The Relationship Between Biometric Features of Trees and the Intensity of Birch Sap Leakage in Various Forest Sites

Abstract

The use of non-wood forest products plays a significant role in sustainable development, especially in the context of regional development. One of the most important and promising raw materials is birch sap, which in European conditions is obtained mainly from silver birch (Betula pendula Roth). Research on the utility value of birch sap and the influence of a number of factors on its efficiency and quality has been carried out in many research centers, but so far, there are not many studies on the variability of such parameters as a function of time, taking into account the entire period of sap leakage. This research was carried out in birch stands of approximately 80 years in three forest site types: mixed coniferous forest, mixed broadleaved forest, and broadleaved forest. In each site, nine sample trees were selected using Hartig’s method. The daily and all-season sap yield obtained from individual trees was statistically characterized. The relationship between birch sap yield and select quantitative (tree height, absolute and relative crown length, and slenderness) and qualitative (forest site type, tree thickness class, and sap harvesting period) variables was examined. For the first time in the literature on the subject, there is a proposal to distinguish the phases of obtaining birch sap, which may bring new knowledge, both in relation to daily productivity and the quality of the sap. As a result, the smallest amount of sap was found in the initial leak phase, a slightly higher amount was found in the final phase, and the largest amount was found in the main phase. Regarding the forest site type and the interaction between the collection phase and forest site type, no statistically significant relationship with the average amount of obtained sap was found.

1. Introduction

Using non-wood forest products plays a significant role in sustainable development, especially in the context of regional development [1,2,3,4,5,6]. Searching for additional (non-timber) sources of income for forest management is one of the priorities listed by the European Union in the New Forest Strategy and an action indicated by social certification systems for sustainable forest management [7,8]. It is also important to establish cooperation with potential business partners from the food and pharmaceutical industries; this is also a postulate in numerous publications related to forest management [9]. In the above context, birch sap deserves special attention. In European conditions, it is obtained mainly from silver birch (Betula pendula Roth). The possibilities of its actual and potential use in practice, including in the food and cosmetics industries, are vast [3,10,11,12], and the profit achieved from its harvest may exceed the profit from the sale of birch timber in the same area [13]. Research on the utility value of birch sap and the influence of different factors on its leakage intensity and quality has been carried out in many research centers [3].

Kostroň [14] stated that the forest site influences the intensity of sap leakage and the quality and health condition of the tree stand; he observed the highest sap production in healthy, 30–40-year-old birch stands. Grochowski [15] observed that in moist habitats, the leakage of birch sap is more abundant, while in drier sites, the amount of sap is smaller—but with higher sugar content.

According to Kostroň [14], sap production depends on the average thickness of the trees. However, Osiak [16], in a 65-year-old tree stand, did not confirm the influence of tree thickness on the intensity of sap flow, pointing to its relationship with the height of the trees.

The exposure of the tree stand was also analyzed: according to Kostroň [14], the highest production of birch sap was recorded on gentle northern slopes and the lowest on southern ones. However, Kaczmarczyk [17] and Zajączkowska et al. [18] state that more sap can be obtained from trees growing on the edge of a forest or open space than from those growing inside a tree stand. Muszyński and Muszyński [19] also note that not only does the microclimate have an impact but also the sunlight reaching the tree trunks.

The period during which birch sap can be obtained varies. Kostroň [14] indicates that harvesting should begin in early spring, during the rapid melting of snow, cold nights, and warm days until the leaves develop. The leakage time is approximately 16–24 days, sometimes longer, and is influenced by climatic conditions. Grochowski [15] states that the sap leakage season of birch begins in mid-April and, in unfavorable climatic conditions, at the beginning of May. Regardless of the location, it lasts 17–24 days—sometimes up to 29 days. The efficiency of sap flow is the highest in the initial period of its harvest. About 3 weeks after sap leakage starts, the yield decreases radically [16]. The most significant leak of birch sap occurs 2–3 weeks after the sap begins to flow in the tree trunk [17]. In the first period of sap obtaining, its amount increases with the temperature, while in the second period, the amount of sap decreases with its increase [14]. This is also confirmed by Grochowski [15], who claims that during the first two-thirds of the harvesting season, the increase in temperature has a positive effect on sap yield, while in the last third, it has a negative effect.

Koroljak and Tomčuk [20] and Kostroň [14] report that the intensity of birch sap leakage varies throughout the day. Bilek et al. [9] found that the volume of sap obtained was not significantly influenced by the temperature on the day of the leak or the night immediately preceding the leak. However, they observed a relationship between the intensity of the leak and the temperature measured several dozen hours earlier. Moreover, they found an almost complete inhibition of birch sap leakage after a night frost.

Differences have been observed in the outflow of birch sap depending on the cardinal directions in which the drilling was performed and the number of holes in the trunk. The results are inconclusive, although in most cases, the highest production is recorded from holes on the north side of the tree [14,21,22]. Kostroň [14] and Teliszewskij [23] provide information on standards specifying the permissible number of holes in one tree: from one to three holes, depending on the DBH. With the increase in the number of holes, the total yield of sap from the trunk increases, but the leakage from individual holes decreases [15].

The chemical composition of birch sap is well known. Apart from water, which usually constitutes over 95% of its mass, the main components of tree sap are sugars, primarily glucose and fructose, and rarely saccharose [24]. The sugar content in birch sap varies; average values, depending on the region, range from 0.7 g/100 mL to 1.2 g/mL [24,25,26,27]. Birch sap is not rich in antioxidants [28]. The beneficial nutritional properties of tree sap are primarily due to minerals [29,30]. According to Janistyn [31], the chemical compound contents also vary depending on the cardinal direction.

Knowledge regarding the use of birch sap is extensive but needs to be completed. The variability of both quality parameters and leak intensity depends on many factors. The research conducted so far has not taken into account their full complexity.

This paper aims to present the relationship between select biometric features of trees and birch sap leakage in different forest sites, considering the entire period of sap leakage.

The following hypotheses were put forward:

–

There is a relationship between the biometric features of trees (such as diameter at breast height, total height, crown base, crown length, relative crown length, and slenderness) and the intensity of birch sap leakage;

–

There is a relationship between the forest site type and the intensity of birch sap leakage;

–

It is possible to optimize the selection of the date and period of birch sap collection in terms of sap leakage efficiency.

2. Materials and Methods

This research was conducted in the Nurzec Forest District (Regional Directorate of State Forests in Białystok) from 14 March to 12 April 2019. The empirical material was obtained from trees growing in the area of 3 forest units with different forest site types. These were mixed coniferous forest (unit: 50 A-f), mixed broadleaved forest (unit: 58 d), and broadleaved forest (unit: 12 d). The ages of the stands were similar and amounted to 78 years, 74 years, and 88 years, respectively.

Hartig’s method was used to designate sample trees representing the stand (in each site, 9 trees). This method is based on taking trees ordered by increasing DBH values and categorizing them into three classes in the same cross-sectional area [32]. Then, three sample trees of the average DBH value from each class were selected: “thin”, “medium”, and “thick” (hereinafter referred to as THN1-3, MED1-3, and TCK1-3). For the mixed coniferous forest site, these values were 30.5 cm, 35.5 cm, and 42.5 cm; for mixed broadleaved forest, they were 23.5 cm, 31.5 cm, and 40.5 cm; and for broadleaved forest, they were 27.0 cm, 34.5 cm, and 42.5 cm.

Then, measurements of the sample trees’ heights and crown heights were taken using a SUNTO PM-5/1520 height gauge (Suunto Oy, Vantaa, Finland). A 50 mm deep, 12 mm diameter hole was made in the trunk of each tree on the north side, at a height of 100 cm, which was then cleaned and rinsed with distilled water. A tube with an internal diameter of 10 mm and an external diameter of 12 mm, made of plastic, approved for contact with food, was placed tightly in each hole. Before installation, each tube was disinfected with 70% ethyl alcohol and washed twice with distilled water. Birch sap was collected in airtight containers of 10 L to 30 L, made of plastic, approved for contact with food (Figure 1). The containers were disinfected similarly to the tubes. The amount of leaking birch sap was measured over 29 days, starting from 15 March 2019. Sap was taken every day at the same time—from each tree and in each habitat—rounded to 10 mL. The containers and tubes were washed with 70% alcohol and distilled water after each measurement.

The sap collection phase has not been precisely defined in the literature so far. A definition has been proposed that includes the following:

Phase 1

—the period when the leakage intensity is below average (initial);

Phase 2

—the period when the leakage intensity is above average (main);

Phase 3

—the period when the leakage intensity is again below average (final).

The amount of birch sap obtained, both daily and seasonally, was characterized statistically, using descriptive statistics measures. The efficiency of sap extracted from individual trees was compared using multifactorial analysis of variance, making the average value of this feature dependent on qualitative variables, which included forest site type, sap collection phase, and tree thickness class in the stand. Hypotheses regarding the influence of these factors and their interactions in the average daily sap efficiency were verified.

A linear model with fixed and random effects was used to model the relationship between the amount of sap obtained and the biometric features of trees and site characteristics. Among the biometric features, the modeling included the following as explanatory variables: tree diameter category within the stand, total height, crown base, crown length, relative crown length, and slenderness. The site characteristic was the forest site type, which, within the classification used in Poland, considers the site’s humidity and fertility. The last variable used in the modeling was the feature related to the date of sap collection, which was assumed to be the sap collection phase. The daily volumes of obtained birch sap were modeled; these variables were explained in the constructed models. The maximum likelihood method, designated ML, was used to search for model parameters. Modeling of the volume of obtained birch sap was performed in the R environment using the lme4 library (Posit PBC, Boston, MA, USA). The AIC information criterion was used to select the best model from all variants with different combinations of explanatory variables. We checked whether the empirical data met the assumptions for building linear mixed models with fixed and random effects. Among others, the following were checked: homogeneity of variance, linearity, collinearity, normality of residuals, and normality of random effects. These analyses were performed in the R environment using the performance library. The sjPlot library was used to visualize and predict the model.

3. Results

3.1. Characteristics of the Amount of Sap Obtained from Individual Trees

3.1.1. Daily Efficiency of Sap Leak

Trees growing in the mixed coniferous forest site were characterized by an average daily birch sap leakage from 3.05 dm³/24 h (MED2) to 11.50 dm³/24 h (MED1) (

Table 1. Summary statistics for daily sap leak volume (dm³/day).

Tree	Mean	Minimum	Maximum	Standard Deviation	Coefficient of Variation
Site: Mixed Coniferous Forest
THN1	6.45	0.12	13.52	4.06	62.9%
THN2	3.49	0.05	6.92	1.80	51.7%
THN3	10.88	1.95	17.15	4.77	43.9%
MED1	11.50	0.07	19.21	5.80	50.4%
MED2	3.05	0.00	7.51	2.14	70.1%
MED3	8.38	0.40	16.24	4.77	56.9%
TCK1	3.80	0.00	11.83	3.71	97.8%
TCK2	3.84	0.80	9.88	3.05	79.5%
TCK3	3.46	0.90	9.21	2.40	69.5%
Site: Mixed Broadleaved Forest
THN1	0.78	0.00	2.91	0.85	109.2%
THN2	4.39	2.30	6.17	1.18	26.9%
THN3	6.05	0.33	9.76	2.55	42.1%
MED1	3.19	0.40	7.39	2.06	64.7%
MED2	6.42	2.71	9.56	1.99	31.0%
MED3	2.83	0.96	4.77	1.08	38.1%
TCK1	3.63	0.53	9.02	2.31	63.6%
TCK2	12.43	1.26	22.50	6.80	54.7%
TCK3	5.99	1.26	12.69	3.55	59.3%
Site: Broadleaved Forest
THN1	5.04	0.38	12.90	3.68	72.9%
THN2	6.36	3.47	9.83	1.71	26.9%
THN3	1.24	0.15	4.94	1.20	97.1%
MED1	4.07	1.36	8.69	2.00	49.2%
MED2	2.15	0.48	5.25	1.48	69.0%
MED3	8.99	1.74	16.61	4.29	47.8%
TCK1	5.87	0.57	13.56	3.67	62.4%
TCK2	2.43	0.70	4.98	1.06	43.6%
TCK3	2.53	0.43	7.39	2.05	81.1%

). The lowest recorded amount of sap was 0.00 dm³/24 h for trees MED 2 and TCK1, which resulted from the later date of the start of leakage by these trees. The most considerable daily amount of sap obtained from one tree at this site was 19.2 dm³/24 h and was observed in the group of medium trees (MED1). The variability of the amount of sap obtained on individual days expressed as a standard deviation ranged from 1.80 dm³/24 h (THN2) to 5.80 dm³/24 h (MED1). The coefficient of variation of daily sap yield was very high and shows values ranging from 43.9% (THN3) to 97.8% (TCK1).

The average daily birch sap leakage from trees growing in the mixed broadleaved forest site ranged from 0.78 dm³/24 h (tree: THN1) to 12.43 dm³/24 h (TCK2) (Table 1). The minimum recorded leakage was 0.00 dm³/24 h (THN1). This was due to the earlier termination of leakage from this tree compared to the others. The maximum leakage was obtained from tree TCK2—22.50 dm³/24 h. The variability of the amount of sap obtained on individual days, expressed by the standard deviation, ranged from 0.85 dm³/24 h (THN1) to 6.80 dm³/24 h (TCK2). The coefficient of variation of daily birch sap yield was the lowest for tree THN2 (26.9%) and the highest for tree THN1 (109.2%).

The average sap leakage from trees growing in the broadleaved forest site ranged from 1.24 dm³/24 h (THN3) to 8.99 dm³/24 h (MED3) (Table 1). The lowest daily leakage was recorded for tree THN3—0.15 dm³/24 h—and the maximum for MED3—16.61 dm³/24 h. The standard deviation of the daily birch sap yield ranged from 1.06 dm³/24 h (TCK2) to 4.29 dm³/24 h (MED3). The values of the coefficient of variation of the daily birch sap yield ranged from 26.9% (THN2) to 97.1% (THN3). The course of changes in the daily birch sap yield during the subsequent days of the season is presented in Figure 2, where a sizeable individual variability is noted.

3.1.2. Seasonal Birch Sap Leak Efficiency

In the mixed coniferous forest site, during the entire study period, the total amount of sap obtained was the highest for trees THN3 (315.58 dm³) and MED1 (333.38 dm³). The lowest seasonal efficiency was characteristic of tree MED2 (88.58 dm³). The average value was 176.76 dm³, the highest value for the compared habitats (

Table 2. Summary statistics for all-season harvested sap volume (dm³/tree).

Site	Number of Trees	Average [dm3]	Minimum [dm3]	Maximum [dm3]	Standard Deviation [dm3]	Coefficient of Variation [%]
Mixed coniferous forest	9	176.76	88.58	333.38	97.70	55.27
Mixed broadleaved forest	9	147.29	22.69	360.61	95.81	65.05
Broadleaved forest	9	124.64	35.92	260.68	72.46	58.14

Two-factor analysis of variance (Table 2) did not show any significant effect of the site (F = 0.753, p = 0.485), tree diameter class (F = 0.151, p = 0.861), or interaction between these factors (F = 0.309, p = 0.309) on seasonal birch sap yield.

).

In the mixed broadleaved forest site, the most sap was extracted from tree TCK2—360.61 dm³—and the least from tree THN1—22.69 dm³. The average amount of harvested sap was 147.29 dm³. In the case of the broadleaved forest site, the highest total amount of leaking sap was found for tree MED3—260.68 dm³—and the lowest for tree THN3—35.92 dm³. The average amount of sap harvested from one tree in the season (124,64 dm³) was the lowest of all compared habitats (Table 2). A considerable variation in the total amount of harvested sap was observed between trees growing in the exact location. The highest variability was found in the mixed broadleaved forest site, where the coefficient of variation was 65.05% (Table 2), and the lowest was found in the mixed coniferous forest site (55.27%). From all trees growing in the individual habitats, the following totals were collected (a total for nine trees): mixed coniferous forest—1590.87 dm³; mixed broadleaved forest—1325.59 dm³; broadleaved forest—1121.72 dm³. From all 27 trees in the three distinguished habitats, a total of 4038.17 dm³ of sap was collected.

3.2. The Relationship Between Biometric Features of Trees and the Amount of Birch Sap Leakage

The relationship between seasonal birch sap yield and the selected biometric features of the trees was examined: diameter at breast height, total height, crown base, crown length, relative crown length, and slenderness (

Table 3. Spearman’s rank correlation coefficient between biometric features of trees and birch sap yield.

Tree Parameter	Site			Total
	Mixed Coniferous Forest	Mixed Broadleaved Forest	Broadleaved Forest
Diameter at breast height	−0.32	0.32	−0.05	−0.05
Total height	0.38	0.44	−0.50	−0.13
Crown base	0.49	0.44	−0.21	0.05
Crown length	−0.12	0.10	−0.31	−0.21
Relative crown length	−0.32	−0.15	−0.17	−0.18
Slenderness	0.55	−0.17	−0.13	−0.01

). The Spearman rank correlation coefficient was not statistically significant for any of the features mentioned, both for all trees together and divided into forest site type groups. The highest absolute values of this coefficient (around 0.50) were observed for the relationship with total height and crown base, although even in this case, the nature of the relationship was not constant in all the studied sites; a positive value of the coefficient was found in the mixed coniferous forest and mixed broadleaved forest sites and a negative one in the broadleaved forest site.

3.3. Average Daily Sap Leakage by Phase

A considerable variation in the amount of birch sap leakage was observed in the individual harvesting stages. The smallest amount of sap was found in the initial stage, for which an average of 2.6 dm³ was obtained from one tree in the mixed coniferous forest and broadleaved forest sites, and an average of 3.2 dm³ was obtained in the mixed coniferous forest site (

Table 4. Summary statistics for daily sap leakage by phase (dm³/tree).

Phase	Number of Trees	Mean	Minimum	Maximum	Standard Deviation	Coefficient of Variation
Site: mixed coniferous forest
Initial	9	2.6	0.8	4.4	1.2	45.7
Main	9	8.0	4.2	14.9	4.2	53.3
Final	9	5.8	0.6	13.4	5.1	88.4
Site: mixed broadleaved forest
Initial	9	3.2	0.3	8.5	2.3	70.4
Main	9	6.3	1.3	14.8	3.8	60.6
Final	9	3.9	0.1	15.4	4.6	118.2
Site: broadleaved forest
Initial	9	2.6	0.4	5.7	1.9	73.7
Main	9	5.3	1.9	12.0	3.4	63.7
Final	9	3.4	0.4	9.3	3.0	88.0

). A slightly higher sap yield was noted in the final phase, for which the average values ranged from 3.4 dm³ in the broadleaved forest site to 5.8 dm³ in the mixed coniferous forest site. On average, the largest amount of sap was obtained in the main phase, from 5.3 dm³ in the broadleaved forest site to 8.0 dm³ per day in trees representing the mixed coniferous forest site.

For individual trees, the extreme values of the average daily birch sap yield ranged from 0.1 dm³ to 15.4 dm³ (the final phase for tree THN1 in the mixed broadleaved forest site; Table 4). The trees showed large variations in the amount of birch sap harvested, even within the same site and phase. The smallest variation was observed in the initial phase, for which the coefficient of variation ranged from 45.7% in the mixed coniferous forest site to 73.7% in the broadleaved forest site. In the main phase, the values of this measure were the closest to each other in the individual habitats. They amounted to 53.3% in the mixed coniferous forest site, 60.6% in the mixed broadleaved forest site, and 63.7% in the broadleaved forest site. The most significant variation in the amount of sap harvested between trees in the site was observed in the final phase, for which the coefficient of variation in the mixed coniferous forest and broadleaved forest sites was around 88%, and the highest value was in the mixed broadleaved forest site (118.2%).

Using a two-factor analysis of variance, we examined whether the forest site type or the sap collection phase affects the variation in the average amount of harvested sap. Statistically significant differences were found in the average amount of harvested sap depending on the collecting phase (F = 7.987, p < 0.0007). For the forest site type (F = 1.468, p = 0.237) and the interaction between the collection phase and forest site type (F = 0.493, p = 0.741), no statistically significant relationship with the average amount of harvested sap was found.

3.4. Modeling the Daily Amount of Birch Sap Collected

A linear mixed-effects model was developed describing the daily birch sap leakage for a single tree. Among the many available variables (diameter at breast height, total height, crown base, crown length, relative crown length, slenderness, forest site type, and phase of sap collection), the following variables were included in the model:

–

Total height—tree biometric feature;

–

Forest site type—stand characteristic;

–

Phase of sap collection—feature related to time.

These variables represented the fixed effect in the model. Tree.Id was assumed to be the variable showing a random effect.

The estimated parameters of the model are presented in

Table 5. Parameters of the mixed-effects model in the prediction of the daily volume of birch sap obtained from one tree.

	Sap Volume
Predictors	Estimates	CI	p
(Intercept)	−3.36	−19.17–12.45	0.677
Height	0.17	−0.35–0.69	0.521
Site [mixed broadleaved forest]	1.22	−1.96–4.40	0.453
Site [mixed coniferous forest]	2.50	−0.90–5.90	0.149
Sap collection phase [phase 2]	3.86	3.38–4.34	<0.001
Sap collection phase [phase 3]	1.67	1.09–2.25	<0.001
Random Effects
σ2	7.54
τ00 Tree.Id	9.48
ICC	0.56
N Tree.Id	27
Observations	783
Marginal R2/conditional R2	0.162/0.629

. The analysis of the estimate values shows that trees growing in the mixed broadleaved forest site are characterized by a daily sap yield higher than 1.22 dm³ compared with the broadleaved forest and 2.50 dm³ compared with trees in the mixed coniferous forest site. The effect of site and height on daily sap volume collection is not statistically significant in that model. However, in terms of research significance, these variables make a difference; e.g., a sap volume of 2.50 dm³ more in the mixed coniferous forest site compared to the broadleaved forest makes a big difference for a person collecting birch sap. The site and height variables turned out to be statistically insignificant due to the small sample (nine trees per site) and the large variability of the amount of sap collected between trees representing the same site.

The effect of the collection phase on the amount of sap collected turned out to be statistically significant. In the main phase, we received, on average, 3.86 dm³ more sap than in the initial phase. In the case of the final phase, this difference was 1.67 dm³.

The prediction of daily birch sap collection against the background of the values of the variables included in the model (tree total height, forest site, and sap collection phase) is presented in Figure 3.

4. Discussion

The average daily birch sap leakage efficiency in individual sites varied. It amounted to 3.05–11.50 dm³/24 h for mixed coniferous forest, 0.78–12.43 dm³/24 h for mixed broadleaved forest, and 1.24–8.99 dm³/24 h for broadleaved forest. According to Kostroň [14], the intensity of sap leakage is influenced by the site and the quality and health of the stand (expressed by the crown size and the trees’ average thickness). He also defines a kind of cause-and-effect chain: the more fertile the soil and the better the quality of the trees, the larger the crown, root system, and average thickness of the tree, which results in more intense sap production. The authors’ research results do not confirm this thesis. Based on the data (Table 1), we determined the average daily birch sap leakage efficiency from all 27 examined trees (all three examined sites) to be 5.16 dm³/24 h. Broken down by site, the values are as follows: mixed coniferous forest—6.09 dm³/24 h; mixed broadleaved forest—5.08 dm³/24 h; broadleaved forest—4.30 dm³/24 h. Kostroň [14] also states that the amount of leaking sap is also influenced by water conditions in the soil, the age of trees, and their exposure. Moreover, on drier soils, the amount of sap is smaller, while the sugar content increases. This is also confirmed by Grochowski [15].

The amount of sap that can be obtained during the entire season is also difficult to determine unequivocally. The average daily sap leakage efficiency is presented in Table 2. During the 29 days of collection in the mixed coniferous forest site, an average of 176.76 dm³ of sap was obtained from one tree; in the mixed broadleaved forest, an average of 147.29 dm³ of sap was obtained, and in the broadleaved forest, an average of 124.64 dm³ of sap was obtained. Taking into account the above data and the high variability of birch sap leakage from individual trees (Table 1), it can be assumed that during a 29-day collection of sap from one tree, approximately 150 dm³ of birch sap can be obtained. This is a larger amount than that provided by many authors in the literature. Hnizdo [33], in his work, obtained a result of 15–20 dm³ from one tree during the season, Kostroň [14] obtained 78 dm³, Koroljak and Tomčuk [20] obtained 80 dm³, and Orlow [34] obtained 120 dm³. Many factors can influence this diversity, including the biometric features of trees, forest sites, the exposure of trees, water conditions, the length of the sap collection season, and the number of holes made and their location in relation to the cardinal directions.

This study did not confirm the statistical significance of the influence of factors such as diameter at breast height, total height, crown base, crown length, relative crown length, and slenderness on the efficiency of birch sap leakage. However, the authors of other publications indicate the influence of certain factors on the amount of sap obtained. It is worth recalling the cause–effect sequence described by Kostroň [14], mentioned earlier: the more fertile the soil and the better the tree quality, the larger the crown, root system, and average thickness of the tree, which results in greater sap production. He also describes how during the day, birches with a large crown yield, on average, more sap (1.932 L) than trees with a medium crown (1.023 L) and trees with a small crown (0.641 L). He also adds that even with a smaller total amount of sap obtained, caused by an unsuitable site and lower rainfall, the size of the crown affects the production of sap in the tree.

Kostroň [14] showed, in his studies, from birches with a breast height of up to 15 cm, the daily leakage was 2.34 dm³; from 16 to 20 cm, it was 2.52 dm³; from 21 to 25 cm, it was 3.14 dm³; and from 26 to 30 cm, it was3.55 dm³. In a study on the efficiency of birch sap flow, depending on the breast height of the tree [16], no statistically significant differences in the efficiency of sap flow were found between individual DBH classes (8–26 cm, 26–30 cm, and 32–42 cm; stand age, 65 years). The author indicates that efficiency is a very variable feature, and height is the best feature based on which we can determine correlations with efficiency. Another aspect worth noting is the difference in the efficiency of sap flow between trees with very similar DBH values and those growing several or several dozen meters apart (in the same habitat with the same microclimate). The experiences of various authors indicate that these differences can even be several times greater. Osiak [16] reported that 5.8 L of sap was obtained from birch with a DBH of 26.9 cm, and only 1.9 dm³ was obtained from a neighboring tree with a DBH of 28.2 cm per day. The authors noticed similar relationships in their studies. In the mixed coniferous forest site, tree THN3 was characterized by an average daily leakage of 10.88 dm³/24 h, while the trees with the largest DBH values (TCK1, TCK2, and TCK3) did not exceed a value of 4 dm³/24 h. Similar relationships can be observed in the other two forest sites (mixed broadleaved forest and broadleaved forest).

In this study, the age of the trees was similar, averaging about 80 years. According to Kostroň [14], healthy, 30–40-year-old birch stands have the highest sap production. Studies conducted in 2017 on the efficiency of birch sap flow depending on the age class of the stand suggest that there are no statistically significant differences in the efficiency of sap flow between stands that significantly differ in age (34 and 84 years), while the daily sap yield is a feature of high variability [35].

The start date of birch sap leakage may be variable but mainly depends on the season and air temperature. Kostroň [14] conducted his research from 3 March to 20 April. He states that the season affects sap production, and the most favorable time is early spring (a period of rapid snow melting with cool nights and warm days). Sap leakage ends after the leaves unfold, and sap production in the tree starts again in autumn after the leaves fall, but the intensity is much lower. He defines the early spring leakage period as 16–24 days (with the possibility of extending this period depending on the impact of climatic conditions). According to Kostroň [14], the optimal period for sap collection is 18–20 days. Grochowski [15] states that the birch sap season begins in mid-April and in unfavorable climatic conditions in the first days of May. The collection period (regardless of the location) lasts 17–24 days, sometimes even up to 29 days. Muszyński and Muszyński [19] state that the commencement of sap collection depends on climatic and atmospheric conditions, and in Poland, sap can only be collected in early spring, before the leaves develop, i.e., in the period from mid-March to the end of April. Głowacki and Kalicka [36] provide an approximate date of the beginning of March, and the leak duration is 10–14 days. The authors also add that this time depends on the weather and climatic conditions. Bilek et al. [9] specify the time of sap collection as 2–3 weeks. In this paper, we proposed to divide the birch sap leakage period into three stages: initial, main, and final. No scientific sources were found that would divide the sap leakage period into characteristic stages. However, it is important that other authors also note the variability of the amount of birch sap leaking during collection and the impact of climatic factors on this variability. According to Kostroň [14], in the first period of sap collection, its amount increases with temperature, while in the second, with its increase, the amount of sap decreases. The efficiency of sap leakage is the highest in the initial period of its collection [16], and about 3 weeks after sap movement, the flow even decreases several times. The highest birch sap flow occurs 2–3 weeks after the start of sap flow in the tree trunk [17]. This is also confirmed by Grochowski [15], who claims that during the first two-thirds of the harvesting season, the increase in temperature has a positive effect on the sap yield, while in the last third, it has a negative effect. Koroljak and Tomčuk [20] state that the trees produce the most sap between 12 and 6 p.m. and the least between 0 and 6 a.m. Kostroň [14] reported that the leakage during the day was about 25% higher than at night. He also added that in early March, the sap production increased, except for two days when the temperature at night reached −8 °C and −0.9 °C during the day, and the leakage was interrupted.

The available literature data mostly come from 50–60 years ago, when the industry related to birch sap extraction was more widely developed in Central and Eastern Europe. Since then, not only have economic changes occurred but also climatic and environmental changes. Aspects such as the length of the winter period, average air temperatures in winter and spring, and groundwater levels are just examples of variables that can affect the physiology of trees and, consequently, the yield of birch sap.

In the present work, a model for predicting daily birch sap collection was proposed based on a selected tree feature (total height), a stand feature (forest site), and a time-related feature (sap collection phase). The model indicates which features are related to the amount of sap collected, which allows for selecting the site, tree, and time when sap collection will be most efficient. The presented model can be used to analyze the profitability of birch sap collection.

Studies aimed at determining the amount and modeling of factors influencing the amount of birch sap collected were conducted in Finland on an extensive sample, which was possible thanks to the involvement of a citizen scientist [37]. This model concerned the total amount of sap collected from one tree during the season, and the explanatory variables included the DBH and tree height. The seasonal sap yield of individual birch trees studied in Finland was, on average, 53.4 dm³ in 2019 and 37.3 dm³ in 2020. These values are several times lower than those of the birches examined in this study, where the average values ranged from 124.6 dm³ in the broadleaved forest site to 176.8 dm³ in the mixed coniferous forest site.

The results obtained in this study allow us to state that in the case of large-scale birch sap harvesting, the selection of the forest site type has little practical significance, but it is important to use the high sap yield in the main leakage phase. However, a number of factors that may affect the quality of the obtained sap and, consequently, the safety of its consumption should certainly be taken into account [38,39,40,41,42,43,44,45,46,47,48]. Finally, it is worth emphasizing that the results of multi-aspect studies on the quality and yield of birch sap can contribute to increasing the importance of obtaining this raw material as a source of additional income for state and private forest owners. This aspect is addressed in multifunctional forest management strategies and certification standards [4,7,8].

5. Conclusions

• The coefficient of variation of daily sap yield is very high, especially in trees with higher DBH values. Considerable variation in the total amount of collected sap was observed between trees growing in the same location.
• Analysis of the relationship between seasonal birch sap yield and the biometric features of trees (DBH, total height, crown base, crown length, relative crown length, and slenderness) for any of the above features did not show statistically significant correlations—both for all trees together and divided into groups of forest site types.
• This paper proposes a new method for determining the phases of birch sap leakage based on the intensity of leakage in relation to the average value for the entire season. This variable turned out to be a feature that significantly differentiates the amount of sap obtained, and it is most useful in modeling this feature. The smallest amount of sap was obtained in the initial phase, a slightly higher amount was obtained in the final phase, and the largest amount of sap was obtained in the main phase. Statistically significant differences were found in the average amount of collected sap depending on the collecting phase.
• Regarding the forest site type and the interaction between the collection phase and the forest site, no statistically significant relationship was found with the average amount of collected sap.
• The mixed-effects linear model indicates features related to the amount of sap potentially collected, which allows for selecting the forest site, tree, and time when sap collection will be most efficient. The presented model can be used to analyze the profitability of birch sap collection.
• Taking into account the obtained results, it can be assumed that in the case of birch sap harvesting on an industrial scale, the selection of an appropriate forest site type is of little practical importance, but the use of high sap yield in the main leakage phase is important.