2.1. Study Area
This study was conducted in the Hyrcanian forests of Iran (
Figure 1), a region covering 1.8 million hectares with high environmental, ecological, and economic value. These forests stretch from the southern shores of the Caspian Sea to elevations up to 2700 m above sea level on the northern slopes of the Alborz mountain range. They are characterized by a heterogeneous structure and diverse tree species, notably beech, which constitutes the largest area and volume of standing timber.
The study area is located in the Nav-Asalem forest, at coordinates 37°38′ N and 48°48′ E, with elevations between 700 and 1200 m above sea level (
Table 1). Annual precipitation averages 963 mm, and the mean annual temperature is 8.8 °C. Historically, shelterwood management was practiced until 1990, after which it transitioned to selection cutting, which continued until 2017—the year a logging ban was implemented. Tree harvesting was conducted using chainsaws, and timber extraction relied on ground-based skidders equipped with winches. In the common shelterwood cutting system in the Hyrcanian forests of Iran, the various treatments consist of preparatory cutting (an optional initial treatment to increase tree vigor and seed production within a mature stand), regeneration cutting (a treatment to establish regeneration throughout the stand area) and light felling carried out before final cutting. The main objectives of shelterwood cutting are to develop seed trees by natural regeneration and encourage desirable species to achieve their maximum diameter growth. In forest management, the shelterwood cutting system involves a series of partial harvests conducted over several years, ultimately resulting in the establishment of a new even-aged stand. However, due to its negative impacts on silvicultural properties, biological diversity, and aesthetic values, shelterwood cutting has been largely replaced by selection cutting in these hardwood forests. In most Hyrcanian forests, a single-tree selection system has been adopted as a close-to-nature management approach, with interventions occurring every 10 years. This method aims to maintain uneven-aged tree populations, mimicking natural gap dynamics processes. The primary goal of selection-cutting management is to create mixed and uneven-aged stands that closely resemble natural forest ecosystems.
This study evaluated three forest management approaches: shelterwood, selection cutting, and protection, the latter serving as a control. For this purpose, similar stands in terms of plant community and soil type and almost adjacent were selected (
Figure 1). The selected stands all had an uneven-aged structure and the dominant stand type of fagetum mixed with other broadleaves. The number of stands selected in shelterwood cutting, selection cutting, and protected management was 3, 5, and 4 stands, respectively. The area of each stand varied from 15 to 20 hectares. The choice of these experimental blocks was made through a design-based approach, which is a statistical approach that establishes the methods of choice and use of the sites, allowing possible pseudoreplication problems to be overcome [
19].
Table 1 provides additional details on the harvesting intensity and historical management practices within the study area.
2.2. Sampling Design and Data Collection
This study assessed the impact of three silvicultural methods (shelterwood, selection cutting, and protection) on four ecosystem service components: habitat conservation, soil conservation, timber production, and carbon storage. Habitat conservation was evaluated using indices of stand structural complexity (SCI), species diversity of trees, shrubs, and herbaceous plants, frequency of large-diameter trees (LDT), and volume of deadwood (standing and fallen). Soil conservation was assessed through soil physical and chemical properties under different management practices. Timber production was quantified by measuring the standing timber volume and calculating its economic value, while carbon storage was estimated for both trees and soil.
Sampling followed a systematic plot design with random starting points. Data on tree and stand structure, as well as soil properties, were collected using a 100 × 100-m grid. Circular sample plots with a radius of 17.85 m (1000 m
2 in area) were established in each forest management type. Each parcel had between 45 and 55 plots, with a sampling intensity of approximately 9% (
Table 2). Within each plot, three subplots (5 × 5 m) and one quadrat (1 × 1 m) were systematically positioned to assess vegetation characteristics across tree, shrub, and herb layers.
Tree diameters (≥7.5 cm at breast height) and heights were measured within plots, with tree volume estimated using species-specific volume tables. Large-diameter trees (LDT), defined as those with a diameter ≥60 cm, were recorded separately. Deadwood (DW) volume was calculated using Huber’s formula, with both standing and fallen deadwood measured within the plots. Canopy cover was estimated as the proportion of sample points covered by the forest canopy. Species richness, cover, and abundance were recorded for each layer.
2.3. Measurement of Ecosystem Services
The species diversity of trees and understory, and the stand structural complexity index (SCI) were calculated by Equations (1)–(10).
Tree size diversity index (TDD)
The tree diameter diversity was obtained by Equation (1) [
20]:
where
n is number of diameter classes, P
i is proportion of individual tree in
ith diameter class.
Tree height diversity index (THD)
The tree height diversity was obtained by Equation (2) [
20]:
where n is number of height classes, P
i is proportion of individual tree in
ith height class.
Canopy tree species richness index (TSRC)
The canopy tree species richness was obtained by Equation (3) [
21]:
where n is number of species in canopy layer, P
i is proportion of basal area of
ith species.
Tree species diversity index (TSD)
The tree species diversity was obtained by Equation (4) [
22].
where TSD is tree species diversity index according to Shannon-Wiener index (H’), P
i is the ratio of the number of the
ith species to the overall number of species.
Tree species evenness index (TSE)
The tree species evenness was obtained by Equation (5) [
23].
where TSE is tree species evenness index according to the Pielou’s evenness index (J
sw), P
i is the ratio of the number of the
ith species to the overall number of species, and
S is the number of species.
Tree species richness index of trees (TSR), shrubs (ShSR), and herbs (HSR)
The tree species richness was obtained by Equation (6) [
24].
where
SR is species richness index according to the Margalef richness index (
R’),
S is the number of species, N is the total number of individuals of all species.
Stand structural complexity index (SCI)
The stand structural complexity index was obtained by Equation (7) [
25,
26].
where
SCI is stand structural complexity index,
TDD is tree diameter diversity,
THD is tree height diversity,
TSRC is species richness of canopy layer,
ShSR is species richness of shrub layer,
HSR is species richness of herb layer, and n refers to the number of structural attributes used in the index (
n = 5).
The economic value of tree boles was estimated based on their species and quality as described in
Table 3. The quality of standing tree boles was graded visually assessment based on the guidelines of the Iranian Natural Resources and Watershed Management Organization [
27] and the quality of the first 6 m of tree boles with a diameter at breast height of larger than 42.5 cm was as follows: Q1: smooth, sound, cylindrical, circular cross-section, no rot, no twisting, no thick branches, suitable for veneering; Q2: sound, no rot, non-cylindrical or non-circular cross-section, with twisting, with one thick branch; Q3: with some rot, with a number of thick branches, with twisting; Fuel wood: heavily rotted, heavily twisted, with a lot of thick branches.
To assess soil characteristics, two soil samples were collected from each plot: one from the center of the plot and the other randomly from the center of one of the quarter plots. Samples were taken from the upper 10 cm of soil using a steel cylinder with an inner diameter of 5 cm and a height of 10 cm.
The bulk density of the soil, which is the ratio of the dry weight of the soil to the volume of the soil sample (after drying the soil in the oven for 24 h at 105 °C), was calculated by Equation (8) [
28].
In Equation (8), BD is bulk density in g cm−3, WD is dry weight of soil in g, and VC is volume of cylinder of soil sample in cm3.
The total soil porosity was determined using the Equation (9) [
28].
In Equation (9), TP is the total porosity of the soil in percent, BD is the soil bulk density in g cm−3, and 2.65 is the density of soil particles (g cm−3) measured by a pycnometer on the same soil samples used to determine the bulk density.
The following characteristics were also determined: soil moisture content, using the weight method; soil organic carbon (
OC) using the Walkley and Black method [
29]; soil total nitrogen (
N) using the Kjeldahl method [
30]. The depth of litter was measured using a metal ruler to an accuracy of 1 mm. The C-stock of soil was determined using the Equation (10) [
30].
where
SCS is the soil carbon stock (Mg ha
−1),
BD is the soil bulk density (g cm
−3),
OC is the soil organic carbon (%), and e is the soil depth (m).
Allometric models were used to estimate the branch and leaf biomass of beech and hornbeam trees, as detailed in
Table 4 [
31]. Since the studied forests primarily consisted of pure or mixed beech stands, with beech accounting for over 85% of the total tree population, the bole, branch, and leaf biomass of other species was also estimated using allometric models developed for beech. Root biomass for all tree species was calculated as 20% of the total tree biomass. Carbon storage in tree components was determined by multiplying the biomass of each component by a conversion factor of 0.531 [
32].