Abstract
Bulk density (Db) and subsequent physical properties are determined by the substrate and packing method. Packing method is the way one fills and compresses a substrate within a given volume. Bulk density produced in the laboratory may not align with “expected” published ranges due to variations in packing. Additionally, it is unknown if ranges identified as “typical” using a small volume sample ring reflect Db occurring in larger production size containers packed using commercial potting practices. Therefore, our objectives were to 1) emulate nursery practices and document the Db associated with a potted 2.8-L (#1) container, 2) develop and test the new “shim and compression” method to determine if it consistently packs sample rings to a Db commensurate with that of a 2.8-L nursery container, and 3) demonstrate how static physical properties are affected by the new “shim and compression” sample ring packing method compared with the traditional bench top tap method. When emulating nursery potting practices with 100% pine bark, coir, and peat, and blends of each, Db ranged from 0.08 g⋅cm−3 (coconut coir) to 0.17 g⋅cm−3 (pine bark). We used an espresso tamp and shims to pack the aforementioned substrates in sample rings. The Db achieved using a range in number of presses and discs was largely dependent on the substrate, but the desired Db was consistently achieved for each substrate. There was no effect of disc number on Db (P = 1.000) for any substrate. There was no effect of tamp number (P ≥ 0.0602) for all substrates except peat-amended, for which five tamps yielded a greater Db than one tamp (P = 0.0324). In an experiment in which a different technician who was accustomed to the conventional benchtop tap packing method performed both methods, method influenced Db (P < 0.0001), and the conventional method more closely attained the target Db. To our knowledge, this is the only report of Db observed in commercial container production facilities (i.e., “native”).
Laboratory testing is often conducted to characterize static physical properties followed by controlled environment or field experiments to develop and refine substrate recommendations for crop producers. Porometers are used to quantify maximum water-holding capacity and minimum air space (AS; Fonteno and Hardin 2010). Static physical properties including container capacity, air space, and total porosity are directly affected by bulk density (Db), the dry mass of substrate per given volume. Nursery substrates are composed of 70% to 90% (v/v) void space (i.e., total porosity), which empirically decreases with increasing Db, subsequently affecting the ratio of water and air that can be retained in a substrate. Even small changes in Db due to substrate aging or compaction during potting can affect the shape and distribution of pores and subsequent amount of plant available water (Fernandes and Corá 2004; Fonteno et al. 1981).
Considering the substantial impact that Db has on substrate physical properties, laboratory packing methods for soilless substrates are intended and assumed to replicate the potting and post-potting practices of a commercial nursery potting line and resultant Db. Yet scant data exist on the physical properties of substrates packed on commercial potting lines with which to validate laboratory Db. Fonteno and Hardin (2010) described in detail the predominate US method (drop-settle, hereafter referred to as bench top tap) for preparing soilless substrate sample cores in the laboratory. In brief, the sample ring is filled and the substrate leveled, then dropped five times from a height of 3 to 6 cm to pack the substrate. The sample core is often used in conjunction with a packing column. In the United States, a less common method is to press the substrate manually into the sample ring to compress it. The precision of either method can be poor due to substrate components, moisture content of material, and laboratory personnel, thus affecting accuracy when comparing data. Furthermore, laboratory packing may not represent that which occurs when commercially producing containerized crops.
Substrate sample rings must be consistently packed in a way that produces a Db commensurate with that of commercial nursery production for the results to be relevant to nursery producers. Therefore, our objectives were to 1) document the Db associated with a 2.8-L (#1) container potted replicating commercial nursery potting line practices, 2) develop a reliable method that uses easy-to-source materials to pack sample rings consistently to a Db commensurate with that observed at commercial production facilities, and 3) compare static physical properties of traditionally packed rings (i.e., bench top tap or manual pack) vs. the experimental method using metal shims (i.e., discs) and a tamp to imitate nursery worker handling of substrates.
Materials and Methods
Experimental overview.
The methods used to fill the container with soilless bark-based substrate, plant the liner, and pack the Db of containers before being transported and placed in their respective production areas were observed at multiple nurseries and later emulated on 2.8-L containers using soilless substrates comprising 100% pine bark (aged; see Supplemental Materials for particle size analysis), 100% Sphagnum peat (Pro-Moss TBK; Premier Horticulture Inc., Rivière-du-Loup, QC, Canada) treated with a surfactant (Aqua-Grow L; Aquatrols Corp of America, Paulsboro, NJ, USA; peat), 100% coconut coir (superwashed; Lynx Associates, Inc., Oakland, NJ, USA; coir), and blends of pine bark and peat [60% pine bark:40% peat by volume (peat-amended)] or coir [60% pine bark:40% coir by volume (coir-amended)] to determine Db representative of typical commercial nursery substrate and their 100% component for comparison. The peat- or coir-amended pine bark were mixed by volume in a 0.099 m3 cement mixer (CME 35, Pro-Series; Buffalo Tools, O’Fallon, MO, USA) for 3 min (± 10 s) each. The respective Db for each substrate component or blend was used to inform packing procedures and establish the target Db in subsequent experiments.
A combination of metal disc shims and an espresso tamp (Calibrated Pressure Tamper, 58 mm diam., 447.8 g; LuxHaus, Sofia, Bulgaria; Fig. 1) were used to achieve the target, representative Db of the aforementioned five soilless substrates in a 251.3-cm3 sample ring. We also examined physical properties via porometers (Fonteno and Hardin 2010) with a 347.3-cm3 sample ring using this new metal disc and shim compression method and the conventional bench tap method.
Metal shim discs and tamp used in the new “shim and compression” packing method.
Citation: HortScience 59, 10; 10.21273/HORTSCI17953-24
Initially, each of the substrates was brought to an appropriate mass wetness (Table 1), which was calculated using the methods described by Fonteno and Hardin (2010). Substrates were then stored in sealed 19-L buckets placed in contractor bags, closed with zip ties, and stored in a walk-in cooler at 4 °C to ensure mass wetness remained constant between uses.
Mass wetness (g water per g substrate) of substrates used for replicating bulk density (g⋅cm−3) of 2.8-L (#1) container in production and for the disc and tamp sample ring packing experiment.
Emulating Db occurring in nursery production.
The potting procedure at Hale and Hines Nursery (McMinnville, TN, USA, 35.72487°, −85.74368°) and Blankenship Farms and Nursery (McMinnville, TN, USA, 35.67547°, −85.89461°) was observed and recorded on video on (17 Mar 2022) to later replicate the Db at commercial nurseries during potting experiments. Subsequently, the observed procedure was emulated at the University of Tennessee Nursery Research Complex (Knoxville, TN, USA, 35.94614°, −83.93820°) using 2.8-L (16.2 cm upper diameter × 18.4 cm height) nursery containers (C300; Nursery Supplies, Chambersburg, PA, USA) and each of the five previously described substrates (n = 3). Two methods were used for potting to explore further the effect of nuances in potting practices on Db:
- 1) Three containers were filled halfway with substrate, then manually compressed until firm as documented at the commercial nurseries. Additional substrate was added and again packed manually until firm. This was repeated until substrate reached a height of 145 mm within the container.
- 2) Three containers were filled halfway with substrate and compressed by hand until firm. Then the container was dropped with moderate force from approximately 5 cm height onto the potting bench three times, rotating the container between drops. The container was then filled to the top with substrate, and firming and dropping were repeated. The container was topped off with substrate so that substrate reached 145 mm height in the container.
Two containers from each soilless substrate component or blend and packing method combination were not irrigated (n = 2). One container from each treatment combination (n = 1) was watered for 20 s by hand using a hose and water wand, replicating the irrigation observed post-potting at the two previously mentioned nurseries, then allowed to drain for 1 h before being weighed again. All containers were transported on a two-wheeled nursery cart, then loaded into a vehicle and driven to the University of Tennessee East Tennessee Research and Extension Center, Plant Sciences Unit (Knoxville, TN, USA, 35.9030°, −83.9533°), where they were dried at 50 °C. Containers were weighed until they stopped losing weight and Db was calculated as dry weight (g) ÷ volume (cm3).
Developing and testing the new sample ring packing method.
A tamp exerting 5.0 N per cm2 pressure and up to five metal disc shims (7.5 cm diam., 0.2 mm thick, 9.0 g each) were used in combination to pack a 5 cm tall × 8 cm wide (inner diameter) metal sample ring (251.3 cm3) (Table 2). For each substrate component or blend, the metal sample ring was filled halfway (2.5-cm height), and then a designated number (one, three, or five) of metal disc shims were placed on the substrate surface. The tamp was placed on the appropriate number of metal shims, and force was exerted by hand either one, three, or five times. The tamp and metal shims were then removed, and a small spatula was used to loosen the top half (∼1.5-cm depth) of the compacted substrate. The sample ring was filled to the top (5-cm height) with more substrate, and the same number of metal shims were placed on the substrate surface and pressed with the tamp the same number of times. The top half (∼1.5-cm depth) of the newly compressed substrate was again fluffed with the spatula. For the third time, the sample ring was filled to the top with substrate and again pressed with the designated number of metal shims and tamps. The top half (∼1.5 cm depth) of the newly added substrate was fluffed, and if necessary, the sample ring was leveled off at the rim with additional substrate. After packing, substrate was emptied from the sample ring into a preweighed aluminum tray and dried for ∼2 d at 50 °C until no further weight loss was detected. The Db for each sample was determined by dividing the dry weight of the substrate by the sample ring volume (251.3 cm3). The tested combinations of discs and tamps are indicated in Table 2; n = 3 for all disc and tamp combinations except five discs and one tamp for peat-amended for which n = 5.
Combinations of metal shims (discs) and espresso tamp compressions tested on two nursery substrate blends and their 100% components for comparison.
For each substrate, a combination of metal discs and tamps were identified that consistently packed a 5-cm-tall, 251.3-cm3 sample ring to the target Db observed in the aforementioned 2.8-L (18.4 cm tall; 2780 cm3) nursery container for use in subsequent experiments (Table 2).
Static physical properties derived from new and conventional packing methods.
For each of the substrates, a 7.6-cm-tall × 7.6-cm-wide (inner diameter) sample ring (347.3 cm3) was prepared using its respective metal disc shim and tamp combination and the traditional bench top tapping method with a packing column. The technician who performed this experiment did not have experience with the new disc and tamp packing method and instead was accustomed to the traditional packing method, whereas all prior experiments using the disc and tamp method were performed by the technician who contributed to the development of that technique and thus was familiar with it. Static substrate physical properties, including Db (g⋅cm−3), total porosity (TP, total pore space; air space + container capacity), air space [AS, maximum pore space filled with air at container capacity (CC); TP – CC], and CC (maximum water holding capacity; TP – AS) were measured on three replicates (n = 3) of each substrate using the 7.6-cm-tall core and inserted into a porometer for analysis (Fonteno and Bilderback 1993).
Experiments were arranged in completely randomized design. A t test or one-way analysis of variance was used to compare means of different substrates, packing methods, and their interaction. Contrast statements were used to compare the effect of metal shims and tamps on Db for each substrate individually; significance was determined at alpha = 0.05 and P values were adjusted using Bonferroni to reduce Type 1 error. For comparison of packing methods on static physical properties, mean separation was conducted using Tukey’s honestly significant difference test.
Results and Discussion
Emulating Db occurring in nursery production.
We found that the nursery-emulated Db for each of the substrates, regardless of how they were packed (i.e., dropped and manually compressed or only manually compressed), irrigated or not, were close in value (Table 3). For pine bark, Db was 0.175 g⋅cm−3 for substrates that were only compressed by hand, whereas those that were dropped and firmed had a Db of 0.170 g⋅cm−3 (Table 3). Containers filled with peat or coir had an average Db of 0.109 and 0.079 g⋅cm−3, respectively, regardless of potting technique or irrigation.
Mean bulk density (Db; g⋅cm−3) ± SD of five substrates derived using four potting techniques that emulate commercial nursery potting and post-potting practices to fill 2.8-L (#1) containers. Components of blends are reported as percent by volume.
For peat-amended substrate, a Db of 0.156 g⋅cm−3 was achieved for manually compressed substrate, regardless of irrigation treatment. Containers potted with peat-amended substrate that were dropped as well as manually compressed had a Db of 0.162 g⋅cm−3, and the container that was dropped, compressed, and irrigated achieved a Db of 0.156 g⋅cm−3.
For the coir-amended substrate, Db of 0.147 and 0.144 g⋅cm−3 were achieved for containers firmed by hand only and firmed by hand and irrigated, respectively. Containers with coir-amended substrate that were dropped as well as firmed had a Db of 0.152 g⋅cm−3, and that which was firmed, dropped, and irrigated had a Db of 0.159 g⋅cm−3.
Substrate composition and packing method impact physical properties as do the dimensions of the container they fill, particularly height (Bilderback and Fonteno 1987; Owen and Altland 2008). Therefore, a substrate sample in a 5-cm-tall × 8-cm-diameter ring may not have the same Db as when in a more typical (i.e., larger) nursery container. The Db documented in the present study varied by substrate and did not always align with Db listed by Fonteno and Hardin (2010). According to Fonteno and Hardin (2010), the Db for pine bark should be between 0.20 and 0.25 g⋅cm−3. However, in this experiment, the Db for PB was 0.170 to 0.177 g⋅cm−3 (Table 3). Bilderback et al. (2005) suggested that pine bark Db should be ∼0.19 g⋅cm−3. For peat, a Db between 0.06 and 0.10 g⋅cm−3 is suggested; in this study the Db was slightly higher for some packing methods, 0.107 to 0.111 g⋅cm−3. In the present study, the Db of coir, 0.079 to 0.084 g⋅cm−3, was very similar to that of Fonteno and Hardin (2010). Fonteno and Harden (2010) stipulate that the Db provided for substrates are for reference; actual values are expected to vary. Moreover, the Db achieved in the potting experiment should be representative of those found in commercial nursery production. To our knowledge, there are no other reports of nursery-emulated Db using container sizes typical of nursery production. Therefore, this information may allow irrigation and soilless substrate scientists to tailor container potting and soilless substrate packing procedures to align better with both commercially relevant potting practices and container height, thereby generating results more reflective of industry conditions. In the future, additional tests with a range of container sizes and a greater number of replications should be conducted using a commercial potting line at several nurseries to establish the “native” Db more comprehensively.
Developing and testing new sample ring packing method.
We found that using metal shims (discs) and compressions was an effective method for preparing sample rings to our nursery-emulated, target Db (Tables 4 and 5). For each of the substrates, the target Db was achieved using one, three, or five discs; there was no effect of disc number (P = 1.000). For each of the substrates except peat-amended, there was no effect of tamp number (P ≥ 0.0602). For these substrates, the target Db was achieved whether using one, three, or five tamps (Tables 4 and 5). For peat-amended, the number of tamps influenced Db, with five tamps yielding a greater Db than one tamp (P = 0.0324), but neither was different from three tamps (Table 5). For all substrates, the variation was low with the standard deviation ranging from 0.001 to 0.006 (Tables 4 and 5).
Mean bulk density (Db; g⋅cm−3) ± SD for sample rings (251.3 cm3) packed with four soilless substrates using a subset of the possible combinations of metal shims (discs) (1, 3, or 5) and espresso tamp compressions (1, 3, or 5). Substrate component is reported as percent by volume.
Mean bulk density (Db; g⋅cm−3) ± SD for sample rings (251.3 cm3) packed with peat-amended substrate using a subset of the possible combinations of metal shims (discs) (1, 3, or 5) and espresso tamp compressions (1, 3, or 5). Substrate component is reported as percent by volume.
The combinations selected for packing sample rings to generate static physical properties were as follows: pine bark, three discs and three tamps; peat, one disc and one tamp; coir, one disc and three tamps; peat-amended, three discs and three tamps; coir-amended, five discs and one tamp (Table 2); however, a number of combinations were possible given the nonsignificance of disc number and the significance of tamps for peat-amended only.
Static physical properties derived from new and conventional methods.
There was an interaction of substrate × packing method for each of the four physical properties (P < 0.0032; Table 6).
Main effect, interaction, and subsequent simple effects of substrate and packing method on static physical properties for sample rings (347.3 cm3 volume; 7.62 cm height) packed with five soilless substrates, reported as percent by volume, using metal shims and tamp compressions compared with the traditional method in which the sample ring is tapped on the laboratory bench to achieve desired bulk density. Mean bulk density (Db; g⋅cm−3; substrate oven dry weight ÷ 347.3 cm3), total porosity (TP; total pore space; air space + container capacity), air space (AS; pore space filled with air at container capacity, total porosity – container capacity) and container capacity (CC; maximum water holding capacity, total porosity – air space).
Db for primary substrate components within a given packing method followed previously established trends (Fonteno and Hardin 2010; Table 6). Pine bark had the greatest Db, followed by peat, and then coir.
Pine bark, a nonfibrous substrate with wide size distribution of platy or rod-like particles, had a 12% greater Db (0.021 g⋅cm−3) when being packed via compression with metal discs and tamp vs. being tapped on the bench (P = 0.0221; Table 6). Pine bark TP increased by 14% (by volume) when compressed vs. tapped (P = 0.0087), resulting in an approximate 10% (by volume) increase in AS (P = 0.0016). This unexpected result could be due to migration of fine particles and reorganization of pores when tapped from the surface, allowing particles to reorganize and settle or “nest” within pores. Packing method did not influence CC (P = 0.2567).
For peat, a fibrous component, Db was 43% greater when packed by compression (P = 0.0001; Table 6). TP was greater when packed by tapping (0.882; P = 0.0192) as was AS (0.022; P = 0.0020) compared with compression, 0.818 and 0.007 for TP and AS, respectively. Container capacity was not influenced by packing method (P = 0.0504).
For coir, also a fibrous component, Db increased 34% when packed via compression vs. tapping (P < 0.0001; Table 6). Coir TP, AS, and CC were unaffected by packing method (P ≥ 0.3254).
For peat-amended substrate, Db was 24% greater when packed by compression, 0.212 g⋅cm−3, than tapping, 0.171 g⋅cm−3 (P = 0.0052; Table 6), but AS was 1.7% (by volume) greater from tapping (P = 0.0335). Neither TP nor CC was affected by packing method (P ≥ 0.3629).
The Db of coir-amended substrate was 24% greater from compression packing compared with traditional benchtop tapping (P = 0.0003). Benchtop tapping resulted in an 8% (by volume) increase in AS (P = 0.0260) and 11% (by volume) decrease in CC (P < 0.0001) of coir-amended substrate. There was no effect of packing method on TP (P = 0.1253).
Shrinkage was a challenge in peat cores, possibly because of particle redistribution during the saturation and drainage phase of the procedure. Peat Db, TP, and AS were influenced by packing method. However, for coir, also a fibrous substrate, TP, AS, and CC were unaffected by packing method (P ≥ 0.3254), suggesting that in this unamended form, coir is less sensitive than peat to variation in packing method. With the exception of coir-amended, CC was unaffected by packing method (P ≥ 0.0504).
For all five substrates, Db was affected by packing method (P ≤ 0.0221; Table 6), but in the present experiment, shim and compression were not always effective at reaching the target Db (Table 3) for a given substrate. This, in part, could be a result of the person’s mastery of packing the sample ring using a new method vs. the traditional method used daily, as was the case for the technician performing this experiment. More importantly, the effect of applying downward force through the sample ring when being dropped or tapped resulted in changes in TP and subsequent air to water ratio in fibrous materials alone or when blended with bark substrate compared with exerting pressure of force to the surface. Greater attention should be given to commercial nursery potting methods to better mimic them in the laboratory. More testing will be needed to ensure replicability across those methodologies and to elucidate their influence on substrate hydrology. Finally, future research should examine the potential to achieve a target Db with substrates across a range of moisture levels because this would give researchers greater flexibility and increase the utility of this packing method.
Limitations and opportunities.
These results suggest that there is potentially great variation in Db due to packing method for at least some substrates and some packing methods. Additional Db tests using a range of substrates and container sizes and potted at nurseries are needed to further establish native Db and generate relevant values for substrate scientists serving the nursery production industry. Technician variation is another factor that needs to be explored with all methods. Further, the combination of discs and tamp compressions needed to achieve accurate Db may vary between technicians, requiring each technician to experiment with packing the sample ring to determine the combination of discs and tamps that will achieve the correct Db. Because of the potential to achieve greater consistency and limit technician-introduced variation, the development of a mechanized or automated method to pack sample rings that is adjustable for a range of substrate components should be explored.
Conclusions
Our results suggest that Db of a nursery-sized container (i.e., 2.8-L) packed to emulate commercial nursery practices may vary from published values (Fonteno and Hardin 2010). The disc and tamp method presented in this article has several advantages when preparing soilless substrate sample cores. The necessary supplies are low-cost and readily available, and no machining or other custom work is required. Additionally, the method achieved the Db associated with common nursery substrates in a highly consistent, repeatable fashion. In a comparison of the new and conventional packing methods, method influenced physical properties, and the target Db was not always achieved. Future research that is needed to address limitations in the current body of knowledge regarding native Db and packing technique is described.
References Cited
Bilderback TE, Fonteno WC. 1987. Effects of container geometry and media physical properties on air and water volumes in containers. J Environ Hortic. 5(4):180–182. https://doi.org/10.24266/0738-2898-5.4.180.
Bilderback TE, Warren SL, Owen JS Jr, Albano JP. 2005. Healthy substrates need physicals too! HortTechnolgy. 15(4):747–751. https://doi.org/10.21273/HORTTECH.15.4.0747.
Fernandes C, Corá JE. 2004. Bulk density and relationship air/water of horticultural substrate. Sci Agric. 61(4):446–450. https://doi.org/10.1590/S0103-90162004000400015.
Fonteno WC, Bilderback TE. 1993. Impact of hydrogel on physical properties of coarse-structured horticultural substrates. J Am Soc Hortic Sci. 118(2):217–222. https://doi.org/10.21273/JASHS.118.2.217.
Fonteno WC, Cassel DK, Larson RA. 1981. Physical properties of three container media and their effect on poinsettia growth. J Am Soc Hortic Sci. 106(6):736–741. https://doi.org/10.21273/JASHS.106.6.736.
Fonteno WC, Hardin CT. 2010. North Carolina State University horticultural substrates lab manual. North Carolina State University, Raleigh, NC, USA.
Owen JS Jr, Altland JE. 2008. Container height and douglas fir bark texture affect substrate physical properties. HortScience. 43(2):505–508. https://doi.org/10.21273/HORTSCI.43.2.505.