Could Forage Palm and Stone Barrier Be as Effective as Native Vegetation in Controlling Runoff and Erosion in the Brazilian Semiarid Region?

Abstract

Most lands in the Brazilian semi-arid region are covered with native vegetation (Caatinga) with limited agricultural practices due to chronic water-scarcity issues. However, clearing the native vegetation and using the land for agriculture is becoming a common practice. The objective of this study was to evaluate the effectiveness of forage palm and stone barrier in reducing runoff and erosion compared to native vegetation. The study was conducted in the Jatobá experimental basin, Brazil, using runoff plots with three surface covers: Caatinga, forage palm barrier, and stone barrier. Data collection includes runoff and erosion rates, and soil moisture dynamics at two depths (0–20 and 20–40 cm) in response to rainfall events. Rainfall characteristics were also recorded and analyzed for 30 min for intensity (I30) and erosivity (EI30). Results showed that stone barrier and forage palm treatments were not as effective as the native vegetation Caatinga in reducing soil loss. Stone barrier was the least effective in reducing runoff and soil loss. However, runoff from forage palm was not statistically different from Caatinga. In addition, forage palm improves soil moisture dynamics at two depths. The study findings highlighted the importance of the Caatinga for soil, water and biome conservation in the region. However, the study also suggested that in the places where agricultural practices are conducted, using forage palm as a soil-conservation strategy could be a good alternative. Additional benefits of forage palm include its suitability for intercropping with other crops, and that it can serve as an alternative for animal feed in the region. Information from this study could be used to inform land management and soil- and water-conservation efforts in the semi-arid region of Brazil.

1. Introduction

The Brazilian Northeast is known for its vulnerability to climatic changes, especially in its semi-arid region. The semi-arid region is known for its low water availability, mainly due to uneven rainfall distribution that is concentrated during a short rainy season and often lasts from January to May [1]. During the rest of the year, long drought periods are common. Rainfall is characterized by high intensity and short duration.

Sparacino et al. [2], in their study in semi-arid northeastern Brazil, observed a strong seasonality of rainfall with 90% of rainfall events occurring during the rainy season, which on average lasts for 152 days. However, the onset and end dates of the rainy season also shows high interannual variability. The high interannual variability also characterizes the annual and monthly rainfall. Other characteristics of the semiarid region, are the intermittent or ephemeral rivers and streams, with limited water availability during most of the year [3].

Droughts are considered natural disasters, as the impacts of semi-aridity represent a risk framework for the population [4]. Land use changes and environmental degradation are becoming more common due to population increase. Deforestation, burning, selective logging, mining, agriculture, and livestock are some of the major land-use-change practices. Miranda et al. [5] evaluated the dynamics of land cover in the Pontal River Basin in the semi-arid region of Pernambuco, Brazil. They showed that deforestation was very aggressive with a loss of 30% of vegetation cover over the 38-year study period, and the conversion of the natural forest was primarily through transformation into rainfed pastures or agriculture, followed by irrigated agriculture. The same process has occurred in many other semiarid regions of Brazil. Deforestation over time shows an increasing threat to the natural environment. This has resulted in the clearing of native vegetation and the conversation of land to agriculture.

The original vegetation of the semi-arid region is a dry deciduous forest, predominantly composed of short trees and shrubs with a low stratum of annual herbs, locally called “Caatinga” [6]. The Caatinga presents great physiognomic variation, mainly in terms of the density and size of the plants [7]. Many of the unique properties of seasonally dry tropical forests like Caatinga depend on their rainfall regimes [8], and have efficient mechanisms of drought resistance, such as the loss of leaves during the dry season [9].

It should also be noted that small-scale farmers in the region are facing complex and interconnected issues and are struggling to survive primarily due to problems attributed to climatic and social factors. As a result, there is a growing interest in incorporating soil- and water-consecrated measures to allow farmers to grow crops without affecting the environment [10]. Therefore, the adoption of soil- and water-conservation practices becomes particularly important for farmers to perform small-scale agriculture in areas where cultivation is permitted without affecting conservation areas. Soil moisture response to precipitation is an important component of the hydrological cycle. The response of soil moisture to precipitation in arid areas has thus attracted full attention from researchers worldwide [11].

Infiltration, surface run-off, and evapotranspiration have been identified as the key processes affecting soil moisture content at fine scales, probably because at large scales, there is an increased opportunity for runoff to infiltrate through vegetation patches along the slopes [12]. However, the soils in the semi-arid region of Brazil are very stony and one way to increase soil moisture is to implement stone barriers on the ground surface to control surface runoff. The investigation of hydrological processes in this area contributes to a better scientific understanding of rainfall regimes and water availability. Borges et al. [13] used stone dams in Argisol in the semiarid region, for the cultivation of corn (Zea mays L.)and found that stone dams efficiently reduced runoff losses compared to bare soil, contributing to better use of rainwater and greater grain productivity.

Sousa et al. [3] also studied the effect of different soil covers on soil moisture dynamics in tropical dry regions and highlighted that there is a clear need to investigate the complex spatio-temporal distribution of soil in northeast Brazil. Forage cactus (Opuntia spp.) spinless is a ground cover widely used in the Brazilian semiarid region, used for livestock feed, and is one of the main sources of income in the region.

The Caatinga biome’s importance is unquestioned. It is an ecosystem with the largest vegetation coverage in northeastern Brazil [14]. The main challenge to be faced in the biome is the social vulnerability of the people in the Caatinga, leaving the preservation and conservation of biodiversity as a lower priority [15]. The unsustainable continuous exploitation of natural resources promotes environmental degradation and threatens Caatinga preservation [16]. Farming systems in semiarid regions need to be convenient for the local population, and well adapted to the local environment. The use of techniques that promote the permanent protection of soil is a promising method for semi-arid regions [17]. Thus, the objective of this study was to investigate the effects of forage palm and stone barrier in reducing runoff and erosion compared to Caatinga and improving the soil moisture dynamics in response to rainfall. The innovation of our research is the association of relevant temporal measurements of soil moisture, runoff and infiltration rates and sediment loss in natural plots.

2. Materials and Methods

2.1. Description of the Study Area

This study was conducted in the Jatobá experimental basin, which is located in the Agreste region of the Pernambuco State, Brazil (Figure 1). According to Köeppen classification, the basin has a very hot and semi-arid steppe-type climate. The average annual precipitation in the region is 607 mm, the average temperature is 23 °C, and the total potential evapotranspiration is around 2000 mm (29 years; from 1991 to 2020). The predominant vegetation is the hypoxerophilous Caatinga [18]. The basin’s drainage area is 12.56 km², with an average slope of 17.92% (Wavy Relief) and elevations ranging from 600 to 1019 m, with a runoff concentration time of 58.71 min [12].

The experiment was conducted at a hillslope of the Jatobá Experimental Basin. Before installing the experiment, a 150 cm deep trench was opened in the field for visualization and classification of the soil profile (Figure 1c). From each soil layer, both disturbed and undisturbed tree samples were collected, using auger for disturbed ones and a volumetric ring for undisturbed. The analysis following the methodology proposed by Teixeira [19] is described in

Table 1. Physical and chemical characteristics of the soil.

Parameter	Soil Profile Depth (cm)
	Ap (0–13)	A1 (13–37)	AB (37–70)	Bt (70–86)	BC (86–100)	C (100–150)
Sand (%)	23.91	17.57	12.85	12.24	13.52	22.91
Clay (%)	33.09	38.76	44.43	47.09	46.15	39.43
Silt (%)	43.00	43.67	42.72	40.67	40.33	37.67
Natural clay (%)	17.87	19.20	20.87	18.53	22.53	18.20
Pd (g/cm3)	2.56	2.59	2.33	2.61	2.67	2.66
BD (g/cm3)	1.41	1.36	1.29	1.32	1.43	1.52
OM (g/kg)	26.9	17.6	9.4	7.1	6.3	3.6
Al (cmolc/kg)	0.10	0.95	1.83	1.90	1.90	1.17
Na (cmolc/kg)	0.15	0.13	0.17	0.30	0.28	0.39
K (cmolc/kg)	0.46	0.17	0.12	0.14	0.15	0.16
H + Al (cmolc/kg)	1.05	3.41	3.88	3.77	3.33	1.95
pH	5.87	5.33	4.97	5.17	5.30	6.00
P (mg/kg)	0.11	0.00	0.03	0.05	0.05	0.05
Ca (cmolc/kg)	1.76	0.85	0.36	0.63	0.49	0.76
Mg (cmolc/kg)	0.94	0.49	0.34	1.01	1.21	2.93

.

2.2. Weather Data

Climatic data were recorded using an Automatic Weather Station (Campbell Scientific^®, Logan, UT, USA) installed nearby the experimental site (Figure 1). The station consists of a set of an anemometer, a rain gauge, a temperature sensor, and the air relative humidity and a pyranometer. Data were recorded hourly and stored in an R1000 datalogger. Adjacent to the runoff plots, an automatic rain gauge, model TB4-L Rain Gauge from Campbell Scientific, was installed, with a 0.254 mm resolution. The device was equipped with a datalogger, programmed to record precipitation every 5 min. Maximum precipitation intensity at 30 min (I30) and erosivity (EI30) were calculated according to Wischmeier and Smith [20]. In addition, historical weather data from 2006 to 2015 were analyzed.

2.3. Experimental Plots

Adjacent runoff and erosion plots were constructed on a field with a 9% slope. Plots were constructed with the following dimensions: 4.5 m wide by 11 m long (Figure 2), and plot boundaries were delimited by masonry. Each plot comprises eight PVC access tubes installed 2 m apart for soil moisture readings at two depths of 20 and 40 cm, using neutron probe equipment. The lower end of the plots has a runoff collector system consisting of a gutter connected by a PVC pipe, to a first masonry tank, located immediately downstream of the plot. This tank, in turn, was connected to the second masonry tank through a PVC pipe, through which the excess runoff was directed to the second one.

The runoff and erosion plots had three surface cover treatments: (i) Caatinga, (ii) forage palm cultivated in a double row to form barriers to contain surface runoff, at a spacing of 0.25 × 0.5 × 3.0 m, and (iii) stone barriers arranged at 1 m spacing (Figure 1b).

2.4. Soil Moisture Measurements

Soil moisture measurements were carried out biweekly using neutron probe equipment, model CPN 503 DR, from February 2009 to December 2010 (36 readings in the total). The equipment was previously calibrated, adopting a total of ten points at two soil depths (20 and 40 cm), where the neutron probe readings were transformed into volumetric moisture (cm³ cm⁻³), based on calibration curves established for each depth: y = 0.2528x + 0.0824 (R2 = 0.97) and y = 0.2604x + 0.0706 (R² = 0.97), for 20 and 40 cm depths, respectively.

2.5. Runoff and Sediment Loss

Runoff and sediment data were collected from a total of 35 rainfall events that generated surface runoff. After each natural rainfall event, the total runoff volume inside the tanks was measured, and three samples from each plot were collected in plastic bottles after mixing the runoff volume inside each tank. Then, the material collected in the plastic bottles was weighed and kept for 24 h for sediment to settle. After 24 h, the supernatant was pipetted, leaving the sediment at the bottom of the containers, which was then oven dried at 65 °C for 72 h to measure the dry soil weight [21]. Total surface runoff, the runoff coefficient, sediment concentration, and total soil loss were determined by weighing the material collected in the plastic pots. A local automatic weather station was used to measure the total rainfall.

2.6. Soil Infiltration Rate

Soil infiltration was measured using a Guelph-type constant head permeameter. This device makes it possible to monitor the infiltration rate in a borehole, under a constant hydraulic head. In this investigation, two heads H₁ and H₂ were used at 5 and 10 cm, respectively. Flow rates Q₁ and Q₂ were obtained by multiplying the measured flow rates by a coefficient corresponding to the ratio of the reservoir and borehole areas. The saturated hydraulic conductivity values were calculated following the method by Reynolds and Elrick [22], which is recommended by the soil moisture equipment corporation [23]. To determine the hydraulic conductivity, the following equations were used:

$$Ksat=G_2{Q}_2-G_{1 }{Q}_1$$

(1)

$${\mathrm{G}}_1=\frac{{\mathrm{H}}_2 {\mathrm{C}}_1}{\mathrm{\pi }\left(2{\mathrm{H}}_1{\mathrm{H}}_2\left({\mathrm{H}}_2-{\mathrm{H}}_1\right)+{\mathrm{a}}^2\left({\mathrm{H}}_1{\mathrm{C}}_2-{\mathrm{H}}_2{\mathrm{C}}_1\right)\right)}$$

(2)

$${\mathrm{G}}_2=\frac{{\mathrm{H}}_1 {\mathrm{C}}_2}{\mathrm{\pi }\left(2{\mathrm{H}}_1{\mathrm{H}}_2\left({\mathrm{H}}_2-{\mathrm{H}}_1\right)+{\mathrm{a}}^2\left({\mathrm{H}}_1{\mathrm{C}}_2-{\mathrm{H}}_2{\mathrm{C}}_1\right)\right)}$$

(3)

$$Q_1=X{R}_1$$

(4)

$$Q_2=X{R}_2$$

(5)

where Ksat is saturated hydraulic conductivity (cm/s); G₂ and G₁ are functions calculated from H₁ and H₂ hydraulic head; X is reservoir area (cm²); R₁ and R₂ are constant head (cm/s); C₁ and C₂ are form factors dependent on the H₁/a and H₂/a ratio; and a is hole radius (cm).

2.7. Soil Moisture Characteristic Curves

Soil moisture characteristic curves were determined for soil samples collected from the research site. Undisturbed samples were collected using metal rings at 20 and 40 cm depths, with three replications. Soil moisture retentions were measured at five pressure levels of 10, 33, 100, 500, and 1500 kPa using the Richards pressure chamber and fitted using the van Genutchen model [24].

2.8. Statistical Analysis

The data involve time series measurements, and the dataset collected in this study did not meet the normality assumption. The data-normality-test result using the Kolmogorov–Smirnov test confirmed that the dataset was not normal. As a result, the Kruskal–Wallis non-parametric test was employed to evaluate treatment effects. When treatment effects were significant, mean comparison tests were performed using the Dunn’s Test at a 5% significance level. All data analyses were done using the R programming language. In addition, Pearson’s correlation coefficient was used to evaluate the relationship between rainfall characteristics and runoff and sediment losses. Correlation coefficients were calculated using the SAS program [25].

3. Results

3.1. Weather Characteristics

Total annual precipitation during the study period was 1049 and 1014 mm for 2009 and 2010, respectively. The months between January and June accounted for 78% and 84% of the rainfall events in 2009 and 2010, respectively. During the study period, the temperature ranged between 21.59 °C and 35.51 °C, with an average of 23.24 °C (Figure 3).

3.2. Soil Properties

The soil in the study site has Ap and A1 soil horizons with clay loam and silty clay loam textures at 20 and 40 cm depth, respectively (Table 1). The soil of the area was classified as Argissolo Vermelho Amarelo Distrófico típico, according to the Brazilian Soil Classification System [26] and by Ultisol according to USDA Soil Taxonomy.

Soil moisture characteristic curves showed that the soil has on average a 16% field capacity measured at 33 kPa while the average permanent wilting point was 10%. This shows that the average available water content for the two depths is 6% (Figure 4).

On the other hand, the top layer has a 26% moisture content at saturation, while at 40 cm depth, the corresponding value was 27% (

Table 2. Soil characteristic curve fitting parameters based on the van Genuchten model.

Parameter	Depth (cm)
	20	40
Θs	0.26	0.27
Θ33	0.16	0.16
Θ1500	0.10	0.09
α	0.01	0.03
n	2.39	3.16
m	0.10	0.14

Θ_s: saturated soil moisture; Θ₃₃: moisture content (field capacity) at 33 kPa; Θ₁₅₀₀: moisture content (permanent wilting point) at 1500 kPa; α, n, and m are model fitting parameters.

). Saturated moisture content was estimated using the van Genuchten equation (Table 2).

In general, up to 100 cm depth, clay content increases, while send decreased as the profile depth increases (Table 1). Organic matter content was highest for the top surface horizon.

3.3. Soil Moisture Content

Results showed that treatment effects were significantly different (p < 0.05). Significant differences were observed between the Caatinga treatment and the other two treatments (stone barrier and forage palm) (Figure 5). Median soil moisture levels for the two years were 0.17 cm³/cm³, 0.18 cm³/cm³ and 0.18 cm³/cm³ for the Caatinga, stone barrier and forage palm treatments, respectively.

3.4. Runoff and Soil Loss

During the studied period, there were 35 rain events that generated surface runoff (

Table 3. Monthly and total soil loss and runoff for 35 rainfall events during the study period. Monthly values are the sums of values of all runoff and soil loss during each month.

		Treatment
		Stone Barrier	Caatinga	Forage Palm	Stone Barrier	Caatinga	Forage Palm
Month	#Events	Runoff (mm)			Soil Loss (t/ha)
Feb/09	3	23.82	8.29	6.06	5.76	1.06	0.80
Mar/09	3	1.58	0.42	0.34	0.31	0.01	0.01
Apr/09	4	74.16	24.29	40.92	151.71	53.11	109.40
May/09	9	85.41	44.49	70.24	10.70	0.58	16.61
Jun/09	3	11.11	11.11	18.36	0.13	0.13	0.22
Jul/09	1	0.46	0.30	0.04	0.01	0.01	0.00
Aug/09	3	12.45	11.01	10.48	0.08	0.07	0.12
Dec/09	3	0.51	0.67	3.30	0.01	0.04	0.05
Jan/10	4	4.70	1.39	1.23	0.17	0.05	0.15
Feb/10	2	26.90	9.70	30.63	0.11	0.06	0.68
Total	35	241.1	111.7	181.6	169.0	55.1	128.0

). The highest number of events with erosive rainfall potential compared to other months were recorded in May 2009. However, the highest runoff and soil loss were recorded in April 2009.

Caatinga treatment resulted in reduced runoff and soil losses compared to stone barrier and forage palm while stone barrier resulted in the highest runoff and sediment losses of all treatments. Forage palm was intercropped with other crops to constituent as an efficient way for agricultural cultivation and soil conservation. It was observed that the runoff coefficient varied from 0.48 to 0.94 with the highest runoff coefficient being from stone barrier treatment (

Table 4. Summary of runoff coefficient.

Treatment	Max	Mean
Forage Palm	0.48	0.18
Caatinga	0.46	0.12
Stone Barrier	0.94	0.25

).

Pearson correlation coefficients between observed variables showed that runoff had a significant positive correlation (p < 0.01) with total rainfall, erosivity (EI30) and maximum 30 min of rainfall intensity (I30) (

Table 5. Pearson correlation coefficients of the variables related to the erosion process at the plot scale.

	EI30	I30	R	TR	RC	TS	SC
EI30	1
I30	0.88 *	1
R	0.72 *	0.63 *	1
TR	0.40 *	0.38 *	0.46 *	1
RC	0.11	0.12	0.09	0.83 *	1
TS	0.07	0.05	0.13	0.62 *	0.62 *	1
SC	0.03	0.002	0.05	0.34*	0.79 *	0.48 *	1

EI30—erosivity (Mj mm/ha h). I30—maximum intensity in 30 min (mm/h). R—rainfall (mm). TR—total runoff (mm). RC—runoff coefficient. TS—total sediment (t/ha). SC—sediment concentration (g/L). *—(p < 0.001).

).

Similarly, total sediment (TS) and sediment concentration (SC) also had significant correlation with total runoff and runoff coefficient.

Soil loss and sediment concentration were the lowest from plots with Caatinga treatment (p < 0.05) (

Table 6. Summary of average soil loss and sediment concentration.

Treatment	Sediment Concentration (g/L)	Soil Loss (t/ha)
Forage Palm	39.80	3.77
Caatinga	25.98	1.62
Stone Barrier	33.26	5.12

).

The surface runoff and runoff coefficient showed higher variation for the stone barrier treatment compared to Caatinga and forage palm (Figure 6). The Caatinga treatment reduced the average surface runoff value by 43.3% compared to the stone barrier treatment (p < 0.05).

Moreover, sediment load was lower for the Caatinga treatment compared to the other treatments (p < 0.05) (Figure 7).

4. Discussion

Regarding the weather variables, 84% of annual rainfall is concentrated between January and May [1]. Soil in the region has high infiltration rates with an average of 32.6 mm/h [27]. Santos and Montenegro [28], in their study based on 29 years of rainfall data for the region where this study was conducted, found that the most erosive rainfall events of the year were concentrated in just three months (February, March and April).

Soil moisture readings showed more dynamics on the top layer compared the lower soil depth. Similar results were observed, of the upper layers as more dynamic than the lower layers, by Queiroz et al. [10]. The authors investigated soil moisture dynamics for soil layers between 5 and 55 cm under three different surface conditions and found that the seasonal moisture fluctuation observed was greater near the soil surface.

The stone barrier treatment has no vegetation while the forage palm is distributed on the soil surface with vegetation in contours at regular spacings. For these two treatments, the average values of soil moisture levels were slightly higher compared to the Caatinga, which suggests that stone barrier and forage palm could be used as soil management to be implemented for cropping in the region, as a soil moisture conservation strategy. Queiroz et al. [10] found higher moisture levels from soils under cacti compared to bare soil. They attributed increases in soil moisture due to cactus root configuration, which in turn results in improved soil moisture retention. The soil moisture values found in the present work are close to the values found by Sousa et al. [3], who evaluated soil moisture for the Caatinga in the semi-arid region as close to 20%.

Slight improvements in soil moisture retention in arid and semi-arid regions is critical for crop production. In work developed by Santos et al. [29], who analyzed the influence of precipitation on soil moisture, comparing it with different soil surface cover conditions, it was found that the Caatinga presented significantly lower soil moisture values than the mulch condition, being attributed to the evapotranspiration factor, which during the dry season is accentuated in this biome. As the data presented in Figure 6 average over a long period, including the wet and dry seasons, evapotranspiration changes may be the cause of lower values for this surface condition.

Observed reduction of moisture from Caatinga plots can be attributed to plant interception that prevents water reaching the ground resulting in subsequent loss of intercepted water through evaporation. Wang et al. [30] reported a similar finding regarding the evaluation of soil moisture under different soil cover conditions in a semi-arid region of China, where soils under shrubs and forests were characterized by lower soil moisture compared to soil moisture levels under corn.

Observed runoff and soil losses were in agreement with reports from several studies including Sousa et al. [1] and Santos and Montenegro [28] and are consistent with the study areas’ rainy season, which is concentrated in the first half of the year.

Measured runoff in this study was consistent with that reported by Anache et al. [31]. The authors in their study in Brazil reported from experimental plots of Caatinga an average of 115 mm/year of runoff, which is comparable with the 111.7 mm/year of runoff observed in this study. However, observed soil loss in this study was higher than reports by Anache et al. [31] for similar soil conditions. Their values of soil loss were 0.7 Mg ha⁻¹ year⁻¹ and 18 Mg ha⁻¹ year⁻¹ for shrubland (similar to Caatinga) and cropland (similar to forage palm), in Northeast Brazil.

In relation to the Pearson correlation, similar results were reported by Santos et al. [26] who found significant correlation between sediments and rain characteristics when they were evaluated at the watershed scale. This suggests that soil loss could be limited by the transport capacity of rainfall. However, sediment transport processes at watershed scale may not be directly applicable to plot level due to heterogeneity (soil and vegetation cover characteristics) at watershed scale, which affects the generation of runoff and sediment transport. Sadeghi et al. [32] found that the smaller the plot, the larger the hydrological disconnection within the system, and the lower the energy flows due to short distances, and the quicker the response to runoff due to an artificial decrease of concentration times for continuous flow. Findings by Liu et al. [33] and Santos and Montenegro [28] suggest a linear positive relationship between annual rainfall amount and annual rainfall erosivity. Rainfall intensity affects several surface hydrological processes including soil erosion in agricultural environments and drylands [34]. Rainfall intensity is an essential variable affecting runoff, and its magnitude directly affects runoff amount [35]. Radatz et al. [36] also found a positive correlation between I30 and runoff. But, the values of runoff decreased by increasing soil coverage in a direct plant system, distributed along a watershed with different crops like oat, corn and alfafa. Significant amounts of organic residues in varying stages of decomposition that accumulate on the soil surface in the direct plant farming system may be related to these differences by increasing infiltration and percolation of precipitation.

The 9% runoff coefficient in the Caatinga treatment is close to the average values of the 15% runoff coefficient reported by Santos et al. [12] from their study using 20 m2 plots with Caatinga. However, cutting Caatinga vegetation for firewood and opening areas for livestock and agriculture cause changes in hydrological and sedimentological processes [37]. Forage palm can also be considered an efficient alternative, which was not statistically different from the Caatinga treatment. During the rainy season, Caatinga reduces surface runoff since it is represented by small trees, shrub species and grass that improve infiltration [10].

5. Conclusions

This study investigated the effects of forage palm and stone barrier in reducing runoff and erosion and improving the soil moisture dynamics compared to a native plant, Caatinga, in response to rainfall. The soil moisture dynamics, runoff and erosion events were evaluated in an Argisol located in a semi-arid area of Brazil. Study findings showed that that approximately 17% of the rainfall events resulted in surface runoff. Total sediment load showed significant correlation with the runoff but not with rainfall characteristics (EI30, I30 and total rainfall). Stone barrier and forage palm treatments were not as effective as the native vegetation Caatinga in reducing soil loss. Stone barrier was the least effective in reducing runoff and soil loss. However, runoff from forage palm was not statistically different from Caatinga. In addition, forage palm improves soil moisture dynamics at two depths. The study findings highlighted the importance of the Caatinga for soil, water and biome conservation in the region. However, the study also suggested that in the places where agriculture practices are conducted, using forage palm as a soil conservation strategy could be a good alternative. Additional benefits from forage palm include its suitability for intercropping with other crops and that it could also serve as an alternative for animal feed in the region. Information from this study could be used to inform land-management and soil- and water-conservation efforts in the semi-arid region of Brazil. We suggest for future studies, testing should be conducted using the parameters used in our experiment for different types of soil with the aim to calibrate the proposed model in other environmental conditions. In addition, the knowledge of palm spacing and consortium would provide relevant information for farmers and scientists.