Pecan wood collection and processing.

Pecan wood was obtained from a commercial orchard located ≈10 km north of Las Cruces, NM, in Doña Ana County. The mature orchard was hedge pruned in Dec. 2011 and the branches were windrowed. After windrowing, the branches were shredded using a brush shredder (Pow’rTrak model WS910; Flory Industries, Salida, CA), and the shreds were cut with an orchard flail mower (Vrisimo Model LP2144; Valley Tool Manufacturing Inc., Hughson, CA). The wood was then a heterogeneous, elongated mixture of coarse branch bark and wood with individual pieces ranging in length from ≈5 to 12 cm and a thickness of ≈2 cm or less.

A sufficient volume of pecan wood was allowed to stand on the orchard floor in full sun until 27 Apr. 2012, when it was raked and bagged in a dry condition and stored at the New Mexico State University Fabian Garcia Science Center (FGSC) in Las Cruces, NM, where the study was conducted. On 31 May, the wood was chipped by a five-horsepower garden chipper-shredder (MTD Yard Machines model 245-645B; MTD Products, LLC, Cleveland, OH). The chipping procedure produced a mixture of elongated particles from ≈2 to 5 cm long and ≈0.2 to 0.3 cm wide, resembling wooden match sticks with finer chips intermingled. The chipped wood was passed through a 3.2-mm screen to recover the finer material that was bagged and stored for ≈12 months until use in the June 2013 study. The dominant size of the latter, elongated final chip preparation was around 1 cm long and 0.1 cm wide resembling a narrow version of rice hulls.

Pecan wood substrate preparation.

The substrates were prepared 13 June 2013 at the FGSC using a local commercial grower’s peat-based formulation that contained 67% peatmoss, 21% rice hulls, and 12% rockwool by volume. Wood chip mixing rates are defined as the amount (by volume) of peatmoss substitution in the grower’s substrate: 0% (control, i.e., nonwood treatment), 25%, 50%, 75%, and 100%. Volumetric proportions of wood were, respectively, 0%, 17%, 33%, 50%, and 67%, with reciprocating proportions of peatmoss and constant proportions of rice hulls and rockwool to complete all physical components. All substrates were thoroughly mixed before use. To a 1 m³ equivalent of each of the premixed substrates, the following components were added and mixed thoroughly: 0.6 kg Talstar granular insecticide (FMC Professional Solutions, Philadelphia, PA), 1.7 kg calcitic limestone, 1.1 kg dolomitic limestone, and 0.3 kg each of KNO₃, FeSO₄, and Micromax Micronutrients (Everris NA, Inc.). Substrate preparation volumes were downscaled from 1 m³ for experimental purposes, as described in Picchioni et al. (2014).

Plant culture and greenhouse conditions.

Two-liter pots were filled with the substrates on 13 June in a greenhouse at the FGSC. Tap water served as the only irrigation source and contained (in meq·L⁻¹) Na (2.8), Ca (2.4), Mg (1.0), K (0.1), SO₄ (4.0), Cl (0.5), and HCO₃ (1.8). The tap water also contained P at 1.2 mg·L⁻¹ but no detectable NO₃ or NH₄, and its EC and pH were 0.6 dS·m⁻¹ and 7.9, respectively. Three WSF rates, supplied by 20N–9P–17K general purpose fertilizer (Scotts Co., Marysville, OH) lacking Ca, Mg, and SO₄, but containing all trace elements except Ni and Cl, included N (in mg·L⁻¹ of tap water) at 0 (0.6 dS·m⁻¹), 200 (1.4 dS·m⁻¹), and 400 (2.2 dS·m⁻¹), hereafter referred to as 0–N, 200–N, and 400–N, respectively. The pots were laid out as a split plot in a randomized complete block design with three blocks. Each block was comprised of a single pot for a given wood substitution × WSF treatment combination (further statistical details are forthcoming).

Rooted plants of ‘Carpino’ garden chrysanthemum (liner size 102) were obtained from Zylstra Growers (Kalamazoo, MI) on 18 June, moved to the greenhouse, irrigated with tap water as needed, and planted on 20 June at the rate of three plants per pot. Pots were immediately irrigated with tap water (800 mL/pot), and after 15-min drainage, the initial container capacity weights were recorded. Initial shoot height at planting ranged from 5 to 7 cm. Initial dry weights of roots and shoots (leaves plus stems) from five 3-plant samples were determined after drying to constant weight at 60 °C in a forced-draft oven (Model 1370D, Sheldon Manufacturing Inc., Cornelius, OR). Before weighing the roots, the original liner substrate was gently broken apart and the roots carefully isolated. Initial root dry weight was 0.28 ± 0.01 g per three plants and initial shoot dry weight was 0.60 ± 0.05 g per three plants. Plants were never pinched during the study.

During the 60-d greenhouse growth period (20 June to 19 Aug.), climate data were collected and analyzed using a Watchdog 2475 Plant Growth Weather Station and SpecWare 9 Basic software (Spectrum Technologies Inc., Aurora, IL). Maximum day temperature averaged 34 °C and minimum night temperature averaged 21 °C. Daily relative humidity averaged 53%. Maximum photosynthetically active radiation was 706 µmol·m⁻²·s⁻¹ and the daily light integral averaged 11 mol·m⁻²·d⁻¹. Photoperiod was held constant at 16 h with natural daylength and daylength extension lighting from 5 to 9 am and from 5 to 9 pm provided by metal halide lamps.

Overhead irrigation began on 25 June with WSF applied at each irrigation using cups calibrated to known volumes. To avoid osmotic stress, the initial fertigation for the 400–N treatment was applied at the rate of 200–N and then at 400–N for all subsequent fertigations. Plants were fertigated when the pooled, three-pot wood percentage × WSF treatment weights (recorded daily) declined to a value that represented 50% of the total substrate moisture depletion, and in an amount to cause a targeted leaching fraction of 10%. At planting, we found no differences in container capacity weights between the five wood substitution treatments, but the weights were re-evaluated during the experiment and when increases were detected due to plant fresh biomass accumulation, adjustments to the fertigation timing and volumes were applied.

Leachate was pooled from the three pots per treatment, the volume measured, and the EC and pH were determined using a Myron Ultrameter (model 6P; Myron L. Co., Carlsbad, CA). Leachate collection dates were at the approximate midway and ending points of fertigation (22–25 July and 14–16 Aug., respectively). The EC and pH averages of the two measurements are reported in the text, as is a representative leaching fraction midway through the study. Starting the day after planting, ET was recorded as the daily change in weight per three-pot treatment. The total cumulative ET per three pots up to termination was then statistically assessed as described below.

Plant harvest, processing, and assessments.

The experiment was ended on 19 Aug., which was 60 d after transplanting and 55 d after start of fertigation. Crop height was measured from the pot substrate line to the tallest leaf. Leaves, stems, and roots were then separated, washed three times in deionized water, and blotted dry. The total leaf area per pot was determined using a LI-3100C area meter (LI-COR Biosciences, Lincoln, NE). The total number of leaves per pot was counted and the average area per leaf calculated. Leaf, stem, and root tissues were dried to a constant weight at 60 °C in the forced-draft oven previously described, and weighed. Roots were first isolated from the liner substrate as previously described.

The dried leaves were ground in a Wiley mill to pass a 40-mesh screen, and total N was determined in 100 mg subsamples by combustion in a C-N furnace (Leco TruSpec CN; Leco Corp., St. Joseph, MI) following the methods of Horneck and Miller (1998). In 500 mg subsamples, leaf P, K, Ca, Mg, S, Fe, Mn, B, Cu, and Zn were extracted in nitric and perchloric acid, and determinations were made using an inductively coupled plasma emission spectrometer (Optima 7300DV; Perkin Elmer, Inc., Waltham, MA) using the methods in Miller (1998).

A subsample of the pecan wood chips was ground in the Wiley mill, from which a 5 mg subsample was analyzed for N and C by combustion in a CHNS/O elemental analyzer (2400 Series II; Perkin Elmer, Inc.) by the methods in Wang et al. (2015). The C:N ratio was then determined. In addition, P and K concentrations in a 500 mg subsample of the ground wood were determined as previously described for leaves.

Experimental design and statistical analysis.

Cumulative ET per three pots at Day 59 was statistically assessed as nonreplicated data using a response surface and mean shift method (Cook and Weisberg, 1982). Data were analyzed by developing a regression model with linear WSF rate (0–N, 200–N, and 400–N, coded, respectively, as 0, 2, and 4 for the model fitting), linear wood substitution percentages (0, 25, 50, 75, and 100), the linear WSF X wood substitution interaction, and a mean shift parameter (Cook and Weisberg, 1982) fitted for WSF at 200–N and wood substitution at 0%. This model allowed testing the hypothesis that for 0% wood substitution, growth (as inferred as ET) would be as much at 200–N as at 400–N, thus making the 200–N treatment an exception to continuous, “stepwise” increases in ET with increasing WSF rate for plants receiving wood at 25% to 100% (i.e., incremental increase in ET between 200–N and 400–N). The mean shift parameter was intuitively based on wood substrate studies on chrysanthemum by Wright et al. (2008) and Still et al. (1972), and our preliminary graphical assessment did, in fact, identify the 200–N and 0% wood mean (high ET) as the only outlying data point among all of the other treatment means. The ET data were analyzed using SAS version 9.3 software (SAS Institute, Cary, NC), and significance within wood substitution rates was determined using t tests to compare fitted values at P ≤ 0.05.

Harvest data were subjected to analysis of variance (ANOVA) for a split plot with WSF rate as main plot and wood substitution level as subplot. The WSF rate was randomized within each of three blocks, and wood percentage was randomly assigned to subplots (single pots) within main plots. The ANOVA for crop height, average area per leaf, root:shoot dry weight ratio, leaf mineral concentrations, and the total leaf number, leaf area, leaf dry weight, stem dry weight, and root dry weight per pot was performed using the SAS PROC MIXED software version 9.3 (SAS Institute). Orthogonal polynomial contrasts were used to assess trends in WSF rate within wood substitution percentage and trends in wood substitution percentage within WSF rate.

Pot leaching characteristics.

The target leaching fraction was 10% but the actual leaching fraction exceeded the target with increasing wood proportions. Midway through the study (22–25 July), the leaching fractions in the 0%, 25%, 50%, 75%, and 100% wood substitution treatments ranged, respectively, as follows depending on the WSF rate: 9% to 13%, 17% to 23%, 24% to 28%, 32% to 39%, and 38% to 47%. Previous studies implicate higher air space and lower water holding capacity that lead to higher leaching in wood relative to nonwood substrates (Jackson et al., 2008; Wright et al., 2006, 2008).

In our conditions with soilless substrates, leachate pH and EC vary by less than 5%, thus leachate data provide a close approximation of pour-through analyses. The recommended pour-through leachate pH and EC ranges for greenhouse chrysanthemum production are 5.7–6.2 and 2.6–4.6 dS·m⁻¹, respectively (Cavins et al., 2000). As expected for the 0–N treatment and depending on the wood substitution percentage, average leachate EC was unacceptably low (0.8–1.2 dS·m⁻¹) and average leachate pH was unacceptably high (7.1–7.4). At 200–N, the average leachate EC and pH for the 0% wood substitution treatment were within acceptable ranges (3.1 dS·m⁻¹ and 5.7, respectively), whereas for the 25% to 100% wood substitution treatments, average leachate EC was below the accepted range (1.5–1.6 dS·m⁻¹) and average pH above the accepted range (6.9–7.1). At 400–N, each of the 0%, 25%, and 50% wood substitution treatments provided an average leachate EC in the accepted range (4.5, 3.4, and 3.0 dS·m⁻¹, respectively), whereas the average leachate EC in the 75% and 100% wood substitution treatments was below the accepted range (2.3–2.5 dS·m⁻¹). The average leachate pH in the 400–N treatment was within range for 0% wood substitution (6.1), only slightly above range for 25% wood (6.4), and increasingly above range for 50% to 100% wood (6.6–6.8).

The N composition of the WSF used in our study is dominated by NH₄ and urea, which produce an acidifying effect in substrate solutions (Fisher et al., 2014). At 200–N, the acidifying effect was considerably reduced by any addition of pecan wood, and at 400–N, the acidifying effect was reduced by relatively high wood substitution levels of 50% to 100%. Jackson et al. (2008, 2009) reported increases in pine-amended substrate pH at given fertilizer rates and concluded that those substrates are more weakly buffered against pH change than peat-based substrates. This also seems to apply to pecan wood substrates. Also, any addition of pecan wood strongly limited the nutrient supplying capability of the 200–N WSF treatment as inferred by leachate EC, although WSF at 400–N counteracted that effect at wood substitution levels of 25% to 50% to allow adequate pot nutrition. These findings agree with those of Jackson et al. (2008) who concluded that higher fertilization rates are necessary to maintain wood-based substrate fertility and pH levels similar to those in a conventional nonwood substrate.

Cumulative crop ET.

For cumulative ET per three pots at 59 d, the linear regression model fit the data well (R² = 0.988; F_4,10 = 214.5, P < 0.0001). The fitted model (with se of parameter estimates in parentheses) was cumulative ET = [11.1087 (0.4350)] − [0.0319 (0.0070) Wood %] + [2.7905 (0.1607) WSF-N] − [0.0164 (0.0026) wood % × WSF-N] + [4.5003 (0.6559) mean shift for 200–N and 0% wood]. Estimated model parameters indicated that with pecan wood substitution, cumulative ET increased with WSF rate but the increase depended on the wood substitution level. Without wood chips, once adequate WSF was provided at either 200–N or 400–N, the ET did not differ at around 21 to 22 kg per three pots, but did differ from the 0–N mean of ≈11 kg per three pots (Table 1). At every other wood substitution level (25% to 100%), cumulative ET increased steadily from the 0–N to the 400–N WSF levels, which resulted from similar incremental increases in total shoot biomass and leaf area to be discussed in the following section. Depending on wood substitution level, the total number of fertigation events ranged from 14 to 19 (200–N) and from 18 to 20 (400–N) with frequency generally decreasing with increasing wood percentage (data not shown).

Table 1.

Model-based estimates at each sampling value on cumulative evapotranspiration at harvest (kilograms per three pots) for ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as nitrogen (N) concentration.^z

View Table

The fitted model parameters indicate that for 25% wood substitution, a 100 mg·L⁻¹ unit increase in WSF-N (1 on analysis scale) produced a 2.38 ± 0.11 kg increase in ET per three pots. However, at the 50% wood substitution level, a 100 mg·L⁻¹ unit increase in WSF-N produced only a 1.97 ± 0.09 kg increase in ET per three pots, and at 75% and 100% wood substitution, ET increases corresponding to a 100 mg·L⁻¹ unit increase in WSF-N were only 1.56 ± 0.11 and 1.15 ± 0.16 kg per three pots, respectively. The negative interaction parameter implies that adding WSF was more effective in counteracting wood-related suppressive effects on ET at lower levels of wood substitution.

Leaf, stem, and root growth.

At harvest, final crop height, total leaf number per pot, and total leaf area per pot increased with increasing WSF rate, but the pattern of increase varied among wood substitution levels (Table 2; column-wise trends). The quadratic effect of WSF rate at 0% wood substitution resulted from lack of increases in height, leaf number, and leaf area at 400–N. Conversely, within all wood substitution levels of 25% to 100%, WSF affected height, leaf number, and leaf area linearly and, in some cases, quadratically due to substantial increases from 200–N to 400–N.

Table 2.

Harvest crop height, total leaf number per pot, and total leaf area per pot of ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as nitrogen (N) concentration. Substrate details in Table 1.

View Table

Within the WSF rates of 200–N and 400–N, increasing wood substitution level decreased crop height, total leaf number, and total leaf area (within-row trends in Table 2). In the 0–N treatment, any addition of wood reduced crop height to about one-half that of the 0% wood substitution level. Even with no WSF, plants grown in the nonwood control were as tall as or taller than with 200–N and wood substitution at 50% to 100%. With WSF at 200–N, quadratic responses to wood substitution percentage were associated with steep falloffs in height (50% wood), and in leaf number and area (25% and 50% wood), with small drops at higher wood percentages. At 400–N, increasing the wood percentage resulted in significant linear declines in height, leaf number, and leaf area, but this high WSF rate minimized the suppressions at 25% or 50% wood that occurred at 200–N (Table 2; Fig. 1). Leaf area declines with increasing wood resulted from leaf number reductions, since for average area per leaf (data not shown), neither the wood substitution main effect nor the wood substitution level × WSF interaction were significant (P = 0.3020 and 0.0798, respectively). However, area per leaf did increase linearly with increasing WSF rate (P = 0.0033) with averages of 3.7, 5.4, and 6.0 cm²/leaf for the 0–N, 200–N, and 400–N rates, respectively.

View Full Size

The 60-d harvest appearance of ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as nitrogen (N) concentration in mg·L⁻¹. One complete replication for each of the WSF rates of 0–N, 200–N and 400–N appear from top to bottom, respectively. Wood substitution percentages shown below pots; vertical scale line ≈10 cm. Substrate details in Table 1.

Citation: HortScience 51, 2; 10.21273/HORTSCI.51.2.177

• Download Figure
• Download figure as PowerPoint slide

The wood percentage × WSF interaction was also observed for total leaf and stem dry weights at harvest (Table 3). Namely, the quadratic WSF rate effect at 0% wood substitution was attributed to lack of leaf and stem weight responses at 400–N, whereas there were considerable increases in leaf and stem weights of wood-treated plants at 400–N, especially at 50% and 75% wood substitution. Throughout a WSF-N concentration range of 50 to 400 mg·L⁻¹, Wright et al. (2008) reported a linear increase in shoot dry weight of garden chrysanthemum in a 100% pine wood chip substrate compared with a curvilinear (“tailing off”) shoot growth response in a peat-based substrate, similar to our findings on garden chrysanthemum with pecan wood chips. In addition, the interaction we observed between fertilizer rate and wood percentage on top growth was also observed by Jackson et al. (2008) for shoot growth of several woody crops. They reported greater shoot growth stimulation on a pine wood substrate than on a conventional pine bark substrate with increasing slow-release fertilizer rate.

Table 3.

Harvest leaf, stem, and root dry weight, and harvest root:shoot dry weight ratio of ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as nitrogen (N) concentration. Substrate details in Table 1.

View Table

Quadratic declines in leaf dry weight and stem dry weight occurred across wood substitution levels for plants receiving 200–N, with large reductions at 25% and 50% wood substitution and smaller reductions at higher wood substitution. By contrast, at 400–N, the quadratic effect of wood substitution on leaf dry weight occurred from the relatively small reductions for as much as 50% wood substitution. In addition, the declining linear trend for stem dry weight at 400–N contrasted the steep quadratic stem weight decline at low wood percentages for its counterpart 200–N treatment just mentioned.

Another side of nutrient limitations is on the root response. With 25% and 50% wood substitution, harvested root dry weight was 61% to 91% higher at 200–N than at 0–N, with no additional increases at 400–N (Table 3). At 200–N, a strongly positive quadratic effect of wood substitution on root weight occurred that was weaker at 400–N. Thus, chrysanthemum root weight increased significantly with as much as 50% pecan wood (200–N) or 75% pecan wood (400–N).

The root:shoot dry weight ratio at harvest was relatively high for all wood substitution levels receiving no WSF (Table 3). Different types of quadratic declines in root:shoot ratio with increasing WSF rate were observed, depending on the wood substitution percentage. For the 0% and 25% wood substitution levels, the major declines were at 200–N whereas with the higher wood substitution levels, the major declines were at 400–N. With WSF at 200–N, wood substitution had a quadratic effect on root:shoot ratio, which increased by ≈6-fold at 50% wood and not much more with higher wood levels. With WSF at 400–N, root:shoot ratio increased linearly with increasing wood percentage, but not to such high values as at 200–N. This may indicate that high WSF at 400–N partially circumvented an adaptive need for chrysanthemum to favor roots in dry matter partitioning.

Preferential dry matter allocation to roots is a signature trait of plants in nutrient-limiting conditions, and Wilson (1988) cited numerous reports showing an increase in root:shoot ratio of plants under nutrient deficiency, particularly of major nutrients. Under P-limiting conditions, for example, plants simply grow more roots (Schachtman et al., 1998). Increased root:shoot ratio noted previously raises the possibility that moderate amounts of pecan wood could improve outplanting survival of outdoor crops. Data on root developmental plasticity in wood-based substrates are limited, and are needed to further address this possibility.

Leaf nutrient concentrations.

Within-column polynomial contrasts for leaf N concentration showed highly significant increasing quadratic trends that differed among the wood substitution levels (Table 4). For 0% and 25% wood substitution, 200–N provided the largest increase in leaf N with comparatively small increments at 400–N. Conversely, the quadratic trends for leaf N at 50% to 100% wood substitution resulted from the large increases (2-fold or more) from 200–N to 400–N. A descending quadratic response of leaf N to increasing wood substitution level at 200–N was observed, with the sharpest decline at 50% wood substitution and smaller declines at higher wood levels. At 400–N, a more gradual linear decline in leaf N across the wood substitution levels occurred.

Table 4.

Harvest leaf nitrogen (N) and phosphorus (P) concentrations of ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as N concentration. Substrate details in Table 1.

View Table

A quadratic WSF effect on leaf P concentration without wood chips was, once again, the result of a large increase from 0–N to 200–N and only a slight increase from 200–N to 400–N (Table 4). Leaf P in all of the wood-supplemented treatments increased linearly with WSF rate, although the size of the increase was progressively smaller with each increase in wood substitution level. Declining quadratic trends for leaf P in the 200–N and 400–N WSF rates across the wood substitution levels differed. At 200–N, leaf P declined considerably at 25% and 50% wood substitution with smaller declines at 75% and 100% wood substitution. At 400–N, leaf P declined only with 50% to 100% wood substitution. Thus, for both leaf N and P, WSF at 400–N counteracted much of the effect of increasing wood proportions observed with WSF at 200–N.

Only main effects of wood substitution and WSF rate affected leaf K concentration (P ≤ 0.0246) with no interaction (P = 0.5931; K data not presented). Leaf K decreased with increasing wood proportions of 0% to 100% (3.9%, 3.0%, 3.0%, 2.9%, and 2.9% of dry weight, respectively), and increased with increasing WSF rate of 0–N to 400–N (2.2%, 3.5%, and 3.7% of dry weight, respectively). Thus, increasing the WSF rate to 400–N did not specifically boost leaf K status in wood-supplemented substrate. Increasing the rate of WSF (which lacked Ca and Mg) did not increase leaf Ca and Mg concentrations within given wood substitution levels, and increasing wood proportions did not reduce leaf Ca and Mg concentrations within WSF rates (data not shown). At 200–N, a quadratic decline in leaf S concentration occurred (P = 0.0104), from 0.30% to 0.31% of dry weight with 0% and 25% wood substitution to 0.21% to 0.23% of dry weight with 50% to 100% wood substitution (data not shown). However, at 400–N, leaf S concentrations were 0.33% to 0.35% of dry weight and were independent of the wood substitution level. This may indicate that the increased top growth of plants receiving pecan wood and with WSF at 400–N (lacking SO₄) increased SO₄ demand and uptake from the tap water and preplant fertilizer.

Data in Table 5 suggest that increasing pecan wood substitution levels limited availability of Fe, Mn, Cu, and Zn. Any addition of wood prevented increases in leaf Fe concentration with increasing WSF rate that were observed without the addition of wood (Table 5). Quadratic reductions in leaf Fe within WSF rates of 200–N and 400–N resulted from large declines with 25% wood substitution (200–N) and 25% to 50% wood substitution (400–N), with relatively little change at higher wood proportions.

Table 5.

Harvest leaf iron (Fe), manganese (Mn), copper (Cu), and zinc (Zn) concentrations of ‘Carpino’ garden chrysanthemum subjected to five pecan wood chip substitution rates and three water soluble fertilizer (WSF) rates expressed as nitrogen (N) concentration. Substrate details in Table 1.

View Table

Leaf Mn concentrations were highest in the 0% wood substitution level even with no WSF (Table 5). At 200–N, wood substitution had a quadratic effect in that low or high wood percentages reduced leaf Mn concentration by ≈3- to 4-fold. With wood substitution at 25% and 50%, high WSF (400–N) partially counteracted the steep declines in leaf Mn observed at 200–N, unlike the leaf Fe response.

For all wood substitution levels except 100%, leaf Cu concentration increased linearly with increasing WSF rate (Table 5). Wood substitution had a quadratic effect on leaf Cu at 200–N, with large reductions at the 25% and 50% wood levels and little or no reductions at the higher wood levels. Conversely, at 400–N, wood substitution had a linear effect with leaf Cu decreasing more gradually up to 75% wood.

Increasing the WSF rate increased leaf Zn concentration quadratically in all but the 75% wood substitution level for which a linear increase occurred (Table 5). The largest increases in leaf Zn came at 200–N with smaller increments at 400–N. Within WSF rates of 200–N and 400–N, increasing wood substitution level reduced leaf Zn linearly with relatively little indication that the high WSF rate of 400–N counteracted the declines at lower WSF of 200–N, unlike leaf Mn and Cu. Leaf B concentration (data not shown) ranged from 43 to 85 mg·kg⁻¹ dry weight and was unaffected by wood substitution level (P = 0.0617), WSF rate (P = 0.5825), or the wood substitution × WSF interaction (P = 0.2798).

According to collective leaf nutrient diagnostic criteria for chrysanthemum in Dole and Wilkins (2005) and Nelson (2012), most of the leaf N and P concentrations at 0–N (Table 4) were at deficient levels as expected, although among the trace metals (Table 5), only leaf Cu at 0–N with 50% or 75% pecan wood substitution would be considered as inadequate. At 200–N, leaf N with 50% to 100% wood substitution was insufficient. Leaf K was deficient with any addition of wood. The generalized criteria do not account for unknown cultivar-specific responses to leaf nutrient concentrations (Paparozzi, 2003). With high WSF at 400–N, increasing wood proportions reduced leaf N, P, and trace metal concentrations although not to insufficient levels according to the general criteria. This leaves unanswered the question as to whether those minerals were deficient for this cultivar.