Optimization of the process of seed extraction from the Larix decidua Mill. cones including evaluation of seed quantity and quality

Cone characteristics: the entire set and individual variants

Cones used in all the test variants did not differ from each other in terms of height (coefficient of variance in the Student t-test–F = 1.33 at p = 0.23), diameter (F = 1.77 at p = 0.08), or initial weight (F = 0.86 at p = 0.55). Analysis of variance revealed a significant difference for cone humidity (F = 2.52 at p ˂ 0.05).

Cone parameters such as height, diameter and initial weight are factors that can determine the course of the extraction process. Therefore, the relationship between diameter and height for all cones used in the study was described using a linear regression equation ((y=0.2794x+8.3195)), which means that cone diameter increased by 0.28 mm per 1 mm of cone height, ((R=0.778>0.104-{R}_{kr})).

The initial weight of cones may be associated with their harvest time or storage conditions. A linear regression equation was also used to describe the relationship between the height and initial weight of the examined cones (y = 0.238x–3.918), which means that initial weight increased on average by 0.238 g per 1 mm of height, (R = 0.795 > 0.104).

Table 2 shows means with standard deviations, the minimum and maximum values of the measured parameters, the range of variance, the coefficient of variation and the standard error for the entire set of studied cones and seeds. The Shapiro–Wilk test showed that the examined characteristics had a normal distribution.

The cones used in the study had a height of 21.4–44.1 mm and a diameter of 12.5–24.3 mm. The mean height of a cone was 33.8 (± 3.4) mm and the mean diameter was 17.8 (± 1.6) mm. The initial weight of cones ranged from 2.137 to 9.111 g, with a mean of 4.144 (± 1.019) g. The initial moisture content of cones was from 27.6 to 57.1%, with a mean of 40.4 (± 4.5)%. Analysis was performed for individual extraction variants. The mean values of cone height h, diameter d, initial weight m₀₁, and moisture content W were calculated (Table 3).

The HSD Tukey test revealed one homogeneous group for cone height encompassing all variants and two homogeneous groups for diameter. The first group consisted of all variants except 7, and the second group included all variants except 2. One homogeneous group was obtained for initial weight. Two homogeneous groups were found for moisture content, one consisting of all variants except 7, and the other one containing variants 1, 4, 5, 6, 7, 8, and 9.

Seed extraction results for the studied steps

Seed extraction conditions and time

The change in cone weight in each step of the extraction process depended on its duration, temperature and humidity conditions in the extraction cabinet, as well as on the initial moisture content of the cones.

Humidity inside the drying chamber decreased to an average of 30% after 2 h of the process in each step as a result of increasing temperature. Over the subsequent 4 h of the process, after increasing the temperature, the humidity inside the chamber declined significantly, and then (over 2 and 4 h) it decreased further only slightly, stabilizing at approx. 5% for the 10 h variants, 6% for the 8 h variants, and 8% for 6 h variants on average.

Moisture content changes in cones during the seed extraction process

The initial moisture content (u₀₁) of the studied cones was much greater than 0.20 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}), which means that special care must be taken during seed extraction, which should be conducted at a temperature of up to 50 °C⁸.

The relatively high moisture content of the cones could be attributed to the absence of preliminary drying in airy storage places prior to seed extraction (which is typically the case in commercial practice) and the early date of cone harvest, at the beginning of the extraction season. The initial (u_0x) and final (u_kx) moisture content of cones used in each process variant is given with standard deviation in Table 4.

The initial moisture content of cones (u_0x) in most variants increased with each extraction step due to immersion. In most variants, the final moisture content (u_kx) was the highest in the first extraction step and decreased or remained at the same level with each subsequent step.

The mean initial moisture content for the three process variants with 10 h of drying was 0.411 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}). After 10 h of drying, the mean moisture content decreased to 0.130 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}). The mean initial moisture content in the fifth extraction step was 0.437 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}), and the final moisture content in that step was 0.071 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}) . Cones dried for 10 h reached on average 7% moisture content after extraction steps 4 and 5.

The mean initial moisture content for the three process variants with 8 h of drying was 0.412 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}). After 8 h of drying, the mean moisture content decreased to 0.128 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}) . The mean initial moisture content in the fifth extraction step was 0.440 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}), and the final moisture content in that step was 0.064 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}) . Cones dried for 8 h reached on average 7.1% moisture content after extraction step IV and 6.4% after step V.

The mean initial moisture content for the three process variants with 6 h of drying was 0.389 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}). After 6 h of drying, the mean moisture content decreased to 0.129 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}) . The mean initial moisture content in the fifth extraction step was 0.415 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}), and the final moisture content in that step was 0.084 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{mathrm{d}.mathrm{w}.}^{-1}) . Cones dried for 6 h reached on average 8.9% moisture content after extraction step IV and 8.4% moisture content after step V, which means that their final moisture content was higher than that of cones dried for 8 h and 10 h.

The cones with the longest immersion time (15 min) were characterized by the highest initial moisture content in each extraction step as compared to the other two variants (immersion of 5 min and 10 min) with the same drying time. The final moisture content in a given extraction step differed between cones with different immersion times. Cones with an immersion time of 15 min were characterized by the highest final moisture content in individual extraction steps, and those with 5 min immersion revealed the lowest final moisture content.

The Tukey HSD test revealed homogeneous groups in terms of initial moisture content (u₀₁, u₀₂, u₀₃, u₀₄, u₀₅) and final moisture content (u_k1, u_k2, u_k3, u_k4, u_k5) in each step, as shown in Table 4. For instance, four homogeneous groups were found for the final moisture content after extraction step V (u_k5): the first one consisted of all variants except for 7, 8, and 9, the second one included variants 1, 2, 3, and 7, the third one comprised of variants 7 and 8, while the fourth one was constituted by variant 9 alone.

Using Eq. (1), changes in moisture content were described for each of the tested cones over all five steps of each variant. The equation included the initial and final values of moisture content and the b coefficient for individual cones. The average values of the b coefficient and standard deviations for each extraction step are presented in Table 5 for individual extraction variants.

The lowest value of the b coefficient was recorded for the first step of the 10h_15min variant (b₁ = 0.34), while the highest value was obtained for the fifth step of the 8 h_15 min variant (b₅ = 0.60). In the process variants involving 10 and 8 h of drying , the b coefficient increased with each extraction step until the third one; in the fourth step it slightly decreased and in the fifth step it remained constant. In the variants with 6 h of drying the b coefficient almost peaked in the second extraction step and remained at a similar level until the fifth step. In the first steps of the variants with 6 h of drying, the mean value of the b coefficient was 0.54 and did not differ significantly from the coefficients obtained during the other steps. It was noted that in the 8 h_15 min variant, the b coefficients increased over successive steps.

Figures 2–3 show examples of curves of actual and model changes in moisture content and the rate of extraction for sample cones, one each for variants 10 h_15 min and 8 h_15 min.

Equations for changes in moisture content and extraction rate in consecutive extraction steps are given below for the graphically for the cone shown in Fig. 2 (no. 32 in the 10 h_15 min variant):

Step I: ({u}_{1}=0.264cdot {mathrm{e }}^{left(-0.38 cdot {tau }_{i}right)}+0.107) ,(frac{d{u}_{1}}{d{tau }_{1}}=-0.100cdot {mathrm{e }}^{(-0.38 cdot {tau }_{i})})

Step II: ({u}_{2}=0.372cdot {mathrm{e }}^{left(-0.44 cdot {tau }_{i}right)}+0.095) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.164cdot {mathrm{e }}^{(-0.44 cdot {tau }_{i})})

Step III: ({u}_{3}=0.397cdot {mathrm{e }}^{left(-0.49 cdot {tau }_{i}right)}+0.086) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.195cdot {mathrm{e }}^{(-0.49 cdot {tau }_{i})})

Step IV: ({u}_{4}=0.536cdot {mathrm{e }}^{left(-0.44 cdot {tau }_{i}right)}+0.080) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.236cdot {mathrm{e }}^{(-0.44 cdot {tau }_{i})})

Step V: ({u}_{5}=0.485cdot {mathrm{e }}^{left(-0.46 cdot {tau }_{i}right)}+0.076) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.223cdot {mathrm{e }}^{(-0.46 cdot {tau }_{i})})

Equations for changes (Fig. 3) in moisture content and extraction rate in consecutive extraction steps are also given for this cone (no. 17 in the 8 h_15 min variant):

Step I: ({u}_{1}=0.304cdot {mathrm{e }}^{left(-0.53 cdot {tau }_{i}right)}+0.113) ,(frac{d{u}_{1}}{d{tau }_{1}}=-0.161cdot {mathrm{e }}^{(-0.53 cdot {tau }_{i})})

Step II: ({u}_{2}=0.292cdot {mathrm{e }}^{left(-0.55 cdot {tau }_{i}right)}+0.085) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.161cdot {mathrm{e }}^{(-0.55 cdot {tau }_{i})})

Step III: ({u}_{3}=0.369cdot {mathrm{e }}^{left(-0.70 cdot {tau }_{i}right)}+0.077) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.258cdot {mathrm{e }}^{(-0.70 cdot {tau }_{i})})

Step IV: ({u}_{4}=0.379cdot {mathrm{e }}^{left(-0.71 cdot {tau }_{i}right)}+0.059) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.269cdot {mathrm{e }}^{(-0.71 cdot {tau }_{i})})

Step V: ({u}_{5}=0.428cdot {mathrm{e }}^{left(-0.77 cdot {tau }_{i}right)}+0.060) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.330cdot {mathrm{e }}^{(-0.77 cdot {tau }_{i})})

Finally, equations for changes in moisture content and extraction rate in consecutive extraction steps are given for cone no. 5 in the 6 h_15 min variant:

Step I: ({u}_{1}=0.308cdot {mathrm{e }}^{left(-0.58 cdot {tau }_{i}right)}+0.0904) ,(frac{d{u}_{1}}{d{tau }_{1}}=-0.179cdot {mathrm{e }}^{(-0.58 cdot {tau }_{i})})

Step II: ({u}_{2}=0.346cdot {mathrm{e }}^{left(-0.63 cdot {tau }_{i}right)}+0.1070) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.218cdot {mathrm{e }}^{(-0.63 cdot {tau }_{i})})

Step III: ({u}_{3}=0.368cdot {mathrm{e }}^{left(-0.63 cdot {tau }_{i}right)}+0.0837) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.232cdot {mathrm{e }}^{(-0.63 cdot {tau }_{i})})

Step IV: ({u}_{4}=0.387cdot {mathrm{e }}^{left(-0.68 cdot {tau }_{i}right)}+0.0838) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.263cdot {mathrm{e }}^{(-0.68 cdot {tau }_{i})})

Step V: ({u}_{5}=0.396cdot {mathrm{e }}^{left(-0.65 cdot {tau }_{i}right)}+0.0743) , (frac{d{u}_{1}}{d{tau }_{1}}=-0.257cdot {mathrm{e }}^{(-0.65 cdot {tau }_{i})})

Figures 2a, 3a show the curves of actual changes in the moisture content of three sample cones subjected to different drying times (10 and 8 h) but the same immersion time (15 min); the curves were fitted to a model which is widely used in descriptions of drying at constant temperature (mostly for vegetables). The present study used variable temperature, which may have influenced the fit of the model, in addition to the input variables (drying and immersion times). The best fit was found for the cone subjected to the variant with 8 h of drying (Fig. 3), with a slight deviation in the first three extraction steps, and with a very good fit in the fourth and fifth steps. The lowest fit was found for the cone subjected to 6 h drying, which may be caused by insufficient drying time (the cone was exposed to 35 °C for 2 h, and to 50 °C for only 4 h).

Figures 2b, 3b show diagrams for cone extraction rates at different drying times (10 h and 8 h) at the same immersion times (15 min). As can be seen, extraction rates decreased in the very beginning, which is characteristic of the so-called second period of solid drying (Pabis⁴⁴).

Seed extraction dynamics

Table 2 presents data on the number of scales and seeds for the studied cones. There were from 33 to 70 open scales per cone, with an average of 48 (± 6). From 1 to 76 seeds were extracted per cone, with an average of 36 (± 18). Finally, each cone contained from 5 to 97 seeds, with an average of 52 (± 19). The weight of the extracted seeds ranged from 0.001 g to 0.651 g, on average 0.193 (± 0.109) g.

Cones obtained from different process variants did not differ in terms of the number of seeds extracted (F = 0.862 at p = 0.55) or their weight (F = 0.720 at p = 0.674). However, ANOVA did reveal significant differences in the number of scales (F = 3.561 at p ˂0.05) and the total number of seeds per cone (F = 2.93601 at p = 0.003645). Table 6 gives mean scale and seed numbers per larch cone (with standard deviations) for the various extraction variants and homogeneous groups.

On average, 70% of the seeds were extracted from cones used in all nine study variants, with 30% of the seeds remaining in the cones. Table 7 shows the number of seeds extracted in individual variants and the number of seeds remaining in the cones, expressed as a percentage.

The greatest number of seeds was obtained in process variants 2–73%, closely followed by variants 3, 1, and 7 (72%), and 8 (70%). The lowest seed yield was obtained from variant 4 (65%).

In all study variants, some of the seeds were obtained in the process of extraction in the chamber and some in the process of shaking in the drum (Table 7). The highest number of seeds in the chamber was obtained in variant 2 (69%), and the lowest in variant 9 (56%). On average, the largest quantity of seeds was obtained in the chamber in the 10 h variants, and the lowest quantity in the 6 h variants. Comparing different process variants of the same drying duration, the greatest number of seeds in the chamber were obtained in variants 2, 5, and 7 (and also in variant 8—only 1% fewer). The greatest quantity of seeds extracted by shaking in the drum was obtained in variant 9 (44%), and the lowest in variant 2 (31%). On average, 38% of seeds extracted in all variants were obtained by shaking in the drum.

It can be seen that in each of the variants and their individual steps, the highest number of seeds was obtained after 6 h of the process. Figure 4a–c shows the percentage of seeds obtained during the effective extraction time, where the number of seeds extracted at a given step was added cumulatively to those from the previous steps.

The diagrams in Fig. 4 show the percentage of seeds obtained throughout the entire process. Each step consists of drying, shaking, immersion, and soaking, except for step V, which involved only drying and shaking without immersion or soaking. Analysis of seed yield over 10 h of drying (Fig. 4a) shows that on average 37% of all extracted seeds were obtained in the first step, 26% in the second step, approx. 20% in the third step, 11% in the fourth step, and about 6% in the fifth step.

As regards the 8 h process (Fig. 4b), on average 30% of all extracted seeds were obtained in the in the first extraction step in the 8 h_5 min and 8 h_15 min variants, and as much as 53% in the 8 h_10 min variant. An average of 27% of all seeds were extracted in the second step, 15% in the third step, about 11% in the fourth step, and approx. 5% in the fifth step. The 8 h_10 min variant was characterized by the highest seed yield, beginning in the first step of the process (as compared to the 8 h_5 min and 8 h_15 min variants).

As far as the variant with 6 h of drying is concerned (Fig. 4c), on average approx. 46% of all extracted seeds were obtained in the first step, 24% in the second step, 15% in the third step, approx. 11% in the fourth step, about 4% in the fifth step.

When extracting seeds from larch cones, scale deflection and the number of obtained seeds are not assessed during the process, as is the case with pine and spruce cones due to the difficulties caused by the aforementioned morphology of larch cones (Tyszkiewicz, 1949). The presented diagrams show that a satisfactory seed yield (60%) was obtained in variants with 8 and 6 h of drying already after 10 h of effective extraction time.

The seed yield coefficient, α (3), and the cone mass yield coefficient, β (4), for each extraction variant are presented in Table 8.

The seed yield coefficient was the highest for variants 2 (0.73) and 3 (0.72), and the lowest for variants 4 (0.65) and 6 and 9 (0.67). The cone mass yield coefficient was the highest for variant 5, and the lowest for variant 9.

Seed viability

Table 9 presents germination energy (E) and capacity (Z) for the control seeds as well as for seeds obtained from the various steps of the nine process variants, as well as their corresponding quality classes.

Germination energy and capacity for the control sample were 45% and 57%, respectively, meaning that naturally released seeds, not subjected to any thermal or mechanical treatments, were classified in quality class I¹⁸. Importantly, seeds obtained from all the studied process variants were also placed in the same class; their germination energy ranged from 30 to 59%, and their germination capacity from 35 to 61%. When analyzing each extraction step separately, no correlation was found between decreasing germination energy and successive steps. However, the average germination energy was 46% for seeds obtained in the first extraction step of all nine variants, 45% for those from the second and third steps, 41% for seeds from the fourth step and 40% for those from the fifth one. Thus, in each subsequent step the average germination energy of seeds was equal or lower than in the previous step, which is consistent with literature reports that prolonged drying may reduce the quality of seeds⁸. This is also corroborated by the fact that the highest germination energy and capacity was revealed by seeds from variants with 6 h of drying while the lowest germination indicators characterized seeds from the 10 h variants. Furthermore, seeds from variant 1 exhibited the lowest germination energy and capacity and seeds from variant 8–the highest.

Another reason for the higher quality of seeds from variants with 6 h of drying may be the lower initial moisture content of the cones due to the longer time they were kept at room temperature immediately before the test (u₀₁ = 0.391 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{d.mathrm{w}.}^{-1}) as compared to u₀₁ = 0.411 ({mathrm{kg}}_{mathrm{water}}cdot {mathrm{kg}}_{d.mathrm{w}.}^{-1}) for seeds from variants with 8 and 10 h of drying). These results are in line with the study of Tyszkiewicz⁸, who noted that under the same temperature and humidity conditions, the quality of seeds from cones with a lower moisture content did not deteriorate, in contrast to the quality of seeds obtained from cones with a higher moisture content.

The germination capacity of seeds calculated from the mean capacity of seeds obtained from the same extraction steps of all process variants was similar at 45% for each of the steps.

In summary, in the study the authors investigated a five-step process of extracting seeds from larch cones involving immersion and heat treatment to maximize seed yield. It was found that the two-step process widely used in extractories is insufficient, while a four-step process does not lead to a significantly higher number of obtained seeds. Thus, a three-step process appears to be optimal.

Source: Ecology - nature.com