1. Introduction
Spatial distribution of surface soil water content (SWC) is significant
to the interaction between vegetation and soil moisture in semi-arid and
arid areas (Brocca et al., 2007; Zhao et al., 2020), and reveals how
hydrological dynamics are linked to ecological patterns and processes
(Grayson et al., 1998; Robinson et al., 2008; Seneviratne et al., 2010).
In fact, the amount and spatial distribution of SWC directly determine
the patterns and population abundance of vegetation (A et al., 2019;
Brocca et al., 2010). In semi-arid regions, the presence and absence of
seasonal vegetation showed highest correlation with SWC (Gomez-Plaza et
al., 2000), while areas with little vegetation were mostly affected by
local controls such as slope and soil texture (Gómez-Plaza et al.,
2001).
The Qilian Mountains in northwestern China is an extensive area with
semi- and arid climate. The area is also a source of numerous inland
rivers which support extensive irrigated agriculture in the Hexi River
Corridor. Several afforestation projects have been implemented in the
area to conserve soil and water, control desertification, and to produce
wood (Wang et al., 2010), P. crassifolia is the main constructive
species in the Qilian Mountains (Zhao et al., 2011). Therefore, planting
and management of P. crassifoflia forests is likely to have a
profound impact on the stability and development of the Hexi River
corridor ecosystem (Baroni et al., 2013; Vereecken et al., 2008) due to
increased demands on the SWC.
The spatial pattern of SWC is driven by many environmental factors,
including soil type, topography, vegetation, and the quantity of local
precipitation (Brocca et al., 2007). Liu et al. (2013) analyzed
vegetation patterns and hydraulic properties of surface soils in an
alpine catchment, and showed that the surface SWC pattern was controlled
mainly by terrain-related processes. In a loess hilly region of China,
SWC determines a self-organized vegetation pattern in arid environments,
while the reverse was not true and vegetation patterns did not greatly
affect SWC (Sun et al., 2014).
At a small spatial scale, heterogeneity in land use, vegetation type,
and soil characteristics are the dominant factors in SWC dynamics (Zhao
et al., 2017; Zucco et al., 2014). SWC is also controlled by antecedent
moisture conditions of the sampling plot (Famiglietti et al., 1998; Zhao
et al., 2011; Zucco et al., 2014). Precipitation was the control factor
in vegetation change in arid zones, and an observed decrease in SWC led
to a reduction in evapotranspiration (Guan et al., 2020). According to
Jian et al. (2015), plant water uptake in semi-arid regions occurs
mostly through night-time root uptake and day-time evapotranspiration;
this indicates that SWC in the main zone of plant root systems is
critical to the physiological activities and growth of surface
vegetation in water-limited ecosystems. Thus, SWC supply in arid regions
is the main limiting factor for land-cover and vegetation succession,
and any tradeoff or change in hydrological processes may cause a
significant disturbance to the ecosystem and result in a new landscape
pattern (Brocca et al., 2007; Cosh et al., 2008; Sala et al., 2014; Wang
et al., 1999; Wang et al., 2011).
Forest planting is extensive in China; it has been extensively promoted
during the last 20 years with the “Grain for Green” projects which
aimed to convert croplands to forests, shrublands and grasslands to
prevent ecosystem degradation (Cao et al., 2011; Fu et al., 2011).
Ecosystem restoration in the Qilian Mountains is reaching new levels
with a plan for the “Qilian Mountain National Forest Park”, in which
5.02 million ha are being restored (Liu et al., 2008; McVicar et al.,
2010). However, limited information is available on the response of SWC
to shifts in vegetation types in arid ecosystems, and, to date, the
hydraulic interaction between planted forests and SWC is not clear. A
greater understanding of this interaction is needed and of the spatial
evolution of SWC under different landcover types to ensure success of
ecosystem restoration.
Here, we used a geostatistical spatial analysis of the pattern of
surface SWC in different-aged P. crassifolia forest, native
grassland, shrubland, and mixed landcover in the region; here,
grassland, shrubland and mixed landcover were considered to be the
successive stages along the development of the local ecosystem into
forestland.
We aimed to advance our understanding of forest plantation in arid areas
by providing data support for regional forest-water relationship
research. Our main objectives were to: (a) quantify the evolution trend
of surface SWC during the early stages of plantation development, (b)
characterize the spatial heterogeneity in SWC in reforested areas of
different ages, and (c) determine the best sampling methods for SWC for
spatial pattern analyses in semi- and arid regions.