Published July, 2002

Organic Amendment and Rotation Crop Effects on the Recovery of Soil

Organic Matter and Aggregation in Potato Cropping Systems
A. Stuart Grandy, Gregory A. Porter,* and M. Susan Erich

ABSTRACT der et al., 1994). Water soluble carbohydrates directly

Soil structural degradation is common in intensively cultivated stabilize aggregates (Angers and Mehuys, 1989) and
ecosystems due to the depletion of soil organic matter (SOM). We have been strongly correlated (r ⫽ 0.74) with aggrega-
investigated the mechanisms by which different frequencies of organic tion (Haynes and Swift, 1990).
amendment application and rotation crops restore C, N, and aggrega- Green manure crops increase soil C inputs, but their
tion in gravelly loam soils used for potato production. A single amend- effects on SOM equilibrium and aggregation may occur
ment application [FIRST; 22 Mg ha⫺1 compost and 45 Mg ha⫺1 beef slowly because background levels of soil C are relatively
cattle (Bos taurus L.) manure] did not affect total C in 1996 and in- high and spatially variable (Robertson et al., 1993, 2000).
creased it by 28% in 1997 relative to unamended plots (NONE); light
Long-term annual compost and manure applications in-
fraction (LF) C accounted for 56% of this increase. Plots in which
amendment was suspended for 1 yr (SASP) following 4 or 5 yr of
crease SOM and structure (Sommerfeldt et al., 1988;
annual application had more total C in 1996 (28%) and 1997 (46%)
Gilley and Risse, 2000); however, annual applications
relative to NONE. A green manure crop consisting of oat (Avena sativa are expensive (Araji et al., 2001) and thus not widely
‘Porter’), pea (Pisum sativum ‘Trapper’), and hairy vetch (Vicia villosa utilized in Maine. Applying organic amendments less
Roth) grown in 2-yr rotation with potatoes (Solanum tuberosum L.) than annually would be more cost effective if they could
increased soil C in 1997 (25.9 vs. 23.9 g kg⫺1 ), LF properties in 1996 reverse the negative effects of potato production.
and 1997, and water soluble carbohydrates (WSC) on several sample Minimizing the frequency of organic amendment ap-
dates relative to an oat rotation crop. Large aggregate (2–6.5 mm) plication depends on predicting the number of applica-
stability in 1996 and 1997 and medium aggregate (1–2 mm) stability tions necessary to halt declines in SOM and structure
in 1997 were increased by FIRST relative to NONE. Total soil C was and, once the decline stops and amendments are no
more strongly related to medium (r ⫽ 0.65 in 1997) and large (r ⫽
longer applied, whether the benefits of amendment per-
0.51 in 1997) aggregate stability than LF or water soluble carbohydrate
sist. Increases in soil C and total aggregation have been
fractions. Compost and manure influences occurred rapidly and were
persistent, demonstrating that annual applications are not necessary
reported in potato cropping systems after a single com-
to reverse soil degradation. post and manure amendment application (Porter et al.,
1999), but relationships between organic matter pools
and aggregation still require elucidation. Effects of or-
ganic amendments may last for a century or more on
P otatoes are Maine’s principal agricultural output
and are currently produced on ≈25 000 hectares.
Rapid declines in SOM concentration and structural
abandoned farm land (Compton and Boone, 2000), but
the persistence of their influence on intensively cropped
fields with low structural stability is unknown.
stability are typical in potato production systems (Saini
In this study, we altered organic amendment treat-
and Grant, 1980) and often lead to erosion (Hepler et
ments in a study originally designed to compare un-
al., 1983; Kachanoski and Carter, 1999). Frequent soil
amended potato cropping systems to long-term annually
disturbances, including primary tillage, planting, and in-
amended systems (Porter et al., 1999). Our objectives
tensive cultivation for weed control and harvest aerate
were: (i) to determine if SOM and aggregation could
the soil and break down aggregates, exposing occluded
be increased following first-year amendment applica-
organic matter (Lal et al., 1994; Six et al., 1999). Potato
tion and, more importantly, if the beneficial effects of
crops return only ≈1500 kg ha⫺1 residue to the soil amendments would be retained after annual applica-
(Porter and McBurnie, 1996) and the residue has a low tions had ceased; (ii) to determine if a productive green
C/N ratio (10:1 for potato shoots) and lignin content, manure crop could enhance active SOM components
which increases its decomposition rate (Bending and and aggregation relative to the 2-yr cereal grain rotation
Turner, 1999). commonly used in potato systems; (iii) to establish
One strategy to increase soil structure is to build or- whether soil aggregation was related to soil C and N,
ganic matter pools that facilitate aggregation, such as LF, and/or water soluble polysaccharides in these differ-
LF and WSC. Products of LF decomposition include entially managed soils.
polysaccharides and other materials that stabilize aggre-
gates (Six et al., 1998). Light fraction decomposition MATERIALS AND METHODS
may also be an important source of plant nutrients (Wan-
Site Characteristics
A.S. Grandy, W.K. Kellogg Biological Stn. and Dep. of Crop and The experiment was conducted at the University of Maine’s
Soil Sciences, Michigan State Univ., Hickory Corners, MI 49060-9516; Aroostook Research Farm in Presque Isle, ME, in 1996 and
G.A. Porter and M.S. Erich, Dep. of Plant, Soil, and Environmental
Sciences, 5722 Deering Hall, Univ. of Maine, Orono, ME 04469-5722. Abbreviations: CONT, amendments applied annually for 5 or 6 yr;
Maine Agric. & Forest Exp. Stn. Publ. No. 2543. Received 9 Mar. FIRST, a single amendment application; LF, light fraction soil organic
2001. *Corresponding author (porter@maine.edu). matter; NONE, amendments never applied; SASP, amendment inputs
suspended after four or five consecutive years of application; SOM,
Published in Soil Sci. Soc. Am. J. 66:1311–1319 (2002). soil organic matter; WSC, water soluble carbohydrates.

1311
1312 SOIL SCI. SOC. AM. J., VOL. 66, JULY–AUGUST 2002

repeated in 1997. The soil was a Caribou gravelly loam (fine- Table 2. Characteristics of compost and manure amendments
loamy, isotic, frigid, Typic Haplorthods). Potatoes were grown applied in 1996 and 1997.†‡
on two adjacent sites with identical treatment schemes: Site Growing year
1 was used for potatoes during 1993, 1995, and 1997; Site 2 Amendment and
characteristic 1996 1997
was used for potatoes in 1994 and 1996. Rotation crops were
planted in alternate years. Waste potato compost
pH 6.3 ⫾ 0.1 8.2 ⫾ 0.1
H2O, g kg⫺1 710 ⫾ 19 510 ⫾ 70
Cultural Practices Total N, g kg⫺1 16.9 ⫾ 0.9 11.9 ⫾ 0.4
NH4-N, g kg⫺1 2.1 ⫾ 0.3 1.2 ⫾ 0.1
The potato cultivar planted was ‘Superior’. Potato seed- Total C, g kg⫺1 293 ⫾ 9 176 ⫾ 10
pieces were planted 5 to 7 cm below the soil surface in rows Ca, g kg⫺1 10.73 ⫾ 0.24 8.08 ⫾ 0.03
91 cm apart and at an average within-row spacing of 23 cm. K, g kg⫺1 4.42 ⫾ 0.45 4.37 ⫾ 0.14
Mg, g kg⫺1 5.27 ⫾ 0.27 6.37 ⫾ 0.13
Fertilizer was applied at planting in two bands ≈5 cm to each P, g kg⫺1 4.44 ⫾ 0.17 4.81 ⫾ 0.05
side and slightly below the seedpieces. Potassium was applied B, mg kg⫺1 67 ⫾ 2 61 ⫾ 2
as KCl, P as diammonium phosphate, and N as diammonium Cu, mg kg⫺1 27 ⫾ 1 36 ⫾ 1
phosphate, ammonium sulfate, and ammonium nitrate. Fertil- Zn, mg kg⫺1 159 ⫾ 3 142 ⫾ 1
izer rates (Table 1) were based on soil tests, previous crop, Cattle manure
and amendment analysis. Tillage consisted of chisel plowing in pH 8.5 ⫾ 0.1 8.4 ⫾ 0.0
H2O, g kg⫺1 630 ⫾ 38 660 ⫾ 22
the fall, spring disking after compost and manure amendment Total N, g kg⫺1 17.1 ⫾ 1.9 15.5 ⫾ 1.0
application, and harrowing before planting. Plots were culti- NH4-N, g kg⫺1 3.8 ⫾ 0.6 1.3 ⫾ 0.4
vated once and hilled once during each growing season. Pest Total C, g kg⫺1 – 267 ⫾ 16
control and other management practices were typical of com- Ca, g kg⫺1 24.17 ⫾ 2.47 17.33 ⫾ 1.66
K, g kg⫺1 8.12 ⫾ 0.69 7.29 ⫾ 0.62
mercial practices in the area. Mg, g kg⫺1 13.25 ⫾ 0.92 9.42 ⫾ 0.23
P, g kg⫺1 6.71 ⫾ 1.92 5.10 ⫾ 0.30
B, mg kg⫺1 32 ⫾ 4 54 ⫾ 2
Experimental Design and Analysis Cu, mg kg⫺1 25 ⫾ 2 31 ⫾ 1
Prior to 1996, our study plots were used to test the effects Zn, mg kg⫺1 290 ⫾ 48 163 ⫾ 4
of two organic amendment and rotation crop treatments (Ta- † Data are reported on a dry weight basis.
ble 1) on soil properties and potato yields (Porter et al., 1999). ‡ Mean ⫾ standard error, n ⫽ 3.
Plots (9.1 by 13.7 m) were arranged as a factorial combination
of amendment and rotation crop with four replicate blocks. in 1996 and 1997, respectively). The application rate for ma-
In the current study, amendment treatments were altered in nure was 45 Mg ha⫺1 cattle manure (dry matter equivalents
one half of the experimental plots, resulting in four amendment of 17 and 16 Mg ha⫺1 in 1996 and 1997, respectively). A disk
treatments (Table 1): NONE; FIRST; SASP, amendments cultivator was used to incorporate amendments in the spring
last applied in 1995 after four consecutive years of applica- before potato planting.
tion (Site 2) or in 1996 after five consecutive years of appli- The two different crop rotations consisted of oats or a le-
cation (Site 1); and CONT, amendments applied annually gume-based green manure grown in a 2-yr rotation with pota-
since 1992 for 5 (Site 2) or 6 yr (Site 1). toes. Oats were seeded at a rate of 112 kg ha⫺1 and received
Amendments consisted of cull potato compost supplied by 45 kg N ha⫺1. The green manure consisted of pea, oats, and
a local producer (H. Smith Packing, Westfield, ME) and beef hairy vetch seeded as a mixture at rates of 112, 54, and 34 kg
cattle manure (Table 2). The application rate for compost was ha⫺1, respectively. The green manure was inoculated with Rhi-
22 Mg ha⫺1 in all years except 1996 when the application rate zobium before seeding and received no fertilizer. Oat straw
was 17 Mg ha⫺1 (dry matter equivalents of 5 and 11 Mg ha⫺1 and green manure were incorporated using a chisel plow in
the fall.
Table 1. Rotation crop management, fertilizer rates, and changes Analysis of variance was performed using a factorial ran-
in organic amendment application from 1992 to 1997. Details domized complete block design with four levels of amendment
for the 1992–1995 design are presented in Porter et al. (1999). and two levels of rotation crop. Amendment and rotation crop
1992–1995 Design 1996–1997 Design effects were analyzed separately for each year because of treat-
ment ⫻ year interactions and because the histories of amend-
Treatment Treatment ment application to Sites 1 and 2 were different. Statistical dif-
Fertilizer
Rotation Amendment Rotation Amendment rate† ferences between amendment means were determined using
Fisher’s LSD at P ⬍ 0.05.
N-P-K, kg ha⫺1
Green manure NONE‡ Green manure NONE 157-134-202
Green manure CONT§ Green manure CONT 90-45-67 Rotation Crop and Amendment Sampling
Oats NONE Oats NONE 224-134-202
Oats CONT Oats CONT 157-45-67 Quadrat samples (0.5 m2 each) were collected from each
Green manure NONE¶ Green manure FIRST# 90-45-134 oat and green manure plot to determine total aboveground
Green manure CONT Green manure SASP†† 157-134-134 biomass production. Oat straw yield was estimated from the
Oats NONE Oats FIRST 157-45-134
Oats CONT Oats SASP 224-134-134
difference between grain (determined using a small-plot com-
bine) and total biomass samples. The oat crop planted prior
† At-planting fertilizer rate to potatoes, given for 1996–1997, was based to the 1997 potato crop was not harvested due to unfavorable
on previous crop and amendment history.
‡ NONE, no amendments applied.
weather, so the total aboveground biomass was used to deter-
§ CONT, continuous amendment application since 1992. mine biomass contributions to the soil.
¶ Underscores designate the change in amendment treatments that oc- Rotation crop N content was determined in 16 green ma-
curred with the imposition of treatments used in this experiment. nure plots in 1997 and 16 oat plots in 1996 and 1997. In 1996,
# FIRST, amendments applied for the first time in 1996 (Site 2) or 1997
(Site 1).
the N content of the green manure was determined in eight
†† SASP, amendments last applied in 1995 (Site 2) or 1996 (Site 1) after plots. In 1997, a subsample of the unsorted, pooled sample from
application since 1992. eight green manure and eight oat plots was analyzed for C.
 GRANDY ET AL.: SOIL ORGANIC MATTER AND AGGREGATION IN POTATO SYSTEMS 1313

Table 3. Amendment and rotation crop effects on soil bulk den- for other analyses was spread thinly on plastic on a greenhouse
sity in 1996 and 1997.†‡ bench and stirred periodically until dry. Depending on the
Treatment 1996 1997 analysis conducted, dry soil was sieved to ⬍0.5 mm (WSC, C
and N), ⬍2 mm (LF), or ⬍6.5 mm (aggregates).
g cm⫺3
Two bulk density samples were taken in each plot to the
Amendment
NONE§ 1.15 ⫾ 0.04 0.83 ⫾ 0.04 depth of the Ap horizon (23 cm) in July 1996 and August
FIRST¶ 1.15 ⫾ 0.03 0.81 ⫾ 0.05 1997. Bulk density was calculated using the volume of the cor-
SASP# 1.15 ⫾ 0.05 0.80 ⫾ 0.04 ing device and oven-dry weight of the soil after correcting for
CONT†† 1.05 ⫾ 0.02 0.71 ⫾ 0.04 coarse fragments ⬎2 mm (Blake and Hartge, 1986). There
Rotation crop
Green manure 1.14 ⫾ 0.04 0.78 ⫾ 0.03 were no significant effects of amendment or rotation crop on
Oats 1.11 ⫾ 0.04 0.79 ⫾ 0.05 soil bulk density (Table 3), so C and N data are presented on
a gravimetric basis. Prior to amendment application in the
† Mean ⫾ standard error, n ⫽ 8.
‡ Amendment, rotation crop and amendment by rotation interactions were spring, samples were taken (23 cm) for soil mineral analysis,
nonsignificant (P ⬍ 0.1) in both years. dried, sieved (⬍2 mm) and subsequently analyzed by the Uni-
§ NONE, no amendments applied. versity of Maine Soil Testing Service using standard methods
¶ FIRST, amendments applied only in 1996 (Site 2) or 1997 (Site 1). (Hoskins, 1997).
# SASP, amendments last applied in 1995 (Site 2) or 1996 (Site 1) after
application since 1992.
†† CONT, continuous amendment application since 1992.
Soil Carbon, Nitrogen, and Organic Matter Fractions
Light fraction organic matter was separated using methods
The two residue types had statistically equal C levels (t-test, developed by Strickland and Sollins (1987). Fifty milliliters of
P ⬍ 0.05). The pooled mean (44.4%) was used as the C value NaI with a density of 1.7 g cm⫺3 was added to 25 g of soil in a
for all rotation crop samples in 1996 and 1997 to determine 100-mL centrifuge tube. Shaking for 45 min on a wrist-action
the C contribution to the soil by the rotation crops. shaker dispersed the samples and the heavy fraction was sepa-
Compost and manure samples were taken in triplicate prior rated out by centrifugation at ≈1070 g for 1 h. The LF was as-
to amendment application in the spring and analyzed for C, N, pirated from the surface of each sample into a vacuum flask
pH, and KCl-extractable NH4⫹ (Griffin et al., 1995). Nutrient and then washed with at least 150 mL of 0.01 M CaCl2 followed
analysis was conducted by inductively coupled plasma atomic by at least 200 mL of deionized water. The LF was dried at
emission spectrometry using a method adapted from Kalra
70⬚C for 24 h.
and Maynard (1991).
Water soluble carbohydrates were determined colorimetri-
cally using anthrone sulfuric acid reagent (Grandy et al., 2000).
Soil Sampling and Analysis Soils were incubated in an oven for 24 h at 85 ⬚C using a soil:
Soil used in WSC and aggregate analyses was collected with water ratio of 1:10 (w/v). Solutions were vacuum filtered through
a trowel to a depth of 15 cm three times each year: (i) before 0.3 ␮m glass filters, reacted with anthrone sulfuric acid reagent,
spring tillage and amendment application (May); (ii) after and absorbance was read at 625 nm using a Spectronic 1001
amendment application but before the first hilling (June); and (Milton Roy Co., Rochester, NY). A solution consisting of de-
(iii) before harvest (September). Ten subsamples per plot ionized water and anthrone was used as the blank.
were taken with care to avoid plot edges and rows used for Total C and N concentrations of the soil, LF, organic amend-
implement traffic and then composited. Soil collected in June ments, and rotation crops were measured by combustion (CN-
of each year was also analyzed for C, N, and LF. Soil used for 2000, Leco Corp., St. Joseph, MI). The low pH (Table 4) in
aggregate analysis was dried in paper bags at 26 ⬚C; soil used these plots and time since the last lime application (spring,

Table 4. Rotation and amendment effects on general soil chemical properties in 1996 and 1997, prior to start of FIRST [Amendments
applied only in 1996 (Site 2) and 1997 (Site 1)] and SASP [Amendments last applied in 1995 (Site 2) or 1996 (Site 1) after application
since 1992] treatments.
Treatment† pH P K Mg Ca CEC‡ K Mg Ca
kg ha⫺1 cmolc kg⫺1 % saturation
1996
Amendment
CONT§ 5.8** 38.5*** 812*** 538* 2065*** 7.5*** 12.6*** 26.0 60.4*
NONE¶ 5.5 30.4 455 437 1486 5.7 9.1 27.6 57.2
Rotation crop
Green manure 5.6* 33.5 638 470 1703 6.5 11.1 26.2 57.3
Oats 5.7 35.3 629 505 1848 6.7 10.6 27.4 60.2
1997
Amendment
CONT 5.7*** 47.0*** 887*** 543*** 2047*** 7.7*** 13.1*** 25.6 58.7***
NONE 5.3 32.2 448 345 1190 5.2 9.8 23.8 50.0
Rotation crop
Green manure 5.5 40.8 660 452 1647 6.5 11.1 25.0 54.7
Oats 5.5 38.5 675 436 1591 6.4 11.7 24.5 53.9
* Indicate significant effects within a year, factor (amendment or rotation crop), and variable at P ⬍ 0.05.
** Indicate significant effects within a year, factor (amendment or rotation crop), and variable at P ⬍ 0.01.
*** Indicate significant effects within a year, factor (amendment or rotation crop), and variable at P ⬍ 0.001.
† Rotation crop ⫻ amendment interactions were nonsignificant for all variables.
‡ CEC, cation exchange capacity.
§ CONT, continuous annual amendment application.
¶ NONE, no amendments applied.
1314 SOIL SCI. SOC. AM. J., VOL. 66, JULY–AUGUST 2002

Fig. 1. Amendment (compost and manure) and rotation crop C and N contributions to the soil in 1997.

1992) indicate that there was little carbonate present in the tions of these nutrients were all significantly increased
soil; therefore, the total soil C measured likely originated from by amendment application.
organic C.

Water Stable Aggregates Soil Carbon, Nitrogen, Light Fraction,

and Water Soluble Carbohydrates
Aggregates were obtained by a modification of the wet-
sieving method developed by Yoder (1936) and more recently In 1996 FIRST and NONE had equal C contents, and
described by Porter et al. (1999). Briefly, a nest of screens with in 1997 FIRST increased soil C by 28% (Fig. 2). Amend-
descending mesh size was used to separate aggregates into ments applied annually for 5 or 6 yr increased total soil
three size classes: 2 to 6.5 mm (large), 1 to 2 mm (medium), and C by 23% relative to SASP in 1996, but there were no
0.25 to 1 mm (small). A 25-g sample was completely immersed significant differences between these treatments in 1997.
in water for 10 min on the uppermost sieve, followed by oscilla- More C was present in SASP compared with NONE
tion for 10 min at 35 oscillations min⫺1. After sieving, the
soils in both 1996 (28%) and 1997 (46%). Total soil N
sample collected on each sieve was washed into beakers, oven
dried, and weighed. Sand content was determined following was statistically equal in FIRST and NONE, and in
dispersion in sodium hexametaphosphate. CONT and SASP, in both years of the study (Fig. 2).
SASP resulted in more N than NONE in 1996 (23%)
and 1997 (22%). Rotation crop had no effects on total
RESULTS soil N in either year or on total soil C in 1996 (data not
Amendment and rotation crop C and N contributions shown). In 1997, the green manure crop resulted in
to the soil are presented only for 1997 (Fig. 1) because higher soil C (P ⬍ 0.05) than the oat rotation crop
the C content of the manure and compost amendments (25.9 vs. 23.9 g kg⫺1).
was not available for previous years. Amendments con- Single amendment applications increased all LF prop-
tributed ≈1.8 times more C to the soil than either rota- erties in 1996 and 1997 relative to NONE (Tables 5
tion crop and ≈1.8 and 2.8 times more N than the green and 6). All LF properties were increased by CONT
manure and oat rotation crops, respectively. In 1997, relative to SASP in 1996 (except for N concentration)
the C contributions from the two rotation crops were and 1997, and were greater in SASP than NONE in
similar because the oat crop was not harvested the previ- 1996 (except for N concentration) and 1997. The green
ous fall. The green manure contributed greater N to the manure crop increased LF N concentration, total LF N
soil than the oats (P ⬍ 0.01) and increased the C/N in the soil, and the percentage of total soil C and N in
ratio (9.95 green manure vs. 9.01 oats; P ⬍ 0.01) of the the LF in 1996. In 1997, the green manure crop increased
soil (data not shown). In 1996, the green manure contrib- all LF properties except the percentage of total soil C
uted more C (2600 vs. 1100 kg ha⫺1) and N (137 vs. in the LF.
40 kg ha⫺1) than oats (data not shown). Soil bulk density The May carbohydrate and aggregate samples were
was not significantly affected by amendment or rotation collected before imposition of the revised amendment
crop (Table 3). treatments, so we expected that CONT and SASP would
Soil mineral analysis was done in the spring before be equal because these plots had the same treatment
imposition of the altered amendment treatments (FIRST history, and that NONE and FIRST would be equal but
and SASP), so comparisons among two amendment lower than the other two. In May 1996, WSC in FIRST
(NONE vs. CONT) and the rotation crop programs are was lower than in NONE, and WSC in CONT was
relevant (Table 4). As a result of the high levels of P, higher than in SASP. Consequently, we were not able
K, Mg, and Ca in the amendments (Table 2), concentra- to detect greater WSC in FIRST relative to NONE
 GRANDY ET AL.: SOIL ORGANIC MATTER AND AGGREGATION IN POTATO SYSTEMS 1315

Fig. 2. Amendment (compost and manure) effects on soil: (a) C, and (b) N. Amendment treatments: NONE, no amendments applied; FIRST,
amendments applied only in 1996 (Site 2) or 1997 (Site 1); SASP, amendments last applied in 1995 (Site 2) or 1996 (Site 1) after application
since 1992; and CONT, continuous amendment application since 1992. Means followed by the same letter within years are statistically equal
(P ⬍ 0.05) using Fisher’s LSD.

during June or July, or decreases in SASP compared amendment addition. In September, the content of me-
with CONT at these dates (Fig. 3a). dium aggregates in CONT was significantly higher than
In 1997, FIRST resulted in greater WSC than NONE in SASP, and in SASP relative to NONE. In September
on the June and September sampling dates (Fig. 3b). 1996, the large aggregate content of the soil was greater
There were no differences between CONT and SASP, in FIRST than NONE, in CONT than SASP, and in
and SASP maintained higher WSC than NONE. Green SASP than NONE.
manure resulted in significantly greater WSC than oats In May 1997, the content of small aggregates was
in June 1996 (7%, P ⬍ 0.05), June 1997 (8%, P ⬍ 0.05), greater in SASP than in CONT (Fig. 5). In June and Sep-
and September 1997 (20%, P ⬍ 0.01) (data not shown). tember, all four treatments were statistically equal, indi-
The differences were not significant on the other three cating a relative decrease in SASP, increase in NONE,
sampling dates. and no differences between FIRST and NONE. In June
and September, the content of medium aggregates in
the soil was statistically greater in FIRST relative to
Soil Aggregation NONE. There were no significant differences in medium
In 1996, there were no amendment effects on the small aggregate content between CONT and SASP; SASP
aggregate content in the soil (Fig. 4). The medium aggre- retained higher aggregate content than NONE. Single
gate content of the soil in June was not affected by amendment applications increased the content of large
1316 SOIL SCI. SOC. AM. J., VOL. 66, JULY–AUGUST 2002

Table 5. Amendment and rotation effects on selected properties

of the soil light fraction (LF) in 1996.
LF N C LF N LF C LF N LF C
g kg⫺1 soil g kg⫺1 LF mg kg⫺1 soil % soil N % soil C
Amendment
CONT† 29.4a‡ 17.4ab 317a 511a 9284a 17.0a 29.5a
SASP§ 17.6b 16.9bc 304b 299b 5365b 10.5b 20.9b
FIRST¶ 15.9b 18.1a 318a 289b 5049b 12.1b 22.9b
NONE# 9.7c 16.5c 283c 160c 2743c 6.9c 13.7c
Rotation crop
Green manure 19.0 17.7 306.3 338 5904 12.7 23.1
Oats 17.3 16.7 304.8 292 5317 10.5 20.3
ANOVA
Amendment (A) *** * *** *** *** *** ***
Rotation (R) NS†† *** NS * NS * *
A⫻R NS NS NS NS NS NS NS
* Significant at the 0.05 probability level.
*** Significant at the 0.001 probability level.
† CONT, continuous amendment application since 1992.
‡ Amendment means followed by the same letter are statistically equal
at P ⬍ 0.05 using Fisher’s LSD test.
§ SASP, amendments last applied in 1995 (Site 2) or 1996 (Site 1) after
application since 1992.
¶ FIRST, amendments applied only in 1996 (Site 2) or 1997 (Site 1).
# NONE, no amendments applied.
†† NS, not significant.

Table 6. Amendment and rotation effects on selected properties

of the soil light fraction (LF) in 1997.
LF N C LF N LF C LF N LF C
g kg⫺1 soil g kg⫺1 LF mg kg⫺1 soil % soil N % soil C
Amendment
CONT† 31.4a‡ 17.9a 307a 563a 9652a 18.9a 32.6a
SASP§ 21.1b 17.1b 295b 365b 6319b 13.4b 23.3b
FIRST¶ 22.3b 17.6ab 297b 392b 6594b 15.4b 27.2b
NONE# 12.7c 16.5c 284c 211c 3611c 9.5c 19.4c
Rotation crop
Green manure 23.6 17.8 298 421 7078 15.9 26.8
Oats 20.1 16.6 293 339 5946 12.4 24.3
ANOVA
Amendment (A) *** *** *** *** *** *** ***
Rotation (R) ** *** * ** ** *** NS††
A⫻R NS * NS NS NS NS NS Fig. 3. Amendment (compost and manure) effects on soil water-solu-
* Significant at the 0.05 probability level. ble carbohydrates: (a) 1996 and (b) 1997. Amendment treatments:
** Significant at the 0.01 probability level. NONE, no amendments applied; FIRST, amendments applied only
*** Significant at the 0.001 probability level. in 1996 (Site 2) or 1997 (Site 1); SASP, amendments last applied
† CONT, continuous amendment application since 1992. in 1995 (Site 2) or 1996 (Site 1) after application since 1992; and
‡ Amendment means followed by the same letter are statistically equal CONT, continuous amendment application since 1992. The May
at P ⬍ 0.05 using Fisher’s LSD test. sample was taken before imposition of the altered amendment
§ SASP, amendments last applied in 1995 (Site 2) or 1996 (Site 1) after
application since 1992. treatments. Data were analyzed separately for each sample date.
¶ FIRST, amendments applied only in 1996 (Site 2) or 1997 (Site 1). Means followed by the same letter within sample dates are statisti-
# NONE, no amendments applied. cally equal (P ⬍ 0.05) using Fisher’s LSD.
†† NS, not significant.
that a single organic amendment application at high
aggregates in the soil relative to NONE in September. rates can prevent additional SOM losses and soil struc-
Large aggregate content of the soil was equal in CONT ture degradation. Light fraction C accounted for 56%
and SASP, and higher in SASP than NONE, on all (by weight) of the soil C increase in FIRST. Changes
sampling dates. in C and N concentrations of the LF in FIRST should
There were no significant rotation crop effects on ag- increase the ability of this SOM pool to supply plant
gregation in 1996 (P ⬍ 0.05). In 1997, green manure in- nutrients and provide habitat for microorganisms (Wan-
creased medium aggregates compared with oats in May der et al., 1994).
(4.98 vs. 3.94%) and June (5.51 vs. 4.67%, by weight). Differences in total C between SASP and CONT in
Large aggregates were also increased by green manure 1996 could largely be accounted for by declines in LF C;
in May (7.58 vs. 5.25%) and June (9.86 vs. 7.85%, by 67% of the 5.9 g kg⫺1 C deficit in SASP was due to
weight) (data not shown). differences in LF C. Converting forest soils and grass-
lands to cultivated agriculture often results in a rapid
decline in C as microbes degrade highly labile material
DISCUSSION (Giddens, 1957) such as LF. In our studies, the short
Total soil C increases in FIRST in 1997 are consistent duration of amendment application in the CONT plots
with previous work (Porter et al., 1999), demonstrating resulted in the accumulation of a large proportion of
 GRANDY ET AL.: SOIL ORGANIC MATTER AND AGGREGATION IN POTATO SYSTEMS 1317

Fig. 4. Amendment effects on soil aggregates in 1996: (a) small (0.25–

1.0 mm); (b) medium (1.0–2.0 mm); and (c) large (2.0–6.5 mm). Fig. 5. Amendment effects on soil aggregates in 1997: (a) small (0.25–
Amendment treatments: NONE, amendments never applied; 1.0 mm); (b) medium (1.0–2.0 mm); and (c) large (2.0–6.5 mm).
FIRST, amendments applied for the first time in 1996; SASP, Amendment treatments: NONE, amendments never applied;
amendments last applied in 1995 after application since 1992; and FIRST, amendments applied for the first time in 1997; SASP,
CONT, amendments applied since 1992. The May sample was taken amendments last applied in 1996 after application since 1992; and
before imposition of the altered amendment treatments. Means CONT, amendments applied since 1992. The May sample was taken
within a year and sample date followed by the same letter are before imposition of the altered amendment treatments. Means
statistically equal (P ⬍ 0.05) using Fisher’s LSD test. No grouping within a year and sample date followed by the same letter are
letters are presented for sample dates that had a nonsignificant statistically equal (P ⬍ 0.05) using Fisher’s LSD test. No grouping
(P ⬍ 0.05) F-test for treatment effects. letters are presented for sample dates that had a nonsignificant
(P ⬍ 0.05) F-test for treatment effects.
labile materials, rather than humic substances, that are
rapidly decomposed in intensive potato cropping sys- and Biederbeck et al. (1998) found that the LF as a
tems. This is indicated by the high percentage of total proportion of soil organic C ranged from 9.5 to 11.4%
soil C in the LF (29.5% in 1996 and 32.6% in 1997). In for different cropping systems. Enhancing the conserva-
comparison, Bremer et al. (1994) found that LF ac- tion of LF through tillage reduction or other strategies
counted for 9 to 24% (by weight) of soil organic C would increase C retention in the amended systems.
1318 SOIL SCI. SOC. AM. J., VOL. 66, JULY–AUGUST 2002

Declines in the percentage of total soil C in the LF changes in soil management; macroaggregates (⬎250
with decreasing organic matter inputs suggest that the ␮m), stabilized by roots, fungal hyphae, and polysaccha-
rate of C decline in SASP should slow down in sub- rides, are more transient and sensitive to short-term
sequent years (Hyvönen et al., 1998). This is also sup- treatment effects than microaggregates (Tisdall and
ported by changes in LF quality in SASP relative to Oades, 1982; Haynes and Beare, 1996; Jastrow et al.,
CONT; SASP LF had a lower C concentration than 1996). The small aggregates in our study, which are only
CONT in 1996 and 1997, and lower N content in 1997. slightly larger than microaggregates, may be stabilized
Light fraction materials with a high C and N content in by humic substances or other stable organic fractions
SASP were likely the most rapidly colonized and de- that are unaffected by short-term changes in soil man-
composed substrates, and their decline should result in agement. Several years may be necessary before changes
slower mineralization rates. in organic matter inputs alter organic matter fractions
Differences between SASP and NONE indicate that that stabilize small aggregates.
the benefits of amendment application on total soil C It was expected that WSC, because of their rapid
and LF persisted for at least one growing season after production after C additions to soil, would be involved
amendment application ceased. Eventually, a new equi- in aggregate stabilization. Studies investigating manage-
librium between inputs and losses of organic materials ment-induced, short-term changes in soil structure have
will be achieved, which may have greater C than plots found strong correlations (r ⫽ 0.74) between aggrega-
that never received amendments (Jenkinson and Ray- tion and soil carbohydrates (Haynes and Swift, 1990).
ner, 1977; Sommerfeldt et al., 1988). Although there were significant correlations between
Greater C contributions from the green manure rela- WSC and small and medium aggregates in 1997, our re-
tive to oats since 1992 (except in 1997) likely explain sults suggest that there may be factors or organic matter
soil C increases in 1997 and changes in LF in 1996 and fractions more closely related to macroaggregation than
1997. Additionally, the green manure residue had a low is WSC. Others have suggested that microbial biomass
C/N ratio which may have enhanced its sequestration C or fungal hyphae may exert major effects on macro-
within aggregates (Drinkwater et al., 1998); evidence aggregation (Bethlenfalvay and Barea, 1994; Degens
for this is that the green manure increased medium and et al., 1994). Previous work in these plots showed that
large aggregates in 1997. Relatively high and spatially amendment applications significantly increased root
variable background levels of soil C likely caused the length density and root growth (Porter et al., 1999).
lack of an effect in 1996 and relatively small differences Some of the increase in aggregation may have been due
in 1997. More time may be necessary for the soil C ef- to physical entanglement of soil particles with roots
fects of the green manure rotation crop to become more and production of plant mucilages (or products of their
consistent. Similarity of N among rotation crops and degradation) that directly stabilize aggregates (Perfect
amendment treatments may be due to the addition of et al., 1990; Haynes and Beare, 1996).
more N fertilizer to unamended potato plots and in the
oat rotation compared with the green manure rotation. SUMMARY AND CONCLUSION
Organic materials stabilizing the small aggregates ap-
peared to be different from those stabilizing the larger Our study indicates that recovery of organic matter
aggregate size classes. In contrast to medium and large pools and aggregation can occur in intensively managed
aggregates, small aggregates did not change in response Maine potato production systems with compost and ma-
to FIRST or SASP, and were generally not correlated nure application less frequently than annually. FIRST
with soil C, N, and LF properties (Table 7). Models of effects on medium and large aggregates demonstrate
aggregate formation and stabilization typically separate that soil structural improvements may occur quickly.
aggregates into two classes: microaggregates (⬍250 ␮m) However, slow changes in small aggregates, which rep-
are stabilized by clay-polyvalent cation-humic substance resent the majority of aggregated soil in this study, in
complexes and are relatively resistant to short-term response to altered organic inputs suggests that several
consecutive annual amendments may initially be neces-
Table 7. Correlations of small, medium, and large aggregates sary. Persistence of SOM and soil structure in SASP
with soil C, N, light fraction (LF), and water soluble carbohy- provides evidence that once soil improvements occur,
drates (WSC). periodic amendment applications should maintain them.
Small Medium Large Light fraction represented a high proportion of total C
aggregates aggregates aggregates in amended plots and changed rapidly with changes in
1996 1997 1996 1997 1996 1997 organic amendment frequency. One way to maximize
the effects of amendment application would be to facili-
Total C 0.24 0.28 0.39* 0.65*** 0.48** 0.51**
Total N 0.34 0.24 0.34 0.35* 0.46** 0.39* tate the stabilization of this organic matter pool. En-
LF dry weight ⫺0.13 ⫺0.08 0.31 0.40* 0.33 0.47** hancing soil aggregation by minimizing soil disturbance
LF N concentration 0.04 ⫺0.16 ⫺0.03 0.32 ⫺0.03 0.34 could increase the physical protection of LF and thus
LF C concentration 0.43* 0.16 0.19 0.50** 0.35* 0.44*
Total LF N in the soil ⫺0.12 ⫺0.08 0.30 0.43* 0.31 0.48** increase its residence time in the soil (Six et al., 1999).
Total LF C in the soil ⫺0.09 ⫺0.05 0.33 0.44* 0.35* 0.49** Rotation crop effects were generally small compared
LF N in soil/total soil N ⫺0.25 ⫺0.19 0.21 0.38* 0.23 0.45* with amendment effects, but changes in total C (1997)
LF C in soil/total soil C ⫺0.24 ⫺0.29 0.19 0.20 0.24 0.40*
WSC (June sample) 0.11 0.45* 0.05 0.52** 0.06 0.00 and LF indicate that green manure crops may be used
as part of a soil management program intended to re-
* Significant at the 0.05 probability level.
** Significant at the 0.01 probability level. verse soil degradation. Reducing the frequency of pota-
*** Significant at the 0.001 probability level. toes in the rotation or including a perennial forage may
 GRANDY ET AL.: SOIL ORGANIC MATTER AND AGGREGATION IN POTATO SYSTEMS 1319

have greater effects on soil C pools and structure than storage in meso-thermal, humid soils. In M.R. Carter and B.A.
the green manure used here (Perfect et al., 1990; Angers Stuart (ed.) Structure and organic matter storage in agricultural
soils. CRC Press, New York.
et al., 1999). Haynes, R.J., and R.S. Swift. 1990. Stability of soil aggregates in
relation to organic constituents and soil water content. J. Soil
ACKNOWLEDGMENTS Sci. 41:73–83.
Hepler, P.R., L.H. Long, and J.A. Ferwerda. 1983. Crop yield and
We thank W.B. Bradbury, B. MacFarline, and J.A. Sisson quality relationships with soil erosion. Field appraisal of resource
for their technical assistance. W. Halteman is gratefully ac- management systems (FARMS). Bull. 799. Maine Agric. Exp. Stn.,
knowledged for providing advice about the experimental de- Orono, ME.
sign and statistical analysis of the experiment. We greatly Hoskins, B.R. 1997. Soil testing handbook for professionals in agricul-
ture, horticulture, nutrient and residuals management. 3rd ed.
appreciate the helpful comments provided by the reviewers Maine Forestry and Agric. Exp. Stn, Orono, ME.
of this paper and L.M. Zibilske for his insightful comments Hyvönen, R., G.I. Ågren, and E. Bosatta. 1998. Predicting long-term
throughout the study. We also acknowledge H. Smith Packing, soil carbon storage from short-term information. Soil Sci. Soc.
Inc. of Westfield, ME, for providing the waste potato compost Am. J. 62:1000–1005.
used in these studies. Jastrow, J.D., T.W. Boutton, and R.M. Miller. 1996. Carbon dynamics
Research supported by USDA-CSREES Special Grants of aggregate-associated organic matter estimated by Carbon-13
for Potato Research (94-34141-0040), the University of Maine natural abundance. Soil Sci. Soc. Am. J. 60:801–807.
Jenkinson, D.S., and J.H. Rayner. 1977. The turnover of soil organic
Potato Ecosystem Project, Northeast SARE/ACE Grant matter in some of the Rothamsted classical experiments. Soil
LNE93-36/ANE93.18, McCain Foods Ltd., and the Maine Po- Sci. 123:298–305.
tato Board. Kachanoski, R.G., and M.R. Carter. 1999. Landscape position and
soil redistribution under three soil types and land use practices in
Prince Edward Island. Soil Till. Res. 51:211–217.
REFERENCES Kalra, Y.P., and D.G. Maynard. 1991. Dry ashing (ignition) for Ca,
Mg, K, P, Cu, Na, Ni, Zn, Mn, Fe, and Al. p. 101–103. In Methods
Angers, D.A., L.M. Edwards, J.B. Sanderson, and N. Bissonnette. manual for forest soil and plant analysis. Inf. Rep. NOR-X-319.
1999. Soil organic matter quality and aggregate stability under eight Forestry Canada, Edmonton, AB.
potato cropping sequences in a fine sandy loam of Prince Edward Lal, R., A.A. Mahboubi, and N.R. Fausey. 1994. Long-term tillage
Island. Can. J. Soil Sci. 79:411–417. and rotation effects on properties of a central Ohio soil. Soil Sci.
Angers, D.A. and G.R. Mehuys. 1989. Effects of cropping on carbohy- Soc. Am. J. 58:517–522.
drate content and water-stable aggregation of a clay soil. Can. J. Perfect, E., B.D. Kay, W.K.P. van Loon, R.W. Sheard, and T. Pojasok.
Soil Sci. 69:373–380. 1990. Rates of change in soil structural stability under forages and
Araji, A.A., Z.O. Abdo, and P. Joyce. 2001. Efficient use of animal corn. Soil Sci. Soc. Am. J. 54:179–186.
manure on crop land—Economic analysis. Bioresour. Technol. 79: Porter, G.A., and J.C. McBurnie. 1996. Crop and soil research. p. 8–62.
179–191. In M.C. Marra (ed.) The ecology, economics, and management of
Bending, G.D., and M.K. Turner. 1999. Interaction of biochemical potato cropping systems: A report of the first four years of the
quality and particle size of crop residues and its effect on the Maine Potato Ecosystem Project. Bull. no. 843. Maine Agric. and
microbial biomass and nitrogen dynamics following incorporation Forest Exp. Stn., Orono, ME.
into soil. Biol. Fertil. Soils. 29:319–327. Porter, G.A., G.B. Opena, W.B. Bradbury, J.C. McBurnie, and J.A.
Bethlenfalvay, G.J., and J.M. Barea. 1994. Mycorrhizae in sustainable Sisson. 1999. Soil management and supplemental irrigation effects
agriculture. I. Effects on seed yield and soil aggregation. Am. J. on potato: I. Soil properties, tuber yield, and quality. Agron J. 91:
Altern. Agric. 9(4):157–161. 416–425.
Biederbeck, V.O., C.A. Campbell, V. Rasiah, R.P. Zentner, and Robertson, G.P., J.R. Crum, and B.G. Ellis. 1993. The spatial variabil-
G. Wen. 1998. Soil quality attributes as influenced by annual le- ity of soil resources following long-term disturbance. Oecologia
gumes used as green manure. Soil Biol. Biochem. 30:1177–1185. 96:451–456.
Blake, G.R., and K.H. Hartge. 1986. p. 363–375. In A. Klute (ed.) Robertson, G.P., E.A. Paul, and R.R. Harwood. 2000. Greenhouse
Methods of Soil Analysis. Part 1. Physical and Mineralogical Meth- gases in intensive agriculture: contributions of individual gases
ods. Agron. Monogr. 9. ASA and SSSA, Madison, WI. to the radiative forcing of the atmosphere. Science (Washington,
Bremer, E., H.H. Janzen, and A.M. Johnston. 1994. Sensitivity of total, DC) 289:1922–1925.
light fraction and mineralizable organic matter to management Saini, G.R., and W.J. Grant. 1980. Long-term effects of intensive cul-
practices in a Lethbridge soil. Can. J. Soil Sci. 74:131–138. tivation on soil quality in the potato-growing areas of New Bruns-
Compton, J.E., and R.D. Boone. 2000. Long-term impacts of agricul- wick (Canada) and Maine (U.S.A). Can. J. Soil Sci. 60:421–428.
ture on soil carbon and nitrogen in New England forests. Ecol- Six, J., E.T. Elliot, and K. Paustian. 1999. Aggregate and soil organic
ogy 81:2314–2330. matter dynamics under conventional and no-tillage systems. Soil
Degens, B.P., G.P. Sparling, and L.K. Abbott. 1994. The contribution Sci. Soc. Am. J. 63:1350–1358.
from hyphae, roots and organic carbon constituents to the aggrega- Six, J., E.T. Elliot, K. Paustian, and J.W. Doran. 1998. Aggregation and
tion of a sandy loam under long-term clover-based and grass pas- soil organic matter accumulation in cultivated and native grassland
tures. Eur. J. Soil Sci. 45:459–468. soils. Soil Sci. Soc. Am. J. 62:1367–1377.
Drinkwater, L.E., P. Wagoner, and M. Sarrantonio. 1998. Legume- Sommerfeldt, T.G., C. Chang, and T. Entz. 1988. Long-term annual
based cropping systems have reduced carbon and nitrogen losses. manure applications increase soil organic matter and nitrogen,
Nature (London) 396:262–265. and decrease carbon to nitrogen ratio. Soil Sci. Soc. Am. J. 52:
Giddens, J. 1957. Rate of loss of carbon from Georgia soils. Soil Sci. 1668–1672.
Soc. Am. Proc. 21:513–515. Strickland, T.C., and P. Sollins. 1987. Improved method for separating
Gilley, J. E., and L.M. Risse. 2000. Runoff and soil loss as affected light- and heavy-fraction organic material from soil. Soil Sci. Soc.
by the application of manure. Trans. ASAE 43:1583–1588. Am. J. 51:1390–1393.
Grandy, A.S., M.S. Erich, and G.A. Porter. 2000. Suitability of the Tisdall, J.M., and J.M. Oades. 1982. Organic matter and water-stable
anthrone-sulfuric acid reagent for determining water soluble carbo- aggregates in soils. J. Soil Sci. 33:141–163.
hydrates in soil water extracts. Soil Biol. Biochem. 32:725–727. Wander, M.M., S.J. Traina, B.R. Stinner, and S.E. Peters. 1994. Or-
Griffin, G., W. Jokela, and D. Ross. 1995. Recommended soil nitrate-N ganic and conventional management effects on biologically active
test. In Recommended soil testing procedures for the Northeastern soil organic matter pools. Soil Sci. Soc. Am. J. 58:1130–1139.
United States. 2nd ed. Northeastern Regional Publication No. 493, Yoder, R.E. 1936. A direct method of aggregate analysis of soils
Newark, DE. and a study of the physical nature of erosion losses. J. Am. Soc.
Haynes, R.J., and M.H. Beare. 1996. Aggregation and organic matter Agron. 28:337–351.