Biologia 69/11: 1539—1554, 2014
Section Botany
DOI: 10.2478/s11756-014-0466-5
Sediment content and chemical properties of water runoff
on biocrusts in drylands
Roberto Lázaro1 & Juan Luis Mora2
1
Estación Experimental de Zonas Áridas (CSIC). Carretera Sacramento, s/n; La Cañada de S. Urbano. Almería 04120.
Spain; e-mail; [email protected]
2
Departamento de Ciencias Agrarias y del Medio Natural. Universidad de Zaragoza. C/ Miguel Servet 177. Zaragoza 50013.
Spain; e-mail: [email protected]
Abstract: In drylands, water erosion can be a process with important economic and ecological implications, and is very
dependent on the soil surface cover. There is broad agreement that biocrusts protect the soil from erosion in a wide range
of circumstances. However, there is little information available on the effect of rain and biocrust types on this protective
capacity and there is particularly very little knowledge on the erosive effects of runoff on biocrusts, which are expected to
be larger in larger drainage areas, on the resistance of biocrusts to the combined effect of raindrops plus runoff flow and
on the solute mobilisation by runoff in biocrusts. To answer these questions, we performed 96 rainfall-simulationin situ
factorial experiments, including two biocrust types (cyanobacteria and lichens), three rain types (42, 63 and 77 mm h−1 ,
always 20 min rain), four plot lengths (1, 2, 3 and 4 m long) and four replicates. In each experiment, runoff volume
was measured and a runoff sample was taken to determine (i) the amount of dry matter in runoff, (ii) the amount of
organic matter among the dry matter, (iii) the electrical conductivity, pH and alkalinity in runoff water. The main findings
were: biocrusts strongly protected soil against water erosion, even under the most erosive conditions, and the protection
increased with the successional development. Biocrusts were very resistant to the impact of raindrops and also to runoff
flow, although an emergent hypothesis arose: under the most erosive conditions, a threshold of erodibility could be reached
at the cyanobacterial biocrust. The lichen crust also protected the soil against the removal of soil soluble substances. The
development of a biocrust could change the chemical composition of the solutes in runoff.
Key words: biological soil crust; cyanobacteria; rainfall simulation; solute mobilisation; terricolous lichens; water erosion;
Tabernas Desert
Introduction
Water erosion and soil loss can represent serious problems with important economic and ecological implications in drylands, mainly in developing countries (Rapp
1986), because sparse vegetation has a limited capacity
to deal with erosive processes. Water erosion is produced by raindrop impact and by surface and subsurface runoff, and it depends on rainfall erosivity and
on soil erodibility, which is highly controlled (among
other factors) by the features of the soil surface cover.
Drylands in the sense of water-limited lands (i.e., areas where precipitation is not always enough to cover
potential evapotranspiration), include hyper-arid, arid,
semiarid and dry-subhumid ecosystems, which occupy
about 45% of emerged land, contain about 30% of total
carbon stored in biomass (Allen-Diaz et al. 1996) and
support about 50% of the world’s livestock (Puigdefábregas & Pugnaire 1999). In dry-subhumid and semiarid areas, even sometimes in arid ones, water erosion
can be very relevant.
Part of the soil water erosion is a direct result
of raindrops (splash erosion), but another, often the
c 2014 Institute of Botany, Slovak Academy of Sciences
most important, is produced by surface runoff (rills,
gullies) or sub-surface processes (piping, mass movements). Therefore, water erosion depends on rainfall
erosivity (mainly determined by the amount and intensity) (Angulo-Martinez & Begueria 2009; Wei et al.
2010), soil erodibility, which is controlled by the soil
texture, the amount and stability of the aggregates and
the organic matter content (De Baets et al. 2006; SoléBenet et al. 1997) and land cover (cover, size and density of the vascular plants, of biocrusts, stones, litter)
(Boix-Fayos et al. 2007; Zavala et al. 2009; Cantón et al.
2011). Erosion also depends on topography (including
surface roughness, as well as the slope angle and the
size of the drainage areas) and antecedent soil moisture (Bowker et al. 2008). The size of the drainage area
appears to be crucial, as larger amounts of runoff are
expected from large areas, particularly where runoff is
concentrated, and a positive correlation between runoff
and erosion has been reported (Herrick et al. 2010; Wei
et al. 2010). Boix-Fayos et al. (2007) obtained larger
sediment concentrations at coarser scales (1.32 g L−1 ,
catchment; 0.30 g L−1 , 30 m2 ; 0.17 g L−1 , 1 m2 scales),
despite the decrease in runoff coefficients. On the other
Unauthenticated
Download Date | 6/18/17 6:30 PM
1540
hand, the larger the degree of erosion is, the greater is
the runoff density, and therefore, the greater its erosive
and transport power.
The hydrological behaviour of pure soil surfaces at
the micro-plot scale is insufficient to understand the
hydrological behaviour at the slope or small catchment
scale (Cantón et al. 2002); because runoff coefficients
considerable decrease as drainage area increase (compare Calvo-Cases et al. 2003 with Yair et al. 1980, and
Cantón et al. 2001; Boix-Fayos et al. 2007). This change
in runoff coefficients with scale occurs because of the
limited slope length that is travelled by runoff (Kidron
2011). This must have consequences on erosion, and it
is essential to understand these under specific circumstances, to up-scale the erosion results from rainfall simulation on microplots to slopes or small catchments.
The erodibility of biological soil crusts (hereafter,
biocrusts) is not well known: Although differences according to the biocrust composition are expected (Belnap 2006; Chamizo et al. 2012), the effects of rain
type and particularly the effect of progressively larger
slope length on water erosion in biocrusts is far from
well-known. Besides, biocrusts are considered as runoff
sources and to understand to what extent they decrease
erosion in relation to the runoff connectivity will provide insights towards the understanding of sediment dynamics at the slope scale in semiarid areas. Little available information is also known about the resistance of
different types of biocrust to the drop impact (but see
Bevan 2009; Rodriguez-Caballero et al. 2013; Zhao et
al. 2014), and less still is known concerning their resistance to runoff flow. This is important because although biocrust resist drop impacts, if it becomes frequently broken by runoff flows, then its ability to protect the soil from water erosion will be limited. Besides,
if runoff increases with slope length, this might increase
biocrust detachment. Furthermore, nothing or very little is known about solute mobilisation by runoff on hills
lopes with biocrust cover.
In addition to the removal and transportation of
sediments, another form of erosion results from the dissolution of soluble soil minerals which can be driven out
of the system by runoff. This chemical erosion is more
uniform in space and time than soil-particle erosion
and is often considered negligible in drylands (Alekin
& Brazhnikova 1968; Meybeck 1976; Yair et al. 1991).
However, studies on the chemistry of the solutes mobilised by runoff in drylands can be used to complement
volumetric measurements of runoff to provide greater
insight into the paths taken by the various runoff components (Chandler & Bisogni 1999), which is essential
to understand the chemical properties of flood- and
ground-water and of soils and their effects on vegetation
(Yair et al. 1991).
Biocrusts are communities of cyanobacteria, fungi,
algae, lichens and mosses on the soil surface and
within its few upper mm (Belnap & Lange 2003).
They are globally common in drylands and occupy
sites that are inhospitable for vascular plants (Büdel
2003). Biocrusts play several relevant ecological roles
R. Lázaro & J.L. Mora
and are sensitive to climatic change (Maestre et al.
2013); they affect hydrology and erosion (Alexander &
Calvo-Cases 1990; Eldridge & Greene 1994; Yair 2003),
physical–chemical soil properties (Belnap & Gardner
1993; Castillo-Monroy et al. 2010) and plant recruitment (Harper & St. Clair 1985; Pendleton et al. 2003).
Biocrusts are pioneers (Lazaro et al. 2008), can be used
to restore/stabilise soil (Bowker 2007) and have indirect stabilisation effects by facilitating plant growth by
improving soil fertility (Pendleton et al. 2003). In the
Iberian semiarid southeast, biocrusts are often dominated by terricolous lichens or by cyanobacteria and
can locally cover large areas (Lázaro et al. 2000), constituting a natural stabilisation mechanism (Alexander
et al. 1994; Lázaro et al. 2008) and are widespread
on undisturbed sites (Lázaro et al. 2008). Some studies have found that biocrusts increase infiltration (Eldridge 1993; Pérez 1997), runoff (Bromley et al. 1997;
Eldridge et al. 2000) or are neutral (Eldridge et al.
1997). According to Warren (2003), this controversy
is due to the influence of soil texture. Belnap (2006)
suggested that differences in flora and morphology also
contribute to these differences, which was confirmed
by Chamizo et al. (2012). According to the literature,
these biocrust have runoff coefficients comparable with
bare soil and larger than those from the rest of surface
classes (Alexander & Calvo-Cases 1990; Cantón et al.
2011; Li et al. 2005; Solé-Benet et al. 1997). Despite
these high runoff coefficients, evidence for the ability
of biocrusts to reduce erosion was found long ago, as
reported by Booth (1941), who reviewed studies since
1888 and proposed a role for soil algae in controlling
erosion. Currently, there is widespread agreement on
the role of biocrusts in protecting the soil against erosion (Bowker et al. 2008; Mücher et al. 1988; Alexander
& Calvo-Cases 1990; Eldridge & Greene 1994; Eldridge
1998; Eldridge 2003); biocrusts decrease water erosion
between 5 and 10-fold (Chamizo et al. 2012) and wind
erosion by 46-fold (Bu et al. 2013). On the other hand,
biocrusts represent a good model system (Bowker et al.
2010; Bowker et al. 2014) and the size of their organisms allows a series of plots with different lengths (the
smaller being large enough to include most of the community variation) to be treated with a portable rainfall
simulator.
Our main hypothesis was that water erosion will
increase with slope length only to a certain point, due
to the limited length of the paths travelled by runoff
(Kidron 2011) and due to the biocrust-driven surface
roughness (Chamizo et al. 2010). Greater roughness
leads to a larger tortuosity of the runoff path, as well
as more microdepressions to be filled by runoff, which
slow water and facilitates sediment deposition. Erosion
is expected to have a negative relationship with biocrust
successional development, and a positive relationship
with the amount and intensity of rainfall. The rate of
organic matter in dry matter transported by runoff is
expected to be larger in larger biocrust cover and under
higher rainfall, but is still expected to be low.
Our aim was to experimentally quantify the chemUnauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
1541
ical and particle erosion by water in situ on biocrusts
in a range of conditions, to investigate to what extent
biocrusts can act as a natural protection against erosion. The concrete objectives were: (i) to identify differences between the two main types of biocrusts in
erodibility and sensitivity to solute mobilisation; (ii)
to analyse differences in erosivity and dissolving power
among rainfall types, as well as their interaction with
soil erodibility and sensitivity to dissolution; (iii) to find
out to what extent the sediment yield and the amount of
mobilised solutes increase when slope-length increases;
(iv) to examine the relationship between both chemical
and particle erosion and runoff; and (v) to identify the
degree of detachment in each biocrust type and circumstance.
Methods
The experimental site was in the Tabernas Desert (Almería,
SE Spain), on deeply dissected Tortonian mudstone under semiarid Mediterranean climate conditions. The parent
material is mainly composed by silt-size (> 60%) gypsumcalcareous and siliceous particles; fine sand ranges from 20
to 35%, and clay ranges from 5% to 10%. The average mineralogy is muscovite 35%, calcite 20%, gypsum 20%, paragonite 10%, quartz 10%, chlorite + smectite 3%, and dolomite
2% (Cantón et al. 2003). Eroded landforms occupy a third
of the territory, another third is covered by short vascular
vegetation with biocrusts in the interspaces and the rest is
covered by biocrusts (Lázaro et al. 2000; Cantón et al. 2004).
Bevan (2009) found four biocrust communities, forming two
classes: Cyanobacterial crust includes the community dominated by cyanobacteria, with a sizable diversity of usually
small lichens such as Collema spp., Endocarpon pusillun,
and Fulgensia spp., which occupy the most sun-exposed
habitats. Lichen crust includes the other three communities, with different ratios of lichens with large, often thick
thalli, such as Squamarina lentigera, Diploschistes diacapsis,
Buellia zoharyi, Lepraria isidiata, etc. The Cyanobacterial
crust is primary a coloniser and is progressively replaced
by Lichen crusts over time (Lázaro et al. 2008) except in
the sunniest sites. The study area has a ‘Hortonian’ (Horton 1933) hydrological response (Solé-Benet et al. 1997) see
Calvo-Cases et al. (2014) for a general description.
We selected two areas: C, where Cyanobacterial crust
(C) dominates, and L, dominated byLichen crust (L). Both
biocrust classes coexist in each area, although in very different proportions. The roughness at C is smaller than in L
(although higher than on bare soil according to Chamizo et
al. 2012).
We used a 2-m-high simulator that provided rainfall
over a 1.9 m × 5.8 m area by 12 Hardy nozzles. An aluminium structure supported the pipe with the nozzles running parallel to the soil surface. The nozzle models 4680 15E,
4680 21E and 4680 25E provided rectangular 0.5 × 1.9 m
rainfall fans perpendicular to the pipe and three intensities
(42, 63 and 77 mm h−1 , respectively), at a water pressure of
1.0 bar. The characteristic drop-size diameter ranged from
0.5 to 1 mm; the mean kinetic energy was 1 J m−2 mm−1 ;
both were measured with a laser distrometer. A plastic
awning covered the simulator to prevent drops from being
blown away by the wind. These intensities, which were high,
but within the natural range (Lázaro et al. 2001), were determined by technical reasons, as a compromise between
Fig. 1. A block formed by eight plots. The upper four plots are,
from left to right, 1, 4, 2 and 3 meters long, respectively. The
lower four plots have the complementary lengths (4, 1, 3 and 2
m, respectively). Dotted lines represent the drainage pipes. Solid
lines represent the galvanized-iron sheets enclosing the plots.
drop size, intensity and water distribution among the available fan-type nozzles on the market.
The experimental design included three factors: biocrust type (Crust, C or L), rainfall type (Rain, small,
Unauthenticated
Download Date | 6/18/17 6:30 PM
1542
R. Lázaro & J.L. Mora
Fig. 2. Photographs providing information on the experimental setup: a) – Mounting the simulator over a way, to avoid damage of
the biocrust; b) – Testing the functioning of the nozzles before the experiment, over a way; c) – Covering the simulator, after put it
over a block of plots; d) – Detail of plots in area L showing the biocrust, the galvanized-iron sheets enclosing the plots, the pipes to
collect runoff and the cables of the soil moisture probes; e) – Running the experiment; runoff pipes finish into test tubes, other test
tubes are prepared to be changed; one person handle 4 test tubes and other the other 4; some bottles to collect runoff are prepared; a
third person is noting the readings of the test tubes; f) – A block of plots in area C after the experiments, showing all the plots, pipes
and cables of soil moisture probes and the wetted area.
medium or large; these did not represent rain intensity per
se because a minimal event duration was necessary to reach
a steady runoff and, on the other hand, the water consumption must be minimized by logistical reasons; therefore, all
experiments had the same duration and larger intensities
had a larger rainfall volume) and plot length (Length, 1, 2,
3 and 4 m). Four replicates were performed for each combination, totalling 96 experiments, which always started with
dry soil (the most common condition in the area). Plots were
0.4 m wide, enclosed by a galvanised-iron sheet buried up
to 15 cm and arranged in semi-blocks of four, with a plot of
each length randomly positioned in every semi-block. These
semi-blocks were paired, separated by 50 cm, and together
constituted a block (Fig. 1). Rainfall was sprinkled simultaneously over the eight plots within each block.
Different categories of Crust, Rain and Length were
tested on different plots. therefore, we established 48 plots
arranged in six 1.75 m × 5.5 m blocks for each crust type,
using two blocks for each rain type, placed in such a way
that they accounted for the possible range of slope angles of
the experimental area. The soil moisture was recorded during each experiment by ECH2 O probes (Decagon Devices,
USA) previously installed immediately below the biocrust
and by HOBO data loggers (Onset, USA), to ensure that
no significant difference in antecedent soil moisture existed
between experiments. From each experiment, we measured
runoff volume and collected (when it was possible) a 2-L
runoff sample for laboratory analyses.
Before the experiments, we checked that (i) the duration of the experiments (20 min) was sufficient to reach a
stable runoff rate, (ii) the infiltration on the upper semiblock did not increase moisture in the paired downslope
semi-block by sub-surface flow, and (iii) a 4-m plot length
was long enough to reach the maximum runoff and erosion
for these surfaces and rainfall intensities, or values near the
maxima, as in preliminary assays the absolute runoff from
a 4-m plot was greater than that from a 1-m plot, but was
much lower than four times the latter value.
Selected photographs show the experimental setup in
Figure 2.
The plots were located in areas with a representative
slope angle, biocrust and plant cover. The slope and the
cover of each biocrust type and vascular plants were taken
in each plot.
From the runoff samples, in the laboratory we measured the electrical conductivity (EC, mS/cm) and pH
(pH, dimensionless), and assessed the alkalinity (Alkali,
mequiv/L HCO−
3 ) by titration with sulphuric acid to an
end point of pH 4.5. We then extracted the total dry residue
from each runoff sample (Eros, g) and analysed these dry
residues for organic carbon by oxidation with potassium
dichromate in acid, followed by back titration with ferrousammonium sulphate in the presence of a ferroin indicator,
and we converted the organic carbon content to organic
matter (OM, %) by multiplying by 1.72. From the EC,
we derived the solutes in the runoff in g/L by multiplying
the mS/cm by the factor 0.64 according to Hanson (2006)
and, from the solutes in g/L and the measured runoff volume, we calculated the total solutes per experiment (TDS,
g). To compare with Alkali, we also calculated the solutes
in runoff in mequiv/L by multiplying the mS/cm by 0.10
(Hanson 2006). Given the pH values below 8.0 of the runoff
water, we assumed that the alkalinity was almost entirely
in the form of bicarbonate (HCO−
3 ). Therefore, we transformed the alkalinity into mg/L bicarbonate by multiplying
the mequiv./L by 6 L and then converting it to the total
number of grammes per experiment by dividing by 1000
and multiplying by the runoff in litres to obtain the amount
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
of total dissolved bicarbonates (TDB, g).
We studied seven dependent variables (Eros, OM, EC,
pH, Alkali, TDS and TDB) in relation to three experimental categorical factors (Crust, Rain and Length), and seven
continuous independent variables: slope angle (Slope, degrees), lichen crust cover (Lichen cover, %), cyanobacterial
crust cover (Cyanobacterial cover, %), total biocrust cover
(Biocrust cover, %), vascular plant cover (Plant cover, %),
total runoff volume (Runoff Vol, L) and runoff coefficient
(Runoff Coef, dimensionless). We used Generalised Linear
Modelling to test for differences in each dependent variable
according the factors (Crust, Rain and Length), all their interactions and the main covariates (Slope, Biocrust cover,
Plant cover and Runoff Vol). The other covariates were not
included because they were correlated with the previously
included ones. Gamma distribution fitted the best our data,
Log was used as link function, and the Akaike Information
Criterion (AIC) allowed selecting the best model.
The relationships among each dependent variable and
the continuous independent ones not included in the models (i.e.: lichen crust cover, cyanobacterial crust cover and
runoff coefficient) were explored by non-parametric correlations (Spearman), except with EC and Alkali, where parametric correlations could be used. The possible differences
in slope or cover between crust and rain types or among plot
lengths were tested by non-parametric ANOVA (Kruskal–
Wallis test). Kruskal–Wallis tests were also used to verify
whether there was a significant difference (i) in antecedent
soil moisture among treatments, (ii) in runoff amount between biocrust types, and (iii) in environmental features
among sampling blocks. Spearman correlations were also
used to test the relationships among the dependent variables.
The statistical analyses were performed using Statistica 7.1 (Stat.Soft. Inc., Tulsa, USA).
Results
All the experiments started on dry soil, with no significantly different antecedent soil moisture among plots
(global mean soil moisture = 2.06 %), as shown by the
Kruskal–Wallis test (H = 29.69; p > 0.1).
There were differences in slope among blocks belonging to different Crusts (p < 0.001) and also among
blocks into every Crust (used for different rain types)
(p < 0.01); but not according to Length. There were
no differences in plant cover with regard to Crust, Rain
or Length. Obviously, both lichen cover and cyanobacterial cover were significantly different according to Crust
(p < 0.001 in both cases). However, they were not different according to Length. The biocrust cover resulted
in slight differences with Crust (mean 80.6% in area L;
73.9% in area C) and was negatively correlated with
plant cover (p < 0.001).
Despite the careful selection of the sites for the
plots, the slope angle was significantly different among
the blocks of plots (Kruskal–Wallis test p < 0.001). L
and C biocrust types have different habitats (the habitat of C being wider, including most of the L habitat; in
this last lichen crust progressively replaces cyanobacterial crust over time; but in the most sunshine part of C
habitat this hardly occurs or it takes long time; Lazaro
1543
et al. 2008) and the blocks of plots were installed to ensure that each biocrust type was well represented. The
consequence was that the results concerning the relationship of the different variables with Slope must be
taken with great caution, as we cannot effectively distinguish the effects of slope angle from those of Rain or
Crust.
Runoff volume resulted considerably larger (H =
14.63, p = 0.0001) on cyanobacterial crust (mean =
8.45 L, Std. Error = 0.97) than on lichen crust (mean
= 4.18 L, Std. Error = 0.49).
None of the relationships among the dependent
variables was significant, except that between the organic matter in sediments and the amount of sediment
(r = 0.6688; p < 0.0001). However, more than one half
of the variance of OM was due to factors other from
sediment amount.
The dry matter in runoff
Figure 3a shows the real values of Eros (the particle erosion measured as the dry matter contained in the runoff
collected in every experiment) for every treatment. Eros
differed according to the interaction between Crust and
Rain: (Wald stat. 14.93; p < 0.001; Table 1). The erosive effects of the different rain types were not the same
in every biocrust. Fig. 3a shows that, on lichen biocrust,
Eros reached smaller values and often with smaller dispersion; dispersion increased with rain, though erosion
values continued being small. However, at cyanobacterial biocrust, Eros was larger and both erosion values
and value dispersions progressively increased with increasing rainfall and plot lengths.
The variation of Eros with regard to the plot length
only was visible at cyanobacterial biocrust (Fig 3a), and
almost exclusively for the largest rains, though the interaction between Crust and Length was not significant,
nor that between Crust, Rain and Length (Table 1).
Eros significantly depended on runoff volume and
on total biocrust cover (Table 1, Fig, 4 a and b, respectively) and was also significantly correlated with
other two independent continuous variables: Lichen
crust cover and runoff coefficient (Table 2). However,
Eros resulted not significantly associated to the slope
angle, the vascular plant cover nor the cyanobacterial
crust cover. Runoff was the main driver controlling the
particle erosion. The lichen crust was the main defence
of the soil against this erosion. Although runoff is expected to increase in amount and speed with slope angle, Eros resulted independent from Slope maybe due
to the strong positive correlation between lichen cover
and Slope (r = 0.6760, p < 0.0001). The cyanobacterial
cover, with a lower protection capacity, decreased with
increasing slope (r = –0.6724, p < 0.0001), whereas the
biocrust cover did not vary.
The organic matter contained in sediments
Figure 3b shows the real values reached in every treatment by OM (the total organic matter contained in
the dry matter mobilised by runoff). The pattern of
variation of OM was not very different from that of
Unauthenticated
Download Date | 6/18/17 6:30 PM
R. Lázaro & J.L. Mora
1544
Table 1. Relationship between every dependent variable (columns) and every factor, interaction and covariate included in the Generalized Lineal Models (rows).
Eros
Indepen.
variable
Wald st.
Intercept
3.4018
Slope
0.3172
biocrust-cover 5.8402
Plant-cover
0.6013
RunoffVol,l
5.2646
Crust
1.3844
Rain
2.4978
Length
1.4739
Crust*Rain 14.9295
Crust*Length 4.0995
Rain*Length
6.2447
Crust*Rain*
4.2821
Length
OM
p
0.0651
0.5733
0.0157
0.4381
0.0218
0.2393
0.2868
0.6883
0.0006
0.2509
0.3963
0.6386
Wald st.
0.4218
0.0284
1.1123
0.0704
5.6572
1.3938
1.5852
1.0973
5.5130
1.9380
1.8745
1.3888
EC
p
0.5161
0.8661
0.2916
0.7908
0.0174
0.2378
0.4527
0.7777
0.0635
0.5854
0.9309
0.9665
Wald st.
55.7811
4.7662
0.2289
0.0402
0.8108
9.4990
39.2729
9.1528
0.5027
4.5712
12.2900
4.8299
pH
p
0.0000
0.0290
0.6323
0.8411
0.3679
0.0021
0.0000
0.0273
0.7778
0.2060
0.0558
0.5658
Wald st.
254.595
0.0787
18.3821
4.9982
1.3505
1.3472
6.0329
0.5319
5.4790
2.3541
6.9151
4.9904
Alkali
p
0.0000
0.7791
0.0000
0.0254
0.2452
0.2458
0.0490
0.9118
0.0646
0.5022
0.3288
0.5450
Wald st.
0.5612
2.6683
0.3067
0.2432
0.0063
2.1752
3.2922
0.6228
0.7207
0.3332
1.5931
0.7083
TDS
p
Wald st.
TDB
p
Wald st.
p
0.4538 4.6162 0.0317 3.0559 0.0804
0.1024 2.9628 0.0852 0.4803 0.4883
0.5797 0.0001 0.9905 0.2444 0.6211
0.6219 2.0208 0.1552 0.7452 0.3880
0.9366 109.902 0.0000 113.267 0.0000
0.1402 1.0586 0.3035 1.9772 0.1597
0.1928 0.6606 0.7187 13.1901 0.0014
0.8912 0.3898 0.9423 0.5898 0.8988
0.6974 6.8364 0.0328 11.6850 0.0029
0.9537 4.4627 0.2156 1.9180 0.5896
0.9531 5.7300 0.4541 4.9187 0.5543
0.9943 5.7818 0.4481 7.3483 0.2898
Fig. 3. Mean value of dry matter in runoff for every treatment (Eros; graph a) and of organic matter within the dry matter in runoff
for every treatment (OM, graph b). The labels of the treatment are composed as follow: The two first characters identify the biocrust
type (Lc, lichen crust; Cc, cyanobacterial crust); the following two characters identify the rainfall type (Sr, small rain; Mr, medium
rain; Lr, large rain) and, the two last characters identify the plot length. These pairs of characters are separated by points.
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
1545
Table 2. Correlation (in the cases of EC and Alkali) or Spearman correlation between every dependent variable (columns) and three
additional continuous independent variables (rows).
Eros
Indep.
variable Spear. r
OM
p
Spear. r
EC
p
correl. r
pH
p
Spear. r
Alkali
p
correl. r
TDS
p
Spear. r
TDB
p
Spear. r
p
Lichen –0.2814 p<0.05 0.0098 p>0.05 0.1602 p>0.05 0.0972 p>0.05 –0.2476 p<0.05 –0.2589 p<0.05 –0.3278 p<0.05
c. cover
Cyano
0.1603 p>0.05 –0.1248 p>0.05 –0.1663 p>0.05 0.0474 p>0.05 0.2180 p<0.05 0.2342 p<0.05 0.3013 p<0.05
c. cover
Runoff 0.6055 p<0.05 0.4426 p<0.05 –0.0823 p>0.05 –0.1848 p>0.05 –0.0293 p>0.05 0.7763 p<0.05 0.7582 p<0.05
coefficient
Fig. 4. Relationship between Eros and OM with the runoff volume (graph a), and relationship between Eros and the total biocrust
cover (OM was not related to any kind of cover).
Eros, but the difference in OM according to the interaction Crust*Rain was only marginally significant (Wald
stat. 5.5130; p = 0.0635; Table 1). In a similar way
to than for Eros, the OM also increased with rainfall
and with plot length, and the differences were larger
in cyanobacterial crust (compare the graphs a) and
b) of Fig. 3). But the differences of OM according to plot length or to the interaction Crust*Length
Unauthenticated
Download Date | 6/18/17 6:30 PM
1546
R. Lázaro & J.L. Mora
Fig. 5. Mean values of electrical conductivity (EC, graph ‘a’), pH (pH, graph ‘b’) and alkalinity (Alkali, graph ‘c’) in runoff water for
every treatment. The labels of the treatment are composed as follow: The two first characters identify the biocrust type (Lc, lichen
crust; Cc, cyanobacterial crust); the following two characters identify the rainfall type (Sr, small rain; Mr, medium rain; Lr, large rain)
and, the two last characters identify the plot length. These pairs of characters are separated by points.
were not significant. The maximum of OM reached
in C for case of the highest rainfalls together with
the longest plots was 1.075 g, and it did not coincide with the maximum sediment yield, although it
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
corresponded to one of the larger values (18.24 g).
No other interaction among factors was significant for
OM.
No covariates had significant effect on OM except runoff volume (Table 1) and runoff coefficient (Table 2). Runoff is also the main driver that controls OM
(Fig. 4a), which was independent from the slope angle
and from every class of soil cover.
Electrical conductivity, pH and alkalinity in runoff
The runoff water obtained from the experiments was
saline, slightly basic and moderately hard, with bicarbonate constituting only a minor part of the total dissolved salts. The values of EC, pH and Alkali for every treatment are shown in Fig. 5, graphs a, b and c,
respectively.
When EC was directly measured in runoff, it
showed significant relationship with the three experimental factors, Crust, Rain and Length (Table 1)
but not with any interaction (though the effect of
Rain*Length was marginally significant). EC decreased
with increasing Rain, increased with increasing Length,
was higher under L than under C crust, and resulted
dependent on Slope.
With regard to the factors, pH showed association only with Rain, increasing with it (the effect
of Crust*Rain was marginally significant); however, it
showed significant higher values with increasing cover
of biocrusts and vascular plants (Table 1).
Alkali was not affected by the factors and their interactions (Table 1) nor by the covariates included in
the model (including the total biocrust cover) but was
found to be positively correlated with cyanobacterial
cover and negatively correlated with lichen cover (Table 2; due to the experimental design, the lichen cover
was inversely related to the cyanobacterial cover). None
of the Spearman correlations resulted significant with
EC or pH (Table 2).
Solutes yield from runoff (TDS,TDB)
The real values of total solutes (TDS) and dissolved
bicarbonates (TDB) mobilised by runoff are shown in
Figs 6a and 6b, respectively. The results of the GLZ
for TDS and TDB were quite similar and are shown in
Table 1. Both dependent variables resulted controlled
by the interaction Crust*Rain and, above all, by the
runoff volume (Fig. 7a). TDB was associated also to
the factor Rain itself.
Both TDS and TDB tended to be larger at
cyanobacterial crust, with larger rainfalls and with
larger plot lengths. The differences in TDS and TDB
according Rain or Length were larger in cyanobacterial crust, although only the differences according
Crust*Rain were significant, whereas the differences according Length or any of its interactions were not significant for any of these variables. The differences according the factor Rain were significant only for TDB
(the larger the rain, the larger the amount of dissolved
bicarbonates).
The independence of TDS and TDB from Slope
1547
(only for TDS the effect of Slope resulted marginally
significant) is counter-intuitive, taking into account the
probable increase in runoff with slope. This relationship was due to the increase in the lichen cover with
slope within these experiments, since the correlations
between solutes (both TDS and TDB) and lichen cover
were negative and highly significant (Table 2, Fig. 7b),
whereas the correlations with the cyanobacterial cover
were positive and significant (Table 2, Fig. 7c). This
provides evidence that the lichen crust not only protects against the erosion of inorganic or organic matter, but also against soil solutes mobilisation by runoff.
Both TDS and TDB showed also a strong positive correlation with the runoff coefficient (Table 2).
Discussion
The results support the hypothesis that the increase
in length of the drainage area very often has no effect
on the sediment yield. An important difference in erosion among plot lengths was only observed, both for
Eros and OM at C under the largest rainfall (Fig. 3) although the effect of Length still was not significant due
to the considerable variance. This is possibly because
the runoff on C was significantly longer, particularly
with larger rainfall. The dense net of microdepressions
developed by the biocrust, particularly at L, probably
acts as a sediment trap by slowing the runoff flow, and
multiple small sediment deposits can occur along the
runoff paths, giving rise to non-significant differences in
sediment yield at the plot outlet according to Length.
Rodriguez-Caballero et al. (2012) found that although
crust roughness slightly reduces the amount of material
exported by runoff, more research is needed to distinguish between the direct and indirect effects of biocrustdriven microtopography on the erosive response to rainfall. According Lazaro et al. (2008), the lichen crust
is late-successional with regard to the cyanobacterial
crust. As expected, Eros decreased with the successional development of the biocrust, as it had a negative
correlation with the lichen crust cover.
The inorganic and organic dry matter in runoff
The most erosive conditions of the experiments combined the largest rainfall (both in intensity and water volume) with the longest plots. Large rainfall has
more kinetic energy and besides, for a given runoff coefficient, it produces a larger runoff volume and larger
runoff connectivity, which facilitates sediment transport. This occurs particularly in sites where the mechanism of runoff generation is infiltration excess, as in
the studied area (Solé-Benet et al. 1997; Calvo-Cases
et al. 2003). Besides, as biocrusts are runoff sources
(Rodriguez-Caballero 2013), longer slopes are expected
to allow larger runoff connectivity, also facilitating sediment transport. The results showed that biocrusts, particularly the lichen crust, could protect soil from erosion, even under the combination of these two circumstances that produced the most erosive conditions.
Many studies agree that biocrusts reduce the sedUnauthenticated
Download Date | 6/18/17 6:30 PM
1548
R. Lázaro & J.L. Mora
Fig. 6. Mean value of total solutes (TDS, graph ‘a’) and dissolved carbonates (TDB, graph ‘b’) in runoff water for every treatment. The
labels of the treatment are composed as follow: The two first characters identify the biocrust type (Lc, lichen crust; Cc, cyanobacterial
crust); the following two characters identify the rainfall type (Sr, small rain; Mr, medium rain; Lr, large rain) and, the two last
characters identify the plot length. These pairs of characters are separated by points.
iment yield, regardless of the biocrust or soil type
(Booth 1941; Mücher et al. 1988; Alexander & Calvo
1990; Eldridge & Kinnell 1997; Zhang 2005; Lazaro et
al. 2008; Chamizo et al. 2012; Rodriguez-Caballero et
al. 2013). This is due mainly to three mechanisms:
i. The biocrust intercept the raindrops, absorbing
their kinetic energy and reducing splash erosion. Data
on splash erosion from our study area are shown by Li
et al. (2005). Zhao et al. (2014) reported experimental
laboratory tests showing the resistance of biocrusts to
raindrops.
ii. The biocrust decreases the runoff amount and
speed, by increasing the surface roughness and soil
porosity and decreasing surface sealing. The larger the
successional development of the biocrust, the larger the
surface roughness (Chamizo et al. 2010) and the in-
filtration (Chamizo et al. 2012). As the length and
tortuosity of the runoff paths increase with roughness, the water remains for a longer time over the
soil, facilitating infiltration (Belnap et al. 2006), and
the velocity of runoff water decreases reducing its erosive force (Kidron et al. 1999). Moreover, infiltration
also increases with the successional development of
the biocrust due to the increase in porosity (Miralles
et al. 2011; Gao et al. 2012) and to the decrease
in soil-surface sealing as the biocrust develops (Assouline 2004; Chen et al. 2013); although Kidron et
al (1999) found that the formation of algae biocrusts
favours pore clogging on sandy soils because of the
water absorption and swelling of the exopolysaccharides.
iii. The soil aggregate stability increases with the
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
1549
Fig. 7. Relationships between total solutes in runoff per experiment (TDS) and total dissolved carbonates/bicarbonates in runoff
(TDB) with: runoff volume (graph a), lichen crust cover (graph b) and cyanobacterial crust cover (graph c).
Unauthenticated
Download Date | 6/18/17 6:30 PM
1550
development of the biocrust, as the amount of extracellular polymer secretions and of cyanobacterial filaments and lichen and moss anchoring structures increases (Zhang 2005; Malam Issa et al. 2001; Warren
2003). Zhang (2005) found that the development of a
cyanobacterial biocrust from an incipient one to another more developed one increased the pressure that
biocrust can resist from 13.42 Pa to 32.53 Pa. Zhao et
al. (2014) reported that while cyanobacterial crust resisted a kinetic energy ca.15-fold higher than that of
bare soil, a late-successional moss-dominated biocrust
reached values up 342-fold higher than that of bare soil.
The lichen crust in this study is late-successional with
regard to the cyanobacterial crust (Lazaro et al. 2008)
and the results showed that the erosion in the lichen
crust was effectively lower, although the difference only
was significant for part of the rainfalls.
Early studies on water erosion in badland areas
in southeastern Spain (Calvo-Cases et al. 1991) found
that the most important factors controlling erosional
response related to cover and, that “the most effective cover appears to be a lichen cover”. The data here
for dry matter in runoff, expressed in g/L, showed an
isolated absolute maximum of 3.8, but the mean was
0.62, which is quite low compared with other results
in Mediterranean region from sites without biocrusts:
1.40–9.26 g L−1 for a simulated rainfall of 41 mm (Boix
Fayos et al. 1995); 11.30 g L−1 after 25 mm simulated
rainfall (Imeson et al. 1998); and 1.2–6.1 g L−1 after
30 mm simulated rainfall (Lasanta et al. 2000). Following monitoring over three-years under natural rainfall, Cantón et al. (2001) found an erosion rate of 21.4
g m−2 of surfaces covered by a lichen crust, whereas
bare soil with a physical crust showed a rate of 308.1
g m−2 . Our data had a mean of 16.79 g m−2 for high
rainfall and 1 m-long plots at C, and 3.37 g m−2 for
the same combination at L. Rodriguez-Caballero et al.
(2013), found sediment yields of 15.2 g m−2 in C and
11.5 g m−2 in L in our study area, using 1-m long plots
under high natural rains (78 mm), but 31.2 and 17.5,
respectively, in plots where biocrust was developing,
because it was removed five years before the measurements. Chamizo et al. (2012) simulated 55-mm-rainfalls
on plots 0.56 m diameter, and found sediment yields of
26.0 and 10.7 g m−2 in C and L, respectively and, 128.8
and 744.6 g m−2 , respectively, following removal of the
biocrust. Therefore, the data of sediment yield here
are similar to those found by other authors and much
lower than those in soils without a biocrust. When the
biocrust is removed, erosion is greater where there was
a lichen late-successional crust than where there was a
cyanobacterial crust (Chamizo et al. 2012), as soil has
a higher porosity and lower bulk density below lichen
crusts (Miralles et al. 2011; Gao el al. 2012). According
to Zhao & Xu (2013), biocrusts in their early successional stage (cyanobacteria) reduced sediments by 92%,
and no sediments were collected in late-successional
biocrusts (mosses) under the same simulated rainfall,
whereas vascular plants only reduced sediment yield by
45%.
R. Lázaro & J.L. Mora
The correlation between erosion in g/m2 versus the
biocrust cover was significant and can be fitted to an exponentially decreasing curve (r = –0.4223; p < 0.0001),
as proposed by Eldridge and Greene (1994), although in
our case, the exponential fit is not better than a simple
linear fit. However, the results here are consistent with
the quoted model: our biocrust covers are high (ranging
between 50 to 100%; mean 77%) and correspond to the
three last points in the graph of those authors, where
erosion decreased linearly.
The capacity of the biocrusts to protect the soil
from erosion will be important whether the biocrust
itself is resistant enough. The fact that biocrusts are
widespread in drylands (Büdel 2003), and can locally
cover large extensions in stable enough areas (Lazaro
et al. 2000), provides evidence that the biocrust can
extend despite splash erosion. However, the resistance
of the biocrust to the impact of raindrops and runoff
flow is expected to be highly species-specific and there
is not much information concerning this; particularly
on possible detachment due to the runoff flow. Kuske
et al. (2012) found that despite the dramatic negative
impacts of trampling on the cyanobacterial biocrust,
Microcoleus vaginatus could still be detected after 10
years of annual trampling. Zhao et al. (2014, see above)
reported that the resistance of the cyanobacterial and
moss crusts to raindrops. Qin & Zhao (2011) found,
by simulating single-drop rain, that the cyanobacterial
biocrusts and the moss biocrusts can resist 0.99 J and
75.56 J of cumulative rain-drop kinetic energy, respectively, and those biocrusts with a larger biomass had a
considerably higher resistance. (When the chlorophyll
content of cyanobacterial biocrusts increased from 3.32
to 3.73 µg g−1 , the resistance of the biocrusts increased
from 0.99 to 2.17 J; and when the moss biomass increased from 2.03 to 4.73 g dm−2 , the resistance increased from 6.08 to 75.56 J). Sempere (1994) found
that the kinetic energy of natural rainfall ranged between 15 and 25 J m−2 for an intensity of about
50 mm h−1 . Taking all this into account, a rainfall
with an extraordinary volume and intensity would be
required to dismantle a biocrust. According to Gao et
al. (2012), the cohesion of biocrusts was six- or sevenfold higher than that of subsurface soil (0–2 cm), and
the impact of biocrusts on soil physical properties was
closely related to the biocrust biomass. However, although some studies exist on the resistance of vegetation to runoff (for example, Xiao et al. 2013), none
exists on the biocrust resistance to runoff. In our experiments, the highest rainfall, together with the longest
plot possibly reached a possible threshold of erodibility at C, where the sediment yield reached the value
of 38.99 g in one case; approximately double that of
other larger values; and the behaviour of OM was similar (Fig. 3).
In our opinion, the amount of OM can be considered as a rough estimate of the degree of detachment of
the biocrust due to both raindrops and runoff flow, since
the soils of the area are very poor in humus (according
to Cantón et al. 2003, the organic matter content in
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
soils of the study area ranged from 0.15% to 1.4% near
the parent material, and from 0.4% to 2.4% in some surface horizons), and litter from vascular plants is scarce
and randomly distributed in the experimental site. The
measured absolute maximum OM was only 11.8% in
the biocrust with larger biomass (L) and under the most
erosive conditions; however, the mean OM was only 4%.
This suggests that the OM coming from biocrust detachment can be ca. 2.6% of the total sediments (4%
– ca. 1.4% of soil organic matter; without subtraction
of the litter) and is evidence that erosion is strongly
skewed towards the detachment of mineral matter due
to the resistance of the biocrust. Although OM showed
a strong correlation with Eros, it did not depend on any
cover (lichen crust, cyanobacterial crust, total biocrust
or plant cover), it only depended on the runoff volume.
This suggests that the detachment is controlled by the
water amount and by the mobilisation of mineral particles and is not limited by the biocrust cover when this
cover is larger than 50% (the minimum value in our experiments). The differences in OM were also larger in
C, for larger rains (marginally significant) and for larger
plot lengths (no significant). This treatment (77 mm/h
for 20 min, 25.6 mm, and drainage areas of 3-4-m long),
the most favourable for runoff, might be near or beyond
an emergent threshold of erodibility of the C biocrust.
Solutes in runoff
In addition to the solid mineral and organic matter removed by rain-splash and runoff, runoff mobilised a
significant amount of soil solutes. The concentration
of solutes in runoff water (as seen by the EC values)
apparently increased with longer contact between the
runoff water and substrate, which was allowed by small
Rainfall, which produces lower runoff and increased by
the larger roughness of lichen crust (Chamizo et al.
2010). The solute concentrations in our experiments
were independent of the runoff volume, suggesting that
the dissolving water was not the factor limiting their
concentrations. This indicates that solutes mainly derived from the dissolution of minerals present in large
amounts in the soil and parent material and that the
main limiting factor was the solubility of these minerals.
Since bicarbonate was found in small rates, dissolved solutes were plausibly dominated by sulphate: According
Cantón et al. (2003), calcite (source of bicarbonate) and
gypsum (source of sulphate) each constitute ca. 20% of
the parent material of these marine marls; other salts
such as chloride are rare in the soils of the study area.
Although biocrusts can modify the ion composition of
their close soil environment, e.g., via the secretion of
acidifying or alkalinising compounds and the concentration of cations and anions in the top-soil (Pointing
& Belnap 2012), this limited bio-weathering is insufficient to explain the high salt concentrations that were
observed.
In turn, the pH of runoff water increased with the
coverage of biocrusts and of vascular plants and with
increasing Rain. Water pH is reported to rise as result of
increased infiltration depth and, thus, of a longer time
1551
of contact of water flow with calcareous substrate leading to high saturation of carbonates in water (Chandler
& Bisogni 1999). In agreement with this, our results for
pH suggest an increased carbonate saturation resulting
from a larger and deeper water infiltration with higher
rainfall amounts and under biocrust and vascular plant
cover. These results were not shared with Alkali, which
represents the amount of carbonates dissolved by runoff
water, but discordance can be expected between the
behavior of bicarbonate and pH values when different
degrees of carbonate saturation occur in water as result of variable water deepening and contact time with
the calcareous material (Chandler & Bisogni 1999). The
negative correlation of Alkali with the lichen cover could
be explained because the lichen thalli protect minerals
from weathering or from liberation of weathering products. Moreover, this finding can also be explained by
lower content of carbonates under lichen crust, which
can result from increasing infiltration under this surface, as highlighted by Cantón et al. (2003) in our
area for sites with the largest infiltration, mainly covered by vascular plants. Ladyman et al. (1993) found
that a greater infiltration under biocrusts could transfer
more calcium carbonate to depth horizons in calcareous
desert soils; but to date no evidence is available that it
might also affect carbonate mobilisation by runoff.
These effects on the solute concentration in runoff
become irrelevant when we consider the total yield of
solutes, which was mainly determined by the runoff volume. Although EC remains independent of the runoff
volume, TDS and TDB showed strong a positive correlation with runoff, and is therefore influenced by those
factors that affect runoff, as the crust type, with TDS
and TDB being larger at C. Both increased as the
cyanobacterial cover increased and the lichen cover decreased. The comparison of the figures 3 and 6 shows
that both the erosion of solid particles (organic and inorganic) as well as the solute mobilisation (carbonates
and total) have similar roughly behaviour. Both TDS
and TDB seem to be affected by the possible threshold of erodibility reached at the cyanobacterial crust
under the extreme runoff conditions (the largest rainfall together with the longest plots). This is consistent
with our results with Eros and OM and with studies on
chemical erosion in arid basins, where solute mobilisation is reported to be controlled in the first instance by
the runoff, except for carbonates when their abundance
is limited (Abbas & Subramanian 1984; France-Lanord
et al. 2003).
Conclusions
Biocrusts strongly protected soil against water erosion
in our experiment, even under the heaviest rainfall combined with the longest plots. The successional development of biocrust enhanced soil protection, the protection provided by the lichen crust was greater than that
provided by the cyanobacterial crust, particularly under
extreme conditions. These biocrusts were very resistant
to the impact of raindrops and also to detachment by
Unauthenticated
Download Date | 6/18/17 6:30 PM
1552
runoff flow. The plot length often had no effect on erosion, due to the decrease in erosion as a result of the protective role of biocrusts, the limited runoff lengths and
the increase of surface roughness, which facilitated sediment deposition. However, a new hypothesis emerged:
under the most erosive conditions of our experiment, a
threshold of erodibility was reached at the cyanobacterial crust, which affects similarly both the erosion of
solid particles (both organic and inorganic) as well as
the mobilisation of solutes. The lichen crust not only
protected the soil from solid erosion, but also against
the removal of soluble soil substances. The increase in
biocrust cover can drive chemical changes in the solutes
mobilised by runoff by modifying the pH. In our experiment, solute mobilisation (mainly sulphates and carbonates) was mainly affected by conditions allowing a
closer contact between runoff water and substrate. The
solute yields were proportional to runoff volume, due to
the occurrence of large sources of salts with low solubility (gypsum, limestone) in soils and parent material.
The type of biocrust affected the solute mobilisation
in two alternative ways: the solute control by solubility together with the greater runoff volume generated
by the cyanobacterial biocrust produced higher yields
of carbonates and total solutes in runoff. On the other
hand, the greater infiltration produced by lichen crusts
possibly decreased the carbonate contents in soils and,
therefore, their yield in runoff.
Acknowledgements
This study was undertaken in the context of the research
projects PECOS (REN2003-04570/GLO) and PREVEA
(CGL2007-63258/BOS), both funded by the Spanish National Plan for RD&I and by the European ERDF Funds
(European Regional Development Fund), as well as the
projects COSTRAS (Excellence project RNM-3614), funded
by the Junta de Andalucia (Autonomous Government of Andalusia, Spain), and SCIN (Soil Crust Inter-National, PRIPIMBDV-2011-0874, European project of ERA-NET BIODIVERSA); the Spanish team was funded by the Spanish
Ministry of Economy and Competitiveness.
We are especially grateful to Emiliano Pegoraro, who
participated full-time in the design and conduct of the experiments but became ill and died shortly after.
We are grateful to the Viciana brothers, owners of the
property where the experiments were carried out, Thomas
Iserloh (Trier University, Germany) for calibrating the drop
sizes and kinetic energy of our simulator with a laser distrometer, Montserrat Guerrero for the laboratory analyses,
and to Alfredo Durán, Mónica Gutiérrez, Amparo Castillo,
Rafael Poyatos, Ana Rey and Florentino Mostaza for their
assistance with the fieldwork. We are also grateful to the
anonymous reviewers for their constructive comments.
References
Abbas N. & Subramanian V. 1984. Erosion and sediment transport in the Ganges river basin (India). J. Hydrol. 69: 173–182.
Alexander R.W. & Calvo-Cases A. 1990. The influence of lichens
on slope processes in some Spanish badlands, pp. 385–398.
In: Thornes J.B. (ed.), Vegetation and Erosion. John Wiley
& Sons, Chichester.
R. Lázaro & J.L. Mora
Alexander R.W., Harvey A.M., Calvo-Cases A., James P.A. &
Cerdá A. 1994. Natural stabilisation mechanisms on badland
slopes: Tabernas, Almeria, Spain, pp. 85–111. In: Millington A.C. & Pye K. (eds), Environmental Change in Drylands: Biogeographical and Geomorphological Perspectives.
John Wiley & Sons, Chichester.
Alekin O.A. & Brazhnikova L.V. 1968. Dissolved matter discharge
and mechanical and chemical erosion. Geochemistry, Precipitation, Evaporation, Soil Moisture, Hydrometry – Proceedings of the General Assembly of Bern, vol. IV, pp. 35–41:
IAHS Publ. no. 78.
Allen-Diaz B., Chapin F.S., Diaz S., Howden M., Puigdefábregas J. & Stafford-Smith M. 1996. Rangelands in a changing
climate: impacts, adaptations and mitigation, pp. 131–158.
In: Watson R.T., Zinyowera M.C. Moss R.H. & Dokken D.J.
(eds), Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge.
Angulo-Martínez M. & Beguería S. 2009. Estimating rainfall erosivity from daily precipitation records: A comparison among
methods using data from the Ebro Basin (NE Spain). J. Hydrol. 379: 111–121.
Assouline S. 2004. Rainfall-induced soil surface sealing: a critical review of observations, conceptual models and solutions.
Vadose Zone J. 3: 570–591.
Belnap J. 2006. The potential roles of biological soil crusts in
dryland hydrologic cycles. Hydrol. Process. 20: 3159–3178.
Belnap J. & Gardner J.W. 1993. Soil microstructure in soils of the
Colorado Plateau: the role of the cyanobacterium Microcoleus
vaginatus. Great Basin Nat. 53: 40–47.
Belnap J. & Lange O.L. 2003. Biological Soil Crusts: Structure,
Function and Management. Springer–Verlag: Berlin, Heidelberg.
Bevan J.L. 2009. Dynamics of Lichen Dominated Biological Soil
Crusts in the El Cautivo Badlands, Southeast Spain. PhD
Thesis, University of Liverpool, UK.
Boix Fayos C., Calvo Cases A., Imeson A.C., Schoorl J.M., Soriano Soto M.D. & Tiemessen I.R. 1995. Properties and erosional response of soils in a degraded ecosystem in Crete
(Greece). Environ. Monitor.Assessm. 37: 79–92.
Boix-Fayos C., Martínez-Mena M., Calvo-Cases A., ArnauRosalén E., Albaladejo J. & Castillo V. 2007. Causes and underlying processes of measurement variability in field erosion
plots in Mediterranean conditions. Earth Surf. Proc. Land.
32: 85–101.
Booth W.E. 1941. Algae as pioneers in plant succession and their
importance in erosion control. Ecology 22: 38–46.
Bowker M.A. 2007. Biological soil crust rehabilitation in theory
and practice: an underexploited opportunity. Restor. Ecol.
15: 13–23.
Bowker M.A., Belnap J., Chaudhary V.B. & Johnson N.C. 2008.
Revisiting classic water erosion models in drylands: the strong
impact of biological soil crusts. Soil Biol. Biochem. 40: 2309–
2316.
Bowker M.A., Maestre F.T. & Escolar C. 2010. Biological crusts
as a model system for examining the biodiversity-ecosystem
function relationship in soils. Soil Biol. Biochem. 42: 405–517.
Bowker M.A., Maestre F.T., Eldridge D., Belnap J., CastilloMonroy A., Escolar C. & Soliveres S. 2014. Biological soil
crusts (biocrusts) as a model system in community, landscape
and ecosystem ecology Biodivers. Conserv. DOI:10.1007/
s10531-014-0658-x.
Bromley J., Brouwer J., Barker A.P., Gaze W.R. & Valentine C.
1997. The role of surface water redistribution in an area of
patterned vegetation in a semi-arid environment, south-west
Niger. J. Hydrol. 198 (1-4): 1–29.
Bu C., Wu S., Xie1 Y. & Zhang X. 2013. The study of biological
soil crusts: hotspots and prospects. Clean – Soil, Air, Water
41: 899–906.
Büdel B. 2003. Synopsis: Comparative biogeography of soil-crust
biota, pp. 141 -152. In: Belnap J. & Lange O.L. (eds), Biological Soil Crusts: Structure, Function and Management.
Springer Verlag, Berlin Heidelberg.
Unauthenticated
Download Date | 6/18/17 6:30 PM
Erosion on biocrusts according slope length
Calvo-Cases A., Harvey M.A., Paya-Serrano J. & Alexander R.W.
1991. Response of badland surfaces in South Easth Spain to
simulated rainfall. Cuaternario y Geomorfologia 5: 3–14.
Calvo-Cases A., Boix-Fayos C. & Imenson A.C. 2003. Runoff generation, sediment movement and soil water behaviour on calcareous (limestone) slopes of some Mediterranean environments in Southeast Spain. Geomorphology 50: 269–291.
Calvo-Cases A., Harvey A.M., Alexander R.W., Cantón Y.,
Lazaro R., Sole-Benet A. & Puigdefabregas J. 2014. Badlands in the Tabernas Basin, Betic Chain, pp. 197–212. In:
Gutiérrez F. & Gutiérrez M. (eds), Landscapes and Landforms of Spain, World Geomorphological Landscapes, chapter
17. Springer Science, Dordrecht.
Cantón Y., Domingo F., Solé-Benet A. & Puigdefábregas J. 2001.
Hydrological and erosion response of a badlands system in
semiarid SE Spain. J. Hydrol. 252: 65–84.
Cantón Y., Domingo F., Solé-Benet A. & Puigdefábregas J. 2002.
Influence of soil-surface types on the overall runoff of the
Tabernas badlands (south-east Spain): field data and model
approaches. Hydrol. Process. 16: 2621–2643.
Cantón Y., Solé-Benet A. & Lázaro R. 2003. Soil-geomorphology
relations in gypsiferous materials of the Tabernas Desert
(Almería, SE Spain). Geoderma 115: 193–222.
Cantón Y., Del Barrio G., Solé-Benet A. & Lázaro R. 2004. Topographic controls on the spatial distribution of ground cover
in a semiarid badlands area. Catena 55: 341–365.
Cantón Y., Solé-Benet A., de Vente J., Boix-Fayos C., CalvoCases A., Asensio C. & Puigdefábregas J. 2011. A review of
runoff generation and soil erosion across scales in semiarid
south-eastern Spain. J. Arid Environ. 75: 1254–1261.
Castillo-Monroy A.P., Maestre F.T., Delgado-Baquerizo M. &
Gallardo A. 2010. Biological soil crusts modulate nitrogen
availability in semi-arid ecosystems: Insights from a Mediterranean grassland. Plant Soil 333: 21–34.
Chamizo W., Rodríguez-Caballero E., Miralles-Mellado I., Afana
A., Lázaro R., Domingo F., Calvo-Cases A., Solé-Benet A.
& Cantón Y. 2010. Características de las costras físicas y
biológicas del suelo con mayor influencia sobre la infiltración
y la erosión en ecosistemas semiáridos. Pirineos 165: 69–96.
Chamizo W., Cantón Y., Lázaro R., Solé-Benet A. & Domingo
F. 2012. Crust composition and disturbance drive infiltration through Biological Soil Crusts in semiarid ecosystems.
Ecosystems 15: 148–161.
Chandler D.G. & Bisogni Jr., J.J. 1999. The use of alkalinity as
a conservative tracer in a study of near-surface hydrologic
change in tropical karst. J. Hydrol. 216: 172–182.
Chen L, Sela S, Svoray T & Assouline S. 2013. The roles of soilsurface sealing, microtopography and vegetation patches in
rainfall-runoff processes in semiarid areas. Water Resour. Res.
49: 1–15.
De Baets S., Poesen J., Gyssels G. & Knapen A. 2006. Effects of
grass roots on the erodibility of topsoils during concentrated
flow. Geomorphology 76: 54–67.
Eldridge D.J. 1993. Cryptogam cover and soil surface condition:
effects on hydrology on a semiarid woodland soil. Arid Soil
Res. Rehab. 7: 203–217.
Eldridge, D.J. 1998. Trampling of microphytic crusts on calcareous soils and its impact on erosion under rain-impacted flow.
Catena 33: 221–239.
Eldridge D.J. 2003. Soil crusts and water relations in Australia.
In: Belnap J & Lange OL (eds) Biological Soil Crusts: Structure, Function and Management. Springer-Verlag.
Eldridge D.J. & Greene R.S.B. 1994. Assessment of sediment
yield by splash erosion on a semi-arid soil with varying cryptogam cover. J. Arid Environ. 26: 221–232.
Eldridge D.J. & Kinnell P.I.A. 1997. Assessment of erosion
rates from microphyte-dominated calcareous soils under rainimpacted flow. Aust. J. Soil Res. 35: 475–489.
Eldridge D.J., Tozer M.E. & Slangen W. 1997. Soil hydrology
is independent of microphytic crust cover: Further evidence
from a wooded semiarid Australian rangeland. Arid Soil Res.
Rehab. 11: 113–126.
Eldridge D.J., Zaady E. & Shachak M. 2000. Infiltration through
three contrasting biological soil crusts in patterned landscapes in the Negev, Israel. Catena 40: 323–336.
1553
France-Lanord C., Evans M., Hurtrez J.-E. & Riotte J. 2003.
Annual dissolved fluxes from Central Nepal rivers: budget of
chemical erosion in the Himalayas. C. R. Geoscience 335:
1131–1140.
Gao L.Q., Zhao Y.G., Qin N.Q., Zhang G.X. & Yang K. 2012.
Impact of biological soil crust on soil physical properties in
the hilly Loess Plateau region, China. J. Natur. Resour. 201208.
Hanson B.R. 2006. Electrical conductivity. In: Hanson B.R.,
Grattan S.R. & Fulton A. (eds), Agricultural Salinity and
Drainage (Revised edition). Division of Agriculture and Natural Resources Publication 3375. University of California.
Harper K.T. & St. Clair L. 1985. Cryptogamic soil crusts on arid
and semiarid rangelands in Utah: effects on seedling establishment and soil stability. Dept Botany Range Sci, Brigham
Young University, Provo, Utah.
Herrick J.E., Van Zee J.W., Belnap J., Johansen J.R. & Remmenga M. 2010. Fine gravel controls hydrologic and erodibility responses to trampling disturbance for coarse-textured
soils with weak cyanobacterial crusts. Catena 83: 119–126.
Horton R.E. 1933. The role of infiltration in the hydrologic cycle.
Trans. American Geophysical Union 14: 446–460.
Imeson A.C., Lavee H., Calvo A. & Cerdŕ A. 1998. The erosional
response of calcareous soil along a climatological gradient in
Southeast Spain. Geomorphology 24: 3–16.
Kidron G.J. 2011. Runoff generation and sediment yield on homogeneous dune slopes: scale effect and implications for analysis
Earth. Surf. Proc. Land. 36: 1809–1824.
Kidron G.J., Yaalon D.H. & Vonshak A. 1999. Two causes for
runoff initiation on microbiotic crusts: hydrophobicity and
pore clogging. Soil Sci. 164: 18–27.
Kuske C.R., Yeager C.M., Johnson S., Ticknor L.O. & Belnap J.
2012. Response and resilience of soil biocrust bacterial communities to chronic physical disturbance in arid shrublands.
ISME J. 6: 886–897.
Ladyman J.A.R., Muldavin E. & Fletcher R. 1993. Pattern and
relationships of terrestrial cryptogam cover in two pinonjuniper communities in New Mexico, RM-236. US Department of Agriculture, Forest Service, Santa Fe, NM.
Lasanta T., Garcia-Ruiz J.M., Perez-Rontome C. & SanchoMarcen C. 2000. Runoff and sediment yield in a semi-arid
environment: the effect of land management after farmland
abandonment. Catena 38: 265–278.
Lázaro R., Alexander R.W. & Puigdefabregas J. 2000. Cover distribution patterns of lichens, annuals and shrubs in the Tabernas Desert, Almería, Spain, pp. 19–40. In: Alexander R.W. &
Millington A.C. (eds), Vegetation Mapping: from Patch to
Planet. Wiley, Chichester.
Lázaro R., Rodrigo F.S., Gutiérrez L., Domingo F. & Puigdefábregas J. 2001. Analysis of a thirty year rainfall record
(1967–1997) from semi-arid SE Spain for implications on vegetation. J. Arid Environ. 48: 373–395.
Lázaro R., Cantón Y., Solé-Benet A., Bevan J., Alexander R.,
Sancho L.G. & Puigdefábregas J. 2008. The influence of competition between lichen colonization and erosion on the evolution of soil surfaces in the badlands (SE Spain) and its
landscape effects. Geomorphology 102: 252–266.
Li X.Y., González A. & Solé-Benet A. 2005. Laboratory methods for the estimation of infiltration rate of soil crusts in the
Tabernas Desert badlands. Catena 60: 255–266.
Maestre F.T., Escolar C., Ladron de Guevara M., Quero J.L.,
Lazaro R., Delgado-Baquerizo M., Ochoa V., Berdugo M.,
Gozalo B. & Gallardo A. 2013. Changes in biocrust cover
drive carbon cycle responses to climate change in drylands.
Glob. Change Biol. 19: 3835–3847.
Malam Issa O., Le Bissonnais Y., Defarge C. & Trichet J. 2001.
Role of a cyanobacterial cover on structural stability of sandy
soils in the Sahelian part of western Niger. Geoderma 101:
15–30.
Meybeck M. 1976. Total mineral dissolved transport by world
major rivers. Hydrological Sciences-Bulletin-des Sciences Hydrologiques XXI, 2: 265–284.
Miralles I., Cantón Y. & Solé-Benet A. 2011. Two-dimensional
porosity of crusted silty soils: indicators of soil quality in
semiarid rangelands? Soil Sci. Soc. Am. J. 75: 1289–1301.
Unauthenticated
Download Date | 6/18/17 6:30 PM
1554
Mücher H.J., Chartres C.J., Tongway D.J. & Greene R.S.B. 1988.
Micromorphology and significance of the surface crusts of
soils in rangelands near Cobar, Australia. Geoderma 42: 227–
244.
Pendleton R.L., Pendleton B.K., Howard G.L. & Warren W.D.
2003. Growth and nutrient content of herbaceous seedlings
associated with biological soil crusts. Arid Land Res. Manag.
17: 271–281.
Pérez F.L. 1997. Microbiotic crusts in the high equatorial Andes,
and their influence on Paramo soils. Catena 31: 173–198.
Pointing S.B. & Belnap J. 2012. Microbial colonization and controls in dryland systems. Nat. Rev. Microbiol. 10: 551–562.
Puigdefábregas J. & Pugnaire F.I. 1999. Plant survival in arid
environments, pp. 381-406. In: Pugnaire F.I. & Valladares F.
(eds), Handbook of Functional Plant Ecology. Marcel Dekker
Inc. New York.
Qin N.Q. & Zhao Y.G. 2011. Responses of biological soil crust to
and its relief effect on raindrop kinetic energy. J. Appl. Ecol.
22: 2259–2264.
Rapp A. 1986. Introduction to soil degradation processes in drylands. Climatic Change 9: 19–31.
Rodriguez-Caballero E., Cantón Y., Chamizo S., Afana A. &
Solé-Benet A. 2012. Effects of biological soil crusts on surface
roughness and implications for runoff and erosion. Geomorphology 145-146: 81–89.
Rodriguez-Caballero E., Cantón Y., Chamizo S., Lázaro R. &
Escudero A. 2013. Soil loss and runoff in semiarid ecosystems:
A complex interaction between biological soil crusts, microtopography and hydrological drivers. Ecosystems 16: 529–
546.
Sempere D. 1994. La lluvia como agente erosivo: formación, distribución, erosividad e intercepción. Ingeniería Hidráulica en
México 9(2): 5–18.
Solé-Benet A., Calvo-Cases A., Cerdá A., Lázaro R., Pini R. &
Barbero J. 1997. Influences of micro-relief patterns and plant
cover on runoff related processes in badlands from Tabernas
(SE Spain). Catena 31: 23–38.
R. Lázaro & J.L. Mora
Warren S.D. 2003. Synopsis: influence of biological soil crusts on
arid land hydrology and soil stability, pp. 349-360. In: Belnap J. & Lange O.L. (eds), Biological Soil Crusts: Structure,
Function, and Management. Springer-Verlag, Berlin.
Wei W., Chen L., Fu B. & Chen J. 2010. Water erosion response
to rainfall and land use in different drought-level years in a
loess hilly area of China. Catena 81: 24–31.
Xiao P.Q., Yao W.Y., Li L. &Yang C.X. 2013. Experimental study
of overland flow resistance of different vegetation covers. J.
Natur. Resourc. 2013–03.
Yair A. 2003. Effects of biological soil crusts on water redistribution in the Negev desert, Israel: a case study in longitudinal dunes, pp. 303–314. In: Belnap J. & Lange O.L. (eds),
Biological soil crusts: structure, function and management.
Springer–Verlag: Berlin, Heidelberg.
Yair A., Lavee H., Bryan R.B. & Adar E. 1980. Runoff and erosion processes and rates in the Zin Valley badlands, Northern
Negev, Israel. Earth Surf. Process. 5: 204–225.
Yair A., Karnieli A. & Issar A. 1991. The chemical composition of
precipitation and runoff water on an arid limestone hillside,
northern Negev, Israel. J. Hydrol. 129: 371–388.
Zavala L.M., Jordán A., Gil J., Bellinfante N. & Pain C. 2009.
Intact ash and charred litter reduces susceptibility to rain
splash erosion post-wildfire. Earth Surf. Proc. Land. 34:
1522–1532.
Zhang Y. 2005. The microstructure and formation of biological
soil crusts in their early developmental stage. Chinese Science
Bulletin 50: 117–121.
Zhao Y. & Xu M. 2013. Runoff and soil loss from revegetated
grasslands in the hilly Loess Plateau region, China: Influence
of biocrust patches and plant canopies. J. Hydrol. Eng. 18:
387–393.
Zhao Y., Qin N., Weber B. & Xu M. 2014. Response of biological
soil crusts to raindrop erosivity and underlying influences in
the hilly Loess Plateau region, China. Biodivers. Conserv.
DOI 10.1007/s10531-014-0680-z
Received June 12, 2014
Accepted July 10, 2014
Unauthenticated
Download Date | 6/18/17 6:30 PM