Formulation of potting substrate.

Seven substrates were formulated by mixing 60% (by vol.) pine bark (Florida Potting Soils, Orlando, FL) and 10% (by vol.) sand (Sakrete multipurpose sand; Bonsal American, Charlotte, NC) with Canadian sphagnum peat (S), Florida reed-sedge peat (R), and/or composted dairy manure solids (C) (30% by vol.) in the following ratios: 1) 3S:0R:0C, 2) 0S:3R:0C, 3) 2S:0R:1C, 4) 0S:2R:1C, 5) 1S:0R:2C, 6) 0S:1R:2C, and 7) 0S:0R:3C. Canadian sphagnum peat was purchased from Fafard (Apopka, FL) and Florida reed-sedge peat was purchased from Reliable Peat Co. (Okahumpka, FL). Li et al. (2009) reported a pH of 3.9 and 6.9, an electrical conductivity (EC) of 0.32 and 0.31 dS·m⁻¹ and a carbon-to-nitrogen (C:N) ratio of 56.8 and 18.3 for Canadian sphagnum peat and Florida reed-sedge peat (respectively) from the same source purchased at about the same time as our material. Composted dairy manure solids (Agrigy Co., Clearwater, FL) generated at a Lake Okeechobee facility were screened, composted in a large rotary drum composter at temperatures up to 65 °C for 3 d, and allowed to cure in static piles (Nordstedt and Sowerby, 2003). Li et al. (2009) reported a pH of 6.9, an EC of 4.8 dS·m⁻¹, and a C:N ratio of 15.1 for composted dairy manure solids from the same source that were received at about the same time as our material. A bulk sample of sphagnum peat, reed-sedge peat, and composted dairy manure solids from the sources noted was ground and sieved to pass a 0.42-mm screen and analyzed for total N and P. The pH and EC of each substrate were measured using a 1:2 substrate-to-deionized water extract (Mylavarapu, 2009). Before planting, commercially available dolomitic lime was used to adjust substrate pH of substrates containing 3S:0R:0C, 0S:3R:0C, 2S:0R:1C, and 1S:0R:2C from pH 3.5, 5, 4.2, and 4.7, respectively, to pH 6.5.

Plant materials and experimental design.

Three landscape shrub species, walter's viburnum, sandankwa viburnum, and japanese privet, were transplanted on 21 June 2007 from plugs (received 12 June 2007) into 3-gal pots. A single plug of walter's viburnum (one cutting per plug), sandankwa viburnum (three cuttings per plug), and japanese privet (three cuttings per plug) were transplanted per pot. Japanese privet were grown in the greenhouse for a period of 124 d after planting (DAP); walter's viburnum and sandankwa viburnum were grown in the greenhouse for 200 DAP. Additionally, each substrate treatment included four pots containing substrate mixes only (no plant material). Plant and substrate only pots were fertilized (as a topdress) with 71 g/pot of a 15N–3.9P–10K, coated, slow-release, 4-month fertilizer (Osmocote Plus; Scotts, Marysville, OH) on 25 June 2007. Plants were irrigated with potable well water [no measureable NO₃–N, NH₄–N, or dissolved reactive P (DRP)] using a drip irrigation system that was regulated to 15 psi. Each 3-gal pot was outfitted with a single, adjustable-flow dripper (Shrubbler® 360°; Antelco Corp., Longwood, FL) that was set to water at the same rate. The irrigation schedule was created to ensure that plants would have adequate water for optimum growth, with initial rates based on measurements of the volume of water required to saturate the substrates in a 3-gal pot containing no plant material. The initial irrigation run time was set to deliver ≈2.85 ± 0.05 L of water per plant per irrigation event. Emitters and/or irrigation run time were adjusted weekly to increase the irrigation volume delivered to each pot to meet the transpiration needs of the growing plants. Irrigation events were initiated at 0900 hr on days when the substrate surface in most pots (>50%) was dry. All pots were irrigated for the same duration to ensure that irrigation volume was applied consistently to all pots. While the irrigation schedule was not specifically designed to produce leachate, variability in substrate water-holding capacity between treatments and individual pot variability resulted in generation of leachate from pots (mean leachate volume = 182 ± 44 mL per pot per day).

Plant measurements and evaluation.

Plant growth measurements and evaluations were conducted on all plant replicates once plants reached marketable size (124 DAP for japanese privet and 200 DAP for walter's viburnum and sandankwa viburnum) based on the Florida grades and standards for shrubs, trees, and groundcovers (Florida No. 2 minimum grade) (Florida Department of Agriculture and Consumer Services, 1998). Plant quality was visually rated on a scale of 1–5 as follows: 1 = unmarketable, dead plant; 2 = unmarketable, below average quality, significant dieback or pest damage; 3 = marketable, average quality, moderate dieback or pest damage; 4 = marketable, above average quality, minimal dieback or pest damage; and 5 = marketable, excellent quality, no dieback or pest damage. Plant growth index (GI) was used as a quantitative indicator of plant growth and was calculated for each plant as GI (cubic meters) = H × W1 × W2, where H is the plant height (meters), W1 is the widest width of the plant (meters), and W2 is the width perpendicular to the widest width (meters). Shoot biomass was determined by removing all plant tissue at the base and root biomass was collected by washing the substrate from the roots. Shoots and roots were dried to a constant weight at 105 °F (minimum 48-h drying time) and the dry weight (grams) was recorded.

Leachate collection and analysis.

Leachate was allowed to drain freely into a thick plastic collection bucket positioned under each pot. The buckets were tightly fitted to the pots to reduce the potential for the evaporation of leachate between collection events. Pots were also elevated to prevent them from sitting in the standing leachate water. Leachate was collected from each pot one day after planting and then weekly until 17 Sept. 2007 (88 DAP). Total volume of leachate was recorded weekly and a subsample was collected biweekly for nutrient analysis. Leachate that was not saved for nutrient analysis was discarded each week. Leachate was thoroughly mixed and a 100-mL sample of unfiltered leachate was passed through 0.45-μm membrane filters. Filtered samples were stored frozen until they were analyzed for NO₃ + nitrite (NO₂)–N [U.S. Environmental Protection Agency (USEPA), 1993a], NH₄–N (USEPA, 1993b), and DRP (Pote and Daniel, 2009) using a discrete analyzer (AQ2; Seal Analytical, Burgess Hill, UK). Nutrient loads from each pot (in milligrams) were determined by multiplying the concentration of NO₃ + NO₂–N, NH₄–N, and DRP with the volume of leachate collected from each pot.

Statistical analysis.

The experiment was designed as a randomized complete block design with seven substrate treatments applied randomly to five plant replicates within each species. Data were analyzed separately for each plant species. Plant quality data were analyzed using the PROC GLIMMIX program in SAS (version 9.2; SAS Institute, Cary, NC) with the multinomial distribution and the cumulative logit link function. Plant GI, dry root weight, and dry shoot weight data were analyzed using the PROC MIXED procedure in SAS (version 9.2). Pairwise comparisons for GI, dry root weight, and dry shoot weight data were completed on the least-squares means of the data using Tukey's honestly significant difference (hsd) test with a significance level of α=0.05.

Leachate samples with concentrations of NO₃ + NO₂–N, NH₄–N, and DRP below the detection limit (MDL × 2.5) were assigned a value of 0.5 × the detection limit. Leachate DRP loads were cube-root transformed and leachate NH₄–N and NO₃ + NO₂–N were log transformed. All transformed load data were analyzed using PROC GLIMMIX in SAS (version 9.2) using a repeated measures model with substrate treatment as a fixed effect and block as a random effect; this model pooled load data from each pot overall sampling dates. A block-diagonal covariance matrix with identical first order autoregressive structure [ar(1)] on each block was specified. This structure incorporated correlation for all observations occurring from the same experimental unit. All data were checked for normality by examining histogram and normality plots of the conditional residuals. Analysis of transformed load data revealed a significant substrate × collection date interaction; however, further analysis of data by leachate collection date revealed consistent trends related to substrate effect on leachate nutrient loads over time. Consistent trends and the preliminary nature of the data presented indicated that it was appropriate to evaluate how substrate formulation affected average nutrient leaching loads over the entire study period using the repeated measures model. All pairwise comparisons of nutrient load data were completed on the least-squares means of the transformed data using Tukey's hsd test with a significance level of α = 0.05. Arithmetic means are reported for nutrient loads in the text and tables.

Plant quality and growth.

Plant quality ratings were unaffected by substrate composition for any of the evaluated plant species (data not shown). Substrate composition also had no effect on the GI of walter's viburnum (mean GI = 0.14 m³) or shoot or root dry weight of walter's viburnum (mean = 49.2 and 30.5 g, respectively) or japanese privet (mean = 48.3 and 30.5 g, respectively). These results were similar to those of Shober et al. (2010), who reported no differences in the quality, shoot dry weight, or root dry weight of coleus (Solenostemon scutellarioides) and marigold (Tagetes patula) grown for 5 weeks in substrates containing varying percentages of peat (reed-sedge or sphagnum) and/or composted dairy manure solids with perlite and vermiculite. Additionally, Li et al. (2009) reported no differences in shoot or root dry weight of golden pothos (Epipremnum aureum) or ‘Florida Spire’ fig (Ficus benjamina) when grown in substrate containing varying percentages of peat (reed-sedge or Canadian) or composted dairy manure solids with perlite and vermiculite.

In contrast, the GI of sandankwa viburnum and japanese privet was influenced by substrate composition (Fig. 1). Sandankwa viburnum shrubs planted in the 3S:0R:0C substrate had greater GI (0.24 m³) than those shrubs grown in the 0S:3R:0C substrate (0.09 m³) or the 0S:2R:1C substrate (0.08 m³). Shrubs grown in the 2S:0R:1C substrate also produced greater GI (0.23 m³) than those shrubs grown in the 0S:2R:1C substrate. Similarly, japanese privet shrubs grown in the 0S:2R:1C or 0S:3R:0C media exhibited the lowest GI (0.12 or 0.10 m³, respectively); however, the GI of japanese privet grown in these two substrate combinations was only significantly less than shrubs grown in the 1S:0R:2C substrate (0.25 m³). Substrate-induced differences in the shoot and root dry weight occurred only with sandankwa viburnum (Fig. 2), where shrubs grown in the 3S:0R:0C substrate produced greater shoot dry weight (42.4 g) than those shrubs grown in 0S:3R:0C, 1S:0R:2C, or 0S:2R:1C (37.1, 37.2, and 39.6 g, respectively).

Growth index (n = 5) of (A) sandankwa viburnum and (B) japanese privet grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 cm³ = 0.0610 inch³.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

View Full Size

Growth index (n = 5) of (A) sandankwa viburnum and (B) japanese privet grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 cm³ = 0.0610 inch³.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

Growth index (n = 5) of (A) sandankwa viburnum and (B) japanese privet grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 cm³ = 0.0610 inch³.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

Sandankwa viburnum (A) shoot dry weight and (B) root dry weight (n = 5) for plants grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 g = 0.0353 oz.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

View Full Size

Sandankwa viburnum (A) shoot dry weight and (B) root dry weight (n = 5) for plants grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 g = 0.0353 oz.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

Sandankwa viburnum (A) shoot dry weight and (B) root dry weight (n = 5) for plants grown from plugs in substrates containing varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C), each with 60% pine bark and 10% sand. Mean separation by Tukey's honestly significant difference test at P ≤ 0.05; 1 g = 0.0353 oz.

Citation: HortTechnology hortte 21, 2; 10.21273/HORTTECH.21.2.240

• Download Figure

The differences in GI, shoot dry weight, and root dry weight for sandankwa viburnum and japanese privet (GI only) reported in our study did not result in noticeable differences in shrub quality. In all cases where substrate composition influenced plant growth, plant performance was reduced in substrates containing reed-sedge peat (Figs. 1 and 2). Reduced GI of sandankwa viburnum and japanese privet may have been due to differences in the water-holding capacity or porosity of the substrates containing reed-sedge peat as compared with substrates containing sphagnum peat and composted dairy manure solids. Other researchers have reported decreased growth of ornamental bedding plants in reed-sedge peat–based substrates compared with plants grown in substrate formulated with sphagnum peat or coconut coir (Meerow, 1994) or a media combining reed-sedge peat and sphagnum peat (Abad et al., 1989). Abad et al. (1989) reported that air-dried reed-sedge peat had low plant available water (11.24% by vol.), which is lower than the recommended optimal range for bedding plant substrates (24% to 40% by volume). Similarly, Meerow (1994) found that reed-sedge peat based potting substrate had higher air porosity and lower water-holding capacity than media containing the same percentage of sphagnum peat or coconut coir at the beginning of their study. However, Meerow (1994) also noted that air-porosity had decreased and water-holding capacity had increased in the reed-sedge peat–based media by the completion of the study (3–5 months depending on species).

In general, the results of shrub growth and quality that we reported were similar to those of other studies that have evaluated the growth of ornamental plants on a substrate comprising some percentage of composted cattle manure as a substitute for sphagnum or reed-sedge peat (Li et al., 2009; Raviv et al., 2005; Shober et al., 2010; Wang et al., 2004). Differences in shrub growth between substrate treatments that did not result in significant visible differences in aesthetic plant quality would not likely be of concern to growers and nursery professionals. Therefore, we suggest that all substrates examined as part of this study provided acceptable growth and would be suitable for use during the commercial production of ornamental landscape shrubs.

Leachate N loads.

Substrate effects on mean NO₃ + NO₂–N loads leached over the 88-d sampling period were minimal, with no reported substrate effects on mean leachate NO₃ + NO₂–N loads for japanese privet, sandankwa viburnum, or containers filled with substrate only (no plant) (Table 1). The only substrate effect was noted for walter's viburnum, where the 0S:1R:2C substrate leached slightly less NO₃ + NO₂–N on average than other substrates (Table 1). Based on mean total leachate data from the 88-d sampling period, no substrate was more or less likely to leach NO₃ + NO₂–N during the first 88 DAP.

Table 1.

Mean leachate nitrate + nitrite nitrogen (NO₃ + NO₂–N), ammonium N (NH₄–N), and dissolved reactive phosphorus (DRP) loads (n = 5) collected over 12 weeks from woody ornamental plant species grown in 3-gal (11.4-L) containers with substrates composed of varying proportions of Canadian sphagnum peat (S), Florida reed-sedge peat (R), or composted dairy manure solids (C).

View Table

Mean NH₄–N loads leached over the 88-d sampling period was significantly different among treatments as a result of substrate composition (Table 1), where substrate containing the highest proportion of composted dairy manure solids (i.e., 0S:0R:3C) on average leached more NH₄–N than other substrates for all species (Table 1). However, the mean NH₄–N load leached from pots containing the 0S:0R:3C substrate was not significantly different from the mean load leached from the 1S:0R:2C substrate when sandankwa viburnum was grown or containers were without plants (substrate only) (Table 1). Additionally, there were no significant differences between the NH₄–N load collected from pots containing the 0S:0R:3C or 0S:2R:1C substrates when growing walter's viburnum (Table 1).

We hypothesize that the trend for higher leachate NH₄–N loads from the 0S:0R:3C substrate was due to higher concentrations of total N in the composted dairy manure solids (12.2 g·kg⁻¹) compared with the sphagnum or reed-sedge peat (3.8 and 9.2 g·kg⁻¹, respectively). Higher concentrations of total N in the composted dairy manure solids may mean that the 0S:0R:3C substrate contained more potentially mineralizable N than the other substrate formulations. Because organic N must first be converted by microbes to NH₄⁺, it is probable that NH₄–N in substrate exceeded plant needs and led to leaching of NH₄–N from the pots. Additionally, storage of leachate at ambient greenhouse temperatures for 7 d before collection may have increased the NH₄–N concentration and load leached from each pot. An increase in NH₄–N concentrations was reported by Busch and McGinley (2010) when agricultural runoff samples were stored at 35 °C before analysis. The authors suggested the increased NH₄–N was due to greater mineralization of organic N due to microbial activity in the warm temperatures. Despite the increased leaching of NH₄–N from pots containing the 0S:0R:3C substrate, NH₄–N loads from all substrates were an order of magnitude lower than NO₃ + NO₂–N leached.

Leachate dissolved reactive P loads.

Trends in average DRP leachate load over the 88-d collection period were consistent for all evaluated species and pots containing substrate only (no plant). On average, the 0S:0R:3C substrate produced greater leachate DRP load than all other substrates through 88 DAP (Table 1). Over the study period, average DRP loads from leached substrates containing a mix of composted dairy manure solids and sphagnum peat or reed-sedge peat were intermediate between the 0S:0R:3C substrate and the 0S:3R:0C and 3S:0R:0C substrates (Table 1). The trend for higher average DRP loads from substrates containing higher proportions of composted dairy manure solids was likely due to higher concentrations of total P in the composted dairy manure solids (3.9 g·kg⁻¹) compared with the sphagnum or reed-sedge peat (1.1 and 2.2 g·kg⁻¹, respectively). The elevated concentrations of total P in the composted dairy manure solids can likely be attributed to the use of high P dietary supplements (specifically calcium phosphate minerals) in dairy feed, which are excreted from the animal and end up in manure (Shober et al., 2006; Toor et al., 2005).