Effects of rootzone composition and irrigation regime on performance of velvet bentgrass putting greens. I. Turf quality, soil water repellency and nutrient leaching

Abstract

The use of velvet bentgrass (Agrostis canina L.) on putting greens is limited by sparse knowledge on optimal maintenance. The objectives of this study were to determine the effects of rootzone composition (SS, straight sand; or GM, sand amended with 20% v/v garden compost) and irrigation regime (LF, light and frequent; or DI, deep and infrequent) on turfgrass visual quality, soil water repellency and nutrient leaching. The study was conducted from August 2007 to October 2009 on an experimental USGA green seeded in June 2007 with velvet bentgrass ‘Legendary’ at a coastal location in Norway (Landvik, 58° N). Better turf performance on GM than on SS was associated with less water repellency at most investigated depths in the 10-cm top layer, 88% lower nitrogen loss in the form of nitrate/nitrite during the first year after sowing, less injury caused by Microdochium nivale, and quicker recovery during both spring periods. A decrease in turf visual quality on SS receiving light and frequent irrigation in the second year was associated with strong water repellency in the mat layer. Compared with light and frequent irrigation, deep and infrequent irrigation resulted in better turf quality and lower drainage volumes.

Keywords:

• Golf greens
• turfgrass
• turfgrass management

Abbreviations

DI	=	deep and infrequent
GM	=	organic amendment ‘Green Mix’
LF	=	light and frequent
SS	=	straight sand.

Introduction

Due to water, fertilizer and pesticide restrictions, there is increasing interest for velvet bentgrass (Agrostis canina L.) as an ideal species for integrated pest management of putting greens in the USA, Canada, and Europe. Velvet bentgrass performs better at low nitrogen rates (Skogley 1975, Espevig et al. 2012), is more resistant to dollar spot (Sclerotinia homoeocarpa) and brown patch (Rhizoctonia spp.) (Brilman and Meyer 2000), needs less irrigation water (DaCosta and Huang 2006a, 2006b) and exhibits lower leaching of nitrates (Paré et al. 2006) than other bentgrass species. This very fine-textured turfgrass is also more tolerant to shade (Reid 1933) and wear stress (Murphy et al. 2009), and it competes better against annual bluegrass (Poa annua L.) (Samaranayake et al. 2009) than creeping bentgrass (Agrostis stolonifera L.).

Despite these advantages, the use of velvet bentgrass on putting greens is limited. Among the reasons is few experimental data on optimal maintenance of this turfgrass species regarding requirements to rootzone and irrigation regimes. Globally there is focus on water conservation, and much work has been devoted to effects of reduced irrigation on turf quality and plant health (Qian and Fry 1996, DaCosta and Huang 2006a, 2006b, McCann 2008). Irrigation influences leaching and runoff of nutrients and pesticides from putting greens (Mancino and Troll 1990, Brauen and Stahnke 1995, Starrett et al. 1995, 2000, Barton and Colmer 2006). Studies showed that infrequent irrigation may lead to development of soil water repellency causing fingered flow and leaching from sandy soils (Bauters et al. 1998, Nektarios et al. 1999, Larsbo et al. 2008). Therefore, in spite of a mostly lower soil water content, deep and infrequent irrigation is often considered to cause more drainage and nutrient leaching than light and frequent irrigation (Kenna and Snow 2000, Barton and Colmer 2006). This situation may, however, be different in a coastal climate where natural rainfall often results in oversaturation and thus drainage from the turfgrass rootzone.

The objective of this study was to determine the effects of rootzone composition and irrigation regime on turfgrass visual quality and nutrient leaching from velvet bentgrass putting green. Our hypotheses were: (1) Velvet bentgrass greens can be maintained at higher turf quality levels and less leaching if compost is included in the rootzone; and (2) in temperate coastal climates with high annual precipitation, deep and infrequent irrigation, as opposed to light and frequent irrigation, will provide better turfgrass quality and reduced leaching losses on velvet bentgrass greens.

Materials and methods

Site and weather conditions

The study was conducted from 21 Aug. 2007 to 1 Oct. 2009 on an experimental USGA green (USGA Green Section Staff 2004). The depth of rootzone was 30 cm and the depth of underlying gravel was 10 cm. The experimental green contained 16 stainless steel lysimeters 1 m×2 m, earlier described by Aamlid et al. (2009), at Bioforsk Øst Landvik, Norway (58° 9′ N; 8° 30′ E, 5 m a.s.l.). The surface area of each plot including lysimeter and boundary area was 6 m². In the spring of 2007, the sod from all plots was removed and replaced with a new 4 cm top layer with slightly finer texture but the same type and amount of organic matter as in the initial rootzone. The green was seeded with velvet bentgrass ‘Legendary’ at a rate of 6 g m⁻² on 8 June 2007, and observations and irrigation treatments started once plant coverage was close to 100% in late August.

Being located on the Norwegian south coast, the experimental site has a moderate climate with fairly mild winters and relatively high annual precipitation. Twenty-three and 70 days of snow cover were registered during the winters 2007–08 and 2008–09, respectively. The mean monthly temperature, monthly precipitation, and monthly evaporation (Thorsrud 2500 evaporation pan; Riley and Berentsen 2009) for the grow-in period and the experimental period are shown in Table I. In 2007 and 2008, the total precipitation records for June–August were, in turn, 178 mm and 169 mm higher than the 30-yr normal value. The highest evaporation values were recorded in June 2008 and 2009 (87 mm and 89 mm respectively). In 2009 a mobile rain-out shelter was constructed; this came into operation on 1 July, but unfortunately malfunctioned during a heavy rainfall on 18 and 19 July. From 20 July 2009 the rain-out shelter worked well for the rest of the experiment. Apart from the period covered by the rain-out shelter, the longest continuous period without rainfall was from 2 May until 13 June 2008.

Table I. Mean monthly air temperature, monthly precipitation and monthly (only for the irrigation period) pan evaporation at Landvik during the grow-in period (8 June–21 Aug. 2007) and experimental period (21 Aug. 2007–1 Oct. 2009) compared with 30-yr normal values.

	Air temperature				Total precipitation				Total pan evaporation
Month	2007	2008	2009	Normal†	2007	2008	2009	Normal†	2007	2008	2009	Normal‡
	° C				mm				mm
January	–	2.8	0.5	−1.6	–	358	178	113	–	–	–
February	–	4.3	−2.2	−1.9	–	82	65	73	–	–	–
March	–	2.3	2.7	1.0	–	184	116	85	–	–	–
April	–	6.0	7.9	5.1	–	91	28	58	–	–	–
May	–	11.9	11.3	10.4	–	25	70	82	–	69	48	65
June	15.9	14.7	14.8	14.7	109	75	59	71	–	87	89	84
July	15.5	17.3	16.8	16.2	213	129	55§	92	–	75	73	88
August	16.2	15.6	15.9	15.4	132	241	Rain-out shelter	113	28	68	54	70
September	12.0	11.6	13.0	11.8	59	129	Rain-out shelter	136	50	31	38	40
October	7.7	7.9	–	7.9	53	153	–	162	–	–	–
November	3.6	3.7	–	3.2	69	123	–	143	–	–	–
December	0.9	0.6	–	0.2	155	80	–	102	–	–	–
Year		8.2		6.9		1670		1230
†Reference period 1961–1990.
‡Reference period 1968–2007.
§Rainfall on 18–19 July due to malfunctioning of the rain-out shelter. The total precipitation in July was 244 mm.

Treatments and experimental design

The experimental treatments were arranged in a completely randomized block design with three blocks. Each block contained four lysimeter plots representing the four combinations of rootzone and irrigation regime. The rootzones were either straight sand (SS) or SS amended with 20% v/v garden compost (GM, ‘Green Mix’, Høst AS, Grimstad, Norway).

Chemical analyses in March 2008 and October 2009 indicated higher pH and higher content of plant-available P, K, Mg and Ca in the GM than in the SS rootzone (Table II). Soil physical data from undisturbed soil cores taken at 13–50 and 150–187 mm depth in October 2007 are given in Table III.

Table II. Chemical analyses of soil samples taken in March 2008 and in October 2009.

Rootzone	pH	P-AL	K-AL	Mg-AL	Ca-AL
	mg per 100 g dry soil
	2008
SS	6.5	1.0	1.0	1.0	5.5
GM	7.2	2.5	2.3	3.3	40.8
p	0.006	0.014	0.015	0.003	0.004
	2009
SS	6.0	1.2	4.2	2.8	15.5
GM	6.6	4.2	5.0	5.0	59.0
p	0.002	<0.001	0.455	<0.001	<0.001
SS, straight sand; GM, ‘Green Mix’.

Table III. Physical characteristics of soil cores from the SS and GM rootzones at two soil depths, sampled on 15 Oct. 2007.

	Capillary porosity
					Plant-available water
Rootzone	Depth	Total porosity	Air-filled porosity	Unavailable water, pF > 4.2	pF 1.5–3.0	pF 3.0–4.2	Total	Field capacity, 0–20 cm†	Soil density	Ignition loss	Hydraulic conductivity
	mm	% (v/v)						mm	g cm^−3	%	mm h^−1
SS	13–50	37.4 b	17.9 b	0.6 b	17.4 a	1.5 b	18.9 a	24.8 b	1.58 a	0.60 b	226
	150–187	36.4 b	25.2 a	0.4 b	9.4 b	1.4 b	10.8 b		1.56 a	0.48 b	237
GM	13–50	42.7 a	18.1 b	3.3 a	15.0 a	6.2 a	21.3 a	41.8 a	1.44 b	2.44 a	200
	150–187	44.0 a	20.8 b	2.4 a	15.1 a	5.7 a	20.8 a		1.39 b	1.84 a	220
SS, straight sand; GM, ‘Green Mix’.
†The calculation of field capacity was based on the fact that the 4-cm top layer had been replaced with new sand of finer texture. The given depths correspond to the depths of TDR probes used to measure soil water content.
Mean values followed by the same letter in the same column are not significantly different based on Fisher's protected least significant difference (LSD) test (α = 0.05).

Assuming a root depth of 20 cm, the field capacity was estimated to 24.8 mm on the SS rootzone and 41.8 mm on the GM rootzone (Table III). Irrigation treatments were either light and frequent (LF) or deep and infrequent (DI). In the LF regime, plots were irrigated back to field capacity once water deficit exceeded 5 mm (19.5% depletion of field capacity) in SS plots and 10 mm (23.9% depletion of field capacity) in GM plots. The corresponding thresholds for DI irrigation were 10 mm (39.0% depletion of field capacity) and 20 mm (47.8% depletion of field capacity), respectively. Accumulated water deficits for each of the four treatments were calculated five days per week using daily rainfall and values from the evaporation pan. Each year the soil water status prior to the initiation of irrigation treatments corresponded to field capacity. Individual plots were irrigated very precisely using a wagon with drip nozzles at 20 cm×20 cm distance, the ordinary sprinkler system only being used for application of 3 mm water to the whole experiment after fertilizer inputs every second week. Irrigation treatments were carried out from 21 Aug. to 1 Oct. 2007 and from 1 May to 1 Oct. 2008 and 2009. Until the rain-out shelter was installed on 1 July 2009, the trial was open for natural precipitation.

Daily rainfall and evaporation, irrigation amounts, soil water deficits, and soil moisture values are illustrated in Figure 1 (Espevig and Aamlid 2012), while seasonal number of irrigations and irrigation amounts are summarized in Table IV.

Figure 1.  Effects of rootzone composition and irrigation regimes on performance of velvet bentgrass putting green in 2007, 2008, and 2009. Vertical bars are LSD values indicating significant differences between treatments at 5% probability level. SS, straight sand; GM, ‘Green Mix’; LF, light and frequent irrigation; DI, deep and infrequent irrigation.

Table IV. Seasonal number of irrigations and total water use in various treatments. Figures include irrigation of the whole experiment after fertilizer application every second week. Treatment abbreviations: SS, straight sand; GM, ‘Green Mix’; LF, light and frequent irrigation; DI, deep and infrequent irrigation.

	Number of irrigations			Total irrigation water, mm
Period	SS + LF	SS + DI GM + LF	GM + DI	SS + LF	SS + DI GM + LF	GM+ DI
Total 2007†	10	5	2	50	45	25
Total 2008‡	45	23	16	239	184	174
Total 2009§	49	29	18	227	213	173
† 21 Aug – 1 Oct.
‡ 1 May – 1 Oct.
§ May – 1 July.

Maintenance of plots

During the period from 21 Aug. to 1 Oct. 2007, all plots recieved 45, 7, and 44 kg ha⁻¹ of N, P, and K, respectively. To equalize growing conditions on the two different rootzones, NPK inputs on SS plots were higher than on GM plots in 2008 and 2009. The total input of N, P, and K on GM plots in 2008 (15 Apr.–1 Nov.) amounted to 131, 14, and 117 kg ha⁻¹ and in 2009 to 108, 20, and 138 kg ha⁻¹, respectively. The corresponding amounts of NPK on SS plots were 192, 28, and 185 kg ha⁻¹ in 2008 and 130, 30, and 206 kg ha⁻¹ in 2009. Fertilization interval was always 2 weeks. Except for four applications of liquid fertilizer Arena® Crystal (Yara International ASA, Norway) in the fall of 2008, fertilizers were mostly given as inorganic granular Arena® products. Due to high pH (Table II) ammonium sulfate 21-0-0 (Yara International ASA, Norway) or Anderson 13-2-13 (Andersons Lawn Fertilizer Division Inc., Maumee, OH) was also included in the fertilization plan on GM plots. The green was mowed using a John Deere 220A walk-behind mower (Moline, IL) three times a week at 3 mm except for the periods April–June and September–October when mowing height was raised to 3.5–4.5 mm. Starting in July 2008, the green was groomed with a groomer attached to the mower once a week and brushed twice a week. In 2008 and 2009, monthly vertical cutting was performed to 2 mm depth using an Aztec verticutter pod mounted on an Aztec drive unit (Allett mowers LTD, Arbroath, Scotland). On 12 June 2008 the green was aerated with a John Deere Aerator 800 (Moline, IL) using 8 mm solid spikes to 8 cm depth. At the end of each season, the green was core-aerated using 6 mm hollow tines to 8 cm depth. From May to October 2008 and 2009, the green was exposed to artificial wear from a friction drum with golf spikes corresponding to 20000 rounds of golf per year. As a result of topdressing every 1 to 2 weeks, the green received 4 mm and 9 mm of sand (no organic matter; grain size 0.2–0.8 mm; Baskarp, Sweden) in 2008 and 2009, respectively.

Registrations and statistical data analyses

Visual assessments were conducted every second week for turf quality (scale from 1=uneven and very bad turf to 9=even and very good turf; acceptability level=5) and monthly for tiller density (scale from 1=very thin to 9=very dense), color (scale from 1=very light to 9=very dark), diseases (% of plot covered with diseased turf) and turf coverage (% of plot covered with undiseased turf of the sown species) from 26 Aug. to 30 Oct. 2007 and from 15 March to 15 Oct. 2008 and 2009. Data from each experimental year were pooled into three consecutive periods with 2–5 registrations each: spring (March–May), summer (June–August), and fall (September–October).

Potential soil water repellency was determined on 13 June 2008 and 20 Sept. 2009 using the water drop penetration time (WDPT) test (Dekker et al. 2001). One soil sample (11×12×2 cm) was taken from each plot with a spade sampler. After 48 h of air drying in the laboratory, three drops of water were placed at 0.5-cm depth (in mat), just under the mat (at 1-cm depth in 2008 and 2-cm depth in 2009), and at 3-, 5-, and 10-cm depth from the green surface, and the times until drops had penetrated were measured. The following classification was used for interpretation of results: Wettable (not water repellent) if WDPT<5 s, slightly water repellent if 5 s≤WDPT<60 s, strongly water repellent if 60 s≤WDPT<600 s, and severely water repellent if 600 s≤WDPT (Dekker et al. 2001).

Leaching water from the lysimeters was collected monthly during the following periods: August–September 2007, May–September 2008, and May–June 2009. The samples were kept at 4 °C prior to analyses. Leaching in May and June 2008 were pooled due to little rainfall. Samples were analyzed individually for nitrate/nitrite (standard EN ISO 13395) and total N (standard EN ISO 13395), P (standard EN ISO 15681-2), and K (standard NS EN ISO 11885) concentrations at AnalyCen laboratory (Norway). The total nutrient leakage was calculated based on monthly concentrations and amount of leachate.

The data were analyzed by the SAS procedure PROC ANOVA using statements providing one-way analysis for a block design (SAS Institute 2008). In the case of water drop penetration times, ANOVA was performed on data transformed to ln(y + 1) due to deviation from normal distribution. The Fisher's protected least significant difference (LSD) at the 5% probability level was used to identify significant differences among treatments.

Results

Turf quality, color, shoot density, and diseases

Regardless of irrigation treatment, velvet bentgrass produced significantly better turf quality (Figure 1a) and higher shoot density (Figure 1b) on GM plots than on SS plots in the spring of 2008 and 2009. A similar tendency (p=0.08) was observed in the fall of 2007. The effect of rootzone composition on turf quality was less conspicuous during the summer and fall of 2008. Starting in the spring of 2009, DI irrigation produced better turf quality and higher shoot density than LF irrigation on the SS rootzone, and this effect became highly significant by the end of the experiment. Color was improved during the first year after sowing from a score of 5.2 in the fall of 2007 to 6.2 in the fall of 2008. There were no differences in color among the treatments except for the higher score (6.4) on SS DI-irrigated plots in the summer of 2008. In 2009, the color trends for the treatments were similar to those for turf quality (data not shown).

A severe snow mold attack, caused by Microdochium nivale, was observed on the SS rootzone in the spring of 2008 (Figure 1c). On that rootzone, light and frequent irrigation led to 30% less injury than DI irrigation. The area covered by diseased turf was lower (14%) and did not depend on irrigation treatments on GM plots. In the spring of 2009, snow mold attack was much lower than in 2008, but again GM plots had less injury (<1%) than SS plots (3%); no significant effect of irrigation treatment could be detected in this case. In September 2009, DI irrigation reduced a sudden outbreak of M. nivale compared with LF irrigation on both rootzones. For LF and DI irrigation the affected area was 14 vs. 5% on the SS rootzone and 8 vs. 4% on the GM rootzone (data not shown).

In August 2009, irregular patches with enhanced shoot growth were observed on some GM plots independent of irrigation treatments. Inspection of the mat layer indicated that the dark-colored thatch, most likely lignin, had been degraded, and microscopic observation suggested that it could be caused by a fungi belonging to Basidiomycota, probably white-rot fungi contained in the compost (Tuomela et al. 2000).

Potential soil moisture repellency

In general, GM plots were wettable (water drop penetration time, WDPT<5 s) or slightly water repellent (5 s≤WDPT<60 s) at all investigated depths except for the mat layer on plots receiving LF irrigation in 2008 (Table V). Both in June 2008 and September 2009, the mat layer was more water repellent under LF than DI irrigation. At 5- and 10-cm depth, water drop penetration time was generally higher under DI (slightly water repellent, WDPT varied from 11 to 19 s) than under LF irrigation (wettable, WDPT varied from 3 to 6 s). In contrast to GM plots, SS plots showed strong (60 s≤WDPT<600 s) and severe water repellency (600 s≤WDPT) at certain depths. In 2008, DI irrigation resulted in higher water repellency immediately under the mat layer and at 1 cm under the mat layer than LF irrigation. In 2009, strong water repellency was observed at these depths regardless of irrigation regime. Deep and infrequent irrigation led to the higher repellency at the 10-cm depth in 2008 and at 5- and 10-cm depth in 2009. Like on GM plots, measurement in September 2009 showed that the mat layer on SS plots was potentially more water repellent under LF irrigation than under DI irrigation.

Table V. Water drop penetration time (WDPT) at different depths of SS and GM plots and LF or DI irrigation, measured after drying samples at room temperature for 48 h in 2008 and 2009.

	June 2008					September 2009
Treatment	In mat	Immediately under mat	1 cm under mat	5 cm from green surface	10 cm from green surface	In mat	Immediately under mat	1 cm under mat	5 cm from green surface	10 cm from green surface
	WDPT, s
SS + LF	74 a	35 a	9 ab	5 ab	6 a	111 d	76 b	123 b	11 b	6 b
SS + DI	217 c	208 b	1282 c	1 a	192 b	31 c	79 b	136 b	24 c	69 c
GM + LF	107 b	7 a	5 a	4 bc	5 a	17 b	8 a	9 a	6 a	3 a
GM + DI	51 a	11 a	14 b	12 c	19 a	12 a	7 a	9 a	11 b	11 b
p	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001
SS, straight sand; GM, ‘Green Mix’; LF, light and frequent irrigation; DI, deep and infrequent irrigation.
Mean values followed by the same letter in the same column have no significant difference based on Fisher's protected LSD test (α = 0.05).

Leaching

During the periods of the leachate collection in 2007, 2008, and 2009, SS plots with LF and DI irrigation received totally 362 mm and 292 mm irrigation water, respectively. The corresponding amounts on GM plots were 292 mm and 222 mm. The total rainfall for the collection periods amounted to 790 mm. Because leachate was collected during periods of different duration in 2007, 2008, and 2009, leaching data are presented on a daily basis (Table VI).

Table VI. Effects of the treatments on leaching rates during the water collection periods in 2007, 2008, and 2009.

Treatment	20 Aug.–1 Oct. 2007	1 May–1 Oct. 2008	1 May–1 July 2009
	mm day ^1
SS + LF	1.7 b	3.1	1.1 d
SS + DI	1.5 ab	2.7	0.8 b
GM + LF	1.5 ab	2.8	1.0 c
GM + DI	1.3 a	2.7	0.7 a
p	0.023	0.105	<0.001
SS, straight sand; GM, ‘Green Mix’; LF, light and frequent irrigation; DI, deep and infrequent irrigation.
Mean values followed by the same letter in the same column have no significant difference based on Fisher's protected least significant difference (LSD) test (α = 0.05).

In 2009, LF irrigation caused significantly higher leaching rates than DI irrigation from both SS plots and GM plots. In 2007 and 2008, differences showed similar trends but were not significant.

Significant differences in concentrations and average daily loss of nutrients occurred between treatments, but they were usually due to rootzone rather than irrigation frequency (Table VII). The highest concentrations were measured of potassium (12.6–31.7 mg L⁻¹) followed by total nitrogen (0.7–3.0 mg L⁻¹) and phosphorus (0.01–0.19 mg L⁻¹).

Table VII. Nutrient concentrations in leachate and total nutrient losses from velvet bentgrass green as affected by rootzone type and irrigation regime and expressed for 2007, 2008 and 2009 on a daily basis.

	20 Aug.–1 Oct. 2007		1 May–1 Oct. 2008		1 May–1 July 2009
Treatment	Concentration, mg ^− 1	Leaching rate, mg m^−2 day^−1	Concentration, mg L^−1	Leaching rate, mg m^−2 day^−1	Concentration, mg L^−1	Leaching rate, mg m^−2 day^−1
	NO ^− 2 /NO ^− 3
SS + LF	1.8 b	3.3 b	0.7 b	2.3 ab	0.5	0.6
SS + DI	2.2 b	3.3 b	1.5 a	4.1 b	0.5	0.4
GM + LF	0.3 a	0.5 a	0.3 b	0.9 a	0.9	0.9
GM + DI	0.3 a	0.3 a	0.4 b	1.0 a	0.7	0.5
p	0.020	0.030	0.028	0.052	0.811	0.639
	Nitrogen (N)
SS + LF	3.0	5.2	2.0	6.4	0.7 a	0.9 a
SS + DI	2.5	3.9	2.7	7.3	0.7 a	0.6 a
GM + LF	1.8	2.8	2.7	7.6	2.4 b	2.4 b
GM + DI	1.9	2.5	3.0	7.9	2.0 ab	1.3 ab
p	0.476	0.293	0.165	0.630	0.057	0.052
	Phosphorus (P)
SS + LF	0.02 a	0.04 a	0.02 a	0.07 a	0.01 a	0.01 a
SS + DI	0.02 a	0.03 a	0.03 a	0.07 a	0.01 a	0.01 a
GM + LF	0.19 b	0.30 b	0.16 b	0.44 b	0.15 b	0.15 b
GM + DI	0.19 b	0.26 b	0.15 b	0.41 b	0.15 b	0.10 b
p	0.001	0.004	<0.001	<0.001	0.003	0.002
	Potassium (K)
SS + LF	14.4 a	25.1 a	12.6	39.5	14.3	16.5 b
SS + DI	12.7 a	19.3 a	14.7	39.4	16.8	13.6 ab
GM + LF	31.7 b	48.2 c	14.4	39.7	17.1	16.7 b
GM + DI	29.2 b	39.1 b	13.5	35.9	15.4	10.4 a
p	<0.001	0.001	0.251	0.821	0.548	0.076
SS, straight sand; GM, ‘Green Mix’; LF, light and frequent irrigation; DI, deep and infrequent irrigation.
Mean values followed by the same letter in the same column have no significant difference based on Fisher's protected least significant difference (LSD) test (α = 0.05).

On average for irrigation treatments the loss of nitrate/nitrite through the drainage system was eight and three times higher on SS plots compared with GM plots in 2007 and 2008, respectively. In 2009, differences in nitrate/nitrite leaching rate from the two rootzones were no longer significant. By contrast, total nitrogen leaching was higher from GM plots than from SS plots in 2009 only. Leaching of phosphorus was eight, six, and 11 times higher on GM plots compared with SS plots in 2007, 2008, and 2009, respectively. Potassium leaching from GM plots was two times higher than from SS plots in 2007, but it did not differ between rootzones in 2008 and 2009.

As compared with DI irrigation, LF irrigation resulted in more leaching of potassium from GM rootzones in 2007 and 2009. Effects of irrigation regime on nutrient losses from SS rootzones were never significant, but a significantly higher concentration of nitrate/nitrite was detected in leachate from plots with DI than from plots with LF irrigation in 2008.

Discussion

The lower turf quality and shoot density on SS vs. GM plots in fall 2007 and during both spring periods is in accordance with earlier investigations (Gibbs et al. 2000, Joo et al. 2001). Nitrogen loss in the form of nitrate/nitrite from the 2–4-month-old green was eight times higher from SS plots than from GM plots and corresponded, in turn, to 4.5% and 0.5% of total N applied in fertilizer. Our results are in agreement with earlier studies showing amendment with compost to improve the water holding and lessen nitrogen leaching, and thus turf quality of sand-based rootzones (McCoy 1992, Brauen and Stahnke 1995, Murphy et al. 2004, Aamlid 2005). Better turfgrass performance on GM compared with SS plots in the spring of 2008 and 2009 was also due to less injury caused by Microdochium nivale. This was later confirmed by a sudden outbreak of the microdochium patch in September 2009; again, the affected area was lower on GM plots compared with SS plots. These results are substantiated by earlier reports showing suppressive effects of compost on soil-borne turfgrass diseases (Boulter et al. 2002b, Nelson and Boehm 2002, Tilston et al. 2002). Boulter et al. (2002a) found a reduction in the development of M. nivale and Typhula ishikariensis on a 4–5-yr-old creeping bentgrass green after topdressing with compost. They also observed a quicker green-up and recovery from dormancy on compost-amended plots.

In our study, the period from 2 May to 13 June 2008 was without any natural rainfall, and the turf obviously performed better and recovered more quickly from the disease with LF than with DI irrigation. Better turf performance of young greens under more frequent irrigation was reported also by Fu and Dernoeden (2009a).

The rapid improvement in turfgrass turf quality and shoot density on SS plots in the summer and fall of 2008 was facilitated by a 46% higher nitrogen input on SS than on GM plots. After achieving maximum scores in the fall of 2008, turf quality declined in response to LF irrigation on SS plots. This is in accordance with earlier research by Jordan et al. (2003) and Fu and Dernoeden (2009a). Thus, Fu and Dernoeden (2009a, 2009b) reported better turf quality with DI irrigation than with LF irrigation by the end of a 2-yr study on creeping bentgrass green (97% sand, 1% silt, 2% clay, and 10 mg g⁻¹ organic matter), and they attributed this to, among other things, a higher chlorophyll content and better adaptation to wilt stress after DI irrigation.

The higher water repellency on SS plots compared with GM plots at most of the investigated depths corroborates earlier studies showing water repellency to be an important property of sand which develops when the soil water content is below a certain critical threshold (Bauters et al. 1998, Dekker et al. 2001, Larsbo et al. 2008). Our results suggest that development of water repellency on SS plots was delayed but not totally prevented by LF irrigation. However, contrary to our expectation, higher soil water repellency in the mat layer on the GM rootzone in June 2008 and on both rootzones in September 2009 was observed on plots receiving LF irrigation as opposed to DI irrigation. We have no good explanation for this phenomenon but suggest that moisture or/and oxygen conditions provided by LF irrigation could derive microbial water repellency (Wilkinson and Miller 1977, Hallet et al. 2001) or repellency induced by the products of organic matter decomposition or by the exudates of grass roots (Doerr et al. 2000).

Infrequent irrigation (DI) did not lead to increased drainage (Bauters et al. 1998, Kenna and Snow 2000, Fry and Huang 2004, Barton and Colmer 2006, Nektarios et al. 2007). Starrett et al. (1995, 1996, 2000) also found more drainage from fine-loamy soil columns with Kentucky bluegrass turf receiving DI irrigation (four applications of 25.4 mm) than LF irrigation (16 applications of 6.4 mm). They also showed more nitrogen (Starrett et al. 1995), pesticide (Starrett et al. 1996), and herbicide leaching (Starrett et al. 2000) with DI vs. LF irrigation. In our study, the higher amounts of leachate with LF than with DI irrigation within each rootzone were caused, firstly, by the generally higher amounts of irrigation water received by LF- than by DI-irrigated plots. Secondly, plots with LF irrigation were mostly closer to field capacity and therefore had less space for natural rainfall than plots with DI irrigation.

The overall leakage of nutrients in our study was low as has previously been reported from other studies with well-established and properly maintained turf (Mancino and Troll 1990, Frank et al. 2005, Paré et al. 2006, Soldat and Petrovic 2008, Steinke et al. 2009). Thus, the nitrate/nitrite concentrations in leachate were 95, 98, and 99% lower on SS plots and 99%, 99%, and 98% lower on GM plots than the current regulatory standard of 50 mg of nitrates per liter groundwater (European Union, 2010).

In 2007 the nitrogen inputs on SS and GM plots were the same and amounted to 34 kg N ha⁻¹ for the period of leachate collection. Leaching of nitrogen as nitrate/nitrite was, nonetheless, significantly higher from SS than from GM plots. That is not surprising due to lower immobilization, absence of thatch and lower organic matter content in the SS rootzone (Brauen and Stahnke 1995). The higher leakage of organic N (difference between total N and nitrate/nitrite) from GM than from SS plots can be ascribed to more “conversion of mineral fertilizer to leachable organic form” (Paré et al. 2006) or to an initially higher content of organic nitrogen in the GM rootzone. Unlike Djodjic et al. (2004) who found no relationship between soil P content and P losses, differences in P level between soil types in our study were huge and caused more P leaching from the GM than from the SS rootzone. The amount of potassium applied to both rootzones could be reduced as suggested by total leaching of this nutrient.

While this investigation showed that establishment and maintenance of velvet bentgrass is possible on both SS and GM rootzones, the use of GM showed clear advantages in the form of higher visual quality, less disease, longer irrigation intervals, and less risk for development of soil water repellency. Disadvantages of using the compost-amended rootzone was greater risk for nitrogen and phosphorus leakage from established turf. Although the effect of rootzone was mostly dominant to the effect of irrigation, this investigation showed a shift in optimal irrigation strategy from light and frequent irrigation during the first 18 months after the 2 months grow-in period to deeper and more infrequent irrigation on more mature velvet bentgrass greens. On the GM rootzone, with an initial organic matter content of 2%, it is noteworthy that mature velvet bentgrass maintained acceptable turfgrass quality if irrigated to field capacity only at 20 mm deficit, i.e., at approximately weekly intervals.

Acknowledgements

We thank the Scandinavian Turfgrass and Environment Research Foundation (STERF) and the Norwegian Research Council (NFR) for funding. We are grateful to Trond Olav Pettersen, Åge Susort, and Anne A. Steensohn for their excellent technical assistance and to Agnar Kvalbein for helpful discussions.