Energy and Biomass Yield of Industrial Hemp (Cannabis sativa L.) as Influenced by Seeding Rate and Harvest Time in Polish Agro-Climatic Conditions

ABSTRACT Energy produced directly from biomass represents an important part of available alternative energy sources. Industrial hemp has potential energy yields that are higher than those of many other energy crops and, as an annual herbaceous crop, hemp fits into existing crop rotations. The total biomass of hemp per hectare is similar to other energy crops, including mostly grown giant miscanthus, poplar, or willow. Industrial hemp is characterized by a short growing cycle, a decreased need for pesticides, and low plant maintenance. The energetic value of hemp depending on agrotechnical factors, such as sowing density and time of harvest, has been assessed in this study. A three-year field experiment was carried out to prove the possibility of using whole plants and separately hemp straw and panicles for energy purposes. C. sativa was sown at eight sowing rates − 5, 10, 15, 20, 30, 40, 50, and 60 kg · ha−1 and harvested at three developmental stages: at the beginning of panicle formation, at full flowering and at the end of seed setting. The optimal harvest time of hemp grown for energetic purposes was demonstrated to be at full flowering and the optimal sowing rate at 30 kg · ha−1 when hemp yielded 14.65 t · ha−1 equals 275.56 GJ · ha−1.


Introduction
Nowadays our civilization is facing the challenges of feeding a growing population and ensuring the supply of the necessary energy without exhausting the biological and physical resources of the planet (Muscat et al. 2020). These days, energy is considered as one of the most important commodities; however, fossil fuels as main global energy resources due to environmental impacts and overexploitation are under extreme pressure to meet the requirements of the growing human population. Thus, the secure and sustainable energy supply can ensure future socio-economic development of any country (Rehman et al. 2013).
Energy produced directly from biomass or its conversion products represents an important part of available alternative energy sources. As biomass is renewable, abundant and can be used locally at the place of production, the sources of biomass can help the world reduce its dependence on coal, petroleum, and other fossil fuels. Additionally, biomass is a sustainable carbon-neutral resource. Those properties give biomass an enormous potential as the demand for biomass and biofuels is expected to triple by 2035 (Haykiri-Acma 2003;Parvez, David Lewis, and Afzal 2021).
Different sources of biomass are available for energy production. These fall into four main categories: residues from agriculture and forestry, organic wastes, surplus forestry, and energy crops (Beringer, Lucht, and Schaphoff 2011;Berndes, Hoogwijk, and van den Broek 2003).
Energy crops have been regularly reported to have the high potential to become a main source of renewable energy although their large-scale cultivation is also one of the most controversial aspects of bioenergy. The key question is whether enough food, animal feed, and -simultaneously -biomass can be produced to fulfill all the bioenergy needs for our future (Beringer, Lucht, and Schaphoff 2011;Muscat et al. 2020;Prade et al. 2011).
To satisfy the growing, worldwide demand for biomass for food and industrial purpose, the two broad options are available: the area under production can be increased or the productivity on existing farmland can be improved. Of the two options, the increasing productivity on existing agricultural land is preferable as it avoids greenhouse gas emissions and the large-scale disruption of existing ecosystems associated with sourcing new land into production. Despite various concerns, as the arable land available for energy production is limited, its high use efficiency indicated by high energy yields per hectare is of critical importance (Edgerton 2009;Prade et al. 2011). With this perspective, hemp can be seen as promising, as it can be grown as multipurpose crop used as a commodity for high-value products like chemicals, pharmaceuticals, textile fibers, or food, while still straw or processing leftovers could be used for energy production with high efficiency.
At present, grains and seeds from crops like corn, wheat, sugar cane, and cereals are used as feedstocks to produce bioenergy (Bentsen and Felby 2012). These feedstocks (directly) and their arable land (indirectly) compete with other food commodities and pose a threat to the food security. Such crops in which the biomass of the whole plant can be used for energy production can potentially result in higher land use efficiency; however, few dedicated energy crops are currently grown in large-scale cultivation (e.g. maize in Germany) and even fewer of those are annual crops that could fit into a crop rotation with food and feed crops (Fernando et al. 2015;Karin and Nilsson 2006;Prade et al. 2011). In contrast, it has been shown recently that hemp is a good predecessor for wheat and the rotation effects of hemp on the subsequent wheat crops increased the wheat yield and the beneficial effects of hemp on wheat could last for a minimum of 2 years (Gorchs et al. 2017). As reported earlier, the cultivation of industrial hemp as an alternative crop to be used in rotation with cereal crops, such as wheat, barley, and oats, has the potential to be an environmentally friendly and highly sustainable crop and can fit well in a crop rotation with the aim of increasing soil fertility as hemp crop gives conspicuous residual fertility with a consistent quantity of organic residues left on the soil and thus the potential to efficiently suppress weeds (Adesina et al. 2020;Satriani, Loperte, and Pascucci 2021).
Fiber crops are potential energy crops, either when used as a whole for this purpose or their residues from different industrial processes. Lignocellulosic biomass production can increase arable land use efficiency and does not compete with food crops. But of course, those lignocellulosic crops that achieve higher yield and energy per hectare should be preferred (Fernando et al. 2015).
Hemp (Cannabis sativa L.) is resurging as an ideal multipurpose crop worldwide. It can be used to produce different energy products, such as heat and electricity (from briquettes, pellets, and baled biomass) or vehicle fuel (in the form of biogas or ethanol from fermentation). Hemp has potential energy yields that are as high or higher than those of many other energy crops and, as an annual herbaceous crop, hemp fits well into existing crop rotations (Burczyk et al. 2008;Tang et al. 2017).
Hemp is a herbaceous plant that can grow from about 1 up to 6 m tall, depending on factors such as cultivar and agronomic and environmental conditions. It can produce very high amount of biomass. Some authors reported extremely high fresh yields of hemp reaching more than 22 tons per hectare (Struik et al. 2000), however, more typical values are lower and other authors reported yields of fiber hemp on the level of 7−15 tons per hectare (Deleuran and Flengmark 2006;Fike 2016;Grabowska, Rębarz, and Chudy 2009).
Fiber hemp is an industrial crop with a rich tradition and history in many regions of the world. For many centuries hemp has been cultivated mostly for its strong stem fiber and seed oil (Struik et al. 2000). Recently, this species is gaining a lot of attention for its alternative use directions. Currently, more than 30 countries grow industrial hemp as an agricultural commodity with high potential (Food and Agriculture Organization of the United Nations, www.faostat.fao.org). It is considered as a lowcost, ecological, sustainable, and multi-use plant of important advantage that the entire plant, i.e., seeds and plant stems are recoverable. Hemp versatility is another advantage of this crop, as it is useable in various forms like fibers, felts, powders, shives, and seed products, including seeds themselves, oil and oil cake. All these products can be used for thousands of applications as described earlier (Crini et al. 2020). The global market for industrial hemp has more than 25,000 products in the 10 main submarkets: textiles, agriculture, automotive, food and beverages, paper, furniture, construction, recycling, and personal care; these include using hemp for biomass and biofuel production (Crini et al. 2020;Johnson 2014;Mark and Will 2019). Hemp is a rapidly growing plant well tolerating high planting density. The total biomass of hemp per hectare is similar to other energy crops, including mostly grown giant miscanthus, poplar, or willow. Industrial hemp is characterized by a short growing cycle, a decreased need for pesticides, and low plant maintenance. All those make it a good candidate crop for phytoremediation utilization and detoxification of contaminated soils due to its resistance to soil contamination, its ability to accumulate heavy metals, and the possibility of cultivation in different climatic conditions (Husain et al. 2019;Kalousek et al. 2020;Mańkowski et al. 2020;Pudełko, Kołodziej, and Mańkowski 2021). The latter property gives hemp an additional advantage over other high-energy crops by naturally extending the available pool of arable land to contaminated and unsuitable for other crops.
The presented research deals with a selection of optimal agrotechnical treatments that could lead to obtaining the highest hemp mass yield and thus to intensify the energetic value of hemp per hectare of cultivation area. The energetic value of hemp depending on agrotechnical factors, such as the sowing density and the time of harvest, has been assessed in the present study. Experiments were carried out to prove and illustrate the possibility of using whole plants and separately hemp straw and panicles for energy purposes.

Growing of C. sativa
The three-year field experiment was conducted in the Experimental Station of the Institute of Natural Fibres and Medicinal Plants in Pętkowo (52° 12'42"N, 17° 15' 30" W), province Wielkopolska, Poland. The experiments were prepared each year in a randomized complete block design. Tests were quadruplicated in an area of 10.5 m 2 each on light loamy (argillaceous) soils made of sand (the arable layer was approx. 30 cm deep).
Fiber hemp variety Białobrzeskie was sown in the first year of the experiment after spring barley, in the second year after maize and in the third after winter wheat. Fertilization applied was (kg ha −1 ): 80 N, 70 P 2 O 5 , and 110 K 2 O. Every year of the experiment, phosphorus and potassium fertilizers were applied in autumn and nitrogen in spring before the sowing of hemp. Organic fertilization in the form of manure was also additionally applied every year in the amount of 30 t · ha −1 . Agricultural practices used were in accordance with the recommendations for the cultivation of fiber hemp.
The experimental scheme was 8 sowing rates × 3 planned harvesting times × 4 replicates. The weight of 1000 seeds was 15.9-17 g and the sowing depth was 4 cm. The hemp from the experimental plots was harvested manually, and the harvest time was adjusted to the experimental scheme with respect to plant development stage.
The sowing dates were 27.04, 29.04, and 06.05 for 3 years of the experiment, respectively, while I harvest time was on 20.07, 20.07, and 21.07 in the years of the experiment, II harvest time was on 23.08, 22.08, and 22.08 and III harvest time was on 30.09, 27.09, and 21.09 or 3 years of the experiment, respectively.
In the first year of the experiment, the sum of rainfall from January to March was 77.9 mm, which contributed to good soil moisture and quick and even emergence. The total rainfall in April was 31.0 mm, while in May, it was 38.9 mm. In mid-May, there was a lot of rainfall associated with the hailstorm. In June, the total rainfall was 52.8 mm, while in July, it was 26.8 mm. July was a much cooler month than June. The sum of rainfall in the growing season until the first date of hemp harvest was 149.5 mm, which was 72.2% of the total rainfall in relation to the long-term average. The sum of rainfall in the growing season until the second harvest date was 195.7 mm, which was 82.7% of the long-term average. The sum of rainfall by the third harvest date was 223.9 mm, which constituted 78.6% of the long-term average.
In the second year of the experiment, in the second half of April, rainfall of 25.6 mm was recorded. In May, 60.8 mm of rain was recorded. The heat began in June, and the drought lasted about 6 weeks. In June, the rainfall was only 8 mm. The sum of rainfall for the growing season until the first harvest date was 108.7 mm, which was 67.6% of the long-term average. The sum of rainfall until the second harvest date was 183.0 mm (which constitutes 79.1% of the long-term average). In the second decade of September, heavy rainfall was recorded (74.3 mm of rain in September). Therefore, the sum of rainfall by the third harvest date was 267.3 mm, which was 96.0% in relation to the long-term average from the same growing season.
In the third year of the experiment, in April, 57.1 mm of water dropped, which led to good soil moistening. In May, rainfall was recorded at 50 mm. In June and July rainfall was recorded at a level of 25 mm, of which 17 mm of rain fell at the beginning of June. The sum of rainfall in this growing season until the first harvest date was 132.1 mm, which was 85.2% in relation to the long-term average. The sum of precipitation by the second harvest date for this period was 231.4 mm (102% of the long-term average). The sum of rainfall by the third harvest date of the period was 274.9 mm, which constituted 99.2% of the long-term average.
The large spread of the amount of sowing seeds resulted from the intention to determine the best parameters for the production of energy from whole plants and separately from straw and hemp panicles.

Laboratory tests
In the course of the experiment, whole hemp plants were collected, tied into sheaves, and dried naturally, assisted by the warm air blown from the heater. After drying stalks and hemp, panicles were separated and tested. All raw materials were obtained at three harvest times (the beginning of panicle formation, full flowering, the end of seed setting) and at eight assumed sowing rates (5,10,15,20,30,40,50, and 60 kg/ha). The humidity was determined by the dryer-weight method with the use of a laboratory dryer type ED 720. The total yield, straw, and panicle yield were determined each time, converted to 18% humidity. The obtained biomass was then comminuted using a laboratory cuttinggrinding mill, Retsch sm100, to 0.25 mm fractions, and the heat of combustion was determined.
Using a calorimetric bomb kl-12mn, the heat of combustion (HOC) of whole hemp plants, straw, and hemp panicles was determined. The determined HOC values were used to calculate the energetic value of the whole plants and their components. Before the test, the material was comminuted in a cutting-grinding mill with the use of sieves with a mesh diameter of 0.25 mm. The test consisted in burning a weighed sample of material in an oxygen atmosphere under pressure and measuring the increase in water temperature in the calorimetric vessel in which the calorimetric bomb was placed.
After determining the heat of combustion, the energetic value of total hemp, straw, and hemp panicles was calculated.

Statistical analyses
The results were evaluated statistically with the R statistical software version 4.0.5 (R Core Team 2020). ANOVA tests were used to test the main effects of test factors. The data from all years were combined and analyzed post hoc with Fisher's LSD test from agricolae package (de Mendiburu 2013) at a significance level α = 0.05.

Energetic value
With the use of a calorimetric bomb, the combustion heat of plant materials obtained from field experiments was determined (Figure 1).
The HOC during the experiment correlated well with CO 2 emission ( Figure 1A, B) during initial 15 min. of combustion. In the later period, the afterburning indicated by a lower heat release rate (HRR) and an increase in CO emission has been observed. HRR is one of the most important parameters of the process of fire and combustion (Bei, Liwei, and Chang 2012). Peak heat release rate present in the initial stage of the experiment proves the good flammability of hemp biomass and its good energetic properties. The flat curve of mass loss rate is characteristic for high fiber biomass as described elsewhere (Dorez et al. 2014).
Combustion heat of the hemp panicle had been determined at 19.8 MJ · kg −1 , of straw at 17.9 MJ · kg −1 , and of whole plants at 18.8 MJ · kg −1 . Those values are in agreement with those of other authors who reported hemp heating value from 17.5 to 19.24 MJ · kg −1 (Burczyk et al. 2008;Das et al. 2017;. Among the tested hemp yield components, the highest value of the heat of combustion for hemp panicles has been noted in the present study, and our observed value is very similar to that reported previously at 19.7 MJ · kg −1 (Burczyk et al. 2008). This is just an interesting observation when C. sativa is grown for other purposes like fiber extraction, as there is no economic value to separate panicles from straw and use it for energy production. The situation would be different when growing whole hemp for energy purposes. Then, the share of panicles can influence the overall energetic value of the total hemp yield. The reported herein heat of combustion of hemp places this crop among other biomass sources of the highest heating value like willow, miscanthus, or spruce (Parvez, David Lewis, and Afzal 2021;Rheay, Omondi, and Brewer 2021).

Hemp yield and energetic value
The influence of the important agronomic factors such as sowing density and time of hemp harvest on its yield and energetic value yield components was tested (Tables 1-3).
Industrial hemp can be cultivated for several purposes. Considering the directions of utilization, different agronomic practices have been elaborated and applied. C. sativa can be treated as a singlepurpose crop when cultivated for fiber or seeds, or as a dual-purpose crop when grown for fiber and seeds. Both types of cultivation differ in terms of the agricultural technology used, including the sowing density. In the cultivation of hemp for fiber, higher sowing densities are applied. When  growing hemp for seeds, the sowing density is lower (Amaducci et al. 2015;Amaducci, Errani, and Venturi 2002;Grabowska and Koziara 2006;Tang et al. 2017). Due to the fact that the literature lacks data on the optimal sowing density for cultivation for energy purposes, the experiment was established, covering sowing densities from industrial plantations for fiber and seed plantations. In industrial plantations, a seeding density of 50-60 kg · ha −1 is used for fiber, and in seed cultivation 10-15 kg · ha −1 .
A statistically significant influence of both experimental factors on yield and energetic value was observed. However, no significant interaction between them has been seen. The highest overall hemp yield was received when whole plants were harvested in the II term from plots sown with 30 kg · ha −1 ( Table 1). It was equal to 14.65 t · ha −1 or 275.56 GJ · ha −1 . In general, with the II term of harvest, the differences between the sowing densities were statistically significant neither for the crop yield nor for the energetic value. When whole plants were harvested in I and III terms, the differences between plots of different sowing densities were significant. When harvested in I term, the highest yield was also obtained from plots sown with 30 kg · ha −1 of seeds while the lowest yield was at 20 kg · ha −1 of seeds. When harvest was delayed to the III term, the situation was different and the highest yield was obtained with the sowing rate of 10 kg · ha −1 (12.64 t · ha −1 and 237.72 GJ · ha −1 ), while the lowest at the sowing rate of 60 kg · ha −1 (9.87 t · ha −1 and 185.68 GJ · ha −1 ). Regardless of these differences, for each sowing rate a lower yield from harvests on dates I and III compared to the harvest on dates II was observed.  A similar trend was observed when straw was harvested separately from panicles (Table 2). Different observations were made when panicles were harvested separately (Table 3). In this case, the highest overall yield and energetic value of hemp panicles were noted with a very sparse sowing rate of 5 kg · ha −1 harvested the latest, in the III term (3.88 t · ha −1 and 76.87 GJ · ha −1 ).
The presented work focuses on an attempt to answer the following question: which of the described cultivation types would be more beneficial in terms of the usage of hemp for energy purposes?
In the course of the experiment, it has been observed that harvesting time influences the energetic value of the hemp yield to a greater extent than sowing density ( Figure 2).
For both, straw and whole plant yield, the most favorable was the second term of harvest, i.e., full flowering. This observation is in agreement with previous studies showing that increasing the planting density had limited effect on stem yield as a consequence of the high incidence of self-thinning at high planting density (Tang et al. 2017). It has also been shown before that a delayed harvest of hemp can lead to an increase in its energetic value ). Prade concluded that the spring harvest of hemp is the most favorable for obtaining the highest energetic value. This observation is in contrast with ours, as we have observed that indeed the delay of harvest from the beginning of panicle formation to full flowering resulted in the increased energy yield from a hectare of crop, a further delay to the end of seed setting decreased the overall yield. It can be assumed that because there is a significant difference between the HOC of panicles and straw (as 19.8 MJ · kg −1 to 17.9 MJ · kg −1 ) and, on the other hand, the share of panicles to the total yield is changing during hemp growth and ripening -in order to maximize the energetic value of the crop grown solely for energetic purposes it is important to find the right balance between those two values ( Figure 3).
As hemp ripens, energy is transferred and used for seed formation, and thus the total yield and energetic value can decrease at this stage of plants development.

Conclusions
The presented research has shown that the energetic value of hemp biomass is higher than that of many other plants typically grown for energy purposes. The high yield of biomass with high heat of combustion results in the high energetic value of plants. The highest energetic value of hemp was obtained with a slowing rate of 30 kg · ha −1 . In the traditional cultivation of hemp for textile purposes, sowing in quantity 50-60 kg · ha −1 is recommended to obtain the highest quantitative and qualitative parameters. The optimal seeding rate for growing hemp for seeds is 10 kg · ha −1 . The energetic value of whole plants obtained with this rate in the presented experiment was by 9% lower than that obtained for the optimal sowing density. The optimal hemp harvest time grown for energetic purposes was observed to be at full flowering. Hemp is an annual plant, which does not occupy the field for many years, as is the case with the cultivation of perennial energetic plants. The high yield of biomass from a hectare of cultivation and the high energetic value predispose hemp to be cultivated as an energy plant (biomass crop).

Highlights
• Energetic value of hemp biomass is higher than of many other plants typically grown for energy purposes • High yield of biomass with high HOC results in high energy efficiency of C. sativa • Combustion heat of hemp panicle was 19.8 MJ · kg −1 , straw 17.9 MJ · kg, −1 and of whole plants 18.8 MJ · kg −1 • The highest energy efficiency of hemp was with sowing rate of 30 kg · ha −1 • The optimal hemp harvest time grown for energetic purposes was at full flowering