Initial Investigations of Mowing Height and Greens Rolling
on Ball Roll of Penncross Creeping Bentgrass D.M. Kopec, J.Long, D.Kerr, and J.J. Gilbert
Abstract
Increased demands for ball speeds is one of the greatest demands on the modern golf course superintendent. The effects of
mowing height and rolling on ball speed performance were evaluated on a sand based USGA 'Penncross' putting green.
Three tests were conducted addressing mowing height and rolling combinations, frequency of rolling (# of passes), and
surface hardness measurements among treatments using a Clegg surface hardness decelerator.
When plots were rolled at one pass (twice per week) ball speed was generally increased by 8-11% at higher cut 'Penncross'
creeping bentgrass (11/64"). Plots rolled at one pass at low mowing heights (9/64") showed increases in ball speed distances
by 5-10%. While rolling did improve ball speeds for both mowing heights, rolling at the higher height of cut (11/64") did not
increase ball speed over unrolled low cut turf mowed at 9/64".
Double rolled turf did not produce greater ball speeds than single rolled turf, when double rolling was performed once. The
effects of rolling on enhanced ball speed appear to last 48 hours at most. Surface hardness was positively correlated to ball
speed for high cut turf (11/64") which received a single rolling event and was strongly negatively correlated when mowed at
low height (9/64") when double rolled. Introduction
Increased demands on the performance of golf greens has precipitated greater demands for the golf course superintendent
over the last fifteen years. This is in part due to (1) increased popularity of the sport itself among the general public, (2)
increased competition within the sport, (3) better mowing practices and specialized equipment, and (4) more precise greens
construction specifications. One such cultural management practice is the use of rollers to increase ball speed of putting
greens. Rolling of turf is not new, but renewed interest in this specialized application for golf greens as part of a standard
maintenance program has re-surfaced in the last six years or so. The claimed benefits of rolling include (1) increased ball
speeds, (2) a more true ball roll, (3) increased speeds at higher mowing heights, and (4) consistency of performance from
green to green. With these considerations in mind, a study was devised to investigate the effects of mowing height, rolling, rolling frequency
and initial testing of surface hardness levels of greens as they relate to ball (distance) speed.
MATERIALS & METHODS
Three experiments were conducted on a four year old 'Penncross' creeping bentgrass green at the University of Arizona
Karsten Turfgrass Facility in Tucson, Arizona. The green is 60 x 60, was watered nightly with 80 reference Eto replacement
from an on-site weather station.
TEST 1
Test 1 investigated the effects of mowing height and the effect of periodic rolling on ball speed. Two mowing heights were
evaluated, 9/64" and 11/64", both with and without rolling from a 680 lb. 30" wide smooth roller. This provided a linear
contact of 8 psi. Plots were mowed 6x weekly with a Toro Tournament Greens Mower (21" cutting width, smooth roller) and
clippings were removed. Four plots were moved to low height and four plots at the high height. Mowed plots were 12' x 24'.
Each half of the 24" length was split with the introduction of a rolling treatment. Plots were rolled in a single direction (four
roller widths) twice per week on a Wednesday/Friday schedule. Each other half plot remained unrolled. This resulted in the
following four treatments; (1) roll - high cut (11/64"), (2) no roll - high cut (11/64"), (3) roll - low cut (9/64") and (4) no roll low cut (9/64"). Ball speeds were taken with a USGA specification stimpmeter twice per week, immediately following the
mowing and rolling treatments. Either two or three Ping Zing #3 balls were released in each of two directions and averaged to
eliminate any directional effect. Each release point was established by inserting a flag in the green at the base of the
stimpmeter and ball speed was assessed to nearest ½ inch of travel. All ball speed data within a mowing height/roll plot was
averaged and used to determine main treatment and interaction effects.
TEST 2
Test 2 was conducted on the same test green as Test 1. All of the materials and methods remained constant except for the
rolling treatments with the turf roller. The objective of Test 2 was to assess the impact of a single one time double rolling
event on ball speed, and to compare the potential residual effects from the 1x and 2x rolling events. In addition to the four
previous treatments, two additional treatments were added; 5) roll 2x - low cut and 6) roll 2x - high cut. Half of each
previously rolled plot (120" wide strip) was now rolled in an additional direction. This resulted in a one time double roll
treatment (60" wide across all mowing treatments). The new treatments were then immediately measured for ball speed
(along with all other treatments). Mowing continued daily afterwards and all plots were measured for ball speed on 2 and 4
days after the last rolling was done on April 21, 1995.
A field event calendar is provided in Table 1.
TEST 3
From April 25 to May 15, plots were mowed 6x weekly at the identical mowing heights previously. Rolling at 1x and 2x
passes were again repeated exactly as described on May 15.
In addition to ball speeds taken on all mowing height/rolling combination plots, surface hardness measurements were
collected using a Clegg surface impact penetrometer. This unit measures surface hardness via the speed of deceleration of a
cylinder (hammer) upon impact on the surface. Higher unit values represent harder surfaces. Use of this unit in sports turf to
assess hardness has been reported and quantified, but has not been attempted on greens for assessing any potential
relationship with ball speeds. One measurement was made per plot, recording and analyzing the first (1 0) third (30) impact
values, respectively. These values were correlated with actual ball speeds to see if hardness readings were related to rolling
performance. Test three included rolling/mowing treatments along with Clegg penetrometer values on May 15.
RESULTS
TEST 1. On all dates except for 3/31, the 'roll- low cut' treatment was fastest, followed by 'low cut- no roll' on seven of eight
days (Table 1./ Figure 1). Generally, there was a 10% difference between these two treatments on seven of eight days. The
'roll- high cut' turf was faster than the 'no- roll, high cut' treatment on all dates with the exception of 3/31. The mean distance
between these two treatments was approximately 15%. On 3/31/95 environmental conditions such as a suspected early
afternoon syringe condition may have produced an overall mean ball speed of 105". The 3/31/95 ball speeds were
inconsistent with the other seven measured dates and there was no treatment effect on this date. Throughout the eight
evaluation periods, the relative differences between treatments remained constant, while actual mean ball speed values did
not result in a linear accumulation of ball speed with time. Roll treatments did have greater speed than their non-roll
counterparts in all cases (Table 2./Figure1).
An expected benefit of rolling is to maintain ball speed while maintaining elevated mowing heights to reduce mowing stress.
In this study, this was not realized. Our data showed that the 'no roll-low cut' treatment was consistently faster than the high
cut, rolled treatment. This was the case over a four week period in which the turf received eight rollings.
TEST 2.
In all cases, the low cut turf had greater performance than the high cut, regardless of rolling frequency and the one time
double roll. The effect of the double roll was slightly greater than the single roll immediately after on the day of rolling. The
1x and 2x treatments had similar performance at two and four days after rolling showing a quick loss of effect. The 2x 'rolllow cut' treatment increased ball speed by 3% over the 1x 'roll-low cut' turf, and 6% over 'no roll-low' cut turf (Table
2./Figure 2).
'High cut' treatments showed the same general trend relationships as the 'low cut' except they were consistently slower. The
double roll event increased ball speed over the single roll turf on day 1 and at 4 days after rolling (high cut only). However, at
4 days after treatment, the effect of all rolling was drastically reduced, as ball speeds of the rolled turf were close to the
unrolled check (for the high cut turf)(Table 2./Figure 2).
Comparison between heights of cut showed that at the low cut, the increase in speed for the double roll, over the low cut no
roll, was 11% immediately after rolling. At the high cut, the one day increase in speed of the double roll vs. no roll was 11%
also. Differences among unrolled turfs was 17% (85" vs. 102") between mowing heights (Table 2./Figure 2).
In conclusion, height of cut alone had a greater effect on ball speed than rolling alone. However, rolling did increase ball
speeds at each cutting height considerably. The effects of the double roll for a single event did increase speed slightly over
the single roll. Effects essentially lasted for 1 day only, based on a one time double roll event.
TEST 3.
Ball rolls were again significantly different due to management treatment effects, as were the hardness values as well. On
May 15, ball speeds ranged from 73.0 to 99.5 inches (24% difference in ball speed). Plots which received low mowing
(9/64") had higher ball speed values than the higher mowed (11/64") plots. This demonstrated the effect of lower mowing
heights on ball speeds (Table 3).
At the low mowing height, there was no difference between 1x and 2x rolling events (96 and 99 inches, respectively). At the
high mowing practice, the 2x rolled plots had only a slightly larger ball speed mean value then the 1x rolled turf (81.5 vs.
78.4, respectively). However, at each mowing height, the effect of rolling (either 1x or 2x) was significantly greater than the
unrolled turf. After a single reenactment of the rolling practice in general, there was no difference between rolling frequency
events. The same results occurred in Test 2. Long term rolling tests at 1x vs. 2x should be conducted.
Clegg surface penetrometer reading ranged from 114 to 138 (first impact readings)(Table 3). Values were larger for the low
mowed turf, over that of the high mowed turf. For the low mowed turf, the double rolled (2x) treatments had a 6% increase of
no rolled turf (138" vs. 130", respectively). There was no statistical difference for ball speed frequency (1x vs. 2x) at the low
height of 9/64" for surface hardness values (138" vs. 132", respectively).
At the high mowing height (11/64") there were no significant differences for surface hardness values between rolling
frequency treatments, nor were rolling treatments different from the mowed only turf at 11/64". As was the case with ball
speed, low mowed turf had the highest Clegg values showing an association between rolling and surface hardness.
If all rolled turf (both 1x and 2x) plots are compared to non-rolled turf (over both heights), the effect of rolling was highly
evident and significant for surface hardness.
It should be noted that in this pilot test, only one Clegg surface measurement was taken per plot, which was compared to the
average of six stimpmeter values.
The third Clegg value (ASTM standard) increased hardness readings by almost 14% on average. Increased values resulted in
similar rankings among treatments for hardness, although values were now closer between treatments following the third
penetrometer hammer drop. The low mowed (9/64") bentgrass had slightly higher Clegg values than higher mowed turf
(11/64").
Data correlations were performed using simple Pearsons product correlation coefficients. For all surfaces (all mowing heights
and rolling combinations), ball speeds were positively associated with initial (1 0) Clegg impact penetrometer values [r=0.43
p=0.03]. The third impact (30) was not correlated, most likely due to accumulative surface hardness from the 30 measurement.
For individual treatments, different associations occurred between ball speed and penetrometer values. For high mowed turf
11/64", ball speeds were either positively correlated with Clegg values for single rolling events, had no association without
rolling, or produced no, or a negative association when double rolling occurred. Perhaps this is a function of redistribution of
the stolon mass within the mat, which while having a greater surface hardness, may have a more loosely knit shoot/leaf
surface from the double mowing. The 1x rolled treatments were positively associated with ball speeds at 11/64" only. First
and third impacts had correlation coefficients of 0.59 and 0.98, respectively (Table 4).
Among low mowed (9/64") turfs, single (1x) rolled turf had little or no relationship with either Clegg value, nor did the
unrolled turf. A very strong negative relationship occurred between double rolled turf at 9/64", where the simple correlation
between ball speed and both Clegg values were -0.93 and -0.83, respectively (Table 4). . This means that higher impact
values were strongly associated with slower ball speeds. This may be due to a rippling effect of the surface from crown
alteration or leaf changes (which causes a surface dimpling effect) albeit on a "hard" surface. Whatever the cause, the
relationship was quite strong. Once again, note that only one Clegg sample was taken per plot.
Further research investigating hardness effects and canopy interaction appear warranted.
It was encouraging that at this higher mowing height of 11/64", ball speed was strongly associated with hardness values for
the first and third hammer impact (0.59 and 0.98, respectively). Additional research is needed to determine if these initial
effects are continuous on a regular scheduled rolling program for greens.
FN:BALSPEED.RPT DIR:Q296 Table 1. Rolling and ball speed performance of a 'Penncross' Creeping Bentgrass Green 1. University of Arizona, Karsten
Turfgrass Research Facility, 1995.
TEST 1
Frequency Roll Ball Speed
1x 3/22 3/22 1x 3/24 3/24 1x 3/29 3/29 1x 3/31 3/31 1x 4/5 4/5 1x 4/7 4/7 1x 4/12 4/12 1x 4/14 4/14
TEST 2
1x and 2x 4/21 4/21
-- --- 4/23
-- --- 4/25 TEST 3 1x and 2x 5/152 5/15 1
Turf mowed 6x week. Mowing heights included 9/64" and 11/64". Turfs rolled with a 680 lb. 30" roller (8 psi).
Includes Clegg surface penetrometer reading.
2
Table 2. Mean ball speed1 values of a 'Penncross' Creeping Bentgrass Green. University of Arizona, Karsten Turfgrass
Research Facility, 1995. Test 1. Treatments4
Rolling
Date2
3/22 1
3/24 2
3/29 3
3/31 4
4/5 5
4/7 6
4/12 7
4/14 8
Roll
Hi-Cut
96.6
101.75
98.1
102.9
95.9
103
100.5
96.9
No Roll Hi-Cut
85.4
93.9
91.1
108
85.7
90.3
91.5
87.1
Rol
Low
108
116
103
105
114
120
118
117
Test 2
Rolling
Date2
4/21 9(1A)3
4/23 10(2A)
4/25 11(3A)
Roll
Hi-Cut
92.1
84.8
83
No Roll
Hi-Cut
84.9
83
82.8
Rol
Low
112
107
105
1
Mean of either 4 or 6 rolls, averaged over 2 directions and 4 replications.
2
Ball rolls performed immediately after mowing and rolling events on dates 1-8.
3
Double roll event one time only on 4/21 along with concurrent single roll. 4/23 and 4/25 represent ball speed at 3 and 5 days
after rolling was terminated.
4
Hi-Cut = 11/64", Low-Cut = 9/64". Roller = 680 lbs. (8 psi).
Table 3. Mean ball speed and Clegg surface hardness penetrometer values among 'Penncross' Creeping Bentgrass rolled 1x or
2x at two different mowing heights. University of Arizona, Karsten Turfgrass Research Facility, 1995. Treatment Combinations
Height
9/64"
9/64"
9/64"
11/64"
11/64"
Ball Speed Inch
Roll
1x
2x
none
1x
2x
11/64"
none
1
Mean of six stimpmeter readings, three in each of two directions, per plot. Values are means of 24 observations.
2
Surface hardness values, 500 gm. Clegg penetrometer. Initial (1 0) and third (30) hammer drop values, respectively. Values
are means of four replications. Table 4. Simple correlations between ball1 speeds and Clegg penetrometer surface hardness values at different mowing
heights and rolling frequency combinations. University of Arizona, Karsten Turfgrass Research Facility, 1995. Height
All
11/64"
11/64"
11/64"
9/64"
9/64"
9/64"
Roll
All
1x
2x
none
1x
2x
none
1
Plots rolled, measured for surface hardness, and stimpmetered on May 15, 1995.
2
Simple Pearsons product moment correlation. Values closer to 1 or -1 indicate strongest possible associations.
3
Initial (10) and (30) hammer impact, 500 gm. hammer, Clegg impact penetrometer.