Growth and Morphological Responses of Kentucky Bluegrass to Homogeneous and Heterogeneous Soil Water Availabilities

Abstract

Understanding the effect of water heterogeneity on the growth and water integration of clonal plants is important for scientific water management. In controlled field plots, we conducted a study by creating three different soil water treatments comprising four patches with different soil water supplies using Kentucky bluegrass (Poa pratensis L.) as the materials. The objective was to explore the differences in the growth indices and morphological attributes, and the possible water integration in ‘Arcadia’ Kentucky bluegrass in plots and patches with different soil water availabilities. Soil water deficit resulted in the declined biomass, root/shoot ratio and leaf relative water content of the whole genet, decreased leaf length and height of mother ramet, reduced leaf width and height of daughter ramets, and significant changes in the structures of vascular bundles in rhizomes and leaves. However, the leaf length, leaf width, height and leaf relative water content of daughter ramets in heterogeneous water-poor patches were recovered to the levels in water-rich patches. In addition, the diameter of the vessel in the xylem and percentage of the xylem area in leaf vascular bundles of daughter ramets in the heterogeneous water-poor patch were insignificantly different from those in water-rich patches. These results demonstrated that water integration among interconnected ramets promoted the growth status of daughter ramets in heterogeneous water-poor patches. However, the water translocation in interconnected ramets subjecting to heterogeneous water supplies was not investigated using stable isotope labelling in this study. Thus, the directions and patterns of water translocation among ramets necessitate further research.

1. Introduction

Plants are exposed to diverse abiotic stresses that may severely influence their growth and development. Among them, drought is the most serious one that strongly affects plant productivity and crop yields [1]. Water in plants plays a key role in various physiological processes, including many aspects of plant growth, development and metabolism [2]. Drought inhibits the photosynthetic rate of plants, which leads to a reduction in the resources available for investment in reproduction [3,4]. Global climate changes, including global warming, rainfall anomalies and shifts in monsoon patterns, are projected to increase the frequency, duration and severity of drought [5,6]. Plants have evolved different adaptive mechanisms to cope with the adverse effects of drought. Stress avoidance, escape and tolerance are the three main survival strategies utilized by plants when exposed to drought [7]. Among these, it is deemed that changes in the root system (root size, length, density, proliferation, expansion, growth rate, etc.) represent the main approach for drought-tolerant plants to cope with water deficiency [1,8].

The availability of resources, such as water, nutrients and light usually exhibits obviously spatial and temporal heterogeneity in natural habitats, even among different micro-habitats [9,10,11]. Clonal plants exhibit some adaptive advantages in residing in habitats with comparatively dynamic and often extreme resource availability [12]. They produce numerous genetically identical ramets, which were physically connected by rhizomes or stolons [11]. These clonal ramets often experience heterogeneous resource availability. Water, nutrients and carbohydrates were found to be transported and shared among interconnected ramets of clonal plants, which was named clonal integration [13,14]. Clonal integration helped clonal plants exploit ubiquitous heterogeneous resources, improve the growth performance of the whole genet and occupy new habitats [15,16,17]. Extensive studies have been reported on the role of clonal integration in coping with heterogeneous resources, such as water, nutrients, light, space and others for clonal plants [17,18,19,20,21].

Kentucky bluegrass (Poa pratensis L.) belongs to Gramineae, which is the most common perennial cool-season grass and probably the best-known one [22]. It makes beautiful, lush and finely textured lawns for homes and sports. Kentucky bluegrass spreads by rhizomes and ramets, and forms a high-quality and dense sod under desired water and nutrient conditions. Kentucky bluegrass develops a shallow root system that is not drought tolerant and will go dormant during extreme conditions. However, it will recover if given intermittent watering during prolonged drought conditions.

Grass species evolved the best root system to cope with nature’s stresses and assure the longevity of each species. Water and nutrients are the two main factors influencing the growth and ornamental characteristics of turfgrass [23]. Especially, in the early growing stages of turfgrass, the amount of available water and nutrients determines the success or failure of turf establishment, as well as the quality of turf [22]. Moreover, the availability of nutrients depends on water availability, and the heterogeneity of water supply results in changes in concentrations and ratios of nutrients [24,25]. The use of best management practices related to plant population and genotype, and management of soil, water and nutrients were effective measures to alleviate the adverse effects of drought [1,26].

In both laboratory and field trials, the clonal integration and morphological plasticity of clonal plants have been extensively investigated. Most studies used pairs of mother ramet and daughter ramets subjected to homogeneous and heterogeneous water and nutrient treatments as materials. The translocation and sharing patterns of water and nutrients, as well as the photochemical activities and growth performances of mother and daughter ramets, were well documented [17,20,21,27,28]. However, only a few studies were conducted on plot and patch levels to investigate the growth, phenotypic responses and recovery mechanisms of clonal plants to scale-dependent nutrient heterogeneity [22,29,30]. Moreover, research on the effect of heterogeneous water availability on the clonal growth and morphological plasticity of the whole genet of clonal plants growing in experimental plots remains limited.

To understand how clonal integration influences the growth and morphological performances of clonal plants in both heterogeneous and homogeneous habitats, this study investigated the growth patterns of Kentucky bluegrass under different soil water availabilities. The effects of homogeneous and heterogeneous soil water on the growth indices (leaf length, leaf width, height and intercode length of mother and daughter ramets as well as the above-ground biomass, under-ground biomass, root/shoot ratio and leaf relative water content of the whole genet) and the morphological structures of rhizomes and leaves were also studied. The objectives of this study included: (1) To observe the impact of water integration in heterogeneous water treatment on the growth of Kentucky bluegrass daughter ramets in water-poor patches. (2) To examine the influence of clonal integration on the leaf relative water content and biomass allocation to roots in the genet under heterogeneous water treatment. (3) To investigate the effect of heterogeneous water treatment on the anatomical structures of rhizomes and leaves in water-poor daughter ramets, which facilitate the transport and sharing of water among interconnected ramets. The present study provides some useful information on the water integration and adaptation mechanism of clonal plants to heterogeneous soil water availability.

2. Materials and Methods

2.1. Experimental Design

To investigate the effect of homogeneous and heterogeneous soil water supplies on the growth and anatomical structures of Kentucky bluegrass ramets, a field experiment consisting of a total of three water treatments with four replications was designed. At first, 12 plots of the same size and depth were prepared. Each plot was divided into 2 patches of the same size with a polyvinyl chloride plate and filled with nutrient soils. Then, a mother ramet of Kentucky bluegrass was carefully and precisely transplanted into the center of each plot to make sure it was evenly distributed in the left and right patches and allowed to spread within the plots. Three different water treatments were created by applying irrigation at different frequencies (Figure 1). In the homogeneous water-rich treatment (NS+NS+), saturated irrigation was given to both patches once every two days based on the measurement of field capacity. In the homogeneous water-poor treatment (NS−NS−), saturated irrigation was given to both patches once every week. In the heterogeneous water treatment (NS+NS−), left and right patches were respectively watered at the designed frequency. Thus, four different kinds of patches, i.e., homogeneous NS+, homogeneous NS−, heterogeneous NS+ and heterogeneous NS− were formed surrounding Kentucky bluegrass mother ramets. Three water treatments were conducted in a precise way to study the changes in morphological growth and in the anatomical structure of Kentucky bluegrass ramets induced by these treatments.

2.2. Plant Materials

Based on the report of the National Turfgrass Evaluation Program (NTEP) and our previous studies on turfgrass drought resistance, a drought-tolerant Kentucky bluegrass cultivar ‘Arcadia’, provided by Dr. Shaun Bushman from USDA-FRRL, was used in the current study. In the 1995 NTEP trials for Kentucky bluegrass, ‘Arcadia’ ranked 20th in overall turfgrass quality among 103 cultivars [31]. ‘Arcadia’ is a turf-type cultivar displaying a dark-green genetic color, good summer density, good resistance to stem rust, crown rust and chinch bug, and moderate resistance to leaf spot [32].

Plant materials were vegetatively propagated from a single parent plant of ‘Arcadia’ Kentucky bluegrass in a greenhouse (Northeast Agricultural University, Harbin, Heilongjiang, China). During the cultivation period, the day temperature and night temperature were respectively maintained at 25 ± 2 °C and 15 ± 2 °C, the relative humidity was 50 ± 10%, and the supplemental light was provided by metal halide lamps at a 16-h photoperiod. After 4 weeks, ‘Arcadia’ plants with uniform size and growth status were selected and used in the subsequent experiments.

2.3. Experimental Conditions

The field experiments were conducted in an open area with natural light in the Experimental Station for Horticulture at Northeast Agricultural University. Twelve plots for homogeneous and heterogeneous water treatments were created in a sunny flat land by removing the original soils. The length and width of each plot were 1.4 m, respectively, and the depth was 0.5 m. The distance between the adjacent plots was 0.2 m. The bottom and four sides of each plot were covered and sealed with a large piece of plastic cloth to ensure that water did not leach into adjacent plots. Each plot was divided into two patches of the same size with a polyvinyl chloride plate (thickness 15 mm) to prevent the penetration of water between adjacent patches. The plots were then filled with nutrient soils (NS), and the total nitrogen content of which was 1.5–2.0% (Heilongjiang Futian Corporation, Harbin, Heilongjiang, China). Two different habitats were designed based on the watering frequency, i.e., water-rich nutrient soil (NS+, saturated irrigation once every two days) and water-poor nutrient soil (NS−, saturated irrigation once every week).

On 8 June 2019, one mother ramet of ‘Arcadia’ Kentucky bluegrass was transplanted into the center of each plot, which was a little higher than the surrounding ground. Thus, the rhizomes derived from mother ramet could spread into different patches within the plot. Three different water treatments were formed for ‘Arcadia’ (Figure 1): homogeneous water-rich treatment (NS+NS+), homogeneous water-poor treatment (NS−NS−) and heterogeneous water treatment (NS+NS−). After a rejuvenation period of one week, the experiments started on 15 June 2019 and lasted 14 weeks. During the experiments, ammonium nitrate fertilizer was applied to the plots once every month (2.6 g·m⁻²). The plots were covered with a big polytunnel when it was raining to prevent the entering of rainwater. The experiments ended on 21 September 2019, and the whole genets in different plots were respectively harvested. Experiments were conducted in a randomized complete block design, and each treatment was replicated four times.

2.4. Measurements of Growth and Physiological Indices

The total number of ramets was counted once every two weeks. At the end of the 14th week, the leaf length, leaf width and heights of both mother ramet and daughter ramets as well as the internode length between adjacent daughter ramets were measured. The above-ground and under-ground parts of the whole genet were separately harvested and weighed for three different water treatments. The weighed samples were first dried at 105 °C for 10 min and then at 80 °C to constant weight. The root/shoot ratio = under-ground dry weight/above-ground dry weight. The leaf relative water content (RWC) was determined according to a previous study based on fresh weight (FW), turgid weight (TW) and dry weight (DW) [33]. RWC (%) = [(FW − DW)/(TW − DW)] × 100.

2.5. Scanning Electron Microscope Observation on Rhizomes and Leaves

For observing the crosscutting structures of rhizomes and leaves of Kentucky bluegrass, four segments of rhizome (3 mm long) were cut from the first daughter ramet, and four segments of the leaf (3 mm long) were cut from the middle of the second leaf from the top of the first daughter ramet in each patch [22]. The samples were prepared with the method of Adame-González [34] for scanning electron microscope (SEM) observation. Segments of rhizomes and leaves were fixed in formaldehyde-acetic acid-ethanol fixative and dehydrated in an ethanol series (30% to 100%). Samples were then critical point dried through carbon dioxide, mounted on aluminum stubs with carbon tape, and sputter-coated with a 20 mA thick gold layer for two minutes. The crosscutting structures of rhizomes and leaves of daughter ramets in different patches were visually examined and photographed with a SEM (S-4800, Hitachi, Japan) at 15 kW.

2.6. Data Analysis

One-way analysis of variance (ANOVA) followed by Duncan’s multiple range tests (DMRT) was used to test the differences in the leaf length, leaf width, ramet height, length of internode, above-ground biomass, under-ground biomass, root/shoot ratio and leaf RWC of the whole genets as well as the diameter of vascular bundles (VBs), the pore diameter of central parenchyma (PT) in the rhizome, the diameter of the vessel in xylem, percentage of xylem area in VBs and diameter of sieve element in the phloem of daughter ramets in different patches under three water treatments. Statistical analyses were conducted with SPSS 18.0 (SPSS, Chicago, IL, USA).

3. Results

3.1. Clonal Growth Patterns of Kentucky Bluegrass

After plantation, the growth status of Kentucky bluegrass clonal ramets was investigated once every two weeks. The growth directions of daughter ramets are random for all three different water treatments. In NS+NS+ treatment, the number of daughter ramets increased in an incremental quadric-curve mode (R² = 0.9959) with the prolonging plantation time (

Table 1. Dynamic models for clonal growth of Kentucky bluegrass under different water treatments.

Treatment	Simulation Equation	Coefficient of Determination (R2)
NS+NS+	y = 0.0107x2 + 0.1399x + 20.185	0.9959
NS+NS−	y = 0.0041x2 + 0.1770x + 12.425	0.9867
NS−NS−	y = 0.0012x2 + 0.1000x + 2.917	0.9933

x: the days after plantation; y: the number of daughter ramets.

). After 10 weeks, the proliferation rate of daughter ramets notably increased, and there were 120.3 daughter ramets of Kentucky bluegrass in the 14th week (Figure 2).

Under NS−NS− treatment, the proliferation growth of Kentucky bluegrass also showed an incremental quadric-curve trend with the extension of time (R² = 0.9933). However, the proliferation rate was obviously lower than compared to NS+NS+ treatment. In addition, the numbers of daughter ramets under NS−NS− treatment were all significantly fewer than those under NS+NS+ treatment at different investigation periods. After 14 weeks, the number of daughter ramets was 21.5, accounting for 17.9% of that under NS+NS+ treatment (Figure 2). Under NS+NS− treatment, the number of daughter ramets also presented a trend of incremental quadric-curve (R² = 0.9867). The numbers of daughter ramets were between those of NS+NS+ and NS−NS− treatments at different plantation periods. After 10 weeks, the number of daughter ramets increased rapidly, which was 63 at the 14th week.

3.2. Growth Indices of Kentucky Bluegrass Clonal Ramets

The growth indices of Kentucky bluegrass mother and daughter ramets under different water treatments were summarized in

Table 2. Growth indices of Kentucky bluegrass under different water treatments.

Ramet	Treatment	Leaf Length (cm)	Leaf Width (mm)	Ramet Height (cm)	Length of Internode (cm)
Mother ramet	NS+NS+	17.97 ± 1.22 a	4.37 ± 0.20 a	15.29 ± 1.09 a	-
	NS+NS−	14.41 ± 0.95 ab	4.29 ± 0.16 a	14.75 ± 0.73 a	-
	NS−NS−	13.38 ± 1.47 b	3.85 ± 0.17 a	11.58 ± 0.68 b	-
Daughter ramet	Homogeneous NS+	13.33 ± 0.83 a	4.19 ± 0.15 a	7.01 ± 0.69 a	2.73 ± 0.51 b
	Heterogeneous NS+	11.44 ± 0.61 ab	4.14 ± 0.11 a	5.27 ± 0.58 ab	3.00 ± 0.32 b
	Heterogeneous NS−	11.21 ± 0.71 ab	3.97 ± 0.29 a	6.17 ± 0.93 a	3.12 ± 0.37 ab
	Homogeneous NS−	9.68 ± 0.31 b	3.29 ± 0.07 b	3.52 ± 0.52 b	4.51 ± 0.26 a

Different lowercase letters within the same column indicate significant differences at 0.05 level.

. Under NS+NS+ treatment, the growth status of the mother ramet was the best, with the longest leaf length and the tallest ramet height. The growth of the mother ramet was restrained under NS−NS− treatment, and the leaf length and ramet height were significantly shorter (p < 0.05). Under NS+NS− treatment, the leaf length and ramet height were insignificantly different from those under NS+NS+ treatment. However, the ramet height was remarkably taller than that under NS−NS− treatment. Furthermore, the leaf width was insignificantly different among the three water treatments (Table 2).

In the heterogeneous NS+ and NS− patches, the differences in the same growth indices (leaf length, leaf width, ramet height and length of internode) of daughter ramets were insignificant, which were also insignificantly different from those of homogeneous NS+ patch (Table 2). As compared to the homogeneous NS− patch, the daughter ramet leaf widths of heterogeneous NS+ and NS− patches were 18.2% and 15.8% wider, respectively. Moreover, the height of the daughter ramet in the heterogeneous NS− patch was notably taller than that in the homogeneous NS− patch (p < 0.05).

The distance between adjacent daughter ramets (length of internode) in the homogeneous NS− patch was the longest among four different patches, which was significantly different from those in the homogeneous NS+ patch and heterogeneous NS+ patch (p < 0.05). In addition, the lengths of internode are insignificantly different between heterogeneous NS+ and NS− patches (Table 2).

As can be seen in Figure 3A,B, the above-ground fresh weights, dry weights and the under-ground fresh weights, dry weights of Kentucky bluegrass genet under NS+NS+ treatment were the highest, which were notably different from those of NS+NS− and NS−NS− treatments. The above-ground and under-ground biomasses of NS+NS− treatment were significantly higher than those of NS−NS− treatment except for the above-ground fresh weight. Moreover, the root/shoot ratio of Kentucky bluegrass under NS+NS− treatment was remarkably higher than that of NS−NS− treatment, but insignificantly different from that of NS+NS+ treatment (Figure 3C).

3.3. Leaf RWCs of Clonal Ramets

The RWCs of Kentucky bluegrass mother ramet and daughter ramets in the left patch and right patch were insignificantly different for all three different water treatments (Figure 3D). Under NS−NS− treatment, the leaf RWCs of mother ramet and daughter ramets were all the lowest, which were, respectively, significantly different from those of NS+NS+ and NS+NS− treatments. In addition, the leaf RWCs of ramets in NS+NS− treatment were insignificantly different from those of NS+NS+ treatment, with the RWCs of mother ramet and daughter ramets in heterogeneous NS+ (left) and NS− (right) patches ranging 80.2–84.1%.

3.4. Anatomical Structures of Rhizomes and Leaves

As shown in Figure 4, the rhizome of Kentucky bluegrass comprises epidermis (E), VBs and ground tissues. The epidermis consists of a layer of parenchymatous cells. VBs distribute in the ground tissues adjacent to the epidermis. The rhizome anatomical structure of the daughter ramet in the homogeneous NS+ patch is alike to that in the heterogeneous NS+ patch with VBs arranging regularly (Figure 4A,B). Under the observation of SEM, the rhizome diameters of daughter ramets in the homogeneous NS− patch and heterogeneous NS− patch (Figure 4C,D) are significantly smaller than those in water-rich patches. In addition, the diameters of VBs in water-poor patches are obviously smaller (

Table 3. Comparison of the rhizome crosscutting structures of Kentucky bluegrass ramets in different patches under different treatments.

Treatment	Diameter of Vascular Bundles (μm)	Pore Diameter of Central Parenchyma in the Rhizome (μm)
Homogeneous NS+ patch	98.13 ± 3.84 a	35.27 ± 2.86 b
Heterogeneous NS+ patch	97.45 ± 3.26 a	36.45 ± 2.37 b
Heterogeneous NS− patch	83.25 ± 4.12 b	50.89 ± 2.25 a
Homogeneous NS− patch	72.80 ± 4.31 c	49.75 ± 3.48 a

Different lowercase letters within the same column indicate significant differences at 0.05 level.

), and the arrangement of some VBs is disordered. The distances between adjacent VBs (i.e., the number of parenchymatous cell layers) decrease. Moreover, the pore diameters of central parenchyma (PT) notably increase whereas the numbers of PT decrease in rhizomes of water-poor patches as compared to those in the homogeneous NS+ patch and heterogeneous NS+ patch (Table 3). The diameter of VBs in the rhizome of the heterogeneous NS− patch is remarkably bigger than that of the homogeneous NS− patch (p < 0.05).

The leaf of Kentucky bluegrass comprises epidermis (E), mesophyll and veins. As shown in Figure 5, the upper and lower epidermises are respectively composed of a layer of epidermis cells arranged tightly, and the one with a convex vein is the lower epidermis. The lower epidermis cells adjacent to the vein are thick. The leaf surface consists of long and narrow cells, which arrange orderly. The large parenchymatous cells located at the sides of the primary vein in the upper epidermis are motor cells (MC), which are also called bulliform cells. Several MCs often arrange together, forming a fan shape in the leaf crosscutting structure. When water is insufficient, leaves fold and curl due to the water loss of MC to prevent the transpiration of water. When water is sufficient, MC absorbs water and expands, thus the leaves recover. The mesophyll of Kentucky bluegrass leaf is composed of parenchymatous cells, with no differentiation of palisade tissue and spongy tissue. There are slight shrinks on the surface, especially on the sides of the primary vein of Kentucky bluegrass leaves in the heterogeneous NS− patch (Figure 5C) and homogeneous NS− patch (Figure 5D) in comparison to those in the homogeneous NS+ patch (Figure 5A) and heterogeneous NS+ patch (Figure 5B).

The veins of Kentucky bluegrass leaf are in paralleled arrangement. Each vein includes a VB, and the size of the VB is proportional to that of the vein. The vein in the center of the leaf is the primary vein, and the smaller ones on the sides of primary veins are lateral veins. Figure 6 shows the crosscutting structures of primary veins of Kentucky bluegrass leaves. The convex part of the leaf is the primary vein, and there is a cluster of sclerenchymas in the lower epidermis, opposite to the primary vein. Under SEM, the edge of VB is surrounded by sclerenchymas, forming a VB sheath. The VB sheath comprises the xylem and phloem. Xylem consists of 3–4 vessels, presenting a “V” shape in the crosscutting structure. The vessels with bigger pore diameters located at the arms of the “V” shape are pitted vessels (PVs), and the inside 1–2 vessels with smaller pore diameters are ringed vessels or spiral vessels (SVs). With the growth of the leaf, ringed vessels or SVs break, thus forming a cavity. Phloem (P) locates at the side adjacent to the lower epidermis, i.e., the site opposite the opening of the xylem vessel “V” shape.

The diameters of vessels in the xylem, the percentages of xylem area in VBs and the diameters of the sieve element in the phloem of the primary veins in Kentucky bluegrass leaves of different patches were summarized in

Table 4. Comparison of the leaf crosscutting structures of Kentucky bluegrass ramets in different patches under different treatments.

Treatment	Diameter of Vessel in Xylem (μm)	Percentage of Xylem Area in Vascular Bundles (%)	Diameter of Sieve Element in Phloem (μm)
Homogeneous NS+ patch	13.48 ± 0.33 a	52.82 ± 5.63 a	2.34 ± 0.13 b
Heterogeneous NS+ patch	13.25 ± 0.41 a	52.57 ± 6.28 a	2.45 ± 0.14 b
Heterogeneous NS− patch	12.89 ± 0.38 a	50.15 ± 8.22 a	2.78 ± 0.13 ab
Homogeneous NS− patch	12.04 ± 0.32 b	46.95 ± 4.52 b	3.22 ± 0.19 a

Different lowercase letters within the same column indicate significant differences at 0.05 level.

, and the data were mean values of 10 measurements. The results showed that the diameter of the vessel in the xylem and the percentage of the xylem area in VBs of the primary vein in leaves of the homogeneous NS− patch were significantly smaller than those of other patches. However, the diameter of the sieve element in the phloem of the homogeneous NS− patch is the biggest, which is significantly different from those of the homogeneous NS+ patch and heterogeneous NS+ patch. The values of the above three attributes in heterogeneous NS+ and NS− patches are similar, which are insignificantly different from those in the homogeneous NS+ patch (Table 4).

Figure 7 presented the crosscutting structures of the lateral veins of Kentucky bluegrass leaves. The VBs in lateral veins are smaller, the structures of which are similar to those of primary veins. Under SEM, a structure similar to a small vascular bundle (B) is observed among lateral veins, which is likely the crossvein linking the net structure of paralleled veins. The structures of VBs in lateral veins are simple, with no differentiation of xylem and phloem, which are composed of a line of vessels and a line of sieve elements. The vessels and sieve elements arrange tightly in water-poor patches (Figure 7C,D), the distances between which are shorter than those in water-rich patches (Figure 7A,B).

4. Discussion

In heterogeneous-resource habitats, the clonal integration and phenotype plasticity of the clonal plant is the important foundation for its escape from unfavorable habitats with scarce resources, exploration of favorable habitats with abundant resources, and the realization of space occupation, domain extension and species continuation [9,11,29]. Water is the key ecological factor for plants growing in arid and semi-arid regions, which influences their growth and development. Plants respond to different soil water conditions through morphological plasticity to realize maximum resource acquisition.

For clonal plants living in heterogeneous water environments, morphological plasticity exists in their growth pattern and population morphology [22,30]. In the present study, the growth directions of Kentucky bluegrass ramets were random in both homogeneous and heterogeneous water treatments, and the numbers of daughter ramets showed incremental quadric-curve trends with the extension of plantation time (Table 1). The new daughter ramets were derived from the continuously forming primary and secondary rhizomes. The results were in accordance with the study on the clonal growth of Chinese seabuckthorn in soils with different water contents [35]. However, there were differences in the proliferation rates and numbers of daughter ramets among different water treatments. The proliferation rate in NS−NS− treatment was distinctly lower than that in NS+NS+ and NS+NS− treatments and the number of daughter ramets were significantly fewer (Figure 2). Yue [36] found that the number of ramets of Phyllostachys praecox cv. prevernalis declined with the decreasing soil moisture, which was consistent with the result of our study. The above results indicated that clonal plants growing in favorable habitats increased the number of ramets to achieve the maximum acquisition and utilization of resources [29].

Under NS−NS− treatment, the growth of Kentucky bluegrass mother ramet was suppressed, with shorter leaf length, narrower leaf width and shorter ramet height. In addition, the internode between adjacent daughter ramets was the longest in NS−NS− treatment, whereas the shortest in NS+NS+ treatment among three different water treatments (Table 2). A phalanx growth form enables clonal plants to make better use of resource-rich patches, whereas a guerrilla growth form provides them with opportunities to escape from resource-poor sites. Ye [37] proposed a hypothesis that a trade-off between these two growth forms existed in plants, producing both spreading and clumping ramets under different resource levels. Kentucky bluegrass in water-rich habitats shortened the internodes (phalanx growth form) to form clumping ramets to maximize the acquisition of required water. Kentucky bluegrass growing in water-poor environments also lengthened the internodes (guerilla growth form) to explore new habitats and acquired more water for better growth [29,37]. The results of the present study supported the above hypothesis. The growth indices, including leaf length, leaf width and the height of mother ramet in NS+NS− treatment were better than those in NS−NS− treatment. Furthermore, the growth status of daughter ramets was similar in heterogeneous NS+ and NS− patches, which was also insignificantly different from that of the homogeneous NS+ patch. These results showed that the translocation and re-distribution of water occurred among different daughter ramets growing in patches with different water supplies. Thus, water in a water-rich patch was supplied to daughter ramets in a water-poor patch through rhizomes, and the growth inhibition by drought was alleviated [13].

It was also found that the root/shoot ratio of Kentucky bluegrass in NS+NS− treatment was significantly higher than that in NS−NS− treatment, but insignificantly different from that in NS+NS+ treatment (Figure 3C). These findings were in agreement with those of You [16] on Myriophyllum aquaticum L. and Luo [38] on buffalograss. The above results indicated that Kentucky bluegrass under heterogeneous water treatment tended to distribute more water to roots to enhance the growth of roots and absorb more water and nutrients.

Leaf RWC is the key index reflecting the water status of the plant. Under NS+NS− treatment, the RWC of Kentucky bluegrass mother ramet was insignificantly different from those of daughter ramets in heterogeneous NS+ and NS− patches, which were also similar to those of NS+NS+ treatment (Figure 3D). These results indicated that there was water sharing and re-distribution among connected clonal ramets under heterogeneous water treatment. The “source-sink” gradient forming among different ramets promoted the water physiological integration and the full utilization of environmental resources [21]. Thus, the growth and development of plants in heterogeneous habitats were enhanced, which improved the ecological adaptability, reduced the risk of death, and increased the advantages of evolution and competition of the whole population.

When plants are growing in different ecological environments, their morphological structures (phenotypes) change accordingly. These changes occur not only in the external morphological characteristics of plants but also in the internal structures. Some studies found that the differences in the anatomical structures of plant rhizomes and leaves were closely correlated with the water conditions of habitats and the drought resistance of plants [22,34]. In the current study, it was observed that the diameters of rhizomes and VBs significantly decreased in Kentucky bluegrass growing in water-poor treatments, and the arrangements of VBs were disordered in the crosscutting structures of rhizomes (Figure 5, Table 3). The reductions in the root and stem cross-section diameters and xylem vessel diameter were also found in Astragalus gombiformis Pomel., Stipa lagascae and Arachis hypogaea L. under drought stress [39,40,41]. These structural changes were the reflections of the extent of drought stress and the drought resistance of plants.

The PT in the rhizome possesses the filling effect, storage function and water retention capacity. Under water-poor treatment, the pore diameters of PT in the rhizome increased, whereas the numbers decreased (Table 3). The above structure changes might favor the storage of water and nutrients in Kentucky bluegrass and maintain its growth under water-poor treatment. The rhizome VB diameter in the heterogeneous NS− patch notably increased as compared to that in the homogeneous NS− patch. These results indicated that there was water sharing to some extent among the connected ramets under heterogeneous water treatment. The re-distributed water improved the growth of daughter ramets growing in water-poor patches, which was a reflection of the morphological plasticity of clonal ramets [39].

The leaf is the primary site where photosynthesis happens. It is also the most sensitive organ, with the largest exposure area in different habitats. Under water-poor treatment, SEM observation on Kentucky bluegrass leaf crosscutting structure showed a reduced diameter of the vessel in xylem, a decreased percentage of xylem area in VB, and an increased diameter of sieve element in the phloem of the primary vein (Table 4). These results were in accordance with those of Saud [22] and Boughalleb [39]. Vessels are elongated, water-conducting die cells in the xylem and the primary channels for the transport of water and mineral ions. Drought induces air embolism in plant xylem vessels that blocks long-distance water transport and reduces water transport capacity [42]. Xylem resistance to drought-induced embolism appears to be correlated to high frequencies of smaller diameter vessels and low inter-vessel connectivity that restricts embolism spread to short files of multiple vessels and limits an extensive spread of air through the xylem network that is initiated from a single embolized vessel [43]. Sieve elements are thin-walled cells that are alive at maturity and function as the basic photosynthate-conducting cell type in the phloem of vascular plants [44]. It was deduced that the increased diameter of the sieve element in phloem together with the decreased percentage of xylem area in VB (Table 4) favored the transport of photosynthates within the plant and decreased water evaporation. Parts of the photosynthates may serve as osmotic adjusting substances and help the whole genet of Kentucky bluegrass cope with drought. To sum up, the structure characteristics of plant VBs were influenced by the resources in the habitats, which in turn affected the transport and utilization of environmental resources as well as the growth patterns of plants to a great extent.

5. Conclusions

Physiological integration among interconnected ramets through rhizomes or stolons confers clonal plants' better adaptability to habitats with resource heterogeneity. Sharing of water, carbohydrates and nutrients seems especially profitable in environments in which these resources are heterogeneously distributed in space. Soil water heterogeneity generally exists in natural habitats and the knowledge regarding the influence of heterogeneous water on the growth of clonal plants in experimental plots is important. In this study, plot experiments with two homogeneous and one heterogeneous soil water treatments were performed to test the influence of water heterogeneity on the growth and possible clonal integration of water in Kentucky bluegrass.

The experimental results showed that water deficits resulted in notable reductions in the above-ground biomass, under-ground biomass, the root/shoot ratio and leaf RWC of the whole genet, and leaf length and height of the mother ramet in NS−NS− treatment and decreases in leaf width and height of the daughter ramet in the homogeneous NS− patch. However, the leaf width, height and leaf RWC of the daughter ramet were greatly promoted in the heterogeneous NS− patch. Translocation and sharing of water among the interconnected daughter ramets of Kentucky bluegrass favored the growth and drought resistance of ramets in the heterogeneous water-poor patch. In addition, drought severely influenced the anatomical attributes of Kentucky bluegrass in the homogeneous NS− and heterogeneous NS− patches. However, the diameter of the vessel in the xylem and the percentage of xylem area in VBs in leaves of heterogeneous NS− patches were recovered to the levels of homogeneous NS+ and heterogeneous NS+ patches by water integration. In general, the results above demonstrated that physiological integration modified the phenotypic responses of Kentucky bluegrass clonal ramets for the efficient acquisition of water in heterogeneous water conditions. The current research provides a theoretical and practical guide for the scientific water management of Kentucky bluegrass in arid and semi-arid regions.