Forest Ecology and Management xxx (xxxx) xxx–xxx Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco Variability of Tamarix spp. characteristics in riparian plant communities are affected by soil properties and accessibility of anthropogenic disturbance in the lower reaches of Heihe River, China ⁎ Jingyi Ding, Wenwu Zhao , Bojie Fu, Shuai Wang, Hao Fan State Key Laboratory of Earth Surface Processes and Resource Ecology, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, PR China Institute of Land Surface System and Sustainable Development, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, PR China A R T I C L E I N F O A B S T R A C T Keywords: Tamarix spp. Riparian plant community Vegetation-environment interactions Anthropogenic disturbance The lower reaches of Heihe River Tamarix, one of the most widely distributed riparian shrub genera, plays an important role in ecosystem stabilization in arid zones. Various studies have assessed the effect of stressors on Tamarix spp. in riparian plant communities in homogenous habitats. However, less is known about the factors driving variation of Tamarix spp. characteristics across heterogeneous habitats in desert riparian zones, which would constrain effective ecological conservation in such areas. In this study, field experiments were conducted to determine the variability of Tamarix spp. characteristics and its driving factors across various riparian plant communities in the lower reaches of Heihe River, China. Five riparian plant communities containing Tamarix spp. were classified, ranging from tree–shrub–herb community to shrub dune. With the increase of height and crown diameter of Tamarix spp., its capacity to resist drought increased, while community diversity formed a bimodal variation pattern instead of consistently decreasing trend due to variations of vegetation–environment interactions across the heterogeneous habitats. Among environmental factors, accessibility of anthropogenic disturbance and soil properties, explaining 42% of vegetation variance, were deemed the key factors influencing the variation of Tamarix spp. characteristics. By contrast, water availability (e.g., soil moisture and proximity to the river) showed a weak relationship with Tamarix spp. characteristics due to the drought tolerance of Tamarix spp. and its interactions with the extreme environment. As accessibility of anthropogenic disturbance and soil properties were regarded as major environmental factors driving the Tamarix spp. characteristics, multiple measures such as establishing protective zone, controlling cantaloupe planting and developing water-saving irrigation were proposed to secure the vegetation structure of Tamarix spp. and maintain the resilience of riparian plant communities under intensive human activities. 1. Introduction Riparian zone is the interfaces between land and water, characterized by a heterogonous environment and various vegetation communities (Naiman and Décamps, 1997). Tamarix (also called: saltcedar or tamarisk, family Tamaricaceae) is one of the most widely distributed riparian genera because of its ability to adapt to a range of saline and drought conditions. (Ding and Zhao, 2016; Stromberg et al., 2009). It provides multiple ecosystem services to humans and it functions as an ecological protection against desertification, especially in arid ecosystems (Ma et al., 2009). As a result of its strong drought tolerance and adaptive ability, Tamarix spp. grows with other species and develops into different riparian plant communities under long-term vegetation–environment interactions (Yang et al., 2002). Various studies have explored the interaction between Tamarix spp. and environmental factors in riparian plant communities, and the results varied with the habitat investigated (Stromberg et al., 2009). Studies in humid riparian areas showed that the Tamarix spp. was characterized by strong competition for water and the ability to secrete salt into the soil, which would alter the growth condition of native species, form a singlespecies community, and thus, decrease the biodiversity (Natale et al., 2010). By contrast, other studies, mainly in desert oasis areas, reported that interactions between Tamarix spp., shrubs, and environmental factors could provide good wildlife habitats through the ‘fertile island’ effect, resulting in a higher level of biodiversity (Li et al., 2007). Although the interaction between Tamarix spp. and environmental factors ⁎ Corresponding author at: Institute of Land Surface System and Sustainable Development, Faculty of Geographical Science, Beijing Normal University, No. 19, XinJieKouWai St., HaiDian District, Beijing 100875, PR China. E-mail address: [email protected] (W. Zhao). http://dx.doi.org/10.1016/j.foreco.2017.10.003 Received 29 April 2017; Received in revised form 28 September 2017; Accepted 2 October 2017 0378-1127/ © 2017 Elsevier B.V. All rights reserved. Please cite this article as: Ding, J., Forest Ecology and Management (2017), http://dx.doi.org/10.1016/j.foreco.2017.10.003 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Fig. 1. The lower reaches of Heihe River basin (B) in China (A) and the location of sampling sites in the study area (C). Notes: A–E in (C) indicate the sampling transects. greatly to variation in species diversity in desert riparian areas (Dı́Az and Cabido, 2001). Except for natural factors, the impact of anthropogenic disturbances on vegetation has significantly increased in the Anthropocene (Lavorel et al., 2015). The ‘accessibility to anthropogenic disturbance’ is a comprehensive index which could represent the possibility of communities being disturbed by various anthropogenic activities, with mainly negative effects on community characteristics (Liu et al., 2014b; Zeng et al., 2011). Vegetation communities near to roads are more easily disturbed by grazing and tourism, which in turn can alter the environmental conditions (e.g., soil properties and water availability), leading to degradation of vegetation communities. In addition, vegetation communities near road are easily accessed by large agricultural machines and disturbed by other human activities (e.g., collecting firewood, expanding farmland), which may alter community structure and decrease ecosystem resilience significantly (Daryanto et al., 2013). Although previous studies provide some insight into the interactions between Tamarix spp. in riparian plant communities and environmental factors (DiTomaso, 1998; Li et al., 2007; Stromberg et al., 2009), key factors that drive the formation of Tamarix spp. characteristics in desert riparian zones remain unclear. In addition, previous studies mainly concentrated on natural factors, such as water availability and soil properties (Ma et al., 2009; Natale et al., 2010; Xu et al., 2011), which, without consideration of anthropogenic effects, cannot provide a comprehensive view of Tamarix spp. characteristics' formation, especially in desert riparian zones, which are characterized by heterogeneous habitats and intense anthropogenic activities. In the current study, we surveyed multiple vegetation community characteristics (including community floristic composition, community structure indices, and diversity indices) and environmental factors (including water availability, soil properties, and accessibility to anthropogenic disturbance) in the lower reaches of Heihe River. We first classified riparian plant community types that contains Tamarix spp. in in riparian plant communities has long been recognized, these studies mainly focused on homogeneous habitats in the study area. However, vegetation–environment interactions vary with habitat types and habitats in the desert riparian zone are highly heterogonous, varying from humid riparian areas near the riverbank to drought-stressed desert peripheries. Previous studies in desert riparian zones mainly focused on vegetation–environment interactions in shrub dune habitats (Xie et al., 2015). However, less is known about the factors determining the variability of Tamarix spp. characteristics in riparian plant community across heterogeneous habitats in desert riparian zones, which inhibits effective ecological restoration of Tamarix spp. in riparian plant communities in this crucial ecosystem (Lü et al., 2011). The variation in Tamarix spp. community characteristics (e.g., community structure, floristic composition and diversity) results from interactions between vegetation and multiple environmental factors (Flato et al., 2013). Water availability is a key factor that positively influences vegetation characteristics, particularly in arid and semiarid ecosystems (Fang et al., 2016). Tamarix spp., as a drought tolerance species, depends on multiple water resources to maintain its growth. Soil moisture, especially a depth of 40–60 cm is the direct water resource for Tamarix spp. in riparian plant communities; by contrast, groundwater, to a depth of 10 m, representing the survival threshold for Tamarix spp., is its stable water resource (Long et al., 2014). In addition, as both groundwater and soil moisture are recharged by rivers or streams, the distance from river channel can be regarded as a proxy for water availability, given that the influence of river will weaken with the distance from river channel (Hao et al., 2010). Soil properties (e.g., soil physical properties and soil chemical properties), as the foundation of vegetation growth, also shape the community floristic composition and species vitality by influencing ecological and hydrological processes (Stirzaker et al., 1996). Previous studies have reported that soil heterogeneity is one of the main factors determining the dominant species in vegetation communities, whereas changes in soil nutrition contribute 2 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. 2.3. Data collecting the study area and then we explored the variation in community characteristics among the different community types. Based on these results, we determined the key factors driving the variability of Tamarix spp. characteristics. The objectives of this study were to: (1) explore the variation of Tamarix spp. characteristics across heterogeneous habitats; (2) discern key factors shaping Tamarix spp. characteristics formation; and (3) suggest possible management suggestions to protect Tamarix spp. in riparian plant communities against changing environmental conditions. The distance of each site from the river channel and the nearest road was recorded by mobile GPS (version eTrex 301, Garmin Electric Business Management Co., Ltd., Shanghai, China). Three shrub quadrats (10 × 10 m) in each site were established in triangle with 30 m away from each other. In each quadrat, the height and species in each layer were recorded. The crown diameter was measured by averaging the diameter of each crown in an east-west and north-south direction. Coverage of each species was calculated by the percentage of crown area to the sampling area and plus the canopy coverage which was measured by photographing from the base looking up using a fish-eye lens (Lei and Wen, 2008). The specific leaf area (SLA) of Tamarix spp. was measured in each site: 20 Tamarix spp. leaves were randomly sampled from four directions (e.g., east, south, west, and north) on each plant and stored in an ice chest immediately after collection. In the laboratory, leaves were removed from twigs and placed without overlap onto an optical scanner bed. WinSEEDLE image analysis software was used to evaluate leaf area (S/cm2). After scanning, the sampled foliage was dried and weighed (M/g), and then the SLA was calculated using Eq. (1): 2. Material and methods 2.1. Study area The study area was located in the lower reaches of Heihe River (40°20′–42°30′N; 99°30′–102°00′E), in the Ejina Oasis, Northwest China and covered an area of 3 × 104 km2 (Fig. 1). The topography of the basin inclined from the southwest to the northeast, with an average slope of 1–3‰ (Akiyama et al., 2007). The region has a typical continental arid climate, with maximum and minimum temperatures usually occur in July (41 °C) and January (−36 °C), respectively. The long-term annual average precipitation at the study site was less than 42 mm, while the annual pan evaporation rate was approximately 2300–3700 mm (Wen et al., 2005). The Heihe River, rising in the Qilian Mountains, is the main water resource for vegetation in the Ejina Oasis. The plant communities in the Ejina oasis grow mainly along the riverbanks and on the fluvial plain, with dominant vegetation including Populus euphratica in the tree layer; Tamarix spp., Lyceum ruthenicum in the shrub layer; and Karilinia caspica and Peganum harmala in the herbaceous layer (Zhao et al., 2016). Sparse and drought-tolerant desert species, such as Reaumuria soongorica and Zygophyllum xanthoxylon, are mainly distributed in the Gobi desert. The main soil types in the study site are grey desert soils and grey–brown desert soils (Chen et al., 2014). Local agriculture mainly depends on cantaloupe plantations cultivated during May–September in each year. In addition, the Ejina Oasis is one of most famous tourist attractions in China, with almost 200,000 visitors visit per year (Hochmuth et al., 2014). Two primary roads and one highway were built parallel to the river channel and across the south of the oasis to support the development of agriculture and tourism in this region. SLA = S/M (unit: cm2/g) (1) At each sampling site, bulk density (BD) was measured using a stainless-steel cutting ring (100 cm3 in volume) and oven dried at 105 °C to a constant weight. Three replicates of soil moisture samples and soil samples were collected along the diagonal line of each quadrat (i.e. on the two opposite angles and the center of each quadrat). Soil moisture samples were collected using an auger (5 cm in diameter). Soil gravimetric water content (SWC) was determined every 10 cm at a depth of 0–80 cm, and every 20 cm at a depth of 80–200 cm. Samples were collected in aluminum boxes and weighed at the time of sampling as well as after oven drying at 105 °C for 48 h. Soil samples were collected to a depth of 100 cm in each quadrat to determine the soil composition and chemical properties. After any roots had been removed, the soil samples were air-dried and passed through a 2-mm sieve; surface soil samples (from a depth of 0–20 cm) were subsequently analyzed in the laboratory to determine their clay (< 0.002 mm), silt (0.002–0.05 mm), sand (0.05–2 mm), and gravel (> 2 mm) content using a Malvern Mastersizer 2000. Soil organic matter (SOM) was measured using the K2Cr2O7 method (Liu, 1996). Total nitrogen (TN) was determined using the Kjeldahl method (ISSCAS, 1978). Total phosphorus (TP) was determined using a UT-1810PC spectrophotometer (PERSEE, Beijing, China) after H2SO4–HClO4 digestion (ISSCAS, 1978). Total salt content (TS) was determined by oven method (Liu, 1996). 2.2. Sampling design The field experiment was conducted in July 2015 and July 2016. As riparian plant communities containing Tamarix spp. are mainly distributed in land-water transition zone (i.e. riparian zone) and their characteristics vary with environmental gradient (Fu et al., 2014), transect method was used to sample sites in this study under the principle of random sampling (Wheater et al., 2011). To fully cover the distribution pattern of riparian plant communities that contain Tamarix spp., five 3500-m long transects (A–E) were randomly established northward across the study area (Fig. 1C). As the distance from river is the major environment gradient in the low reaches of Heihe river (Fu et al., 2014), transects were established along the distance from river. On each transect, systematic sampling was used to sample each site (Yates, 1981). Sampling started at the distance of about 100 m from river which was randomly sampled across the region and the sampling sites (50 × 50 m) were systematically sampled with about 500 m intervals at each transect. Transects were not strictly in line shape or perpendicular to river as sites were sampled far from villages and large farmlands. A total of 38 sites were sampled during the field sampling. 2.4. Statistical analysis Importance value (IV) is a comprehensive index that used to determine the overall importance of each species in the community structure (Odum and Barrett, 1971; Ruiz and Lugo, 2012; Ross et al., 2016) and it was used in community classification in this study. The importance value (IV) of a tree was calculated based on the average of relative density, relative dominance, and relative height, whereas the IV of shrubs and herbs in each site was calculated based on the average of relative density, relative dominance, and relative coverage (Zhang and Dong, 2010). The total diversity index of each Tamarix spp. community was used to depict the community diversity in each site. Based on the community vertical structure, the total diversity index for each community was measured using the weight of indices in different growth types, using Eq. (2) (Gao et al., 1997): 3 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Wi = (Ci/ C + hi / h)/2 (2) significance. The statistical analyses were performed using SPSS software (version 20.0; SPSS lnc., Chicago, IL, USA). The CCA analysis and Monte Carlo forward selection in RDA were performed using CANOCO (version 4.5; Microcomputer Power, Ithaca, NY, USA). where C was the total coverage of community (C = ∑ Ci ); i = 1, tree layer; 2, shrub layer; 3, herb layer; h was the thickness of the leaf layer for various growth types (h = ∑ hi ), Wi was the weighted parameter of diversity index of ith growth type, Ci was the coverage of the ith growth type, and hi was the average thickness of the leaf layer of the ith growth type. The thickness of the tree leaf layer was calculated as being 33.3% of the height of the tree layer, whereas the shrub layer was 50% and the herb layer was 100%. The total diversity index of the community (A) was calculated using Eq. (3): A= ∑ Wi Ai 3. Results 3.1. Classification of riparian plant communities that contain Tamarix spp. In the lower reaches of Heihe River, five types of riparian plant communities that contain Tamarix spp. were determined based on TWINSPAN classification and DCA ordination (Fig. 2). The composition of each sampling site is detailed in Table 1 and the spatial location of each community is shown in Fig. A.1. Five communities were mainly different in community structure and floristic composition and the characteristics of each communities is listed as below. Community I was Ass. Populus euphratica–Tamarix spp. + herbs. This community type was distributed within 1600 m of the river channel with low coverage (36.09% on average). Characterized by treeshrub-herb layers, this community was dominated by Popolus euphratica in the tree layer. Tamarix spp. were mainly sparsely distributed (coverage 0.67–22%) under the Popolus euphratica canopy, while Sophora alopecuroides was the only herb species in the community, growing sparsely in the understory (average coverage < 5%). Community II was Ass. Tamarix spp.–Lycium ruthenicum + herbs. This community was distributed widely, ranging from 1050 m to 3200 m from the river channel, with a total community coverage of 15.54–85.42%. This community was primarily dominated by Tamarix spp. (importance value = 0.48–0.93), accompanied with Lycium ruthenicum in the shrub layer (importance value = 0.07–0.52). Sophora alopecuroides and Karelinia caspia were the major species in the herb layer with the latter being the more dominant of the two species (importance value = 0.51–0.97). Community III was Ass. Lycium ruthenicum–Tamarix spp. + herbs. This community was distributed within 1020–2870 m of the river channel, with a community coverage of 28.50–88.79%. The shrub layer was dominated by Lycium ruthenicum (importance value = 0.54–0.77), accompanied with a high coverage of Tamarix spp. (3.5–60%). In the herb layer, most sites were dominated by Sophora alopecuroides with importance value range from 0.16 to 0.65. Community IV was Tamarix spp. + sparse herbs. This community was primarily distributed more than 2000 m from the river channel. Tamarix spp. dominated in this community with a high importance value (0.76–1) and coverage (19.67–86%). Lycium ruthenicum and Reaumuria songarica (a typical desert shrub) occurred in some sites, but with low importance value (< 0.25). Herb grows sparsely in only five sites in this community with most of sites dominated by Sophora alopecuroides. Community V was Ass. Tamarix spp. shrub dune. This community was mainly distributed within 960–2780 m of the riverbank. Tamarix spp. were the dominant species of the shrub layer (importance value = 0.55–0.71), and mainly occurred in shrub dune form. Lycium ruthenicum also grows in this community (importance value = 0.29–0.45) and Reaumuria songarica grows in some sites. Sophora alopecuroides herbs dominated the herb layer, but were sparsely distributed and only existed in one site. (3) where A is the total diversity index of the community; i = 1 for the tree layer, 2 for the shrub layer and 3 for the herb layer; Wi is the weighted parameters of the diversity index for the ith growth type; Ai is the diversity index of the ith growth type. Species diversity indices of each layer were determined (Liu et al., 1997) using Shannon-Wiener diversity index (Eq. (4)): s H = − ∑ (Pi lnPi ) i=1 (4) Simpson dominance index (Eq. (5)): s D = 1− ∑ Pi2 i=1 (5) Pielou evenness index (Eq. (6)): Jsw = H /(ln(S )) (6) and Patrick richness index (Eq. (7)): R=S (7) where Pi was the relative importance value of species i, and S was the total number of species in the ith site. All the diversity indices were calculated based on the above equations using Excel software (Microsoft Office, version 2010). We used Two-way Indicator Species Analysis (TWINSPAN, in WinTWINSPAN, version 2.3; a method of hierarchical classification) and Detrended Correspondence Analysis (DCA, in CANOCO, version 4.5; an ordination axes partitioning) to classify the possible Tamarix spp. community types. The importance value data for all plant species were used in these analyses. Least significant difference (LSD) test in the one-way analysis of variance (ANOVA) was used to determine the significance of differences in vegetation characteristics and environmental factors among the classified communities. Pearson correlation analysis and canonical correspondence analysis (CCA) were used to determine the strength of possible relationships between vegetation indices and environmental factors. The statistical significance level was P < .05. Stepwise multiple linear regression analyses between Tamarix spp. characteristics and environmental factors were used to explore the explanation of each environmental factor for the vegetation variance. To further separate the key influencing environment variables, marginal and conditional effects of various variables were calculated via Monte Carlo forward selection in a redundancy analysis (RDA). Marginal effects reflect the effects of environmental variables on the community characteristics, whereas conditional effects reflect the effects of environmental variables on the community characteristics after the anterior variable had been eliminated by the forward selection method. Once the redundant variables were eliminated and a group of key environmental factors was determined through the forward selection, this method enabled key variables to be determined through the strength of their effects and 3.2. Variation of Tamarix spp. characteristics The height and crown diameter of Tamarix spp. showed a similar pattern, increasing from Community I to Community V, with significantly (P < .05 or P < .01) higher values in Communities IV or V compared with Communities I to III. With the increase in height and 4 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Fig. 2. The dendrogram of sampling sites based on the TWINSPAN classification (a) and DCA ordination diagram (b). Notes: Numbers 1–38 represent the number of the sampling sites. D indicates the classification level and N indicates the number of sampling sites in each classification. I–V represent community I–V. Arrows indicate that all the sites were divided into five major groups after the fourth classification. characteristics and environmental factors, CCA (Canonical Correspondence Analysis) ordination diagram was conducted for these variables (Fig. 4). A Monte Carlo permutation test for all canonical axes showed P < .05 which indicated the significance of CCA analysis result. In Fig. 4, the coverage of Tamarix spp. was distributed closely with soil moisture, especially the 30–100 cm soil moisture (SWC2), while the height of Tamarix spp., which was situated opposite to coverage, showed a close relationship with bulk density and coarse particle composition (e.g., sand and gravel). Crown diameter was distributed at the left of diagram and showed a positive relationship with the distance from road. The specific leaf area and diversity indices were grouped together in the diagram, which showed a close relation with bulk density and soil particle composition. Regression analyses between Tamarix spp. characteristics and environmental factors were used to further explore the influence of environmental factors on the Tamarix spp. characteristics (Table 3). The regression analyses indicated the distance from river as predictors for Pielou evenness index. Soil properties (e.g., bulk density, clay, sand, and total phosphorus), explaining 11.9–31.7% variance, were selected as the predictors for coverage, specific leaf area, and community diversity (e.g., Shannon–Wiener diversity index, Simpson dominance index and Patrick richness index). Distance from road, explaining 14.7–38.1% of the variance, was mainly related to vegetation height and crown diameter. crown size of Tamarix spp. (i.e. from Community I to Community V), the community coverage also increased, before a slight decrease in Community V. The specific leaf area, which represents the droughtresistant ability of a plant, decreased from Community I to Community V, being significantly (P < .01) lower in the Community V. Community diversity differed significantly among the five communities (Fig. 3e, f). Shannon-Wiener diversity index, Patrick richness index and Pielou evenness index were all high in Communities II, III, and V. By contrast, Simpson dominance index reached high value in Communities I and IV, whereas the other diversity indices were low. 3.3. The relationship between Tamarix spp. characteristics and water availability, soil properties, and accessibility of anthropogenic disturbance To explore how environmental factors affect the Tamarix spp. characteristics, 14 environmental indices including water availability (i.e., 0–30 cm soil moisture, 30–100 cm soil moisture, 100–200 cm soil moisture, and distance from river), soil properties (i.e., bulk density, clay, silt, sand, gravel, soil organic matter, total nitrogen, total phosphorus, and total salt content) and accessibility of community to anthropogenic disturbance (i.e., distance from road) were selected based on 38 sampling sites in the field experiment. The result of Pearson correlation analysis between Tamarix spp. characteristics and environmental indices is shown in Table 2. Water availability showed a weak relationship with the vegetation characteristics, with only the 30–100 cm soil moisture positively correlated with Tamarix spp. coverage, whereas the distance from river was positively correlated with the diversity indices. Soil properties showed a significant relationship with vegetation characteristics. Bulk density and soil composition (e.g., clay and gravel) negatively correlated with vegetation structure (e.g., coverage, height crown diameter, and specific leaf area) and positively correlated with community diversity (e.g., Shannon–Wiener diversity index and Patrick richness index), whereas total nitrogen was only significantly correlated with crown diameter. Distance from road, representing the accessibility of anthropogenic disturbance, was positively correlated with community structure, especially crown diameter. To visually display the relationship between vegetation 3.4. Key factors affecting Tamarix spp. characteristics To examine key factors that affect the vegetation indices in communities that contain Tamarix spp., redundant variables were eliminated based on the marginal and conditional effects of 14 variables determined using Monte Carlo forward selection (Table 4). All the environmental factors explained 56% of the total variance. Distance from road, total nitrogen and bulk density were regarded as the key environmental factors (P < .1) shaping the characteristics of the Tamarix spp. Distance from road, representing the accessibility of anthropogenic activity, accounted for the most explanation (50%), whereas total nitrogen and bulk density, representing the effect of soil properties 5 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Table 1 The information of sampling sites in riparian plant communities that contain Tamarix spp. Plot number Distance from river (m) Tamarix spp. coverage (%) Community coverage (%) Importance value of major species in tree/shrub/herb layer S1 S2 S3 S4 S5 S6 Community I 1 2 8 14 21 30 220.00 630.00 490.00 60.00 140.00 1530.00 8.00 22.00 20.00 1.67 1.00 0.67 9.00 80.42 41.00 29.00 33.17 23.92 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 – – – – – – – – – – – – 1.00 1.00 – – 0.62 – – – – – – – Community II 3 4 6 7 19 20 1050.00 1380.00 2850.00 3200.00 2700.00 3200.00 43.00 50.00 27.00 35.00 15.00 4.50 85.42 52.25 80.50 84.37 31.70 15.54 – – – – – – 0.93 0.91 0.64 0.71 0.88 0.48 0.07 0.09 0.36 0.29 0.12 0.52 – – – – – 0.10 0.49 0.19 – 0.33 – – 0.51 0.70 0.97 0.59 0.84 0.87 Community III 9 10 11 12 13 16 17 36 1020.00 1470.00 1860.00 2510.00 2870.00 1250.00 1700.00 2870.00 13.00 23.00 7.80 15.67 8.33 60.00 3.50 15.00 32.00 81.54 41.22 82.67 61.35 88.79 28.50 57.20 – – – – – – – – 0.43 0.40 0.33 0.46 0.32 0.44 0.23 0.57 0.57 0.60 0.67 0.54 0.68 0.56 0.77 0.43 – – – – – – – – – 0.38 0.65 – 0.37 0.16 – 0.25 – – – – 0.20 0.23 – – Community IV 5 15 18 22 24 28 29 31 32 33 34 38 2290.00 480.00 2200.00 440.00 1580.00 110.00 580.00 1770.00 2350.00 2560.00 2840.00 2290.00 73.30 26.00 55.00 70.00 86.00 19.67 69.00 78.33 81.33 22.67 49.33 45.00 76.63 95.00 55.00 70.00 86.23 98.67 69.00 78.33 81.33 23.33 49.83 67.94 – – – – – – – – – – – – 0.84 1.00 1.00 1.00 0.87 0.88 1.00 1.00 1.00 0.89 0.82 0.76 0.16 – – – 0.13 0.12 – – – – 0.11 0.24 – – – – – – – – – 0.11 0.08 – 0.50 0.17 – – – 0.51 0.26 – – – – 0.26 0.07 – – – – 0.11 0.10 – – – – – Community V 23 25 26 27 35 37 960.00 1920.00 2370.00 2780.00 1470.00 2510.00 53.33 70.00 48.33 75 40.00 10.00 82.17 98.33 64.83 92.00 34.87 28.34 – – – – – – 0.56 0.55 0.61 0.53 0.64 0.71 0.44 0.45 0.39 0.42 0.29 0.29 – – – 0.06 0.07 – 0.14 – – – – – – – – – – – Notes: S1, Populus euphratica; S2, Tamarix spp.; S3, Lycium ruthenicum; S4, Reaumuria songarica; S5, Sophora alopecuroides; S6, Karelinia caspia. leaf area, reflecting the resource utilization strategy of a species under drought stress (Anyia and Herzog, 2004), decreased with the increase in the height and crown size of Tamarix spp. and reached a significantly low value at the Community V. This indicates that the community became more tolerant towards the drought environment with the increasing dominance of Tamarix spp., especially in Community V which was located at the periphery of oasis and desert. Species diversity is often used to depict variation in vegetation-environment interactions (Isbell et al., 2015). Previous studies showed that the growth of Tamarix spp. could either increase or decrease the biodiversity of community in different habitats (Whitcraft et al., 2007; York et al., 2011). In addition, it has also been reported that species–environment interactions in stressed environments can change the fundamental response shape of the community (Oksanen and Minchin, 2002). In the current study, instead of forming a consistent decreasing or increasing trend with the enhanced dominance of Tamarix spp. accounted for 25% of the total explanation. 4. Discussion 4.1. Variation of Tamarix spp. characteristics Five riparian plant communities that contain Tamarix spp. were classified by TWINSPAN and DCA in the lower reaches of Heihe River. From Community I to III, community mainly varied in community structure and floristic composition, transiting from a tree-dominated community (Community I) to a shrub-dominated community (Community III), with diminishing tree coverage and increasing shrub and herb coverage (Fig. 3a). Community coverage fluctuated slightly from Community III to V, with increasing dominance of Tamarix spp. (e.g., represented by a significant increase in the height and crown diameter) (Fig. 3a, c, d) and a rapid decrease in herb coverage. Specific 6 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Fig. 3. Differences in the coverage (a), specific leaf area (b), height (c), crown diameter (d) of Tamarix spp. and community diversity indices (e, f) in the five riparian plant communities that contain Tamarix spp. Notes: Values represent mean ± SE. Values with the same letter are not significantly different at the .05 significance level (LSD test). Values of an index followed by different lower-case letters differ significantly at the .05 significance level, while, values of an index followed by different upper-case letters differ significantly at the .01 significance level. 7 Forest Ecology and Management xxx (xxxx) xxx–xxx −0.071 −0.182 0.162 Fig. 4. CCA ordination diagram between vegetation characteristics and environmental factors. Notes: CD, crown diameter; SLA, specific leaf area; Shannon, Shannon–Wiener diversity index; Simpson, Simpson dominance index; Patrick, Patrick richness index; Pielou, Pielou evenness index; SWC1, 0–30 cm soil moisture; SWC2, 30–100 cm soil moisture; SWC3, 100–200 cm soil moisture; BD, bulk density; SOM, soil organic matter; TN, total nitrogen; TP, total phosphorus; TS, total salt content; Df River, distance from river; Df Road, distance from road. Table 3 Stepwise multiple linear regression of Tamarix spp. characteristics with environmental factors. Notes: Significant correlations (P < .05) are shown in bold and significant correlations (P < .01) in bold with underline. 0.586 0.22 Distance from road (m) Accessibility of human disturbance 0.359 −0.299 −0.181 0.311 0.077 0.194 −0.226 0.259 −0.01 0.07 0.05 −0.039 0.338 0.113 0.095 −0.155 0.324 −0.242 −0.186 −0.232 −0.055 −0.342 −0.137 −0.199 0.238 −0.22 0.071 −0.016 0.043 0.016 0.428 0.35 −0.346 0.347 −0.327 0.061 0.212 −0.165 −0.287 −0.13 −0.39 −0.149 −0.038 0.154 −0.393 −0.198 0.268 0.034 0.101 0.123 0.099 0.231 0.114 −0.375 0.095 0.121 −0.084 −0.337 −0.118 −0.319 −0.14 −0.095 0.136 −0.523 0.096 0.263 −0.223 −0.246 0.05 −0.218 −0.051 0.133 Bulk density (g/cm3) Clay (%) Silt (%) Sand (%) Gravel (%) Soil organic matter (g/kg) Total nitrogen (g/kg) Total phosphorus (g/kg) Total salt content (%) 0.205 Soil properties −0.06 −0.103 0.35 0.157 0.189 −0.239 0.258 −0.128 −0.042 −0.102 −0.026 0.112 −0.153 −0.044 0.106 0.22 0.154 0.155 0.214 −0.078 −0.307 −0.036 0.041 0.014 0.186 0.328 0–30 cm soil moisture (g/g) 30–100 cm soil moisture (g/ g) 100–200 cm soil moisture (g/g) Distance from river (m) Water availability 0.064 0.132 0.13 0.063 0.289 −0.0003 0.21 0.031 Patrick richness index Simpson dominance index Shannon–Wiener diversity index Specific leaf area (cm2/g) Crown diameter (cm) Height (cm) Coverage (%) Environmental factors Table 2 Pearson correlation analysis of the relationship between community characteristics, water availability, soil properties, and accessibility of anthropogenic disturbance. Pielou evenness index J. Ding et al. Variable Predictor Adjust R2 P value Pielou evenness index Coverage Specific leaf area Shannon–Wiener diversity index Distance from river Bulk density Clay Bulk density 0.131 0.242 0.129 0.128 < .01 < .001 < .05 < .05 Bulk density, Sand Bulk density Bulk density, Sand Bulk density Bulk density, Sand Bulk density, Sand, Total phosphorus Distance from road Distance from road 0.316 0.119 0.302 0.12 0.218 0.317 < .001 < .05 .001 < .05 < .01 < .001 0.147 0.381 < .05 < .001 Simpson dominance index Patrick richness index Height Crown diameter (DiTomaso, 1998), community diversity indices (i.e. Shannon-Wiener diversity index, Patrick richness index and Pielou evenness index) formed a bimodal pattern, peaking in Community II, III, and Community V. In the Communities II and III, the community was mainly characterized by diverse species in the shrub and herb layer (e.g., Tamarix spp., Lycium ruthenicum, Karelinia caspia, and Sophora alopecuroides). Compared with communities dominated by a single, highly competitive species (e.g., Popolus euphratica in Community I or Tamarix spp. in Community IV), less competitive environments enable to balance resource at different layers and make it possible for the growth of other species (Marshall and Baltzer, 2015). In addition, high water availability at all soil depths in Community III (Table A.1) provided an ample water resource for both deep- and shallow-rooted species, thus contributing to the high community richness (Fig. 3f) in this waterlimited ecosystem. Despite the high diversity recorded in Communities II and III, the diversity indices peaked again in Community V. Although this site was characterized by low soil moisture, the frequent interactions between wind erosion and shrubs create a suitable environment for diverse species (Ravi et al., 2010). The large crown diameter of Tamarix spp. in Community V enables sand to be trapped in the foliage (Cheng, 2007), resulting in the deposition of fine soil particles onto the shrub dunes (5.67% clay and 44.42% silt) (Table A.1), leading to 8 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Table 4 Key influencing factors selected based on the marginal and conditional effects obtained from the Monte Carlo test of forward selection. Environmental factors Distance from road Total nitrogen Bulk density Gravel Silt Sand Total phosphorus Soil organic matter Distance from river 100–200 cm soil moisture 30–100 cm soil moisture Clay 0–30 cm soil moisture Total salt content Marginal effects Percentage of variance explained (%) 28.3 13.7 11.8 8.8 3.8 2.3 2.1 1.6 1.1 8 7 5 4 4 Conditional effects Percentage of variance explained (%) P value Relative explanation (%) Environmental factors Distance from road Total nitrogen Bulk density 30–100 cm soil moisture 100–200 cm soil moisture 0–30 cm soil moisture Silt Gravel Distance from river Clay Sand Soil organic matter Total phosphorus Total salt content Total 28 8 6 3 3 2 2 2 2 0 0 0 0 0 56 0.006 0.03 0.064 0.174 0.148 0.278 – 0.308 0.324 0.884 0.742 – – 0.734 50.00 14.29 10.71 5.36 5.36 3.57 3.57 3.57 3.57 0 0 0 0 0 – also have an important role in shaping Tamarix spp. characteristics. Bulk density and soil composition, explaining 24.2% of the variance in coverage and 12.8–30.2% of the variance in community diversity, respectively (Table 3), were key factors that determined soil structure and indirectly influenced its water-holding capacity (Stirzaker et al., 1996). In humid area, soil with high bulk density often composed by high content of fine particle and the water holding capacity usually increases as the texture become finer (Black, 1968; Ravi et al., 2010). However, in hyperarid zone, soil of high bulk density usually characterized by relatively low fine particle as the soil contains a part of coarse fragment in the study area (e.g., the gravel content in the sites of Communities II and III) (Table A.1), which would result in a low water-holding capacity in soil, thus inhibit the growth of dominant species in the community and negatively shaping the Tamarix spp. structure (e.g., coverage, height, and crown diameter). By contrast, inhibiting the growth of Tamarix spp. could alleviate the strong competition and allelopathic effects on soils (Zhang et al., 2002). This would encourage the growth of diverse species and thus promote community diversity and weaken community dominance by a single species. Total nitrogen explained 8% of the variance in Tamarix spp. characteristics in our study and has been reported as one of the main factors influencing riparian vegetation in the Heihe River Basin (Xu et al., 2011). Although it is an essential nutrient supply for vegetation growth, we found that total nitrogen was negatively correlated with crown diameter, which differs from the positive relationship commonly reported between total nitrogen and crown diameter in humid areas (Wang et al., 2010). In hyperarid areas, Tamarix spp., which is often the dominant species in the shrub layer, is characterized by strong competition for water and nutrition resources. With the increasing dominance of Tamarix spp., indicated by the increase in Tamarix spp. coverage, height and crown diameter, the resource competition and utilization of the shrub layer increased, which could lead to a reduction in the quality of the soil resource. This would prevent the growth of the herb layer, the main resource of soil organic matter in the hyperarid zone, thus further reduce the amount of total nitrogen in the soil (Xu et al., 2011). Although water availability has been defined as a crucial factor determining vegetation growth in arid zones (Lü et al., 2014), current study found that the variation in Tamarix spp. characteristics showed a weak relationship with water availability. Soil moisture is the direct water resource for vegetation in arid zones. Approximately 60% of Tamarix spp. roots were concentrated at a depth of 20–60 cm (Imada et al., 2013), indicating the high potential for soil moisture utilization at this depth. However, only Tamarix spp. coverage was positively changes in the deposition of soil in vegetated patches. This can lead to the development of water and nutrient sinks in vegetated patches, rendering them suitable habitats for the growth of xerophytic herbs and shrubs, thus contributing to the accumulation of soil nutrients (8.09 g/ kg soil organic matter) (Table A.1) and high diversity recorded in Community V (D'Odorico et al., 2007). 4.2. Impact of soil properties and accessibility of anthropogenic disturbance on Tamarix spp. characteristics The spatial variation in Tamarix spp. characteristics was the result of interactions between vegetation and environmental factors. Previous study has shown that variation in water and nutrient availability associated with soil properties are likely to be important selective forces shaping ecosystem stability in arid zones (Rosenthal et al., 2005). In this study, we found that accessibility of anthropogenic disturbance (i.e., distance from road) and soil properties (e.g., bulk density and total nitrogen), explaining 28% and 14%, respectively, of the variance in community characteristics and were the key factors affecting the variation of Tamarix spp. characteristics (Table 4). The distance from road had significant influence on the structure of the Tamarix spp., explaining 14.7% and 38.1% of the variance in height and crown diameter, respectively (Table 3) (Liu et al., 2014a; Zenner et al., 2012). Tamarix spp. is often used as fuel wood in the desert riparian zone and the communities distributed near the road, such as Communities I, II, III, and IV (Table A.1), which were more easily being affected by fuel wood collection. The habitats of Tamarix spp. communities, especially Communities III and IV, were also more likely to be utilized as farmland, because they were characterized by suitable agricultural resources (e.g., high soil moisture and fine particle soils) as well as being located near to the road, making it easier to be accessed by humans and agriculture machinery (Yang, 2013). In addition, the herb species in the understory of riparian plant communities containing Tamarix spp. are used for grazing (Thiele et al., 2008) and the Ejina Banner is a predominantly pastoral region, raising 62% of the livestock (e.g., approximately 178,000 sheep livestock unit) of the whole area (Gala et al., 2007). Heavy pastoral pressure threatens the herb layer and the vegetation community structure, especially in those communities located near the road, which makes them more easily accessed by the livestock and herdsman. Thus, being located near a road results in a high risk of disturbance by anthropogenic activities, which can affect the growth of Tamarix spp., especially the crown diameter. Soil properties, which explained 14% of the variance in vegetation, 9 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Heihe River, the farmland also largely expanded (Fig. A.2), especially for cantaloupe cultivation, which is characterized by high water consumption. The conveyance water, which was intended to be used for ecological recovery, is instead used for the expansion of irrigation farmlands. The increasing water requirements of the expanding farmland will results in drought stress and threatens species survival, especially for herb species, which have low drought tolerance but contribute significantly to the community diversity (Thuiller et al., 2006). Thus, in order to maintain ecosystem resilience and avoid decrease in community diversity, measures such as controlling cantaloupe planting and developing water-saving irrigation practice are suggested, particularly around the high diversity communities (e.g., Community II, III and V). correlated with the soil moisture at a depth of 30–100 cm, and this explained only 3% of the variance in vegetation (Tables 2 and 4). This could be due to the variation in water-use strategies in different Tamarix spp. habitats. When located close to the riverbank, Tamarix spp. mainly utilizes a combination of river water and soil water; as the distance from the riverbank increased, Tamarix spp. mainly replies on the groundwater (Fu et al., 2012). This change in water resource is likely to explain the weak relationship between soil moisture and variance in vegetation, as well as between soil moisture and Tamarix spp. characteristics. In addition, although previous studies showed that community diversity decreased with the weakening of water influence, the species richness and evenness were positively correlated with the distance from river. This is likely to result from the variation in vegetation–environment interactions with the distance from river. As the distance from river increased, the interaction between Tamarix spp. and wind erosion became more prominent, leading to finer particles (e.g., clay) being deposited on the surface soil, benefitting the growth of diverse species (D'Odorico et al., 2007). In addition, Tamarix spp. could also benefit the growth of shallow-rooted species (e.g. herbs) by redistributing the deep soil water to the shallow layer (via hydraulic uplift) in drought environments (e.g., Community V) (Yin et al., 2010). Thus, this mutualistic effect of Tamarix spp. as well as vegetation–environment interactions in extreme environments, could result in increased community richness and evenness even in communities located further away from river. 5. Conclusion Based on the variation of Tamarix spp. characteristics, we found that Tamarix spp. characteristics formed significant pattern along heterogeneous habitats investigated in this study. With the increase in height and crown size, Tamarix spp. coverage and drought tolerance increased, whereas community diversity displayed a bimodal pattern due to the variation of species-environment interactions in the resource limited area. Among environmental factors, accessibility of anthropogenic disturbance and soil properties, which explained 42% vegetation variance of Tamarix spp. and accounted for 75% of the total explanation, were deemed to be key environmental factors that affecting the variation in Tamarix spp. community characteristics. The variation of Tamarix spp. community characteristics showed a weak relationship with water availability because of the ability of Tamarix spp. to draw up water from the deep soil and the vegetation-environment interactions in extreme environments. Based on these results, management such as establishing a protective zone, could be put in place to protect the vegetation structure and secure the ecosystem services provided by these communities (e.g. carbon storage, firewood, and sand fixation) against aggravated anthropogenic disturbance. In addition, management that reduce drought stress (e.g. controlling cantaloupe planting and developing water-saving irrigation practice) were also recommended to help secure the vegetation structure of Tamarix spp. and maintain the resilience of riparian plant communities under intensive human activities. 4.3. Applications for management Our study illustrates that vegetation structure of Tamarix spp. (e.g., height and crown diameter), contributing to the ecosystem service provisions (e.g., firewood provision and sand fixation) (Hierro et al., 2000), were negatively influenced by the accessibility of anthropogenic disturbance and soil physical properties. However, with ecological recovery, the seasonal tourism, which is concentrated between September and October, has significantly increased (Hochmuth et al., 2014). Exposure to intense trampling by tourists might potentially reduce the water holding ability in soils and thus negatively affecting the vegetation structure of Tamarix spp. as well as the ecosystem services provision (e.g., firewood provision and sand fixation). Among five communities, Communities I to IV were located nearest to the road, which makes these communities more likely to be affected by the diverse effects of intensive tourism. Thus, for Community I to IV, management measures, such as establishing protective zones near the roads is recommended in future ecological restoration to secure ecosystem service provisions in this desert riparian zone. Riparian plant communities containing Tamarix spp. also strongly influence the resistance of ecosystem against disturbance in hyperarid zone. Communities characterized by high diversity (e.g., Community II, III and V) could maintain ecosystem functions during stress or disturbance due to the complementary of function traits and ecological redundancy (Isbell et al., 2015). Although ecological water conveyance has significantly improved the water condition in the lower reaches of Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 91425301) and State Key Laboratory of Earth Surface Processes and Resource Ecology (No. 2017-FX-01(2)). We are grateful for reviews from the editor, Dan Binkley, guest editors and two anonymous reviewers, who significantly improved the quality of this article. We are thankful to the Alashan Research Station of Cold and Arid Region Environment and Engineering Research Institution, Chinese Academy of Sciences for their support and contributions to the fieldwork. Appendix A A.1. The spatial location of five classified riparian plant communities that contain Tamarix spp. in the lower reaches of Heihe In the lower reaches of Heihe, 38 sampling sites were classified into five types of riparian plant communities that contain Tamarix spp. based on Two-way Indicator Species Analysis (TWINSPAN) and Detrended Correspondence Analysis (DCA). The spatial location of each community was displayed in different symbol based on the classification result (Fig. A.1). 10 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Fig. A.1. The spatial location of five classified riparian plant communities that contain Tamarix spp. in the lower reaches of Heihe. A.2. The temporal variation of sown area in the lower reaches of Heihe during 2000 to 2015 The temporal variation of sown area in the lower reaches of Heihe during 2000 to 2015 is displayed in the Fig. A.2 based on the statistical yearbook of Ejina Banner obtained from National Library, China. The total sown area significantly increased since the implement of ecological water conveyance project in 2000, rising from 1514 hm2 in 2000 to 5136 hm2 in 2015. Among different types of crops, the cantaloupe characterized with high economic value, has been widely planted and accounted for 79.6% of the total sown area after 15 years’ ecological water conveyance. Fig. A.2. The temporal variation of sown area in the lower reaches of Heihe during 2000 to 2015. 11 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. A.3. The characteristics of environmental factors in the five classified riparian plant communities that contain Tamarix spp. Table A.1 summarizes the characteristics of environmental factors in five classified communities. In all communities, the soil moisture increased with the depth of soil layer, while the coefficient of variation value decreased along the depth with the highest variation at the surface layer (88.93%). Community III had the highest soil moisture in all layers, while Community II and V characterized with relative low soil moisture among the five communities. Bulk density and soil particle composition were used to depict the soil physical property and all of these indices generally did not differ significantly among the five communities. Characterized with the highest proportion of fine soil particle, the Community IV had the lowest bulk density, while Community II and III, containing certain percentage of gravel, had higher bulk density. Compared with soil particle composition, there was relative low degree of variability (29.92–37.63%) in soil nutrient and there was no significantly difference in soil nutrient among five communities. Total salt content, did not differ significantly among five communities, and reached the highest value in Community III. The distance from river represents the impact of river on each community. Community I, located nearest to the river bank (511.67 m), differed significantly from other communities. Community III, IV, V located in the same range of distance (1500–2000 m) from the river channel, while Community II located farthest from the river channel (2396.67 m). The distance from road depicted the accessibility of human disturbance to certain communities. Community II, III, IV located near to the road, with the similar distance from the road (217.50–255.00 m). This was then followed by the Community I (578.33 m) and the Community V distributed remotest (861.67 m) from the road. Table A.1 The characteristics of environmental factors in the five classified riparian plant communities that contain Tamarix spp. Factor SWC1 (g/g) SWC2 (g/g) SWC3 (g/g) Clay (%) Silt (%) Sand (%) Gravel (%) BD (g/cm3) SOM (g/kg) TN (g/kg) TP (g/kg) TS (%) Distance from river (m) Distance from road (m) Community CV(%) I II III IV V 0.06 ± 0.01a 0.10 ± 0.01a 0.14 ± 0.01a 5.50 ± 1.63a 18.00 ± 4.77a 76.67 ± 6.10a 0 1.28 ± 0.03a 7.52 ± 0.46a 0.52 ± 0.02a 0.48 ± 0.03a 1.02 ± 0.14a 511.67 ± 90.64a 578.33 ± 58.24a 0.06 ± 0.01a 0.08 ± 0.01a 0.11 ± 0.01a 2.17 ± 0.56a 23.53 ± 6.12a 67.75 ± 5.91a 5.72 ± 1.10a 1.39 ± 0.04a 7.15 ± 0.43a 0.53 ± 0.03a 0.40 ± 0.02a 0.77 ± 0.19a 2396.67 ± 156.97b 255 ± 40.02a 0.08 ± 0.01a 0.11 ± 0.01a 0.23 ± 0.01b 7.81 ± 1.60a 20.69 ± 3.64a 65.90 ± 4.51a 5.63 ± 0.89a 1.41 ± 0.05a 5.77 ± 0.20a 0.48 ± 0.02a 0.51 ± 0.01a 1.93 ± 0.29a 1943.75 ± 90.41b 238.75 ± 24.06a 0.08 ± 0.005a 0.13 ± 0.005a 0.17 ± 0.01a 6.88 ± 0.68a 37.39 ± 3.16a 55.73 ± 3.59a 0 1.19 ± 0.01a 7.09 ± 0.27a 0.47 ± 0.02a 0.43 ± 0.01a 1.12 ± 0.07a 1624.17 ± 80.31b 217.50 ± 16.10a 0.08 ± 0.01a 0.09 ± 0.01a 0.16 ± 0.01ab 5.67 ± 0.83a 44.42 ± 6.20a 49.92 ± 6.83a 0 1.34 ± 0.04a 8.09 ± 0.56a 0.48 ± 0.04a 0.50 ± 0.02a 1.40 ± 0.26a 2001.67 ± 114.85b 861.67 ± 134.02b 88.93 62.85 47.42 140.66 128.19 62.83 121.16 17.76 37.63 36.33 29.92 109.29 55.31 83.47 Notes: Values represent mean ± SE. Values with the same letter are not significantly different at the 0.05 significance level (LSD test), while values of an index followed by different letters differ significantly at the 0.05 significance level (LSD test). CV, coefficient of variation; SWC1, 0–30 cm soil moisture; SWC2, 30–100 cm soil moisture; SWC3, 100–200 cm soil moisture; BD, bulk density; SOM, soil organic matter; TN, total nitrogen; TP, total phosphorus; TS, total salt content. Appendix B. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foreco.2017.10.003. deep soil moisture under different vegetation types and influencing factors in a watershed of the Loess Plateau, China. Hydrol. Earth Syst. Sci. 20, 3309–3323. http:// dx.doi.org/10.5194/hess-20-3309-2016. Flato, G., Marotzke, J., Abiodun, B., Braconnot, P., Chou, S.C., Collins, W.J., Cox, P., Driouech, F., Emori, S., Eyring, V., Forest C., 2013. Evaluation of Climate Models. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Fu, A., Li, W., Chen, Y., 2012. The threshold of soil moisture and salinity influencing the growth of Populus euphratica, and Tamarix ramosissima, in the extremely arid region. Environ. Earth Sci. 66 (8), 2519–2529. Fu, A., Chen, Y., Li, W., 2014. Water use strategies of the desert riparian forest plant community in the lower reaches of Heihe River Basin, China. Sci. China Earth Sci. 57, 1293–1305. Gala, X., Cao, C., Jiao, C., Jiao, W., 2007. Ecological environment and sustainable development of animal husbandry in Ejina, Inner Mongolia. Anim. Husbandry Feed Sci. 28, 95–96. Gao, X.M., Huang, J.H., Wan, S.Q., 1997. Ecological studies on the plant community succession on the abandoned cropland in Taibaishan, Qinling Mountains. I. The community a diversity feature of the successional series. Acta Ecol. Sin. 17, 619–625. Hao, X., Li, W., Huang, X., Zhu, C., Ma, J., 2010. Assessment of the groundwater threshold of desert riparian forest vegetation along the middle and lower reaches of the Tarim River, China. Hydrol. Processes 24 (2), 178–186. Hierro, J.L., Branch, L.C., Villarreal, D., Clark, K.L., 2000. Predictive equations for biomass and fuel characteristics of argentine shrubs. J. Range Manage. 53, 617–621. Hochmuth, H., Thevs, N., He, P., 2014. Water allocation and water consumption of irrigation agriculture and natural vegetation in the Heihe River watershed, NW China. Environ. Earth Sci. 73, 5269–5279. Imada, S., Taniguchi, T., Acharya, K., Yamanaka, N., 2013. Vertical distribution of fine References Akiyama, T., Sakai, A., Yamazaki, Y., Wang, G., Fujita, K., Nakawo, M., Kubota, J., Konagaya, Y., 2007. Surfacewater-groundwater interaction in the Heihe River basin, Northwestern China. Bull. Glaciol. Res. 24, 87–94. Anyia, A.O., Herzog, H., 2004. Water-use efficiency, leaf area and leaf gas exchange of cowpeas under mid-season drought. Eur. J. Agron. 20, 327–339. Black, C.A., 1968. Soil-plant Relationships, second ed. John Wily & Sons, New York. Chen, X., Duan, Z., Luo, T., 2014. Changes in soil quality in the critical area of desertification surrounding the Ejina Oasis, Northern China. Environ. Earth Sci. 72, 2643–2654. Cheng, H., 2007. The Studies of the Function of Resisting Wind, Stabilizing Sand and Fixing Carbon of Vegetation in the Middle and Lower Reaches of Tarim River. Xinjiang Agriculture University. D'Odorico, P., Caylor, K., Okin, G.S., Scanlon, T.M., 2007. On soil moisture–vegetation feedbacks and their possible effects on the dynamics of dryland ecosystems. J. Geophys. Res. Atmos. 112, 231–247. Daryanto, S., Eldridge, D.J., Wang, L., 2013. Spatial patterns of infiltration vary with disturbance in a shrub-encroached woodland. Geomorphology 194, 57–64. DiTomaso, J.M., 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed Technol. 12, 326–336. Dı́Az, S., Cabido, M., 2001. Vive la différence: plant functional diversity matters to ecosystem processes. Trends Ecol. Evol. 16, 646–655. Ding, J., Zhao, W., 2016. Comparing Chinese and international studies of riparian forests: a bibliometric survey (1981–2014). Acta Ecol. Sin. 36, 377–385. http://dx.doi.org/ 10.1016/j.chnaes.2016.05.004. Fang, X., Zhao, W., Wang, L., Feng, Q., Ding, J., Liu, Y., Zhang, X., 2016. Variations of 12 Forest Ecology and Management xxx (xxxx) xxx–xxx J. Ding et al. Ruiz, M.Á., Lugo, A.E., 2012. Landscape effects on structure and species composition of tabonuco forests in Puerto Rico: implications for conservation. For. Ecol. Manage. 266, 138–147. Stirzaker, R.J., Passioura, J.B., Wilms, Y., 1996. Soil structure and plant growth: Impact of bulk density and biopores. Plant Soil 185, 151–162. Stromberg, J.C., Chew, M.K., Nagler, P.L., Glenn, E.P., 2009. Changing perceptions of change: the role of scientists in Tamarix and river management. Restor. Ecol. 17, 177–186. Thiele, T., Jeltsch, F., Blaum, N., 2008. Importance of woody vegetation for foraging site selection in the Southern Pied Babbler (Turdoides bicolor) under two different land use regimes. J. Arid Environ. 72, 471–482. Thuiller, W., Midgley, G.F., Hughes, G.O., Bomhard, B., Drew, G., Rutherford, M.C., Woodward, F., 2006. Endemic species and ecosystem sensitivity to climate change in Namibia. Glob. Change Biol. 12, 759–776. Wang, D.P., Li, J.H., Chen, Y.L., Liu, R.Y., 2010. The soil composition of Pteroceltis tatarinowii natural community and its effect on the growth of plant. Ecol. Environ. Sci. 19, 2318–2324. Wen, X., Wu, Y., Su, J., Zhang, Y., Liu, F., 2005. Hydrochemical characteristics and salinity of groundwater in the Ejina Basin, Northwestern China. Environ. Geol. 48, 665–675. Wheater, C.P., Bell, J.R., Cook, P.A., 2011. Practical Field Ecology: A Project Guide. John Wiley & Sons, Chichester. Whitcraft, C.R., Talley, D.M., Crooks, J.A., Boland, J., Gaskin, J., 2007. Invasion of tamarisk (Tamarix spp.) in a southern California salt marsh. Biol. Invasions 9, 875–879. Xie, Y., Qiu, K., Xu, D., Shi, X., Qi, T., Pott, R., 2015. Spatial heterogeneity of soil and vegetation characteristics and soil-vegetation relationships along an ecotone in Southern Mu Us Sandy Land, China. J. Soils Sediments 15, 1584–1601. Xu, S.S., Sun, G.J., Liu, H.M., Gong, X.P., Hao, Y.Y., Zhang, L.X., 2011. The relationship between riaprian vegetation and environmental factors in Heihe river basin. Acta Ecol. Sin. 09, 2421–2429. Yang, W., Zhang, D., Yin, L., Zhang, L., 2002. Distribution and cluster analysis on the similarity of the Tamarix Communities in Xinjiang. Arid Zone Research. 19, 6–11. Yang, Y.S., 2013. Study on ecosystem service values based on land use change in Ejina Oasis. Chin. Agric. Sci. Bull. 29, 218–224. Yates, F., 1981. Sampling Methods for Censuses and Surveys, fourth ed. Charles Griffin & Company, London. Yin, C.H., Feng, G., Zhang, F.S., Tian, C.Y., Tang, C.X., 2010. Enrichment of soil fertility and salinity by tamarisk in saline soils on the northern edge of the Taklamakan Desert. Agric. Water Manage. 97, 1978–1986. York, P., Evangelista, P., Kumar, S., Graham, J., Flather, C., Stohlgren, T., 2011. A habitat overlap analysis derived from maxent for tamarisk and the south-western willow flycatcher. Front. Earth Sci. 5, 120–129. Zeng, S.L., Zhang, T.T., Gao, Y., Ouyang, Z.T., Chen, J.K., Li, B., Zhao, B., 2011. Effects of road age and distance on plant biodiversity: a case study in the Yellow River Delta of China. Plant Ecol. 212, 1213–1229. Zenner, E.K., Olszewski, S.L., Palik, B.J., Kastendick, D.N., Peck, J.L.E., Blinn, C.R., 2012. Riparian vegetation response to gradients in residual basal area with harvesting treatment and distance to stream. For. Ecol. Manage. 283, 66–76. Zhang, D.Y., Yin, L.K., Pan, B.R., 2002. Biological and ecological characteristics of Tamarix L. and its effect on the ecological environment. Sci. China: Earth Sci. 45, 18–22. Zhang, J.T., Dong, Y., 2010. Factors affecting species diversity of plant communities and the restoration process in the loess area of China. Ecol. Eng. 36, 345–350. Zhao, L., Wang, L., Cernusak, L.A., Liu, X., Xiao, H., Zhou, M., Zhang, S., 2016. Significant difference in hydrogen isotope composition between xylem and tissue water in Populus euphratica. Plant Cell Environ. 39 (8), 1848–1857. http://dx.doi.org/10. 1111/pce.12753. roots of Tamarix ramosissima in an arid region of southern Nevada. J. Arid Environ. 92, 46–52. Isbell, F., Craven, D., Connolly, J., Loreau, M., Schmid, B., Beierkuhnlein, C., Bezemer, T.M., Bonin, C., Bruelheide, H., Luca, E., Ebeling, A., 2015. Biodiversity increases the resistance of ecosystem productivity to climate extremes. Nature 526, 574–577. ISSCAS (Institute of Soil Sciences, Chinese Academy of Sciences), 1978. Physical and Chemical Analysis Methods of Soils. Shanghai Science Technology Press, Shanghai. Lei, W.N., Wen, Z.M., 2008. Research on soil erosion vegetation factor index based on community structure. J. Soil Water Conserv. 22 (5), 68–77. Lü, Y., Ma, Z., Zhao, Z., Sun, F., Fu, B., 2014. Effects of land use change on soil carbon storage and water consumption in an oasis-desert ecotone. Environ. Manage. 53, 1066–1076. Lavorel, S., Colloff, M.J., Mcintyre, S., Doherty, M.D., Murphy, H.T., Metcalfe, D.J., Dunlop, M., Williams, R.J., Wise, R.M., Williams, K.J., 2015. Ecological mechanisms underpinning climate adaptation services. Glob. Change Biol. 21, 12–31. Li, J., Zhao, C., Zhu, H., Li, Y., Wang, F., 2007. Effect of plant species on shrub fertile island at an oasis-desert ecotone in the South Junggar Basin, China. J. Arid Environ. 71, 350–361. Liu, C.R., Ma, K.P., Yu, S.L., Wang, W., 1997. Plant community diversity in Dongling mountain, Beijing, China: effect of sample size on diversity measures. Acta Ecol. Sin. 17, 584–592. Liu, G.S., 1996. Soil Physical and Chemical Analysis and Description of Soil Profiles. Chinese Standards Press, Beijing. Liu, S., Deng, L., Chen, L., Li, J., Dong, S., Zhao, H., 2014a. Landscape network approach to assess ecological impacts of road projects on biological conservation. Chin. Geogra. Sci. 24, 5–14. http://dx.doi.org/10.1007/s11769-014-0651-z. Liu, S., Dong, Y., Deng, L., Liu, Q., Zhao, H., Dong, S., 2014b. Forest fragmentation and landscape connectivity change associated with road network extension and city expansion: a case study in the Lancang River Valley. Ecol. Ind. 36, 160–168. http://dx. doi.org/10.1016/j.ecolind.2013.07.018. Long, X., Sun, Z., Zhou, A., Yu, S., Cao, H., 2014. Using 18O stable isotopes to trace the water sources of riparian Tamarix in middle reaches of Heihe River. J. Arid Land Resour. Environ.. 28, 150–155. Lü, Y., Fu, B., Wei, W., Yu, X., Sun, R., 2011. Major ecosystems in China: dynamics and challenges for sustainable management. Environ. Manage. 48, 13–27. http://dx.doi. org/10.1007/s00267-011-9684-6. Ma, Q., Wang, J., Li, X., Zhu, S., Liu, H., Zhan, K., 2009. Long-term changes of Tamarixvegetation in the oasis-desert ecotone and its driving factors: implication for dryland management. Environ. Earth Sci. 34, 119–148. Marshall, K.E., Baltzer, J.L., 2015. Decreased competitive interactions drive a reverse species richness latitudinal gradient in subarctic forests. Ecology 96, 461–470. Naiman, R.J., Décamps, H., 1997. The ecology of interfaces: riparian zones. Annu. Rev. Ecol. Syst. 28, 621–658. Natale, E., Zalba, S.M., Oggero, A., Reinoso, H., 2010. Establishment of Tamarix ramosissima under different conditions of salinity and water availability: implications for its management as an invasive species. J. Arid Environ. 74, 1399–1407. Odum, E.P., Barrett, G.W., 1971. Fundamentals of ecology, vol. 3 Saunders, Philadelphia. Oksanen, J., Minchin, P.R., 2002. Continuum theory revisited: what shape are species responses along ecological gradients? Ecol. Model. 157, 119–129. Ravi, S., Breshears, D.D., Huxman, T.E., D'Odorico, P., 2010. Land degradation in drylands: interactions among hydrologic–aeolian erosion and vegetation dynamics. Geomorphology 116, 236–245. http://dx.doi.org/10.1016/j.geomorph.2009.11.023. Rosenthal, D.M., Ludwig, F., Donovan, L.A., 2005. Plant responses to an edaphic gradient across an active sand dune/desert boundary in the great basin desert. Int. J. Plant Sci. 166, 247–255 doi:urn:nbn:nl:ui:32-354059. Ross, A.M., Johnson, G., Gibbs, J.P., 2016. Spruce grouse decline in maturing lowland boreal forests of New York. For. Ecol. Manage. 359, 118–125. 13
1/--страниц