Does the Slope Aspect Really Affect the Soil Chemical Properties, Growth and Arbuscular Mycorrhizal Colonization of Centipedegrass in a Hill Pasture? ^†

Abstract

Arbuscular mycorrhizal (AM) fungi (AMF) form a symbiotic association with terrestrial plants and increase growth and productivity. The relationships between the growth of centipedegrass (CG) and AMF are not well understood. We monitored the growth and AM colonization of CG growing on the four slopes (north, east, south, and west) of a pasture, to obtain information on aspect differences in the soil chemical properties–grass–AMF association. Soil properties almost always varied between the slope aspects. The total soil N, C, EC, and moisture tended to be highest on the northern aspect, whereas the soil available P and pH tended to be highest on the western and southern aspects, respectively. Despite the aspect differences in the microclimate and soil properties, CG grew well in all aspects, showing similar dry matter weights (DMW) for the fouraspects. Furthermore, the AM colonization of CG, in any characteristic structures (internal hyphae, vesicles, and arbuscules), was not significantly different between the slope aspects on most measurement occasions, although the colonization usually varied between the seasons and years. There were no relationships between the DMW and AM characteristic structure colonization and between the DMW and soil chemical properties. However, the colonization of the arbuscules and vesicles of the CG had a correlation with some soil chemical properties. The results suggest that AM colonization on CG growing in a hill pasture did not differ between the slope aspects. This may be a factor contributing to the high adaptability of the grass to all slope aspects.

1. Introduction

Generally, grassland is distributed on flat and slope lands. The aspect of a slope can affect some climatic, edaphic, and biotic variables [1,2,3,4,5]. Important climatic variables that require consideration include the average amount and seasonal distribution of precipitation, seasonal temperature fluctuations, and length of the growing season [5]. In the Northern Hemisphere, north-facing slopes tend to be wetter and cooler than south-facing slopes [5], mainly through variation in the soil temperature caused by differences in solar radiation [1,2]. In previous reports [1,4,6,7,8,9], the soil organic C, total N, and moisture content were higher on shady slopes than on sunny slopes, whereas the light intensity, soil temperature, and soil pH were higher on sunny slopes than on shady slopes. The slope aspect also has an influence on the productivity of the herbage [2,10]. Lambert and Roberts [2] reported that, in New Zealand, east-facing slopes have a higher pasture productivity than the others. Reflecting these aspect differences, hilly pastures in the mid-altitude region of Kyushu, Japan are often dominated by cool-season grasses on the northern aspect and warm-season grasses on the southern aspect [11,12].

Centipedegrass (Eremochloa ophiuroides (Munro) Hack.) is a warm-season perennial that is native to central/southern China and is widely distributed in south-east Asia, southern USA, South America, the West Indies, and parts of Africa and tropical north/east Australia [13,14,15]. It is adapted to a range of environments, showing superior cold tolerance compared to most warm-season grasses and can grow on a range of soils (sandy to heavy soils; pH 4.5–8.5; well adapted to infertile soil) [12,14]. Centipedegrass forms a relatively short and dense sward or turf with leafy stolons and short internodes, requiring low maintenance or little care [12,16,17]. Centipedegrass possesses more acceptable or superior characteristics than a warm-season forage grass in terms of the production and quality of herbage and preference/intake by animals [18]. Centipedegrass is considered to be of potential value for use in low-input grassland systems and environmental conservation in Japan [13,19,20]. An experiment conducted in a hill pasture in the mid-altitude region of Kyushu has shown that centipedegrass is well adapted to all slope aspects (north, east, south, and west), despite the aspect differences concerning the environmental conditions [12]. However, the effects of the slope aspect on the growth and dry matter production of centipedegrass are not sufficiently understood.

The role soil micro-organisms play in sustainable agricultural systems is becoming increasingly important, and the use of synthetic inputs such as fertilizers and chemical pest control agents is being reduced or avoided [21]. In conventional high-input systems, there are continual disturbances in soil system (e.g., the addition of chemicals), which probably affect intrinsic abiotic and biotic soil factors, possibly leading to long-term soil degradation [21,22]. Rhizobia fix atmospheric nitrogen via symbiosis with legumes and contribute to low input grassland production systems and environmental conservation, reducing the application of nitrogen fertilizer to soil and companion grass [23,24]. In addition, arbuscular mycorrhizal (AM) fungi (AMF) probably play a critical role in low-input or organic systems because of their role in linking plant and soil processes [21,25]. The use of the soil symbiont is useful for the development of low-input grassland systems and for environmental conservation. However, the relationships between the growth of centipedegrass and belowground microbial symbionts are not well understood.

AMF form a symbiotic association with more than 80% of terrestrial plants and principally benefit their hosts by increasing the uptake of water/nutrients [26,27,28] and improving the soil structure [28,29]. This is particularly important for phosphorus uptake, as fungal extraradical mycelium can access relatively immobile phosphate ions through the ability to grow beyond the phosphate depletion zone that rapidly develops around roots [26]. This symbiotic association is known to promote growth and improve the drought and disease resistance of the host plants [26]. Due to their ability to increase plant nutrient acquisition, AMF are increasingly recognized as having an important role in sustainable agricultural production systems [30,31]. Previous studies have reported that AMF communities are affected by a range of factors, such as soil characteristics, plant species, and climate [27,32,33,34]. According to previous reports [8,9], AM colonizations are affected by the slope aspect, being higher on sunny slopes than on shady slopes. It was considered that sunshine and temperature influence it, and as a result, these studies identified the soil chemical properties it affects [8,9]. However, the regions investigated in these studies have arid, cool summer climates or semi humid alpine climates; they are more than 1000 m above sea level and more than 35º N. Their vegetation types are temperate or alpine plant communities, and only few reports have investigated tropical grass. In addition, the effects of slope aspect on the AM colonization of centipedegrass and the relationships between AMF and soil chemical properties are not sufficiently understood. From previous field studies [35,36,37,38], seasonal variation within the AM colonization rate was confirmed. It was found that the AM colonization rate increases from summer to early fall and then decreases in winter [36,37,38]. AM colonization seasonality is dependent upon the host plant, field conditions [35,38], carbon sinks in mycorrhizal-independent plants [35,37], host plants demand for nutrients [27,38], and air temperature [39]. However, the seasonal variations in AM colonization of centipedegrass are not well understood. It is important to clarify the relationships between AMF and the productivity of centipedegrass in order to develop low-input grassland systems. In addition, it contributes to detailed pasture management, by clarifying the effect of the slope aspects on the growth, AMF colonization, and seasonal variation in centipedegrass. Tobisa et al. [40] reported that effect of slope aspect on arbuscular mycorrhizal colonization of centopedegrass on a hill pasture, but sufficient understanding was not obtained.

In the present study, we monitored the AM colonization of centipedegrass growing on the four slopes of a pasture to examine the aspect differences in the soil chemical properties–grass–AMF association, in reference to a previous report [40]. Furthermore, we examined the seasonal variation in the AM colonization of centipedegrass, affected by the aspect differences.

2. Materials and Methods

2.1. Research Site and Observations

The study was conducted between 2007 and 2009 on the 4 slopes (north, east, south, and west; approximately 25° gradient) in a hill paddock (2.4 ha; 32°42′ N; 131°46′ E; 540 m a.s.l.; Brown Forest Soil) at the Kagamiyama Livestock Farm in Nobeoka City, Miyazaki Prefecture, southwestern Japan (

Table 1. Summary of research method, experimental design, and evaluation parameters.

Item	Contents
Plant species, cultivar	Centipedegrass (Eremochloa ophiuroides (Munro) Hack.) cv. “Common”
Research Site	Hill paddock (2.4 ha; 32°42′ N; 131°46′ E; 540 m a.s.l.; Brown Forest Soil) at the Kagamiyama Livestock Farm in Nobeoka City, Miyazaki Prefecture, southwestern Japan
Treatment	4 slope aspects (north, east, south, and west; approximately 25° gradient)
Experimental design	4 slope aspects (north, east, south, and west), three replicated plots, each 2 m (down the slope) × 4 m (across the slope)
Established dates	mid-May (late spring) 2000
Grassland management and fertilizer application	Paddock was stocked with beef cows from spring (late April–early May) to autumn (October–early November) Compound fertilizer once (spring) or twice (spring and autumn) a year, at average annual rates of 70 kg N ha−1 and 70 kg K ha−1
Sampling dates	June and October 2007 (2 times,); May, July, August, and November 2008 (4 times); and June, July, August, October, and November 2009 (5 times)
Evaluation parameters	AM colonization (internal hyphae, vesicles, and arbuscules), plant dry matter weight (DMW), soil properties (moisture, total N, total C, available P, EC, pH, porosity)

) [12]. In 2000, 3 replicated plots, each 2 m (down the slope) × 4 m (across the slope), were established on each of the 4 aspects and sown with centipedegrass (cv. Common) [12]. In subsequent years, the paddock was stocked with beef cows from spring (late April–early May) to autumn (October–early November), and fertilized with compound fertilizer once (spring) or twice (spring and autumn) a year, at average annual rates of 70 kg N ha⁻¹ and 70 kg K ha⁻¹ [12]. Until our study, Hirata et al. [12] had conducted a vegetation investigation (2000–2003), including coverage, and the above grassland management had been conducted since then.

Meteorological data were obtained from the observation records at the Furue Meteorological Station (32°42′ N, 131°49′ E; approximately 7 km from the study site) [41]. Based on data from the meteorological station and previously observed data, the temperature was estimated using an estimated demand equation. In Japan, the temperature rises from March, reaches its maximum in July–August, and then declines until February. The period from June to July is called the “rainy season,” and is the period with the greatest amount of precipitation. The mean monthly air temperature and monthly rainfall at Kagamiyama between 2007 and 2009 were 1.9–24.6 °C and 9–638 mm, respectively, with annual means and totals of 13.3–14.1 °C and 1787–2699 mm, respectively (Figure 1). The temperature of early spring in 2008 was slightly lower than that in 2007 and 2009. In addition, there was less precipitation during the summer of 2009 compared with 2007 and 2008. This site experienced warm, humid summers (June–August) and cool, dry winters (December–February).

2.2. Sampling and Measurements

Sod samples (vegetation and soil) were collected to a depth of 100 mm using an iron tube (75 mm diameter) in June and October 2007 (2 times); May, July, August, and November 2008 (4 times); and June, July, August, October, and November 2009 (5 times), a total of 11 times in each slope aspect, and 3 replicates per plot. The sampling could not be conducted in the same month of each year due to the management conditions of the study site, weather, etc. The collected sod samples were passed through a sieve with a 2 mm mesh, and the soil that was attached to the residual substance roots on the sieve were carefully removed under running water for the collection of the plant roots. Concerning the plant roots that were included in a collected plant, only the above ground section was used, to prevent contamination with weed plant roots other than centipedegrass. These roots were subsampled for the AM colonization (internal hyphae, vesicles, and arbuscules) measurements, following the Giovannetti and Mosse [42] method. Briefly, plant roots were cleared in 10% KOH (w·v⁻¹), bleached in 1–2% HCl (v·v⁻¹), dyed in 0.05% trypan blue (w·v⁻¹), and then scored for the presence or absence of AM colonization under a microscope at ×400 magnification. To increase inter-rater reliability and within-sample reproducibility, observers were trained in advance, and measurements were repeated for each sample. The AM colonization level was calculated as AM colonization = number of intersections colonized (internal hyphae, vesicles, and arbuscules)/total number of intersections (200) examined. The above ground and residual roots of centipedegrass were dried at 70 °C for a minimum of 72 h for the dry matter weight (DMW) measurement. The root used for the AM colonization observation was converted into DMW based on a dry matter rate and was added to a root dry matter weight. In addition, the data from the above ground part of June 2007 were not collected.

Soil samples were collected in October 2007, May and November 2008, and November 2009 a total of 4 times in each slope aspect, with 3 replicates per plot. The soil samples that had passed through the sieve with the 2 mm mesh were analyzed for moisture, total N, total C (N/C analyser, Sumigraph NC−220F; Sumika Chemical Analysis Service, Tokyo, Japan), available soil P (Bray II method, Bray and Kurtz [43]), pH (H₂O, glass electrode method, D−54, Horiba Ltd., Kyoto, Japan), and electric conductivity (EC, platinum-titanium electrode method, D−54, Horiba Ltd., Kyoto, Japan). Soil moisture was calculated from the difference between the weight before and after drying of some soils dried in a draft dryer (105 °C) for 24 h. Soil porosity was measured using a core sampler (DIK-1601, Daiki Rika Kogyo Co., Ltd., Tokyo, Japan) to a depth of 100 mm and an actual volumenometer (DIK-1130, Daiki Rika Kogyo Co., Ltd., Tokyo, Japan) in November 2009.

2.3. Statistical Analysis

AM colonization (internal hyphae, vesicles, and arbuscules) data were converted into arcsine values, and the data for these, the soil chemical properties, and plant dry weights were analyzed using repeated-measure analysis of variance (ANOVA) considering the block effects (factors include slope aspect, survey date set to repeat). Effect size (partial-η²) and test power (1 − β) were also calculated. Significant ANOVA results were followed by Tukey’s HSD post hoc tests. The least significant difference between the mean values was used to identify statistical differences at the p < 0.05 level. Since the shoot DMW could not be surveyed accurately in June 2007 and data were missing, the data for the shoot DMW and total plant DMW were considered missing for June 2007, and statistical analysis was conducted using data from other survey dates. Seasonal variation was determined by the variation between each survey date from spring to fall of the year. Correlation analysis was used to evaluate the relationships between the total plant DMW and AM structural organ colonization, DMW and soil chemical properties, and AM structural organ colonization and soil chemical properties. For these correlation analyses, all data from the four soil surveys and corresponding DMW and AMF data from the four surveys were used (n = 16; 4 slope aspects × 4 times). Correlation coefficients (r) were deemed statistically significant at p < 0.05. The effect size was set as small; r > 0.10, medium; r > 0.30, large; r > 0.50 (Pearson’s correlation coefficient r) [44] and z-test p value was also calculated. All statistical analyses were performed using STATISTICA (version 10.0; StatSoft Inc., Tulsa, OK, USA).

3. Results

3.1. Soil Chemical Properties

The soil properties almost always varied (p < 0.05) between the slope aspects (

Table 2. Soil properties in centipedegrass swards on 4 slope aspects.

Year	Week †	East		West		South		North		Mean	SE
N (mg g−1 DM)
2007	32	3.26	b	2.13	c	2.47	c	5.73	a	3.40	0.81	A
2008	62	2.46	b	2.87	b	2.86	b	5.22	a	3.35	0.63	A
	85	1.86	a	2.43	a	2.10	a	2.07	a	2.11	0.12	B
2009	142	2.13	c	2.94	b	2.68	b	5.22	a	3.24	0.68	A
Mean		2.42	b	2.59	b	2.53	b	4.56	a	3.03	0.51
SE		0.30		0.19		0.16		0.84		0.31
C (mg g−1 DM)
2007	32	35.6	b	27.9	b	34.4	b	76.7	a	43.7	11.1	A
2008	62	26.8	c	41.1	b	38.9	b	65.8	a	43.1	8.2	A
	85	21.4	b	32.3	a	29.0	ab	27.7	ab	27.6	2.3	B
2009	142	23.9	c	36.8	b	36.8	b	67.1	a	41.2	9.2	A
Mean		26.9	b	34.5	b	34.8	b	59.3	a	38.9	7.0
SE		3.1		2.9		2.1		10.8		3.8
Available soil P (mg kg−1)
2007	32	11.37	b	14.06	a	1.63	c	9.59	b	9.16	2.67	A
2008	62	10.03	a	9.41	a	2.27	b	10.35	a	8.02	1.93	AB
	85	6.96	b	12.53	a	2.22	c	4.47	bc	6.54	2.22	AB
2009	142	5.70	a	7.35	a	1.87	b	6.59	a	5.38	1.22	B
Mean		8.52	ab	10.83	a	2.00	c	7.75	b	7.27	1.88
SE		1.32		1.51		0.15		1.36		0.83
EC (dS m−1)
2007	32	59.5	a	43.6	b	26.2	d	34.8	c	41.1	7.1	A
2008	62	24.1	c	33.3	b	28.8	bc	47.8	a	33.5	5.1	B
	85	16.1	b	24.6	a	24.2	a	30.2	a	23.8	2.9	C
2009	142	17.7	b	15.1	b	18.1	b	29.7	a	20.2	3.2	C
Mean		29.4	ab	29.2	ab	24.3	b	35.6	a	29.6	2.3
SE		10.2		6.1		2.3		4.2		4.7
pH
2007	32	4.66	b	4.73	b	5.13	a	4.51	b	4.76	0.13	B
2008	62	5.19	a	4.73	b	5.10	a	4.73	b	4.94	0.12	B
	85	5.45	ab	5.23	b	5.62	a	4.68	c	5.24	0.21	A
2009	142	5.13	a	4.85	b	5.21	a	4.66	b	4.96	0.13	B
Mean		5.11	ab	4.88	ab	5.26	a	4.64	b	4.97	0.13
SE		0.16		0.12		0.12		0.05		0.10
Moisture (%)
2007	32	31.1	b	29.2	b	18.5	c	53.0	a	33.0	7.2	BC
2008	62	30.9	b	32.0	b	35.2	b	51.3	a	37.4	4.7	AB
	85	35.6	c	41.1	b	32.8	c	53.7	a	40.8	4.6	A
2009	142	26.7	c	26.8	c	31.0	b	45.9	a	32.6	4.5	C
Mean		31.1	b	32.3	b	29.4	b	51.0	a	35.9	5.0
SE		1.8		3.1		3.7		1.8		2.0
Porosity (%)
2007	32	-		-		-		-		-	-
2008	62	-		-		-		-		-	-
	85	-		-		-		-		-	-
2009	142	63.7	a	66.3	a	68.4	a	75.1	a	68.4	2.5
Mean		-		-		-		-		-	-
SE		-		-		-		-		-	-

^† The first week is the one week in April, 2007. Mean values with different lower-case letters are significantly different (p < 0.05) between the aspects. Mean values with different upper-case letters are significantly different (p < 0.05) between each investigation time.

). For soil total N, between the slope aspects (p < 0.001, partial-η² = 0.912, 1 − β = 0.999), surveys (p < 0.001, partial-η² = 0.481, 1 − β = 0.983) and the slope aspects x surveys interactions (p < 0.01, partial-η² = 0.546, 1 − β = 0.954), for soil total C, between the slope aspects (p < 0.001, partial-η² = 0.903, 1 − β = 0.999), surveys (p < 0.001, partial-η² = 0.455, 1 − β = 0.971) and the slope aspects x surveys interactions (p < 0.01, partial-η² = 0.540, 1 − β = 0.950), for soil available P, between the slope aspects (p < 0.001, partial-η² = 0.911, 1 − β = 0.999) and surveys (p < 0.01, partial-η² = 0.376, 1 − β = 0.897), for soil EC, between the slope aspects (p < 0.01, partial-η² = 0.743, 1 − β = 0.941), surveys (p < 0.001, partial-η² = 0.845, 1 − β = 0.999) and the slope aspects x surveys interactions (p < 0.001, partial-η² = 0.819, 1 − β = 0.999), for soil pH between the slope aspects (p < 0.05, partial-η² = 0.589, 1 − β = 0.675) and surveys (p < 0.001, partial-η² = 0.600, 1 − β = 0.999), for soil moisture between the slope aspects (p < 0.001, partial-η² = 0.963, 1 − β = 0.999), surveys (p < 0.001, partial-η² = 0.534, 1 − β = 0.996) and the slope aspects x surveys interactions (p < 0.01, partial-η² = 0.528, 1 − β = 0.938). The soil total N, total C, EC, and moisture tended to be highest on the northern aspect, whereas the soil available P and pH tended to be highest on the western and southern aspects, respectively. Soil total N and C tended to be lowest during the third survey (85 week), while pH and moisture tended to be highest during the third survey. Soil available P and EC tended to be lowest during the fourth survey (142 week). There was no difference in soil porosity between the slope aspects, although only one investigation was conducted.

3.2. Dry Matter Weight of Centipedegrass

Centipedegrass grew well on all 4 aspects, each with similar coverage values (>50%). Major species other than centipedegrass were zoysia grass, tall fescue, Pennisetum alopecuroides, and lichens on the northern aspect; tall fescue, P. alopecuroides, and Cyperaceae species on the eastern aspect; and bahia grass and Cyperaceae species on the southern and western aspects. For shoot DMW of the centipedegrass, between the slope aspects (p < 0.05, partial-η² = 0.625, 1 − β = 0.745), surveys (p < 0.001, partial-η² = 0.413, 1 − β = 0.999) and the slope aspects x surveys interactions (p < 0.001, partial-η² = 0.484, 1 − β = 0.999), for root DMW, between the slope aspects (p < 0.01, partial-η² = 0.798, 1 − β = 0.986), surveys (p < 0.001, partial-η² = 0.859, 1 − β = 0.999) and the slope aspects x surveys interactions (p < 0.05, partial-η² = 0.386, 1 − β = 0.989), for whole plant DMW, between the slope aspects (p < 0.05, partial-η² = 0.672, 1 − β = 0.835), surveys (p < 0.001, partial-η² = 0.586, 1 − β = 0.999) and the slope aspects x surveys interactions (p < 0.001, partial-η² = 0.457, 1 − β = 0.998). Concerning the shoot DMW of the centipedegrass, the southern aspect was higher than the eastern aspect during the autumn of 2007 (p < 0.05). The western aspect was significantly higher than the eastern aspect on average over the survey period. For the root DMW of the centipedegrass, the western aspect was higher than the eastern aspect in the late spring of 2007 (p < 0.05), but for the other measurement occasions there were no significant differences (p > 0.05) between the slope aspects. The western aspect was significantly higher than eastern and northern aspects, and the southern aspect was significantly higher than the northern aspect for the average over the survey period (

Table 3. Dry matter weight of centipedegrass on 4 slope aspects.

Year	Week †	East		West		South		North		Mean	SE
Shoot (kg m–2)
2007	13	-		-		-		-		-
	32	0.34	b	0.60	ab	0.72	a	0.68	ab	0.59	0.09	A
2008	62	0.22	a	0.57	a	0.56	a	0.38	a	0.43	0.08	BC
	68	0.30	a	0.43	a	0.49	a	0.21	a	0.36	0.06	BC
	74	0.31	a	0.43	a	0.50	a	0.27	a	0.38	0.05	BC
	85	0.35	a	0.41	a	0.39	a	0.38	a	0.38	0.01	BC
2009	116	0.44	a	0.39	a	0.25	a	0.33	a	0.35	0.04	C
	123	0.34	a	0.65	a	0.46	a	0.49	a	0.48	0.06	AB
	130	0.33	a	0.49	a	0.46	a	0.29	a	0.39	0.05	BC
	136	0.38	a	0.57	a	0.42	a	0.45	a	0.46	0.04	ABC
	142	0.31	a	0.37	a	0.61	a	0.38	a	0.42	0.07	BC
Mean		0.33	b	0.49	a	0.49	ab	0.39	ab	0.42	0.04
SE		0.02		0.03		0.04		0.04		0.02
Root (kg m–2)
2007	13	0.37	b	0.63	a	0.54	ab	0.40	ab	0.48	0.06	A
	32	0.27	a	0.39	a	0.40	a	0.25	a	0.33	0.04	B
2008	62	0.27	a	0.34	a	0.31	a	0.30	a	0.30	0.01	B
	68	0.30	a	0.35	a	0.42	a	0.20	a	0.32	0.05	B
	74	0.39	a	0.40	a	0.41	a	0.25	a	0.36	0.04	B
	85	0.24	a	0.43	a	0.34	a	0.30	a	0.33	0.04	B
2009	116	0.19	a	0.19	a	0.10	a	0.07	a	0.14	0.03	C
	123	0.09	a	0.15	a	0.11	a	0.10	a	0.11	0.01	C
	130	0.10	a	0.12	a	0.10	a	0.06	a	0.10	0.01	C
	136	0.10	a	0.13	a	0.13	a	0.08	a	0.11	0.01	C
	142	0.11	a	0.08	a	0.12	a	0.07	a	0.10	0.01	C
Mean		0.22	bc	0.29	a	0.27	ab	0.19	c	0.24	0.02
SE		0.03		0.05		0.05		0.04		0.04
Total plant (kg m–2)
2007	13	-		-		-		-		-
	32	0.61	b	0.99	ab	1.12	a	0.93	ab	0.91	0.11	A
2008	62	0.49	a	0.90	a	0.87	a	0.68	a	0.73	0.09	B
	68	0.60	ab	0.77	ab	0.91	a	0.41	b	0.68	0.11	BC
	74	0.70	a	0.84	a	0.91	a	0.52	a	0.74	0.09	AB
	85	0.59	a	0.84	a	0.72	a	0.69	a	0.71	0.05	B
2009	116	0.63	a	0.58	a	0.34	a	0.41	a	0.49	0.07	D
	123	0.43	a	0.80	a	0.56	a	0.59	a	0.60	0.08	BCD
	130	0.44	a	0.61	a	0.56	a	0.35	a	0.49	0.06	D
	136	0.48	a	0.70	a	0.55	a	0.53	a	0.57	0.05	BCD
	142	0.41	a	0.45	a	0.73	a	0.45	a	0.51	0.07	CD
Mean		0.54	b	0.75	a	0.73	ab	0.55	ab	0.64	0.06
SE		0.03		0.05		0.07		0.06		0.04

^† The first week is the one week in April, 2007. Mean values with different lower-case letters are significantly different (p < 0.05) between the aspects. Mean values with different upper-case letters are significantly different (p < 0.05) between each investigation time.

). Concerning the whole plant DMW of the centipedegrass, the western aspect was higher than the eastern aspect in the autumn of 2007 (p < 0.05), and the southern aspect was higher than the northern aspect in early summer of 2008 (p < 0.05), but for the other measurement occasions, there were no significant differences (p > 0.05) between the slope aspects. The western aspect was significantly higher than the eastern aspect on average over the survey period. Annual changes were observed: there was a lesser root DMW in 2009 than that in 2007 and 2008 (p < 0.05), and there was slightly little tendency of total DMW in 2009 compared with that in 2007 and 2008.

3.3. AM Colonization of Centipedegrass

The AM colonization of centipedegrass, in any characteristic structures (internal hyphae, vesicles, and arbuscules), was not significantly different (internal hyphae, p = 0.758, partial-η² = 0.117, 1 − β = 0.102, vesicles, p = 0.311, partial-η² = 0.315, 1 − β = 0.253, and arbuscules, p = 0.164, partial-η² = 0.453, 1 − β = 0.371) between the slope aspects on most measurement occasions, although the colonization usually varied with the surveys (internal hyphae, p < 0.001, partial-η² = 0.840, 1 − β = 0.999, vesicles, p < 0.001, partial-η² = 0.779, 1 − β = 0.999, and arbuscules, p < 0.001, partial-η² = 0.923, 1 − β = 0.999) and the slope aspects x surveys interactions (internal hyphae, p < 0.001, partial-η² = 0.534, 1 − β = 0.999, vesicles, p < 0.01, partial-η² = 0.430, 1 − β = 0.999, and arbuscules, p < 0.001, partial-η² = 0.496, 1 − β = 0.997,

Table 4. Arbuscular mycorrhizal colonization (internal hyphae, arbuscules, vesicles) of centipedegrass on 4 slope aspects.

Year	Week †	East		West		South		North		Mean	SE
Internal hyphae
2007	13	0.97	a	0.95	a	0.94	a	0.89	a	0.94	0.02	A
	32	0.61	a	0.51	a	0.59	a	0.38	a	0.52	0.05	E
2008	62	0.48	a	0.61	a	0.68	a	0.73	a	0.62	0.05	D
	68	0.77	a	0.62	a	0.84	a	0.73	a	0.74	0.05	BC
	74	0.81	a	0.85	a	0.80	a	0.69	a	0.79	0.04	B
	85	0.90	a	0.94	a	0.84	a	0.88	a	0.89	0.02	A
2009	116	0.64	a	0.71	a	0.77	a	0.80	a	0.73	0.04	BC
	123	0.61	a	0.62	a	0.65	a	0.56	a	0.61	0.02	D
	130	0.63	a	0.66	a	0.75	a	0.67	a	0.68	0.03	CD
	136	0.81	a	0.75	a	0.82	a	0.80	a	0.80	0.02	B
	142	0.75	a	0.74	a	0.82	a	0.73	a	0.76	0.02	BC
Mean		0.73	a	0.72	a	0.77	a	0.71	a	0.73	0.01
SE		0.04		0.04		0.03		0.04		0.04
Arbuscules
2007	13	0.89	a	0.78	a	0.87	a	0.85	a	0.85	0.02	AB
	32	0.40	a	0.38	a	0.47	a	0.37	a	0.40	0.02	D
2008	62	0.37	b	0.52	ab	0.65	a	0.55	ab	0.52	0.06	C
	68	0.86	a	0.81	a	0.91	a	0.88	a	0.87	0.02	A
	74	0.92	a	0.88	a	0.92	a	0.88	a	0.90	0.01	A
	85	0.88	a	0.92	a	0.81	a	0.89	a	0.87	0.02	A
2009	116	0.74	a	0.75	a	0.82	a	0.86	a	0.79	0.03	B
	123	0.87	a	0.90	a	0.88	a	0.84	a	0.87	0.01	A
	130	0.85	a	0.83	a	0.87	a	0.86	a	0.85	0.01	AB
	136	0.92	a	0.85	a	0.88	a	0.92	a	0.89	0.02	A
	142	0.84	a	0.86	a	0.89	a	0.80	a	0.85	0.02	AB
Mean		0.78	a	0.77	a	0.81	a	0.79	a	0.79	0.01
SE		0.06		0.05		0.04		0.05		0.05
Vesicles
2007	13	0.46	a	0.56	a	0.40	a	0.32	a	0.44	0.05	C
	32	0.22	a	0.14	a	0.30	a	0.19	a	0.21	0.03	D
2008	62	0.15	a	0.37	a	0.27	a	0.30	a	0.27	0.05	D
	68	0.56	a	0.37	a	0.55	a	0.58	a	0.51	0.05	ABC
	74	0.37	a	0.47	a	0.45	a	0.35	a	0.41	0.03	C
	85	0.57	a	0.58	a	0.47	a	0.54	a	0.54	0.03	AB
2009	116	0.16	a	0.31	a	0.09	a	0.20	a	0.19	0.05	D
	123	0.23	a	0.32	a	0.27	a	0.27	a	0.27	0.02	D
	130	0.34	a	0.42	a	0.53	a	0.44	a	0.43	0.04	BC
	136	0.59	a	0.48	a	0.53	a	0.63	a	0.56	0.03	AB
	142	0.60	a	0.56	a	0.64	a	0.51	a	0.58	0.03	A
Mean		0.39	a	0.42	a	0.41	a	0.39	a	0.40	0.01
SE		0.05		0.04		0.05		0.05		0.04

^† The first week is the one week in April, 2007. Mean values with different lower-case letters are significantly different (p < 0.05) between the aspects. Mean values with different upper-case letters are significantly different (p < 0.05) between each investigation time.

). In 2008, the colonization rate of the internal hyphae, vesicles, and arbuscules increased with the progression of the seasons (p < 0.05). In 2009, the colonization rate showed the same tendency, although a reduction in the colonization rate of the internal hyphae during the summer was recognized (p < 0.05). Concerning the results of 2007, there was a different tendency from the results of 2008/2009, the reason for which we were not able to clarify. The colonization rate increased with the increase in the vegetation growth period, and both of these showed a pattern of decreasing late into the growth period.

3.4. Relationships Between the Total Plant DMW and AM Structural Organ Colonization, Total Plant DMW and Soil Chemical Properties, and AM Colonization of Centipedegrass and Soil Chemical Properties

There was no relationship between the total plant DMW of the centipedegrass and AM structural organ colonization (n = 16; internal hyphae, r = −0.321, p = 0.22; vesicles, r = −0.416, p = 0.11; arbuscules, r = −0.421, p = 0.11), and between the total plant DMW of the centipedegrass and soil chemical properties (n = 16; soil N, r = −0.034, p = 0.90; soil C, r = 0.092, p = 0.73; soil available P, r = −0.013, p = 0.96; soil EC, r = 0.253, p = 0.35; soil pH, r = −0.072, p = 0.79; soil moisture, r = −0.072, p = 0.79). The colonization of the arbuscules and vesicles of the centipedegrass had a negative correlation with the soil EC (n = 16, r = −0.652, p < 0.01, r = −0.656, p < 0.01, respectively, Figure 2b,c,

Table 5. Relationships between the arbuscular mycorrhizal colonization of centipedegrass and soil properties, and between the soil chemical properties, related to Figure 2 and Figure 3.

Relationship Item	Equation	n	r Value	p Value	z-Test p Value
IHC–EC	Total; IHC = −0.0056 × EC + 0.8635	16	−0.426	0.10	0.681
AC–EC	Total; AC = −0.0117 × EC + 1.0077	16	−0.652	0.01	0.805
VC–EC	Total; VC = −0.0095 × EC + 0.6817	16	−0.656	0.01	0.808
	South; VC = −0.0364 × EC + 1.3042	4	−0.971	0.05	0.999
	West; VC = −0.0156 × EC + 0.8687	4	−0.929	0.10	0.994
IHC–pH	Total; IHC = 0.2455 × pH − 0.5236	16	0.495	0.05	0.284
AC–pH	Total; AC = 0.2958 × pH − 0.8103	16	0.436	0.09	0.314
AC–AP	Total; AC = −0.0233 × AP + 0.8301	16	−0.425	0.10	0.681
	East; AC = −0.0975 × AP + 1.4534	4	−0.937	0.10	0.996
SC–SN	Total; SC = 13.09 × SN − 0.072	16	0.984	0.001	0.001
	East; SC = 10.20 × SN − 0.221	4	0.998	0.01	0.001
	West; SC = 13.78 × SN − 0.118	4	0.924	0.10	0.008
	South; SC = 12.97 × SN − 0.198	4	0.998	0.01	0.001
	North; SC = 12.85 × SN + 0.075	4	0.996	0.01	0.001
pH–SN	Total; pH = −1.636 × SN + 5.469	16	−0.626	0.01	0.789
	East; pH = −5.115 × SN + 6.347	4	−0.951	0.05	0.999
SM–SN	Total; SM = 0.4962 × SN + 0.2092	16	0.597	0.05	0.228
pH–SC	Total; pH = −0.120 × SC + 5.441	16	−0.611	0.05	0.780
	East; pH = −0.503 × SC + 6.463	4	−0.956	0.05	0.999
SM–SC	Total; SM = 0.038 × SC + 0.213	16	0.603	0.05	0.225
EC–AP	Total; EC = 1.605 × AP + 17.95	16	0.527	0.05	0.268
pH–EC	Total; pH = −0.016 × EC + 5.456	16	−0.614	0.05	0.782
	East; pH = −0.015 × EC + 5.542	4	−0.927	0.10	0.993

IHC: internal hyphae colonization, AC: arbuscules colonization, VC: vesicles colonization, EC: soil electric conductivity, AP: available soil P, SC: soil carbon, SN: soil nitrogen, SM: soil moisture.

). The colonization of the internal hyphae had a negative correlation tendency with the soil EC (r = −0.426, p = 0.10, Figure 2a) and a positive correlation tendency with the soil pH (r = 0.495, p = 0.05, Figure 2d). The colonization of the arbuscules had a negative correlation tendency with the available soil P (r = −0.425, p = 0.10, Figure 2f) and a positive correlation tendency with the soil pH (r = 0.436, p = 0.09, Figure 2e). Between the soil chemical properties (n = 16), the soil C had a positive correlation with the soil N (r = 0.984, p < 0.001, Figure 3a, Table 5), the soil pH had a negative correlation with the soil N (r = −0.626, p < 0.01, Figure 3b), C (r = −0.611, p < 0.05, Figure 3d), and EC (r = −0.614, p < 0.05, Figure 3g), the soil moisture had a positive correlation with the soil N (r = 0.597, p < 0.05, Figure 3c) and C (r = 0.603, p < 0.05, Figure 3e), and the soil EC had a positive correlation with the soil available P (r = 0.527, p < 0.05, Figure 3f).

4. Discussion

4.1. Slope Aspects and Soil Chemical Properties

Slope aspect has a considerable influence on the edaphic and biotic characteristics of a hilly pasture, mainly through the variation in the soil temperature caused by differences in the solar radiation [1,2,5]. In fact, soil properties almost always varied between the slope aspects (Table 2). As with previous reports [8,9], Liu et al. [8] reported that the soil organic C and moisture content were significantly higher on shady slopes than on sunny slopes, whereas the soil pH was significantly higher on sunny slopes than on shady slopes, in arid ecosystems of the Daqingshan Mountains, Inner Mongolia. Chai et al. [9] reported that the soil moisture, organic C, and total N were significantly higher on the northwest-facing slope than on the southeast-facing slope, whereas the light intensity, soil temperature, and soil pH were significantly higher on the southeast-facing slope than on the northwest-facing slope in an alpine ecosystem. In the present study, the highest soil moisture on the northern aspect is attributed to the lowest incident of solar radiation on this shady aspect in the Northern Hemisphere. Furthermore, the soil moisture had a positive correlation with the soil N and C, and the soil pH had a negative correlation with the soil N, C, and EC (Figure 3). The highest soil total N and C on the northern aspect is also attributed to the lowest incident of soil temperature and the soil organic matter decomposition of soil microorganisms [4], resulting in the lowest soil pH. Soil moisture content was higher and soil N and C contents were lower for the 85-week soil compared to the other survey dates. Since soil moisture content is affected by precipitation immediately prior to the survey, it was considered to be higher due to precipitation prior to the survey. Soil N and C content can also be affected by grassland management, so it is likely that fertilizer application and pasture litter from winter to the following spring increased values in the 142-week survey.

4.2. Slope Aspects and Dry Matter Weight of Centipedegrass

Generally, the slope aspect has an influence on the productivity of the herbage [2,9]. Lambert and Roberts [2] reported that east-facing slopes have a higher pasture productivity than the others in a temperate-type plant community in New Zealand. In the northern hemisphere, vegetation coverage is generally greater on the more shaded north-facing slopes than the drier, sunlit south-facing slopes [7,45]. Chai et al. [9] reported that aboveground net primary productivity was significantly higher on the northwest-facing slope than on the southeast-facing slope in an alpine plant ecosystem. Liu et al. [8] also reported that plant coverage was affected by the slope aspects, being significantly higher on shady slopes than on sunny slopes in the arid temperate plant vegetation ecosystems of the Daqingshan Mountains, Inner Mongolia. On the other hand, Chu et al. [7] reported that herbaceous and shrubby biomass was not affected by slope aspects in a Boreal Forest in China. Despite the aspect differences in the microclimate and soil properties, centipedegrass grew well on all aspects, but varied between the measurement occasions. Because the coverage of the centipedegrass was not significantly different between the four aspects after the fourth year of establishment [12], it was considered that the coverage was maintained after establishment until this investigation period of 8–10 years.

In Japan, warm-season pasture grasses begin to grow in the spring around March, when temperatures are rising, with maximum production during the warmer months of July and August, followed by a decline in production and almost no growth during the winter months of December and February. In 2008 and 2009, there was little seasonal variation in DMW, as described above. In 2009, the June-September root DMW remained lower than in 2008. It was considered that a decrease in the precipitation during the summer influenced this (Figure 1). It is thought that the difference in response of the vegetation studied previously [2,7,8,9,45] depends on the difference in vegetation types (cool-season or temperate type vs. warm-season or tropical type) and climate (semi-arid, cool summer climate or semi humid alpine climate vs. warm, humid summers and cool, dry winters). Since centipedegrass is adapted to a variety of environments [12,13,14], it was thought that the dry matter weight would not vary different slope aspects due to its cold tolerance and capacity to thrive in various soil types.

4.3. Slope Aspects and AM Colonization of Centipedegrass

According to previous reports [8,9], AM colonizations are affected by the slope aspect. However, in the present study, the AM colonization of centipedegrass, in any characteristic structures (internal hyphae, vesicles, and arbuscules), was not significantly different between the slope aspects on most measurement occasions, although the colonization usually varied between the seasons and years (Table 4). Chai et al. [9] reported that AM colonization variables (percentage of root length colonized by AM fungi, percentage of root length colonized by vesicles, and percentage of root length colonized by hyphae) were significantly higher on the southeast-facing slope than on the northwest-facing slope in an alpine plant ecosystem. Liu et al. [8] also reported that AM colonization variables (total root colonization, arbuscule abundance, vesicle abundance, and hyphal colonization), spore density, and AM fungus diversity were significantly higher on sunny slopes than on shady slopes in the arid temperate plant vegetation ecosystems of the Daqingshan Mountains, Inner Mongolia. It was thought that sunshine influences the variation in light intensity, solar radiation, soil temperature, and soil moisture, and, as a result, these studies identified the soil chemical properties it affects [8,9]. On the other hand, Chu et al. [7] reported that the AMF community was not affected by slope aspects in a Boreal Forest in China, though slope aspects affect chemical properties. The insusceptibility of AM colonization to the effects of the slope aspects is thus interpreted as a result of the vigorous growth of host plant, centipedegrass on all aspects. Previous reports [8,9] have shown that solar radiation varies with slope aspects, which in turn affects air temperature, soil temperature, and soil properties, which in turn affects mycorrhizal formation. Because the study site was a private pasture and regularly grazed by cattle, it was not possible to install solar radiation or air temperature measurement equipment; therefore, the relationship between these factors could not be examined in detail.

Furthermore, there was seasonal variation in AM colonization. In the present study, the colonization rate of the internal hyphae, vesicles, and arbuscules in 2008 increased with the progression of the seasons. In 2009, the colonization rate showed the same tendency, although a reduction in the colonization rate of the internal hyphae in the summer was recognized. It was considered that a decrease in the precipitation during the summer influenced this (Figure 1). In previous reports, it was found that the mycorrhiza colonization rate rises from the spring to early fall, and it decreases to winter [36,37,38]. Thus, the present results from 2008 and 2009 show the same change with four slope aspects as these reports [36,37,38]. The colonization rate increases with the increase in the vegetation growth rate, and both of these show a pattern of decreasing late into the growth, which is in accordance with the report of Wearn and Gange [37]. As described before, AM colonization seasonality is dependent upon the host plant, field conditions [35,38], carbon sinks in mycorrhizal-independent plants [35,37], and host plants demand for nutrients [27,38]. From the results of 2008 and 2009, it was thought that the phenomenon of seasonal variation in AM was same as the above reports. Concerning the results of 2007, there was a different tendency from the results of 2008/2009, the reason for which we were not able to clarify.

4.4. Relationships Between the Plant DMW and AM Colonization, Plant DMW and Soil Chemical Properties, and AM Colonization of Centipedegrass and Soil Chemical Properties

In a previous study [46], dry matter yield of Zoysia grass showed a positive correlation with the colonization of the arbuscules and vesicles, available phosphorus, and soil pH and a negative correlation with the soil EC in flat grassland. This was the result that caused a variation in yield by the effect of soil chemical properties and AM colonization with fertilizer application [46]. In the present study, there were no relationships between the total plant DMW of the centipedegrass and AM structural organ colonization, and between the total plant DMW of the centipedegrass and soil chemical properties. As mentioned above, despite the differences in aspect in the soil chemical properties, centipedegrass grew well with the same AM colonization on all aspects. In addition, seasonal variation as well as AM colonization was not observed in DMW whereas seasonal variation was found in AM colonization; this might be why no correlation was observed between DMW and AM colonization of the centipedegrass. However, AM colonization was shown to be influenced by some soil chemical properties.

In previous reports, AM colonization showed a correlation with the soil moisture (negative correlation [36], positive correlation [34,47]), pH [32,39,46,48,49], organic C [8,47], EC [34,46,49], and available phosphorus [8,46,50]. In the present study, the colonization of the arbuscules and vesicles of centipedegrass had a negative correlation, and the colonization of the internal hyphae of centipedegrass had a negative correlation tendency with the soil EC. Soil EC is useful as a relative measure of the total quantity of ions in a soil solution [51] and the amounts of salts in soil (salinity of soil) [49,52]. Soil EC has no direct effect on crop growth or yield, but close relationships are frequently observed between EC and a variety of other soil properties that are closely related to plant growth and yield [53]. Posada et al. [34] reported that small increases in EC in soil were related to arbuscule production. In a previous report [46], the colonization of the arbuscules and vesicles showed a negative correlation with the soil EC, and it was revealed that an increase in the soil EC reduced the AM colonization. Similarly, AM colonization was negatively correlated with soil EC in soil low pH growing rice plant in the Mekong Delta, Vietnam [49], in coastal saline soil of China [54] and in Phaltan tehsil of Satara District, Maharashtra, India [55], and in Songnen saline–alkali grassland [56]. An increase in the soil EC means an increase in total quantity of ions in the soil solution, in some cases, soil salinity, and the symbiotic relationship between the plant and AMF becomes weak, it is because the higher EC can delay AMF spore germination and inhibit the elongation of AMF hyphae and the colonization of plant roots [56]. The soil EC had a positive correlation with the available soil P, and soil EC had a negative correlation with the soil pH, similar to previous reports [46,54]; this change was dependent on the increase in nutrients in the soil with the passage of time after the application of fertilizer [46], or the soil salinity levels [54].

Soil pH also affects AM colonization [39], as the soil pH affects the growth of the fungal mycelium directly [39,48], and AM colonization is high when soil pH is near neutral [32]. The colonization of the internal hyphae and arbuscules of centipedegrass had a positive correlation tendency (effect size medium) with soil pH in the present study (Figure 2d,e), the colonization of internal hyphae and arbuscules in mycorrhizae generally increases as the pH level nears neutrality, same as that in previous reports [32,39,48,49]. Soil properties differed between slope aspects, but there were no differences in mycorrhizal colonization between slope aspects, indicating that these correlations were only weakly related. In the present study, the effect size was insufficient to clarify the relationship between soil properties and AM colonization, and further investigation and study with a larger amount of data are needed.

Generally, a high concentration of available soil P suppresses AM colonization [26,27,28,50]. The colonization of the arbuscules had a negative correlation tendency with the available soil P in the present study, similar to that in previous reports [8,26]. When associated with plant roots, AMF can access sources of phosphorus and other trace minerals unavailable to the plant, and in return, the plant shuttles carbon to the AMF [27]. Under low light intensity conditions, plant photosynthesis is restricted. Allocation of a limited supply of photosynthate to AMF might reduce plant allocation to functions related to its fitness [57]. Liu et al. [8] discussed how in poor water and nutrient conditions, such as sunny slopes, plants may be most dependent on AMF to meet their demands for growth, leading to a stronger association to overcome the harsh conditions. However, in the relatively rich water and nutrient conditions of shady slopes, plants may take up sufficient mineral nutrients from the soil without the help of AMF, leading to a gradual reduction in the dependency of plants on AMF [8]. Additionally, on the shady slopes, low light availability may decrease photosynthesis and carbon translocation to roots, thereby limiting the carbon availability to AMF. In this way, the slope aspect abiotically influences the physical state of the host plant and, as a result, has an influence on the symbiotic relationship between the host plant and mycorrhizal fungi. As above, despite the slope aspect differences in the soil properties, there was no influence of the slope aspects in DMW and AM colonization of centipedegrass. It was thought that this depended on the adaptation abilities of the centipedegrass, adapting to a range of environments, showing superior cold tolerance compared to most warm-season grasses, and can grow on a range of soils [12,13,14]. However, same as previous reports, AM colonization was shown to be influenced by some soil chemical properties. In regard to the environment and the growth of the centipedegrass, and the growth and AM colonization of the centipedegrass, a more detailed study and longer-term field study will be necessary in the future.

5. Conclusions

The slope aspect affected the soil chemical properties; howewer, it did not affect the growth and arbuscular mycorrhizal colonization of centipedegrass in a hill pasture. This may be a factor contributing to the high adaptability of the grass to all slope aspects, under an environmental condition of southern Kyushu, Japan. The insusceptibility of AM colonization to the effects of the slope aspects is thus interpreted as a result of the vigorous growth of host plant, centipedegrass on all aspects. Because there is no difference in the growth of centipedegrass between the slope aspects, it is easy for management of the centipedegrass in a hill pasture and low-input pasture management will be enabled. In addition, seasonal variation was shown in the AM colonization, and increased with the progression of the seasons, with four slope aspects. There were no relationships between the DMW of the centipedegrass and AM characteristic structures colonization, and soil chemical properties. However, the colonization of the arbuscules and vesicles of the centipedegrass showed a correlation with some soil chemical properties, and it was shown that the colonization was dependent upon the soil chemical properties. Thus, centipedegrass symbiosis with AMF may be of potential value to low-input grassland systems and environmental conservation in suitable regions, regardless of slope aspect. However, it is possible that background AM colonization levels in these climatic and soil conditions are not necessarily linked to the centipedegrass yield. Further studies are warranted to examine and clarify the relationships between the growth of centipedegrass and the AMF under a controlled environmental condition.