Abstract
Two common challenges reported by cannabis growers are low yields and small profit margins. Although recent research of cannabis yield has focused on lighting and nutrition, little research has examined how changes in other cultivation practices may be beneficial. The objective of this study was to evaluate the following two techniques to potentially improve yield: fertilizer restriction (FR) to reduce plant size and, thus, increase plant density and shoot number manipulation (SNM) to reduce shoot length and improve biomass partitioning. The FR technique involves leaching the substrate and providing only tap water for 0, 1, or 2 weeks at the start of flower initiation, whereas SNM involves pinching shoot tips 2, 3, or 4 times to produce 4, 8, or 16 shoots/plant, respectively. This study used a full factorial treatment design for a total of nine treatments (three FR × three SNM). Plants were flowered under 12-hour photoperiods for 8 weeks and then destructively harvested for data collection. The results demonstrated that both techniques improve plant productivity in different ways. The FR technique reduced all mass measurements (g/plant) and decreased plant area (m2/plant); therefore, the yield metrics (kg·m−2) increased with the increasing FR treatments. The SNM technique did not affect plant area, but more pinching events resulted in a decrease in reproductive shoot length (cm/shoot) and an increase in inflorescence to trim the dry mass ratio (inflorescence:trim). Shorter shoot lengths are desirable for eliminating trellis support netting, which helps growers reduce material costs and improve labor efficiency during harvest. Increasing the inflorescence:trim may also reduce labor costs related to trimming, which comprise the largest cost of production by many growers. Although both techniques offer advantages, there are trade-offs that must be considered in the context of overall profitability.
An industry survey reported that insufficient yield and shrinking profit margin are the two largest challenges faced by commercial cannabis growers (Hansen 2023). In cannabis production, yield is generally defined as female inflorescence mass per unit area (kg·m−2), and it is referred to here as inflorescence yield. The cannabis inflorescence is rich in glandular trichomes containing many compounds with medicinal and recreational values, including tetrahydrocannabinol (THC), cannabidiol (CBD), and cannabigerol (Saloner and Bernstein 2023). Much of the recent research of cannabis yield has focused on lighting (Hawley et al. 2018; Morello et al. 2022; Rodriguez-Morrison et al. 2021; Westmoreland et al. 2021); however, few studies have evaluated how changes in cultural practices might improve yield in controlled-environment cannabis production.
Planting density is defined as the number of plants grown per unit of cultivation area (plants/m2). For most horticultural and agronomic crops, the yield increases proportionally with planting density until a maximum is reached; at that point, any further increase in planting density negatively affects yield. Improvements in yield at higher planting densities occur because of greater photon capture, which drives photosynthesis and biomass accumulation (Ortega et al. 2004). Optimal planting densities are influenced by environmental factors, such as photoperiod, daily light integral (DLI), and temperature, as well as production factors, such as cultivar, fertilizer regimen, and shoot architecture.
Conventional cannabis production in controlled environments uses relatively large plants grown at relatively low planting densities (1.5 to 3.0 m tall and two to three plants/m2). This approach results in many inefficiencies during production, especially those related to photon capture (Bugbee and Salisbury 1988) and uniformity of photon distribution (Danzinger and Bernstein 2022). At low planting densities, canopy closure may not be achieved until late in the crop schedule, if at all, thus lowering photon capture and biomass growth. Achieving canopy closure early during crop cycles is crucial for maximizing yield (Hoffman 2019; Lu et al. 2021; Solanti and Galeshi 2002). Danzinger and Bernstein (2022) reported that spatial gradients in photosynthetic photon flux density (PPFD) along a reproductive shoot resulted in up to 50% differences in cannabinoid concentrations among inflorescences. The plants in this study were 2 to 3 m tall. Lamps are typically mounted in a fixed position within a commercial facility; therefore, crops that elongate ≥1 m will experience large changes in PPFD as the shoots elongate. To improve inflorescence uniformity, spatial gradients in photon distribution must be minimized.
Manipulating canopy architecture by pruning developing shoots has been demonstrated to improve yield and alter biomass partitioning in many crops such as cotton (Gossypium spp.) (Nie et al. 2021), raspberry [(Rubus occidentals × R. idaeus) × R. idaeus] (Gundersheim and Pritts 1991), and peach (Prunus persica) (Anthony and Minas 2021). In ‘Topaz’ cannabis, repeated decapitation of shoot tips improved inflorescence yield and uniformity (Danziger and Bernstein 2021). In a later study, the researchers evaluated four pruning treatments at a planting density of 1 to 2 plants/m2 and found that regardless of pruning treatment, yield was consistently greater in the higher-density treatment (Danziger and Bernstein 2022). These results suggest that increasing the planting density and modifying canopy architecture can improve the yield of cannabis; however, a major challenge with increasing planting density in cannabis production is the high rate of stem elongation that occurs when plants are placed under short-day photoperiods to initiate flowers. The cannabis industry refers to this phenomenon as “flower stretch.” Cannabis growers have traditionally applied excessively high rates of nitrogen and phosphorus (e.g., 250 and 100 mg·L−1, respectively), which also promote stem elongation (Justice and Faust 2014; Kwon et al. 2019; MacIver 2021). Rapid stem elongation is problematic because this results in long, thin shoots that require physical support to remain upright and may grow too close to overhead lamps, resulting in physical damage to developing inflorescences. If stem elongation is reduced, then supplemental lamps could be placed closer to the plant canopy and increase the incident PPFD.
In floriculture production, plant size is often manipulated by restricting internodal growth with plant growth regulators; however, these products are not labeled for use in cannabis production; therefore, nonchemical strategies must be identified. The objective of this research was to explore two methods of restricting stem elongation and manipulating plant dry mass that may facilitate greater planting densities and improve yield based on area. We hypothesized that restricting fertilizer when stem elongation is most rapid during the first 1 to 2 weeks after the plants are moved from a long-day photoperiod to a short-day photoperiod, and that pinching shoots after the start of short days reduces stem length and eliminates the need to physically support developing shoots. Furthermore, we hypothesized that these strategies may reduce per-plant measurements of dry mass or plant area; however, they may improve mass:area and, therefore, yield. Finally, we hypothesized that these strategies may alter biomass partitioning of reproductive shoots by increasing inflorescence mass and/or decreasing trim (stems, leaves, and bracts).
Materials and Methods
Plant material and propagation.
Seventy-two shoot-tip cuttings (length, 6 cm) were harvested from stock plants of ‘Southern OG’ chemotype III cannabis maintained under a 21-h photoperiod. All leaves with a petiole >2 cm were removed, and the basal 2 cm of each cutting was dipped in a rooting solution consisting of 3000 mg·L−1 naphthyl acetic acid. Then, these cuttings were stuck in a 72-cell propagation tray filled with a peat-based germination mix (Sunshine Mix #5; Sun Gro, Anderson, SC, USA) and placed in a shaded propagation greenhouse equipped with intermittent mist. Mist frequency was regulated with a controller (Water Pro II; MicroGrow, Temecula, CA, USA) that was programmed to mist cuttings for 6 s each time the accumulated vapor pressure deficit exceeded 2.0 kPa. Cuttings were provided a PPFD of 28.1 ± 3.0 µmol·m−2·s−1 from 0500 to 0200 HR daily using light-emitting diode (LED) lamps (Hawk Duo 2.0; Sananbio, Xiamen, China). The average daily temperature during propagation was 26.5 ± 0.5 °C, and the average relative humidity 49.3 ± 12.5%.
Vegetative growth.
After 14 d of propagation, all rooted cuttings were transplanted into 1.3-L plastic containers with a peat-based growing medium (Fafard 3B; Sun Gro, Anderson, SC, USA) and provided a continuous liquid fertilizer program (Peters Excel Cal-Mag Special; 15N–5P2O5–15K2O) at 200 mg·L−1 N with each irrigation. Plants were spaced on a greenhouse bench at a planting density of 36 plants/m2. The LED lamps (Hawk Duo 2.0; Sananbio, Xiamen, China) delivered a PPFD of 390 ± 35 µmol·m−2·s−1 from 0500 HR to 0200 HR daily, resulting in a 21-h photoperiod. Sunlight provided a DLI of 12.2 ± 6.2 mol·m−2·d−1. The DLI delivered for 2 weeks after transplanting was 41.7 ± 5.3 mol·m−2·d−1. The average daily temperature during vegetative growth was 21.5 ± 3.3 °C, and the average relative humidity was 73.3 ± 13.6%.
Flowering.
All plants were flowered in a greenhouse equipped with an automated blackout curtain system for photoperiod control. The plants were evenly spaced on greenhouse benches at a planting density of 8 plants/m2. The blackout system opened at 0800 HR and closed at 2000 HR each day to provide a 12-h photoperiod. During the 12-h photoperiod, LED lamps (GreenPower LED HO; Philips, Somerset, NJ, USA) delivered supplemental lighting with a PPFD of 522 ± 69 µmol·m−2·s−1. Sunlight provided a DLI of 13.1 ± 6.9 mol·m−2·d−1. Sunlight and LED lighting combined resulted in a total DLI of 34.1 ± 5.9 mol·m−2·d−1. The average daily temperature was 22.7 ± 2.0 °C, and the average relative humidity was 48.4 ± 14.3%.
Shoot number manipulation.
Shoot number was modified by pinching the shoot tips to produce 4, 8, or 16 shoots/plant (Table 1). At the time of transplanting, all plants were pinched to leave two nodes on the primary stem. Each node consisted of an axillary shoot subtended by a single leaf. The two secondary shoots were pinched to two nodes 1 week later. After this second pinch, the most uniform 36 plants were selected and treatments were randomly assigned to individual treatments (4 plants/treatment). Plants in the 4 shoots/plant treatment were not pinched after the second pinch. For plants in the 8 shoots/plant treatment, the four tertiary shoots were pinched to two nodes 1 week later, which coincided with the start of the inductive 12-h photoperiod. For plants in the 16 shoots/plant treatment, the eight quaternary shoots were pinched to two nodes 1 week after starting the 12-h photoperiod. Thus, the shoot number manipulation (SNM) treatments altered both the shoot number (4, 8, or 16 shoots/plant) and the timing of the last pinch event, which included 1 week before the start of short days (week −1), the start of short days (week 0), and 1 week after the start of short days (week 1).
Timeline for fertilizer restriction and shoot number manipulation treatments that were provided for a 3 × 3 factorial arrangement, resulting in nine treatment combinations. P = pinching of all shoots to two nodes/shoot; F = fertilizer provided; NF = no fertilizer provided. Weeks are numbered relative to the start of the 12-h photoperiod (week 0). Unrooted cuttings were propagated at week −4 and transplanted at week −2. All plants were provided a 21-h photoperiod (PP) from week −4 to week 0. At the beginning of week 0, the substrate was leached with clear water until the substrate electrical conductivity was <0.2 mS/cm. After the leaching event, only tap water was provided during the weeks indicated with NF. Plants were harvested for data collection during week 8.
Fertilizer restriction.
Fertilizer restriction (FR) treatments were applied for 0 weeks, 1 week, or 2 weeks at the start of the 12-h photoperiod (Table 1). To apply FR treatments, the growing substrate was leached with tap water until the electrical conductivity (EC) of the leachate measured <0.2 mS·cm−1. Three drip emitters inserted into the top of the growing substrate cumulatively passed 3 L of water per hour and were allowed to run for at least 5 h. Leachate measurements were conducted using the Pour-Thru method (Mattson 2008) 30 min to 1 h after termination of the leaching treatment. After the growing substrate was leached, plants were irrigated with clear water as needed until the end of the FR treatment. After the FR treatments, plants were again provided with the initial liquid fertilizer program with each irrigation until the end of the experiment. The 0-week FR treatment was provided fertilizer immediately after the initial leaching of substrate, whereas the 1-week and 2-week FR treatments were provided fertilizer starting 1 and 2 weeks, respectively, after the start of 12-h photoperiods.
Data collection.
The experiment was conducted twice. Environmental data were averaged across two replications of the experiment. Each replication was ended 8 weeks after the start of the 12-h photoperiod; then, data were collected. Plant width was measured as the furthest distance between the outermost terminal inflorescences. A second width measurement was performed perpendicular to the first measurement and, again, was a measure of the furthest distance between the outermost terminal inflorescences along that axis. Plant area (m2/plant) was calculated using the two perpendicular width measurements recorded for each plant assuming an oval shape. The reciprocal of the average plant area (m2/plant) per treatment was the calculated plant density (plants/m2) per treatment.
For biomass partitioning data, the primary stem of each plant was cut 2 cm above the substrate surface. Reproductive mass and vegetative mass were separated, and the fresh weight of each was recorded. Reproductive mass was defined as the sum of the lateral shoots that developed above the final pinch. Vegetative mass was defined as the sum of the stems and leaves above the substrate and below the reproductive mass. Total shoot mass refers to the sum of the reproductive and vegetative mass. All plant mass was dried in a growth room set at 20 °C with 55% relative humidity for 3 weeks. Dry mass data are reported on a per-plant basis, whereas yield data are reported based on plant area.
After drying, the length of each reproductive shoot was measured from the base of the cut stem to the uppermost tip of the inflorescence. Reproductive dry mass was divided into inflorescences and trim, and the mass of each was recorded. Inflorescences were defined as trimmed clusters of flowers resembling commercially acceptable, smokeable product, whereas trim was defined as the sum of the bracts, leaves, and internodes that were separated from the inflorescences. The harvest index was defined as the ratio of the inflorescences to total dry mass.
Statistical analysis.
This study used a completely randomized design. Each plant was considered a separate experimental unit. A statistical analysis of data were performed using JMP Pro (version 16.0) (SAS Institute Inc., Cary, NC, USA). An analysis of variance (ANOVA) was performed to determine the significance of FR, SNM, and the FR×SNM interaction for each of the responses measured or calculated at P < 0.05. Tukey’s honestly significant difference post hoc test was performed to determine differences that occurred within each main effect or interaction for each measured response.
Results
Dry mass (g/plant).
We found that FR, SNM, and FR×SNM were significant for vegetative dry mass (Table 2). The interaction was attributable to differing sensitivity to FR within SNM treatments (Fig. 1A–1C). Vegetative dry mass was insensitive to FR in the 4 and 8 shoots/plant treatments because no differences were observed within these SNM treatments. In the 16 shoots/plant treatment, vegetative dry mass was highly sensitive to FR; vegetative dry mass decreased from 16.0 to 8.0 g/plant as FR increased from 0 to 2 weeks, respectively.
‘Southern OG’ cannabis plants were grown under a 12-h photoperiod for 8 weeks, and data were collected. The analysis of variance (ANOVA) output describes the significance of fertilizer restriction (FR), shoot number manipulation (SNM), and their interaction on mass measurements and yield calculations. FR was applied for 0, 1, or 2 weeks following the start of 12-h photoperiods. Shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively.
(A–C) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. The reproductive mass was defined as the sum of the lateral shoots that developed above the final pinch. Vegetative mass was defined as the sum of the stems and leaves above the substrate and below the reproductive mass. Total mass refers to the sum of reproductive and vegetative mass. Fertilizer restriction was applied for 0, 1, or 2 weeks following the start of 12-h photoperiods. The shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
The main effects of FR and SNM were both significant for reproductive mass and total dry mass, but FR×SNM was nonsignificant (Table 2). Each increase in FR or SNM treatment resulted in a significant decrease of the reproductive dry mass; however, increasing FR had a much larger effect on reducing reproductive dry mass than that of increasing SNM (Fig. 1A–1C). For example, reproductive dry mass decreased from 58.0 to 29.0 g/plant as the FR duration increased from 0 to 2 weeks, respectively. In contrast, reproductive dry mass decreased from 54.3 to 37.2 g/plant as SNM increased from 4 to 16 shoots/plant, respectively. Similar trends were observed for total dry mass.
Additionally, FR and SNM affected inflorescence dry mass, trim dry mass, and inflorescence:trim; however, no interaction occurred (Table 2). Increasing FR or SNM negatively affected both inflorescence and trim dry mass (Fig. 2A, 2B); however, the magnitude of decrease for both responses was greater for FR; inflorescence dry mass decreased from 40.6 to 22.3 g/plant as FR increased from 0 to 2 weeks, respectively, which represents a 45% decrease. Increasing SNM from 4 to 16 shoots/plant reduced inflorescence dry mass from 36.0 to 29.1 g/plant, respectively, which represents a 19% decrease. The trends observed in inflorescence dry mass were consistent with trim dry mass. Inflorescence:trim was more sensitive to SNM than to FR (Fig. 2C, 2D); the inflorescence:trim increased from 2.1 to 4.2 as SNM increased from 4 to 16 shoots/plant, respectively, whereas the inflorescence:trim increased from 2.6 to 3.9 as FR increased from 0 to 2 weeks, respectively. The harvest index was affected by FR and SNM, but not by the interaction (Table 2). For example, the average harvest index was 0.61 in the 0- and 1-week FR treatments; however, it increased to 0.65 in 2-week FR treatment (data not shown). In contrast, the average harvest index was 0.64 in the 4 and 8 shoots/plant treatments; however, decreased to 0.60 in the 16 shoots/plant treatment. The effects of both FR and SNM on inflorescence:trim were much larger than that of the harvest index.
(A–D) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. After drying, reproductive mass was divided into inflorescence and trim mass and data were recorded. (A and B) Inflorescence was defined as trimmed clusters of flowers resembling a commercially acceptable smokeable product. Trim was defined as the sum of internodes, bracts, and small leaves that were separated from the inflorescences. Fertilizer restriction was applied for 0, 1, or 2 weeks following the start of 12-h photoperiods. The shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Furthermore, FR was significant for CBD and the total cannabinoid concentration, whereas only FR×SNM was significant for the THC concentration (Table 2). The CBD concentration increased from 13.7% to 14.8% as FR increased from 0 weeks to 1 week (data not shown). No differences were found between the 1- and 2-week FR treatments. Regarding the THC concentration, the FR×SNM interaction was caused by a slight inconsistency in the effect of FR among SNM; in the 4 and 8 shoots/plant treatments, the THC concentration with 1-week FR was higher than that with 0-week FR, but the THC concentrations with 1-week FR and 2-week FR were similar. In the 16 shoots/plant treatment, no differences in THC concentrations occurred with the FR treatments.
We found that FR, SNM, and FR×SNM were all significant for THC, CBD, and total cannabinoid mass (g/plant) (Table 2). The FR×SNM interaction was caused by differing sensitivity to FR within SNM. In the 4 and 8 shoots/plant treatments, no differences were found with 0-week and 1-week FR; however, 2-week FR resulted in a significantly lower THC (Fig. 3A–3C), CBD (Fig. 3D–3F), and total cannabinoid (Fig. 3G–3I) mass. For example, within the 8 shoots/plant treatment, CBD mass decreased from 5.7 to 3.1 g/plant with 1-week FR and 2-week FR, respectively (Fig. 3E). In contrast, each increase in FR led to significantly lower CBD mass in the 16 shoots/plant treatment. The CBD mass decreased from 5.3 to 4.2 to 3.0 g/plant with 0-week FR, 1-week FR, and 2-week FR, respectively (Fig. 3F).
(A–I) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and then analyzed for tetrahydrocannabinol (THC), cannabidiol (CBD), and total cannabinoid mass(g/plant). Fertilizer restriction was applied for 0, 1, or 2 weeks following the start of 12-h photoperiods. The shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Plant area.
Plant area (m2/plant) was affected only by FR, whereas SNM and FR×SNM were nonsignificant (Table 2). Plant area decreased from 0.09 to 0.08 to 0.04 m2 in the 0-week, 1-week, and 2-week FR treatments, respectively (Fig. 4). The reciprocal of the plant area is the planting density (plants/m2). This translated to 11, 13, and 25 plants/m2 during 0-week, 1-week, and 2-week FR, respectively.
‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. Fertilizer restriction was applied for 0 weeks, 1 week, or 2 weeks following the start of 12-h photoperiods. Plant width was measured as the furthest distance between the outermost terminal inflorescences. A second width measurement was performed perpendicular to the first measurement and, again, was a measure of the furthest distance between the outermost terminal inflorescences along that axis. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Yield (kg·m−2).
No FR×SNM interaction was observed among total, vegetative, reproductive, inflorescence, and trim yields (Table 2). The total dry yield was affected only by FR; it increased from 0.70 to 0.80 kg·m−2 in the 0- and 2-week FR treatments, respectively (Table 3).
‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. Yield (kg·m−2) calculations were performed by dividing mass (g/plant) by plant area (m2/plant). The total is the sum of the vegetative yield and reproductive yield. The reproductive yield is the sum of inflorescences and trim yield. Fertilizer restriction was applied for 0 weeks, 1 week, or 2 weeks following the start of 12-h photoperiods. Shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Lowercase letters indicate significant differences within treatments using Tukey’s honestly significant difference post hoc test (α = 0.05).
The vegetative dry yield increased with FR and SNM, but the effect of SNM was much larger (Table 2). For example, the vegetative dry yield increased from 0.09 to 0.12 kg·m−2 as FR increased from 0 to 2 weeks, respectively, whereas yield increased from 0.04 to 0.18 kg·m−2 as SNM increased from 4 to 16, respectively (Table 3). The reproductive dry yield was affected only by SNM (Table 2); it decreased from 0.75 to 0.56 kg·m−2 as SNM increased from 4 to 16 shoots/plant, respectively. The inflorescence dry yield was affected only by FR; it increased from 0.43 to 0.52 kg·m−2 as FR increased from 0 to 2 weeks, respectively (Table 3). The trim yield decreased as FR or SNM increased, but the effect of SNM was greater; the trim yield was similar with 0-week and 1-week FR at 0.18 and 0.19 kg·m−2, respectively, but it was lower, at 0.15 kg·m−2, with 2-week FR (Table 3). In contrast, the trim yield decreased from 0.24 to 0.11 kg·m−2 as SNM increased from 4 to 16, respectively (Table 3). This represents a 54% decrease in trim yield when SNM increased from 4 to 16 shoots/plant.
Both FR and SNM were significant for THC, CBD, and total cannabinoid yield (g·m−2), but no FR×SNM interaction was observed (Table 2). Additionally, THC, CBD, and total cannabinoid yield increased as FR increased from 0 weeks to 1 week; however, no differences were observed with 1 week and 2 weeks of FR (Fig. 5A, 5C, 5E). For example, the CBD yield increased 18%, from 58.8 to 77.9 g·m−2, as FR increased from 0 weeks to 1 week; however, similar values were observed with 1 week and 2 weeks of FR. Similarly, the THC yield increased from 2.8 to 3.3 g·m−2 as FR increased from 0 to 1 week, respectively; however, similar values were observed with 1 week and 2 weeks of FR. For each SNM treatment, THC, CBD, and total cannabinoid yield were greater in the 4 shoots/plant treatment than in the 16 shoots/plant treatment (Fig. 5B, 5D, 5F). For example, the CBD yield decreased from 75.6 to 64.1 g·m−2 in the 4 and 16 shoots/plant treatments, respectively.
(A–F) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. Yield (g·m−2) calculations were made by dividing mass (g/plant) by plant area (m2/plant). The total yield is the sum of all detectable cannabinoids, including tetrahydrocannabinol (THC) and cannabidiol (CBD). Fertilizer restriction was applied for 0 weeks, 1 week, or 2 weeks following the start of 12-h photoperiods. The shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
As plant area increased, opposing trends for mass (g/plant) and yield (kg·m−2) were observed. For example, a positive linear relationship was observed between plant area and inflorescence dry mass, which indicated that as plant area increases so does inflorescence dry mass on a per-plant basis (Fig. 6A). In contrast, a negative linear relationship was observed between plant area and inflorescence dry yield, which indicated that as plant area increases, inflorescence yield decreases on a per-unit area basis (Fig. 6B). Cannabinoid mass and yield responded similarly to plant area; CBD mass (g/plant) increased whereas yield (g·m−2) decreased as plant area increased (Fig. 6C, 6D). Similar trends of THC and total cannabinoid responses to plant area were observed.
(A–D) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. Plant width was measured as the furthest distance between the outermost terminal inflorescences. A second width measurement was performed perpendicular to the first measurement and, again, was a measure of the furthest distance between the outermost terminal inflorescences along that axis. Plant area (m2/plant) was calculated with the two perpendicular width measurements recorded for each plant assuming an oval shape. Inflorescence dry mass (g/plant) was divided by plant area to calculate inflorescence dry yield (kg·m−2). Cannabidiol (CBD) mass (g/plant) was calculated by multiplying the inflorescence dry weight (g/plant) by the dry weight (%) of CBD for each treatment. CBD mass was divided by the plant area to calculate the CBD yield (g·m−2).
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Reproductive shoots.
Fertilizer restriction, SNM, and FR × SNM were significant for reproductive shoot length (Table 2). The FR×SNM interaction was caused by differing sensitivity to FR within the SNM treatments (Fig. 7A–7C). In both the 4 and 8 shoots/plant treatments, reproductive shoot length decreased as FR increased from 1 week to 2 weeks. For example, in the 4 shoots/plant treatment, the reproductive shoot length decreased from 33.3 to 24.6 cm/shoot as FR increased from 1 week to 2 weeks, respectively (Fig. 7A). In the 16 shoots/plant treatment, the observations with 1-week FR treatment were similar to those with 0-week and 2-week FR treatments (Fig. 7C).
(A–F) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. The reproductive shoot length was measured from the base of the cut stem to the uppermost tip of the terminal inflorescence, and the average reproductive shoot length (cm/shoot) (A–C) was recorded. The average reproductive, inflorescence, and trim mass (g/shoot) (D–F) of dried reproductive shoots were also recorded. Fertilizer restriction was applied for 0 weeks, 1 week, or 2 weeks following the start of 12-h photoperiods. The shoot number was manipulated by pinching the developing shoot tips 2, 3, or 4 times to achieve 4, 8, or 16 shoots/plant, respectively. Error bars represent 1 SE from the mean. Nonoverlapping error bars indicate significant differences.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
An inverse relationship between reproductive shoot length and inflorescence:trim was observed (Fig. 8). Data were segmented into two regions (≤10 cm and >10 cm) to calculate the slope of each. Inflorescence:trim decreased approximately nine-times faster with shoot lengths ≤10 cm than with shoot lengths >10 cm. These results suggest that cultivation strategies that seek to improve inflorescence:trim should prioritize keeping the reproductive shoot lengths as short as possible, which can be achieved with FR and/or SNM.
‘Southern OG’ cannabis plants were flowered under 12-h photoperiods for 8 weeks and harvested. The reproductive shoot length was measured from the base of the cut stem to the uppermost tip of the terminal inflorescence. Inflorescences were defined as trimmed clusters of flowers resembling commercially acceptable smokeable product, whereas trim was defined as the sum of the bracts, leaves, and internodes that were separated from the inflorescences. The inflorescence:trim was calculated based on the dry masses of each group of tissues.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Discussion
Both FR and SNM were effective techniques for improving crop performance per unit area, but in different ways. We found that FR reduced plant width, thus allowing for higher plant densities (Fig. 9A), whereas SNM affected dry mass partitioning by increasing inflorescence:trim (Fig. 9B). Both techniques affected dry mass per plant and reproductive shoot length. As with most horticultural techniques, there are trade-offs that must be considered.
(A–C) ‘Southern OG’ cannabis plants were flowered under 12-h photoperiods, and photos were obtained after 3 weeks (A and B) or 8 weeks (C). (A) Each plant was pinched twice to allow four reproductive shoots to develop. (Left to right) Treatments underwent 0 weeks, 1 week, or 2 weeks of fertilizer restriction (FR) at the start of the 12-h photoperiod. (B) All treatments underwent 1 week of FR at the start of the 12-h photoperiod. Each plant was pinched 2, 3, or 4 times to allow development of 4, 8, or 16 reproductive shoots, respectively. (C) All treatments underwent 2 weeks of FR treatment at the start of the 12-h photoperiod. Each plant was pinched 2, 3, or 4 times to allow development of 4, 8, or 16 reproductive shoots, respectively.
Citation: HortScience 59, 10; 10.21273/HORTSCI17968-24
Fertilizer restriction.
High-density planting enhances photon capture per m2, which drives greater photosynthesis, biomass accumulation, and, ultimately, yield (Bugbee and Monje 1992). In crops with vertically oriented leaves, such as wheat (Triticum aestivum L.), maize (Zea mays L.), and rice (Oryza sativa), individual plants increase in height while maintaining a narrow width, which limits interplant competition for light and maximizes photon capture throughout a crop cycle. However, cannabis leaves are horizontal, and both the height and width of individual plants grow. Therefore, methods that control size must be used to grow cannabis at high planting densities, especially for cultivars that undergo rapid stem elongation after the start of short days. In this study, we found that FR was highly effective for controlling plant width, and narrower plant areas correlated to greater yields of inflorescences and cannabinoids per m2 (Fig. 5). However, extrapolating results from per-plant measurements to whole canopies requires uniform environmental conditions throughout the growing area; gradients in temperature and light often found in greenhouse environments will impact growth and introduce plant-to-plant variation. Therefore, although FR is beneficial as a nonchemical means for controlling plant area and facilitating higher planting densities, further work is needed to test limitations of modeling yield based on per-plant measurements.
A tangential benefit of FR is lowering the fertilizer input per crop cycle, which has economic and environmental benefits. When FR is used in combination with preharvest FR, commonly referred to as “flushing,” it may be possible that cannabis plants can be grown with much lower quantities of fertilizer than currently provided in the industry (MacIver 2021). For plants undergoing FR, leaf chlorosis worsened with each day and was followed by leaf abscission in some plants undergoing 2-week FR (Fig. 9C). This induced leaf abscission may be beneficial for reducing labor invested in defoliating leaves at lower node positions on the plant. After FR treatments ended and the liquid fertilizer program resumed, any apparent nutrient deficiencies were resolved within 1 week, whereas the reduced plant area in the FR treatments persisted until the end of the experiment.
Although high-density planting may improve yield, some challenges should be considered. For example, higher planting densities create higher variable costs because of a greater demand for cuttings, containers, and substrate. Cannabis stock plants are often oversized, with poor cutting yield (cuttings/m2) (Faust and Colenbaugh 2023); therefore, transitioning to high-density planting requires redesigning cultivation strategies at all phases of growth. Furthermore, there may be greater pest and disease pressures at higher planting densities because air flow is reduced within the canopy and sprays are less effective at penetrating to the interior. Ultimately, the relationship between yield and production costs must be considered because a lower yield may be tolerable if production costs decrease to a point where the profit margin increases.
Shoot number manipulation.
The SNM treatments did not affect plant area in this study, but they did affect dry mass partitioning and reproductive shoot length. Because pinching plants is a significant labor cost, end-product goals must be considered (flower buds vs. extracts). We found that inflorescence:trim improved with each additional pinch event, which can substantially reduce time and labor associated with trimming reproductive mass into flower buds. Trimming comprises one of the largest costs of cannabis production (Simakis 2021), and cannabis producers track inflorescence:trim as a key performance indicator. Therefore, the additional labor costs from repeated pinches early during the crop cycle could be justified by reducing labor costs during postharvest processing; however, further work is necessary to quantify this trade-off. When plants are grown exclusively for cannabinoid extraction, all reproductive mass is processed together; therefore, no benefit is gained by improving inflorescence:trim.
Both the absolute number of reproductive shoots and the time from final pinch to harvest were affected by the SNM treatments. For example, in the 16 shoots/plant treatment, the lateral shoots developed for a total of 7 weeks from the final pinch to the end of the experiment. This was the shortest interval that reproductive shoots were allowed to develop, and it corresponded to shorter reproductive shoots. In the 4 and 8 shoots/plant treatment, the lateral shoots developed for a total of 9 and 8 weeks, respectively, following the final pinch, and they corresponded to longer reproductive shoots. The time interval from final pinch to harvest as well as the timing of that final pinch during production impact the reproductive shoot length. Commercial cannabis growers may consider shifting the timing of the final pinch to a slightly earlier or later time when growing fast-developing or slow-developing cultivars, respectively.
Regardless of the end-product goals, a substantial benefit of reducing the shoot length by SNM is the elimination of one to two layers of trellis netting needed to physically support top-heavy reproductive shoots found in conventional production systems. Many cannabis growers report challenges associated with using trellis netting, especially when removing plants from the growing benches during harvest. Furthermore, growers expend labor and material costs to replace the multiple layers of trellis netting for each crop cycle; therefore, reducing or eliminating trellis netting is highly desirable. In this study, plants finished with 16 reproductive shoots did not require any physical support, whereas plants finished with 4 or 8 shoots/plant required support for at least one or more shoots per plant. This means that repeatedly pinching plants early during the crop production cycle can eliminate trellis netting, whereas fewer pinches may require some amount of physical support for reproductive shoots. Increasing the FR duration also reduced the reproductive shoot length; however, the magnitude of reduction was much greater as SNM increased.
Reducing the reproductive shoot length with repeated pinching provides an opportunity to move lamps closer to the canopy and increase incident PPFD. The cannabis inflorescence yield reportedly increases with PPFD up to 1800 µmol·m−2·s−1 (Rodridguez-Morrison et al. 2021). In conventional cannabis production, lamps are typically mounted high above the growing benches in anticipation of rapid internodal growth. For reproductive shoots that elongate ≥1 m, large differences in the incident PPFD of developing inflorescences that will affect dry mass and cannabinoid concentration may occur (Danziger and Bernstein 2021). To minimize PPFD gradients, reproductive shoots are pinched to reduce shoot length and, consequently, increase shoot density (shoots/m2); this approach reduces the longitudinal distance from terminal to basal inflorescences along the reproductive shoot as well as improves the height uniformity among reproductive shoots. Then, lamps may be lowered to increase incident PPFD. An added benefit of lowering lamps is that plants moved into flowering rooms receive a much higher PPFD from the start of flowering, which should increase yield. For growers using sole-source lighting environments, reducing the reproductive shoot length also provides an opportunity to increase the use of multitier production systems. Optimizing canopy architecture to improve yield is the subject of research of many other crops such as maize, cotton, wheat, and peach. Future research should seek to understand the dynamics of labor cost and yield with repeated pinching.
Inflorescence quality.
The quality and economic value of cannabis inflorescences are influenced by a combination of inflorescence morphology, cannabinoid concentrations, and aroma. We found that cannabinoid concentrations of inflorescence samples were affected by some of the experimental treatments. Plumb et al. (2022) reported that the aroma of dried inflorescences was far more indicative of a subject’s perception of quality than the absolute cannabinoid or terpene concentration. The researchers found no correlation between the THC concentration and quality of inflorescences; however, they found a significant correlation between desirable aroma and user enjoyment. Inflorescence morphology (e.g., color, shape, trichome density) also plays a large role in quality perception. Consumers who consider a given product attractive will rate the product as higher-quality, which influences purchasing decisions (i.e., the affect heuristic) (Slovic et al. 2007); therefore, small changes in the cannabinoid concentration when using FR or SNM regarding inflorescence quality are likely irrelevant; however, further work is necessary.
When growing cannabis for extraction, cannabinoid concentration is a key parameter. We found that CBD and the total cannabinoid concentration in the 1-week and 2-week FR treatments were greater than those of the 0-week control. This increase in concentration is likely caused by yield dilution (i.e., the concentration of compounds in plant tissues increases as tissue mass decreases) (Wang et al. 2018). The highest inflorescence mass and lowest CBD concentration occurred in the 0-week FR treatment, and the lowest inflorescence mass and highest CBD concentration occurred in the 2-week FR treatment. Similar to the dry mass and yield responses, the CBD mass (g/plant) increased and CBD yield (kg·m−2) decreased as the plant area increased. It should be noted that the total cannabinoid yield is the sum of THC, CBD, and all other detectable cannabinoids (e.g., cannabinol, cannabichromene, and cannabigerol). Although this study largely focused on THC and CBD because these cannabinoids are currently the most valuable, there may be opportunities to target and improve the yield of other cannabinoids as the market continues to evolve and medicinal effects are better understood.
Conclusions
The FR and SNM strategies are effective for improving productivity; however, each technique includes trade-offs that must be carefully considered. These techniques will affect the yield, cannabinoid content, uniformity, labor, and energy efficiency; therefore, a dynamic understanding of how changes during one phase of production impact other phases is required to maximize the benefits of these techniques. From a yield perspective, higher inflorescence yield and cannabinoid concentrations are desirable, but optimizing these parameters may come with a cost to inflorescence quality. The trade-off between yield and quality has been reported for many crops, such as tomato and alfalfa (Feng et al. 2022; Li et al. 2021). Profitability is always most important in commercial crop production; however, how various factors interact to affect the bottom line are complicated and require thorough evaluation. Future research should consider cultivars that differ in vigor and stem elongation as well as the manipulation of specific nutrients such as nitrogen and phosphorus after the start of short days.
References Cited
Anthony B, Minas IS. 2021. Optimizing peach tree canopy architecture for efficient light use, increased productivity and improved fruit quality. Agronomy. 11(10):1961. https://doi.org/10.3390/agronomy11101961.
Bugbee B, Monje O. 1992. The limits of crop productivity. BioScience. 42(7):494–502. https://doi.org/10.2307/1311879.
Bugbee BG, Salisbury FB. 1988. Exploring the limits of crop productivity: I. Photosynthetic efficiency of wheat in high irradiance environments. Plant Physiol. 88(3):869–878. https://doi.org/10.1104/pp.88.3.869.
Danziger D, Bernstein N. 2021. Shape matters: Plant architecture affects chemical uniformity in large-size medical cannabis plants. Plants (Basel). 10(9):1834. https://doi.org/10.3390/plants10091834.
Danziger D, Bernstein N. 2022. Too dense or not too dense: Higher planting density reduces cannabinoid uniformity but increases yield/area in drug-type medical cannabis. Front Plant Sci. 13:713481. https://doi.org/10.3389/fpls.2022.713481.
Faust JE, Colenbaugh C. 2023. Cannabis stock plant management. https://www.inside-grower.com/Article/?srch=1&articleID=26512&highlight=colenbaugh. [accessed 12 Mar 2024].
Feng Y, Shi Y, Zhao M, Shen H, Xu L, Luo Y, Liu Y, Xing A, Kang J, Jing H, Fang J. 2022. Yield and quality properties of alfalfa (Medicago sativa L.) and their influencing factors in China. Eur J Agron. 141:126637. https://doi.org/10.1016/j.eja.2022.126637.
Gundersheim NA, Pritts MP. 1991. Pruning practices affect yield, yield components, and their distribution in ‘Royalty’ purple raspberry. J Am Soc Hortic Sci. 116(3):390–395. https://doi.org/10.21273/JASHS.116.3.390.
Hansen J. 2023. State of the cultivation industry report. https://www.cannabisbusinesstimes.com/article/2023-state-of-the-cultivation-industry-report/. [accessed 12 Mar 2024].
Hawley D, Graham T, Stasiak M, Dixon M. 2018. Improving cannabis bud quality and yield with subcanopy lighting. HortScience. 53(11):1593–1599. https://doi.org/10.21273/HORTSCI13173-18.
Hoffmann CM. 2019. Importance of canopy closure and dry matter partitioning for yield formation of sugar beet varieties. Field Crops Res. 236:75–84. https://doi.org/10.1016/j.fcr.2019.03.013.
Justice A, Faust JE. 2014. Phosphorus-restriction as a potential technique to control Impatiens stem elongation. Acta Hortic. 1104:9–14. https://doi.org/10.17660/ActaHortic.2015.1104.2.
Kwon SJ, Kim HR, Roy SK, Kim HJ, Boo HO, Woo SH, Kim HH. 2019. Effects of nitrogen, phosphorus and potassium fertilizers on growth characteristics of two species of bellflower (Platycodon grandiflorum). J Crop Sci Biotechnol. 22(5):481–487. https://doi.org/10.1007/s12892-019-0277-0.
Lu J, Liu K, Deng J, Feng X, Xiong X, Huang L, Tian X, Zhang Y. 2021. Evaluating the effect of population density and the contribution of early canopy closure to grain yield of hybrid rice. J Plant Growth Regul. 41(2):830–839. https://doi.org/10.1007/s00344-021-10342-1.
Li H, Liu H, Gong X, Li S, Pang J, Chen Z, Sun J. 2021. Optimizing irrigation and nitrogen management strategy to trade off yield, crop water productivity, nitrogen use efficiency and fruit quality of greenhouse grown tomato. Agric Water Manage. 245:106570. https://doi.org/10.1016/j.agwat.2020.106570.
MacIver B. 2021. State of the cannabis nutrients market. https://giecdn.blob.core.windows.net/fileuploads/document/2021/12/07/dec%20soi%202021.pdf. [accessed 12 Mar 2024].
Mattson N. 2008. Testing Container Media. https://www.css.cornell.edu/courses/260/Media%20testing.pdf. [accessed 19 Aug 2024].
Morello V, Brousseau VD, Wu N, Wu B-S, MacPherson S, Lefsrud M. 2022. Light quality impacts vertical growth rate, phytochemical yield and cannabinoid production efficiency in Cannabis sativa. Plants. 11(21):2982. https://doi.org/10.3390/plants11212982.
Nie J, Li Z, Zhang Y, Zhang D, Xu S, He N, Zhan Z, Dai J, Li C, Li W, Dong H. 2021. Plant pruning affects photosynthesis and photoassimilate partitioning in relation to the yield formation of field-grown cotton. Ind Crops and Prod. 173:114087. https://doi.org/10.1016/j.indcrop.2021.114087.
Ortega RG, Gutierrez M, Cavero J. 2004. Plant density influences marketable yield of directly seeded ‘Piquillo’ pimento pepper. HortScience. 39(7):1584–1587. https://doi.org/10.21273/HORTSCI.39.7.1584.
Plumb J, Demirel S, Sackett JL, Russo EB, Wilson-Poe AR. 2022. The nose knows: Aroma, but not THC mediates the subjective effects of smoked and vaporized cannabis flower. Psychoactives. 1(2):70–86. https://doi.org/10.3390/psychoactives1020008.
Rodriguez-Morrison V, Llewellyn D, Zheng Y. 2021. Cannabis yield, potency, and leaf photosynthesis respond differently to increasing light levels in an indoor environment. Front Plant Sci. 12:646020. https://doi.org/10.3389/fpls.2021.646020.
Saloner A, Bernstein N. 2023. Dynamics of mineral uptake and plant function during development of drug-type medical cannabis plants. Agronomy. 13(12):2865. https://doi.org/10.3390/agronomy13122865.
Simakis M. 2021. 2021 State of the trimming market. https://www.cannabisbusinesstimes.com/article/state-of-the-cannabis-trimming-market-research/. [accessed 12 Mar 2024].
Slovic P, Finucane ML, Peters E, MacGregor DG. 2007. The affect heuristic. J Operational Res. 177(3):1333–1352. https://doi.org/10.1016/j.ejor.2005.04.006.
Soltani A, Galeshi S. 2002. Importance of rapid canopy closure for wheat production in a temperate sub-humid environment: experimentation and simulation. Field Crops Res. 77:17–30. https://doi.org/10.1016/S0378-4290(02)00045-X.
Wang S, Wang Z-H, Li S-S, Diao C-P, Liu L, Hui X-L, Huang M, Luo L-C, He G, Cao H-B, Yu R, Malhi SS. 2018. Identification of high-yield and high-Zn wheat cultivars for overcoming “yield dilution” in dryland cultivation. Eur J Agron. 101:57–62. https://doi.org/10.1016/j.eja.2018.08.008.
Westmoreland FM, Kusuma P, Bugbee B. 2021. Cannabis lighting: Decreasing blue photon fraction increases yield but efficacy is more important for cost effective production of cannabinoids. PLoS One. 16(3):e0248988. https://doi.org/10.1371/journal.pone.0248988.