JARS 43n1 - Controlled-Release Fertilizers And Application Methods

Controlled-Release Fertilizers And Application Methods
A Report On The Effect Of Controlled-Release Fertilizers And Application Methods On Production And Nutrition Of Five Rhododendron Cultivars
C.J. French and J. Alsbury
Agriculture Canada
Sidney, British Columbia, Canada

Three controlled-release fertilizers: Nutricote 16N-4.3P-8.3K ("16-10-10") (NUT), Osmocote 18N-2.6P-10K ("18-6-12") (OSM18) and Osmocote 17N-3P-10K ("17-7-12") (OSM17) and two application methods, incorporation and dibbling, were compared for production of five Rhododendron cultivars, varying in sensitivity to soluble salts.
The cultivars were: 'Anna Rose Whitney' ( R. griersonianum x 'Countess of Derby'), 'Unique', ( R. campylocarpum x), 'Sonata' ('Purple Splendour' x R. dichroanthum ), 'Pink Bountiful' ( R. williamsianum x 'Linswegeanum') and 'Elizabeth' ( R. griersonianum x R. forrestii var. repens ).
NUT and OSM18 stimulated faster growth compared to OSM17. Dibbling fertilizer produced more growth compared to incorporation, the effectiveness being related to cultivar soluble salt sensitivity. Dibbling stimulated flower bud set in 'Pink Bountiful' at the end of the first year. Dibbling in the first year stimulated flower bud set in the second year in 'Sonata' but had no effect on 'Unique'.
There was a strong negative relationship between electrical conductivity (EC) measured in the growing medium and plant growth. Relationships between growth and flowering and leaf tissue analysis were complex and varied considerably between cultivars.

Introduction
Controlled release fertilizers are commonly used in production of many horticultural plants (13). Nursery stock production has increasingly utilized this type of fertilizer, due to advantages in crop management (5).
Earlier studies showed the advantage of using Nutricote 16N-4.3P-8.3K (Type 180: Type 40, 3:1) compared to Osmocote 18N-2.6P-10K for production of 'Anna Rose Whitney' (4). Also, dibbling was superior to uniform incorporation when using controlled-release formulations during production of 'Anna Rose Whitney' (4) and several other soluble salt sensitive plants (8).
Despite being generally classified as soluble salt sensitive, there are considerable differences among various Rhododendron cultivars (2, 6). In view of the large number of Rhododendron cultivars it was of interest to compare the effects of controlled-release formulations and application methods on several broad-leaved Rhododendron cultivars differing in soluble salt sensitivity and flower bud set.
Available data on leaf tissue analysis in broad-leaved Rhododendron cultivars is limited (11, 12, 15). Tissue analysis was conducted in an attempt to relate plant growth to leaf nutrient levels resulting from the various fertilizer and application treatments.

Materials and Methods
Five cultivars of Rhododendron were studied. In approximate order of increasing sensitivity to soluble salts they were: 'Anna Rose Whitney', 'Unique', 'Sonata', 'Pink Bountiful' and 'Elizabeth'. Of these, 'Unique', 'Sonata' and 'Elizabeth' were grown for two seasons to study the effects of fertilizers and application methods on flower bud set.
Cuttings were taken from field-grown stock plants in September 1984, rooted and transplanted into 10 cm plastic containers as previously described (3, 4). Plants were selected for uniformity and transplanted into 2.8 liter containers with milled softwood bark: coarse peat: sawdust: sand (3:1:1:1 by volume) with the following amendments (per m ³ ), 2.1 kg dolomitic lime; 0.75 kg triple superphosphate; 1.1 kg, CaSO ₄ ; 0.15 kg fritted trace elements.
In April 1985, plants were assigned to the following treatments:
1) Nutricote 16N-4.3P-8.3K, Type 180: Type 40, 3:1 by weight (NUT) (Chisso-Asahi Fertilizer Co. Ltd., Tokyo, Japan; distributor, Plant Products Co. Ltd., Bramlea, Ontario, Canada), incorporated at a rate of 4.7 kg. m- ³ ;
2) Osmocote 18N-2.6P-10K, (OSM18), (Sierra Chemical Co., Milpitas, CA), incorporated at 4.2 kg. m- ³ and
3) Osmocote 17N-3P-10K (OSM17), incorporated at 4.4 kg. m- ³ . Dibbling rates (per container) were NUT, 9 g; OSM18, 8g and OSM17, 9g. Rates were calculated to give approximately equal nitrogen content between treatments.
In April 1986, containers were top-dressed with the appropriate fertilizer at the following rates: NUT 9 g, OSM18, 8g and OSM17, 9g. Unfertilized controls, containing mix and amendments only, were observed for each cultivar. Growth was negligible and was not analyzed statistically.
A 3 X 2 factorial design was used with fertilizer type and application method as main effects. Separate complete randomized block designs were employed for each cultivar with 4 blocks and 6 replications per block, one plant as an experimental unit. Statistical treatment was by analysis of variance and single degree of freedom comparisons were made where appropriate. Correlation coefficients were obtained by standard methodology (14).
Plants were grown in a wood-framed, clear fiberglass, unheated greenhouse with 50% light transmission. Large roof and side vents ensured excellent ventilation and temperatures remained within 3°C of ambient. Plants were over wintered in the same building without special precautions. Irrigation was by an overhead suspended sprinkler system, providing 3 cm/day from May-September. Terminal shoots were pinched on 'Anna Rose Whitney' and 'Elizabeth' (first growing season only) to induce branching.
Electrical conductivity (EC) of the growing medium was determined, as described previously (4). pH was determined on a 10 ml aliquot of soil mix by stirring with distilled H ₂ O for 20 minutes, followed by measurement of the pH of the supernatant. Periodically, assessments of shoots, height, mean width and number of flower buds per plant were made. Growth index was defined as height x (mean width) ² , directly related to shoot dry weight (4).
Leaf material was collected from 'Anna Rose Whitney' and 'Pink Bountiful' for tissue analysis in early November 1985. Approximately 50 g fresh weight was collected from mature leaves in the upper-most rosette. After drying at 70°C and grinding to pass a 40 mesh in a Wiley mill, tissue was analyzed for total N (micro-Kjeldahl); P (colorimetry using ammonium molybdate) and K, Mg, Ca, Mn, Fe, Zn, Cu by atomic absorption spectrometry.

Results
First year growth (October 1985)
Shoot growth at the end of the first growing season showed differences due to cultivar, fertilizer and application method (Table 1). In all cultivars, except 'Elizabeth', there was an effect of formulation on number of shoots. In 'Anna Rose Whitney', 'Unique', 'Sonata' and 'Pink Bountiful', NUT and OSM18 produced equal results and were greater than OSM17. In all cultivars, dibbling fertilizer produced more shoots per plant compared to incorporation.
Shoot growth, as assessed by Growth Index, was affected by fertilizer formulation in all cultivars except 'Elizabeth'. In 'Anna Rose Whitney' and 'Unique', the order of growth was NUT > OSM18 > OSM17. In 'Pink Bountiful', NUT = OSM18 > OSM17 and in 'Sonata' OSM18 > NUT > OSM17.
Dibbling fertilizer produced greater growth than incorporation in all cultivars. There was an interaction between fertilizer and application method in 'Sonata' and 'Pink Bountiful', less difference in shoot growth due to application method was observed with NUT compared to the other two formulations.
In certain treatment combinations 'Pink Bountiful' set flower buds at the end of the first growing season. Plants treated with NUT produced more flower buds than OSM18 and OSM17 and dibbling fertilizer produced more flower buds than incorporation. No flower buds were formed on plants treated with incorporated OSM18 and OSM17.

Table 1. Comparison of Osmocote and Nutricote, dibbled (DIB) and incorporated (INC) on first year growth z of five Rhododendron cultivars.
	Osmocote		Osmocote		Nutricote		Significant Effects
	18N-2.6P-10K		17N-3.0P-10K		16N-4.3P-8.3K		Fertilizer	Application	F x A
	INC	DIB	INC	DIB	INC	DIB		Method
'Anna Rose Whitney'
No. shts/plant	5.1	7.5	4.3	6.9	5.2	7.3	* w	**	NS
Growth index y x 10 -3 (cm 3 )	7.8	15.3	4.3	13.5	13.7	20.1	**	**	NS
'Unique'
No. shts/plant	10.4	15.6	7.3	13.9	12.5	15.8	**	**	NS
Growth index x 10 -3 (cm 3 )	2.1	5.4	1.3	4.1	3.4	6.8	**	**	NS
'Sonata'
No. shts/plant	6.8	16.1	4.7	11.7	9.0	14.2	**	**	NS
Growth index x 10 -3 (cm 3 )	2.4	9.4	0.9	5.4	2.5	6.3	**	**	**
'Pink Bountiful'
No. shts./plant	5.0	12.9	4.1	9.3	5.7	10.4	**	**	*
Growth index x 10 -3 (cm 3 )	0.8	3.5	0.5	2.4	1.0	2.6	**	**	*
No. flower buds/plant	0.0	1.2	0.0	1.2	1.4	1.4	*	*	NS
% Plants with flower buds	0	54	0	63	67	63	**	**	**
'Elizabeth'
No. shts/plant	6.8	11.9	4.9	10.0	7.6	11.2	NS	**	NS
Growth index x 10 -3 (ex 3 )	0.8	2.3	0.6	1.8	1.1	3.0	NS	**	NS
z Assessment on October 17, 1985. y Growth index = Ht x (mean width) 2 . x Percentage data subjected to arcs in transformation before analysis. w NS, *, **   Non significant (NS) or significant at the 5% (*) or 1 % (**) level.

Second year growth (October 1986)
Three cultivars were grown on for a second year to study the effects of fertilizer and first year application method on growth and flower bud set (Table 2). 'Unique' and 'Elizabeth' set flower buds readily at the end of two years, whereas 'Sonata' typically sets few flower buds when grown under similar conditions (French, unpublished observations).
Fertilizer type influenced shoot production in all three cultivars. In 'Unique', NUT = OSM18 > OSM17, in 'Sonata', OSM18 > OSM17 > NUT and in 'Elizabeth' the order of shoot production was OSM18 > OSM17 = NUT.
Dibbling fertilizer at the beginning of the first year increased shoot production in all three cultivars at the end of the second year.
Overall growth was affected by fertilizer formulation in 'Unique' and 'Sonata', 'Elizabeth' showing no differences. In 'Unique', NUT = OSM18 > OSM17 and in 'Sonata', OSM18 > NUT > OSM17.
Flower bud set results in 'Elizabeth' were affected by premature opening of the flower buds. During the growing season many flower buds were produced but a substantial number opened before October. Therefore, number of flower buds was not assessed in this cultivar. In 'Unique', flower bud set was not affected by either fertilizer type or application method. In 'Sonata', OSM18 set the most flower buds/plant followed by NUT and OSM17. However, plants treated with NUT showed the greatest percentage of plants setting buds, followed by OSM18 and OSM17. Dibbling fertilizer at the start of the first year in 'Sonata', produced much greater flower bud set compared to incorporation.
In a previous study (4), EC measured in the growing medium was strongly influenced by both fertilizer formulation (NUT and OSM18) and application method and there was a general correlation between plant performance and EC. Since previous work was conducted in full sun the results could have been influenced by direct solar heating of containers with a concomitant acceleration of N-P-K release. It was also of interest to compare EC in medium supplied with OSM17 and OSM18, since the former is designed for a longer release pattern (12-14 months compared to 9 months).
Samples taken at mid-season showed no significant differences in EC due to fertilizer type (Table 3). Dibbling resulted in much lower EC readings than incorporation. At the end of the growing season, there were clear differences due to fertilizer type, NUT producing much lower readings than either OSM formulation. The results were similar to those reported previously (4) for NUT and OSM18, indicating that earlier results were not significantly affected by direct solar heating of containers.
Minor differences in pH resulted from the various fertilizer and application treatments. In view of the limited sensitivity to pH of containerized plants grown in artificial soil mixtures (16), it is unlikely that these differences would be important in determining plant performance.

Table 2. Comparison of Osmocote and Nutricote, dibbled (DIB) and incorporated (INC) in first year; top-dressed in second year, on overall growth z and flower bud set of three Rhododendron cultivars.
	Osmocote		Osmocote		Nutricote		Significant Effects
	18N-2.6P-10K		17N-3.0P-10K		16N-4.3P-8.3K		Fertilizer	Application	F x A
	INC	DIB	INC	DIB	INC	DIB		Method
'Unique'
No. shts/plant	32	43	27	36	33	39	*w	**	NS
Growth index y x 10 3 (cm 3 )	46.5	64.1	33.2	43.1	46.1	58.6	**	**	NS
No. flower buds/plant	4.6	5.3	4.3	6.8	3.9	4.0	NS	NS	NS
% Plants with flower buds x	82	100	96	92	92	96	NS	NS	NS
'Sonata'
No. shts7plant	51	59	39	60	43	42	**	**	**
Growth index x 10 3 (cm 3 )	60.2	91.6	35.9	60.1	59.0	74.0	**	**	NS
No. flower buds/plant	0.8	3.3	0.1	0.7	0.3	2.5	**	**	NS
% plants with flower buds	20	93	5	70	63	93	**	**	NS
'Elizabeth'
No. shts/plant	32	46	28	38	28	35	*	**	NS
Growth index x 10 3 (cx 3 )	44.4	56.7	31.7	41.8	46.5	48.5	NS	NS	NS
z Assessment October 10, 1986. y Growth index = Ht x (mean width) 2 . x Percentage data subjected to arcs in transformation before analysis. w NS, *, **   Non significant (NS) or significant at the 5% (*) or 1% (**) level.

Table 3. Electrical conductivity and pH of growing medium during growth of containerized rhododendrons ('Anna Rose Whitney') in relation to fertilizer and incorporation method.
	Osmocote		Osmocote		Nutricote		Significant Effects
	18N-2.6P-10K		17N-3.0P-10K		16N-4.3P-8.3K		Fertilizer	Application	F x A
	INC z	DIB y	INC	DIB	INC	DIB		Method
Sampling date - July 85
EC x (mmohms. cm -1 )	2.3	0.2	2.1	0.1	0.4	0.1	NS w	**	NS
pH	5.9	6.3	5.9	6.3	6.0	6.3	NS	**	NS
Sampling date - Nov. 85
EC (mmohms. cm -1 )	4.0	0.2	5.9	0.1	0.3	0.1	**	**	**
pH	5.7	6.1	6.1	6.6	6.0	6.2	**	**	NS
z Fertilizer incorporated uniformly in growing medium. y Fertilizer dibbled beneath root ball. x Electrical conductivity of soil extract at field capacity. w NS, *, **   Non significant (NS) or significant at the 5% (*) or 1% (**) level.

Leaf tissue analysis (November 1985)
In 'Anna Rose Whitney', fertilizer type significantly influenced tissue analysis of the following elements and ratios: N, P, K, Mn, Fe, Zn, Mg/Ca, N/ K, P/K (Table 4). NUT produced lower levels of N than the OSM formulations, despite being supplied at equal rates. Mn, Fe, Zn, N/K and P/K were highest in plants treated with OSM18, those treated with OSM17 and NUT being approximately equal. The Mg / Ca ratio was highest in OSM17 and equal in OSM18 and NUT.
Plants treated with dibbled fertilizer had slightly lower values for K and Mg/ Ca compared to incorporation.
In 'Pink Bountiful' (Table 5), only Mn and P/K were significantly affected by fertilizer type. Mn was highest in plants treated with OSM17, followed by OSM18 and NUT. In contrast, application method influenced all elements and ratios considered, with the exception of Ca, Zn and N/K. Dibbling fertilizer compared to incorporation produced higher levels of N, K, and N/P, and lower levels of P and Mg/Ca. Interactions between fertilizer type and application method were noted for Mg, Ca, Mn, Fe, Cu and P/ K. OSM18 produced similar levels of Mg, Ca, Fe and P/K by either application method, whereas concentrations in dibbled plants treated with NUT and OSM17 were lower compared to incorporation. For both OSM18 and OSM17, leaf Cu was lower when fertilizer was incorporated compared to dibbled, whereas in NUT-treated plants dibbling produced higher concentrations compared to incorporation.
In an attempt to identify elements that might be important in influencing plant performance, simple correlation analysis was conducted between plant growth, assessed from growth index and leaf tissue concentrations (Table 6). In both cultivars there was a significant negative correlation between growth and the Mg/Ca ratio. In 'Anna Rose Whitney', growth was positively correlated with Ca, and negatively correlated with K. In 'Pink Bountiful', growth was positively correlated with N, K and N/P. Negative correlations were noted with P, Mg, Fe and P/K.
A similar simple correlation analysis conducted for 'Pink Bountiful' between number of flower buds/plant and element concentrations/ratios revealed a single significant positive correlation with N (r =0.45), all other correlations were not significant at the 5% level.

Table 4. Leaf tissue z of containerized 'Anna Rose Whitney' in relation to fertilizer type and analysis application method.
	Osmocote		Osmocote		Nutricote		Significant Effects
	18N-2.6P-10K		17N-3.0P-10K		16N-4.3P-8.3K		Fertilizer	Application	F x A
	INC y	DIB x	INC	DIB	INC	DIB		Method
N w	1.90	1.80	1.51	1.39	1.36	1.23	** v	NS	NS
P	0.17	0.15	0.12	0.12	0.11	0.11	**	NS	NS
K	1.1	1.0	1.1	1.1	1.0	0.88	**	**	NS
Mg	0.34	0.34	0.32	0.34	0.33	0.35	NS	NS	NS
Ca	1.1	1.2	0.77	1.0	1.0	1.2	NS	NS	NS
Mn	410	370	245	268	294	261	**	NS	NS
Fe	142	157	122	99	111	98	**	NS	NS
Cu	3.0	3.8	3.5	3.0	4.3	5.0	NS	NS	NS
Zn	34	44	28	29	27	24	**	NS	NS
Mg/Ca	0.33	0.27	0.41	0.35	0.33	0.29	**	**	NS
N/P	11.2	12.0	12.6	11.6	12.4	11.2	NS	NS	NS
N/K	1.72	1.80	1.38	1.28	1.38	1.43	**	NS	NS
P/K	0.15	0.15	0.11	0.11	0.11	0.13	**	NS	NS
z Assessment Nov. 1985. y Fertilizer uniformly incorporated through mix. x Fertilizer dibbled below liner root ball. w Values expressed in percentage dry wt (N, P, K, Mg, Ca) and ppm (Mn, Fe, Cu, Zn). v NS, *, ** Non significant (NS) or significant at the 5% (*) or 1 % (**) level.

Table 5. Leaf tissue z of containerized 'Pink Bountiful' in relation to fertilizer type and analysis application method.
	Osmocote		Osmocote		Nutricote		Significant Effects
	18N-2.6P-10K		17N-3.0P-10K		16N-4.3P-8.3K		Fertilizer	Application	F x A
	INC y	DIB x	INC	DIB	INC	DIB		Method
N w	1.70	1.75	1.56	1.75	1.60	1.76	NS v	*	NS
P	0.16	0.15	0.21	0.15	0.19	0.15	NS	**	NS
K	.88	.96	.74	.88	.77	.86	**	**	NS
Mg	0.36	0.34	0.42	0.34	.40	.29	NS	NS	*
Ca	.69	.72	.75	.71	.75	.6	NS	NS	*
Mn	357	409	544	408	370	218	**	*	*
Fe	98	108	127	91	119	74	NS	**	**
Cu	1.5	3.0	1.8	2.5	3.3	1.8	NS	NS	**
Zn	32	42	28	35	33	61	NS	NS	NS
Mg/Ca	0.53	0.47	0.56	0.48	0.54	0.49	NS	**	NS
N/P	10.6	11.6	7.4	11.7	8.4	11.7	NS	**	NS
N/K	1.95	1.83	2.19	2.01	2.15	2.15	NS	NS	NS
P/K	0.18	0.16	.29	0.17	0.24	0.17	**	**	**
z Assessment Nov. 1985. y Fertilizer uniformly incorporated through mix. x F e r t i l i z e r dibbled below liner root ball. w Values expressed in percentage dry wt (N, P, K, Mg, Ca) and ppm (Mn, Fe, Cu, Zn). v NS, *, ** Non significant (NS) or significant at the 5% (*) or 1 % (**) level.

Table 6. Simple correlation analysis of the relationship between growth z and leaf tissue y analysis in 'Anna Rose Whitney' and 'Pink Bountiful'.
	'Anna Rose Whitney'		'Pink Bountiful'
	r value x		r value
N	-0.33		0.54	**w
P	-0.32		-0.49	*
K	-0.61	**	0.44	*
Mg	0.31		-0.55	**
Ca	0.59	**	-0.19
Mn	-0.03		-0.24
Fe	-0.24		-0.43	*
Cu	0.36		0.28
Zn	-0.44		-0.31
Mg/Ca	-0.75	**	-0.71	**
N/P	0.08		0.48	*
N/K	0.015		-0.14
P/K	-0.052		-0.56	**
z Growth index = Ht x (mean width) 2 , assessed Oct. 1985. y Sampled Nov. 1985. x Correlation coefficient, 23 df. w *, **   Significant at 5% (*) or 1 % (**) level.

Discussion
Genetic origin of soluble salt sensitivity
The Rhododendron cultivars used were chosen to provide a range in tolerance to soluble salts. 'Anna Rose Whitney' represents a group of medium/large cultivars that are relatively tolerant to soluble salt levels in the growing medium (2). At the other extreme of sensitivity is 'Elizabeth', where controlled-release fertilizer must be applied with extreme care to avoid leaf burn and plant death (French, unpublished observations). The low tolerance of 'Elizabeth' is apparently due to parentage from R. forrestii , a species reported as particularly sensitive to soluble salt concentrations (2). The other three cultivars form an intermediate group. The Neriiflora sub-section is noted for general sensitivity to soluble salts (2) and includes R. dichroanthum ,a parent of 'Sonata'. The moderate sensitivity of 'Pink Bountiful' is probably due to 25% parentage from R. forrestii . Parentage of 'Unique' is less well-known ( R. campylocarpum x); however, the intolerance of this cultivar to excessive fertilizer has been reported (15).

Growth and flower bud set
In general, 'Unique' and 'Anna Rose Whitney' responded similarly to fertilizers and application methods during first year growth. In 'Sonata' and 'Pink Bountiful', with increased sensitivity to soluble salts, differences between fertilizers became less important and the magnitude of the dibbling effect compared to incorporation became greater. In 'Elizabeth', poor growth was observed during the first growing season in all treatments and there was evidence of leaf tip and margin burn typical of excessive soluble salts, especially in the early stages of growth. In such extremely sensitive cultivars, rates of fertilizer may have to be reduced to avoid setbacks in growth at the liner transplanting stage, regardless of application method.
Rapid early development was stimulated by dibbling in all cultivars except 'Elizabeth' (results not shown). The pattern of initial growth generally corresponded to early release of nutrients from the fertilizers. NUT contained a Type 40 formulation designed to release 80% of N in 40 days at 25°C, in addition to the main Type 180 formulation. Omission of the Type 40 component from NUT results in very poor growth of transplanted Rhododendron liners (Alsbury, unpublished observations). With OSM18 and OSM17, differences in growth must be directly related to the pattern of nutrient release, since the two fertilizers have a similar chemical composition. Early release of N from OSM18 (0-75 days) is considerably faster than OSM17, whereas the reverse is found later in the season (9). These results support the concept that a fast N release component is necessary to obtain maximum growth in the first growing season.
At the end of the second growing season there was a stimulatory effect of dibbling in the first year on vegetative growth in 'Unique' and 'Sonata'. A similar trend was observed for 'Elizabeth', although the results were not statistically significant. These results contrast with those previously found for 'Anna Rose Whitney' (French, unpublished data), in which the stimulatory effects of dibbling in the first year were lost during the second growing season. With the more sensitive cultivars, the magnitude of the first year dibbling effect was apparently sufficient to maintain an advantage during second year growth. In 'Elizabeth', surface-applied fertilizer resulted in much improved growth compared to the first year and tended to obscure the effects of first year dibbling.
Flower bud set in 'Unique' was not affected by either fertilizer type or first year application method. Dibbling in the first year stimulated flower bud set in the second year in 'Sonata', a result similar to previous observations with 'Anna Rose Whitney' (French, unpublished data). Of these three cultivars, only 'Unique' sets flower buds readily in containers at the end of the second year under the experimental conditions (French, unpublished data). Therefore, the beneficial effects of first year dibbling on second year bud set may be confined to cultivars that do not set flower buds readily in containers.
The results confirm previous observations (4, 7) that soluble salt levels with OSM18 greatly exceed those produced by NUT when the fertilizers are incorporated. Similar results were noted with OSM17, despite prill coatings designed to slow diffusion of solutes and extend the life of the fertilizer. These results are in accordance with the release patterns of OSM18 and OSM17 under field conditions (9). There was a strong relationship between EC in the growing medium and vegetative growth/flower bud set, the magnitude of the response being related to the relative sensitivity of the cultivars to soluble salts. This finding re-emphasizes the importance of low EC levels in the growing medium as a factor in determining growth in these plants.
The mechanism of the dibbling effect is unknown, although there appears to be a general beneficial effect on soluble salt sensitive plants (4, 8). Such effects are confirmed in the present study. The relationship between dibbling and low EC values suggests that placement of fertilizer away from feeder roots at the perimeter of the containers provides a favorable environment for nutrient uptake by capillary action, avoiding direct contact between roots and fertilizer prills.

Tissue analysis
Previous studies have shown correlations of N with flower bud set in both field-grown and containerized plants (12) and N was positively associated with plant quality in 4 out of 6 cultivars (15). In the present study, N supplied to the plants was equalized between treatments but tissue levels showed a considerable range of values as a consequence of fertilizer and application methods. N was positively correlated with both vegetative growth and flower bud set in 'Pink Bountiful' but there was no relationship in 'Anna Rose Whitney'. No general correlation between N and growth was found.
Plant quality was positively related to P in 5 out of 6 Rhododendron cultivars (15) and the role of P in Rhododendron flower bud set has been shown in field-grown plants (10). However, later results showed that the P effect could be indirect and related to co-absorption with N (12). N/P was negatively associated with plant quality in 4 out of 6 cultivars (15) and N/P at the end of the first growing season was negatively correlated with flower bud set at the end of the second year (11). In the present study, P was negatively correlated with growth in 'Pink Bountiful' and showed no other correlations. Levels with NUT tended to be low, (4) but this did not adversely affect flowering. No trend was noted for N/P, being positively correlated with growth in 'Pink Bountiful' and not correlated with vegetative growth in 'Anna Rose Whitney'. No general conclusion on the importance of P or N/P can be drawn from these results.
Higher tissue levels of K were associated with poor quality in 5 out of 6 Rhododendron cultivars, the exception being 'Unique', which followed a reverse trend (15). In the present study, 'Anna Rose Whitney' showed negative correlation of growth with K, whereas 'Pink Bountiful' showed a positive correlation, re-emphasizing differences between cultivars in this regard.
The roles and interrelationships between Ca and Mg in Rhododendron nutrition are complex; for example, Mg may function as a counter-ion preventing Ca toxicity on calcareous soils (6) and Ca may be a growth stimulant (1). In previous work the ratio of Mg/Ca was more important than the levels of the individual elements and Mg/Ca was negatively associated with plant quality in all cultivars investigated (15). In the present study there was little variation in Mg between treatments, although Mg showed a negative correlation with growth in 'Pink Bountiful'. Ca was positively correlated with growth in 'Anna Rose Whitney' and in both cultivars there was a strong negative correlation between Mg/Ca and first year growth. The Mg/Ca ratio appears to be an important factor related to growth in Rhododendron, although the physiological role is obscure.
These and previous studies on tissue nutrient levels in broad-leaved Rhododendrons (15) show considerable quantitative differences between cultivars. Upper and lower values for nutrients in relation to growth/flowering are yet to be established and failure to detect a correlation may simply be due to the data falling within a non-critical range. With this background, interpretation of tissue analysis in broad-leaved Rhododendron must be preliminary. Many more analyses are needed before precise nutritional strategies can be formulated for this diverse group of plants.

Literature Cited
1.  Close, F.J. 1967. Gypsum as a Rhododendron growth stimulant. Plant Propagator 12:15-16.
2.  Cox, P.A. 1985. The Smaller Rhododendrons . Timber Press, Portland, Oregon.
3.  French, C.J. 1985. Effect of supplementary lighting on rooting of rhododendrons. HortScience 20:706-708.
4.  French, C.J. and J. Alsbury. 1988. Comparison of controlled-release fertilizers, dibbled and incorporated for production of containerized Rhododendron 'Anna Rose Whitney'. HortScience , (In press).
5.  Johnson, C.R. 1979. How to fertilize plants in containers. Amer. Nurseryman 150:12-13, 105-108.
6.  Leach, D.G. 1960. Rhododendrons of the World . Charles Scribner's Sons, New York.
7.  Lumis, C.P and A.G. Johnson. 1985. Nutricote for container-grown nursery stock. Landscape Trades 7:7-9.
8.  Meadows, W.A. and D.L. Fuller. 1983. Relative effectiveness of dibble applied vs. incorporated Osmocote for container grown woody ornamentals. Proc. S.N.A. Res. Conf. 29:63-66.
9.  Meadows, W.A. and D.L. Fuller. 1983. Nitrogen and potassium release patterns of five formulations of Osmocote fertilizers and two micronutrient mixes for container grown woody ornamentals. S.N.A. Nursery Res. J. 9:28-34.
10. Myhre, A.G. and W.P. Mortensen. 1964.The effect of phosphorus on rhododendron flower-bud formation. J. Am. Rhod. Soc. 18:66-71.
11. Polites, L. 1973. Effects on fertilizer nitrogen and phosphorus on growth and flower bud development of Rhododendron 'Roseum Elegans' and 'America'. Ph.D. Diss., Univ. of Delaware.
12. Ryan, C.F. 1979. Effects of nitrogen and phosphorus on flower bud formation in Rhododendrons . Washington State Univ. College of Agric. Res. Bui. 872.
13. Sharma, G.C. 1979. Controlled release fertilizers and horticultural applications. Sci. Hort. 11:107-129.
14. Steel, R.C.D. and J.M. Torrie. 1960. Principles and Procedures of Statistics . McGraw-Hill, New York.
15. Ticknor, R.L and J.E. Long. 1978. Leaf analysis survey establishes standards for rhododendron production. Am. Nurseryman 148:14-15, 38, 40.
16. Whitcomb, C.E. 1983. Does pH really have an effect on nutrition of container-grown plants? Amer. Nurseryman 158:33-35.

Technical assistance from T. Helmer (tissue analysis), B. Ferrie and J. Paynter is gratefully acknowledged.

Dr. French is currently at Agriculture Canada Research Station, Vancouver, British Columbia, Canada.