.3	7.3	0.6	3.0	-	-	-	-

TCDD—tree crown defoliation degree; H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

. The wood defects that are influenced by growing conditions (sound knots, sweep, ring shake, and eccentric pith) are more common on H trees than on D trees, while wood defects related to biological degradation (rot and insect holes) are more common on D trees than on H, 3A, and 3B trees.

The share of the rot, insect holes, covered knots, and sound knots were statistically significantly different between TCDD classes regarding the number of defects occurring and volume with the particular wood defect (

Table 8. Analysis of variance for wood defect presence.

Wood Defect	p Value for the Number of Wood Defect Occurrence	p Value for the Volume with Wood Defect
Rot	0.0035	<0.0001
Sapwood rot	0.6870	0.7509
Insect holes > 3.00 mm	<0.0001	<0.0001
Insect holes < 3.00 mm	0.0125	0.0118
Covered knots	0.0002	0.0012
Unsound knots	0.4697	0.3576
Sound knots	0.0024	0.0381
Sweep	0.0040	0.0961
Ring shake	0.0554	0.0796
Eccentric pith	0.1873	0.2263
Cracks	0.1043	0.1043
Double pith	0.4335	0.4335

), while the sweep was statistically significantly different (p = 0.0040) between TCDD classes only by the number of wood defect occurrences. Results of the Tukey honest significant difference test (

Table A9. Results of Tukey honest significant difference for wood defect presence between tree crown defoliation degrees.

Homogenous Group by Number of Occurred Wood Defects					Homogenous Group by Wood Volume with Wood Defect
Rot
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.9945	0.2106	0.0467	H		0.9729	0.0527	0.0025
3A	0.9945		0.1040	0.0093	3A	0.9729		0.0112	0.0001
3B	0.2106	0.1040		0.9995	3B	0.0527	0.0112		0.9947
D	0.0467	0.0093	0.9995		D	0.0025	0.0001	0.9947
Insect holes > 3.00 mm
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.7635	0.0088	0.0000	H		0.9056	0.0200	0.0000
3A	0.7635		0.0670	0.0000	3A	0.9056		0.0687	0.0000
3B	0.0088	0.0670		0.0040	3B	0.0200	0.0687		0.0366
D	0.0000	0.0000	0.0040		D	0.0000	0.0000	0.0366
Insect holes < 3.00 mm
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.9852	0.6677	0.0248	H		0.9990	0.6964	0.0332
3A	0.9852		0.8132	0.0377	3A	0.9990		0.7358	0.0255
3B	0.6677	0.8132		0.6158	3B	0.6964	0.7358		0.6396
D	0.0248	0.0377	0.6158		D	0.0332	0.0255	0.6396
Covered knots
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.9023	0.7043	0.0092	H		0.7060	0.9945	0.0968
3A	0.9023		0.2972	0.0001	3A	0.7060		0.8981	0.0009
3B	0.7043	0.2972		0.3985	3B	0.9945	0.8981		0.0946
D	0.0092	0.0001	0.3985		D	0.0968	0.0009	0.0946
Sound knots
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.1779	0.0138	0.0022	H		0.4787	0.0739	0.0478
3A	0.1779		0.5229	0.4441	3A	0.4787		0.5836	0.6636
3B	0.0138	0.5229		0.9948	3B	0.0739	0.5836		0.9744
D	0.0022	0.4441	0.9948		D	0.0478	0.6636	0.9744
Sweep
TCDD	H	3A	3B	D	TCDD	H	3A	3B	D
H		0.1750	0.1826	0.0014	H	-	-	-	-
3A	0.1750		0.9904	0.3695	3A	-	-	-	-
3B	0.1826	0.9904		0.7598	3B	-	-	-	-
D	0.0014	0.3695	0.7598		D	-	-	-	-

TCDD—tree crown defoliation degree; H—tree with crown defoliation less than 25%; 3A—tree with crown defoliation 61%–80%; 3B—tree with crown defoliation 81%–99%; D—tree with 100% defoliated crown.

) indicate that D trees had a statistically significant different share of wood defects (except covered knots) compared to the H trees.

3.3. Assortment Structure

The determined assortment structure is shown in Appendix A. Figure A1 and Figure A2 for each TCDD class according to the HRN standard and Figure A3 and Figure A4 show determined assortment structure according to the HRN EN standard.

In the total measured gross volume on site A, FW (71.6%) had the highest share, followed by saw logs (10.6%) without the presence of veneer logs. Regarding the TCDD class in DBH classes 17.5–27.5, on site A, the highest share of SL was produced from D trees (10.5%), followed by 3A trees (4.1%), H trees (4.0%), and 3B trees (2.7%). The DBH of D trees influenced this quite illogical trend, where 100% of D trees had a DBH of 29.0 cm and the first log could satisfy minimum dimensions for the SLII (under bark diameter of 25.0 cm), while only 25% of H trees could satisfy minimum dimensions. In 3A and 3B trees, the share of SL was influenced by dimensions as well as the presence of wood defects. In contrast to the better assortment structure, D trees had the largest share of waste (21.8%) followed by 3B trees (19.1%), 3A trees (13.6%), and H trees (12.7%). In the total measured gross volume on site B, FW (33.27%) also had the highest share, followed by SLIII (21.27%), SLII (9.67%), SLI (7.25%), VLI (1.79%), and VLII (0.74%). The influence of the TCDD class on the assortment structure was analyzed in DBH classes 32.5–42.5 since all TCDD classes were present. The highest share of SL was present in 3A trees (49.56%), followed by H trees (46.26%), 3B trees (36.81%), and D trees (25.20%), which means that the quality of 3A trees in terms of assortment structure did not significantly decrease. At the same time, the highest share of waste was found for D trees (24.9%) followed by 3A trees (23.9%), H trees (19.2%), and 3B trees (19.1%).

3.4. Average Tree Value

The deviation in average tree value was tested by t-test, which did not show a statistically significant difference between the determined and expected average tree value (

Table 9. t-test parameters of average tree value.

Site	Mean D	Mean E	t-Value	df	p	N D	N E	SD D	SD E	F-Ratio Variances	p Variances
A	39.7742	38.0715	0.9021	102	0.3692	52	52	12.4153	5.5799	4.9506	<0.0001
B	81.4273	84.7133	−0.7630	124	0.4469	63	63	28.1577	19.3813	2.1107	0.0038

D—determined average tree value; E—expected average tree value; SD—standard deviation.

). Despite the statistical test not showing a statistically significant difference, additional analysis was performed to further analyze the reasons for this trend.

The influence of the TCDD class on the average tree value regarding the DBH class is shown in Figure 2a. On site A, the average tree value remains the same up to DBH class 22.5, regardless of the TCDD class. This trend was caused by the small diameter of the produced logs, and despite their quality, they were classified as FW. In DBH class 27.5, D trees had a higher average tree value of 24.6% compared to the H trees. The significantly higher average tree value of D trees is the result of the previously mentioned higher share of SL. On site B, D trees had a lower average tree value in all DBH classes; 3B trees showed a decrease in average tree value in DBH class 32.5 by 24.0%, while the increase in average tree value was determined in DBH class 37.5 by 14.0% and DBH class 42.5 by 23.6%; 3A trees showed an increased average tree value from a minor 1.9% in DBH class 37.5 to 20.5% in DBH class 42.5.

The deviation in the determined average tree value from the expected average tree value is shown in Figure 2b. In DBH classes 17.5 and 22.5, the determined and the expected average tree value was not different since the currently used assortment tables predict only the production of FW. The quite similar overestimation in the expected average tree value of H and 3B trees was in DBH class 27.5 by 7.3% and 7.6%, and in DBH class 32.5 by 1.5% and 1.6%. In DBH 32.5, 3A trees were significantly underestimated by the planned assortment structure by 33.5%. The overestimation of H and D trees and underestimation of 3A and 3B trees in DBH classes 37.5 and 42.5 show significant inconsistency in the expected average tree value.

4. Discussion

Accelerated climate change leads to changes in forest ecosystems in which trees develop, commonly with negative effects on the forest and the trees. Intensifying research related to the assessment of adverse impacts associated with climate change on forest ecosystems is supported by the Croatian climate change adaptation strategy [2]. The motivation for the research was based on the lack of knowledge about the relation between assortment quantity and quality as well as the presence of wood defects during different stages of pedunculate oak dieback in accelerated climate change, which more often requires forestry management adaptation to the new circumstances. Research was focused on the stand that is expected to be exposed to the new natural disturbances in the future (site A, age 62) and on the stand that has already been exposed to the impact of natural disturbances and climate change (site B, age 116). The results of the study provide valuable insight into the quality and quantity of logs that are produced from trees that were affected by dieback but also provide insight into timber grading influencing factors in the different dieback stages.

Generally, the deviation in expected and actual tree gross volume can always be expected. Deviation rate depends on the calculation method, which implies the number of input variables, and the calculation formulae or volume tables used. In Croatian forests, the calculation of expected gross volume using single-entry volume tables is defined by the Regulation on tree marking, classification of forest products, consignment notes, and forest order [52]. This calculation method resulted in an underestimated gross volume on the researched sites by 5.45% and 6.16%. In previous studies, according to Hubač 1982 [53], the deviation between expected and determined gross volume was also found to range from quite low (2.5%) in the beech stand to significantly high (45.78%) in the ash stand [43], while the deviation of ±10% is considered very good [54]. The reasons for the determined deviations could be found in the wrong prescribed single-entry volume or (which is more likely) because of the different measurement methods in single-entry volume table construction and the measurement method conducted in this study. Namely, single-entry volume tables were constructed by measuring 1.0–3.0 m long tree segments, while the determined gross volume in this study was based on the produced logs’ actual dimensions, which commonly had a length of over 3.0 m. Despite the research being conducted on only two sites, results indicate a higher deviation in gross volume for older stands, which was confirmed in a previous study where the average determined deviation in gross volume ranged from 8.8% for stands under 50 years old to 11.0% for stands over 50 years old [53].

The classification of produced logs into quality class is firstly influenced by dimensions and secondly by the presence of wood defects, and both are defined by the standard used. The tree crown defoliation degree is introduced in this study as an indicator of produced log features and an overall indicator of assortment structure. A previous study of oak dieback influence on assortment structure determined a 10.0% lower share of VL and SL and, at the same time, a 10.0% higher share of waste in trees with tree crown defoliated over 25% compared to the nondefoliated trees [55]. In the literature review of this paper, this was the only research related to the oak dieback and quality of produced assortments found, but without any analysis of present wood defects. The results of this study showed the influence of the TCDD class on the reduced net wood volume of the D trees (compared to the H trees) by 9.1% on site A and 5.8% on site B. The share of waste regarding the TCDD was in a logical order on site A (H trees < 3A trees < 3B trees < D trees), while the share of the waste on site B had an illogical trend in the following order: 3B trees < H trees < 3A trees < D trees. Quite an illogical trend on site B could be explained by different calculation methods of net volume for VL and SL compared to FW. Namely, the VL and SL net volume is calculated using the under-bark diameter, while FW is calculated using the over-bark diameter, which means higher net wood volume in case of a higher share of FW.

In general, the highest influence on log declassification was rot, which had the highest presence in logs produced from D trees. In contrast to the results of this study, rot share between three forest stands in Poland ranged from 1.0% to 12.0%, while the highest influence on declassification was knots (75.0%) [36]. On the other hand, the results of this study are comparable in the case of wood defect presence for the H trees, which also had the highest share of knots (52.0%), but 23.0% less than in the above-mentioned study, which could be a result of the different limits set by the standard used. The impact of biotic damage is shown in previous research where 45.7% of oak planks for structural use were rejected because of biotic damage [56]. Another study emphasizes the necessity of increasing the wood hardness produced from dead trees [28]. Despite the direct correlation between wood defect type and ratio along TCDD class not being noticed, it seems that the wood defects related to the growing conditions (e.g., knots) are more common on the H trees than on the D trees. In contrast to that, wood defects that are more related to biological activity (e.g., rot) are more common on dead trees than on live trees.

The last part of the study results showed the absence of a statistically significant difference between the expected and determined average tree value. Further, an additional analysis confirmed our speculations that currently used assortment tables cannot accurately predict average tree value along tree crown defoliation degree classes. Namely, on site B, the expected average tree value of 3A and 3B trees was, in most cases, underestimated, while the expected average tree value of H and D trees was, in all cases, overestimated, which probably resulted in no statistically significant difference since the over- and underestimation were almost mirrored. On the other hand, 3A and 3B trees showed quite illogical trends because, in some cases, they had higher average tree value than H trees. This trend led us to consider decreasing the crown defoliation percentage of H trees as it was in previous studies, e.g., on level of 10% [28] or 5% as it was in some other studies [57].

Since the tree crown defoliation degree is commonly used as an indicator of tree vigor and health, overall results and knowledge derived from this study strongly support the idea of creating models for assortment quality prediction (e.g., assortment tables) based on the tree crown defoliation degree among other variables. Although the study results did not show a consistent correlation between tree crown defoliation degree and average tree value on the research sites, determined discrepancies between expected and determined assortment structure impose the necessity for additional measurements. Further research should be aimed at expanding the limited site conditions of this study, e.g., different stand ages, phytocenosis, management systems, etc. Research related to assortment structure determination is often time-consuming and requires high costs. Thus, automatization of data collection should be considered. Namely, in »Croatian forest« Ltd., data collected (under-bark diameter, length and quality class) during roundwood grading are digitally recorded but they are not directly connected to attributes of trees that logs are produced from. As the tree crown defoliation degree and DBH are already measured during the tree marking process, adding a label on the marked trees would provide an opportunity to collect and pair a large amount of data regarding the quality of produced logs from trees with different tree crown defoliation degrees.

5. Conclusions

The results of the study conducted lead us to the following conclusions:

(a)

The literature review indicates a knowledge gap about tree crown defoliation degree influence on wood quality;

(b)

Currently used single-entry volume tables can predict harvesting volume in pedunculate oak stands with satisfactory accuracy;

(c)

The ratio of present wood defects is significantly influenced by tree crown defoliation degree class;

(d)

Variability in average tree value along tree crown defoliation degree classes indicates the necessity of additional research to clearly define relationships between tree crown defoliation degree class and average tree value;

(e)

Currently used assortment tables in Croatian forestry should be revised for healthy and dead trees and expanded in order to include the tree crown defoliation degree class as an additional variable in harvesting plan preparation.