Ecosphere - 2022 - Galindo - Land Use Conversion To Agriculture Impacts Biodiversity Erosion Control and Key Soil

Received: 1 June 2021 Revised: 30 September 2021 Accepted: 8 October 2021
DOI: 10.1002/ecs2.3979
ARTICLE
Agroecosystems
Land use conversion to agriculture impacts biodiversity,

erosion control, and key soil properties in an Andean
watershed
Víctor Galindo1,2 | Carolina Giraldo1 | Patrick Lavelle3 | Inge Armbrecht2 |

Steven J. Fonte4,5
1
Fundaci
on CIPAV, Cali, Colombia
2 Abstract
Universidad del Valle, Facultad de
Ciencias, Biology Department, Cali, The conversion of natural vegetation to agricultural land uses in mountainous
Colombia Andean landscapes threatens an array of key ecological processes and ecosys-
3
Sorbonne Université, Institute of tem services (ES). In protected areas and buffer regions that provide water to
Ecological and Environmental Sciences,
Paris, France cities, it is critical to understand how interactions between plants and soil
4
Department of Soil and Crop Sciences, communities sustain a range of ecosystem functions associated with nutrient
Colorado State University, Fort Collins, recycling, soil structure, and erosion control. We sought to examine how land
Colorado, USA
5
use conversion within a mountainous tropical forest landscape influences the
Graduate Degree Program in Ecology,
Colorado State University, Fort Collins,
diversity of vegetation and soil macrofauna communities, soil physicochemical
Colorado, USA properties, and hydrological regulation services. Biodiversity and a suite of key
soil-based ES were compared in five major land uses of the Cali River water-
Correspondence
Steven J. Fonte shed: (1) annual cropping systems, (2) coffee plantations, (3) pastures, (4) aban-
Email: steven.fonte@colostate.edu doned shrubland, and (5) secondary forests. The diversity of woody and
herbaceous vegetation, as well as soil macrofauna, was assessed in each land
Funding information
Colciencias - Colombia, Grant/Award use. Soil chemical fertility and aggregate morphology were assessed via labora-
Number: FPA44842-042-2016; Fundaci on tory analyses and visual separation of soil aggregates based on their origin.
CIPAV
Infiltration, runoff, and sediment production were measured using a portable
Handling Editor: Erika Foster rainfall simulator. We found a decrease in the diversity of woody vegetation
across land uses to be associated with lower diversity of soil macrofauna. At
the same time, agricultural management, annual crops in particular, supports
the largest earthworm populations, likely due to increased organic inputs and
low impact tillage, which appears not to diminish soil fertility and water infil-
tration. In contrast, the low soil fertility in pastures was associated with the
lowest values of soil C, poor aggregation, and high bulk density and likely
reflects overgrazing, with negative implications for water infiltration and ero-
sion. Associations between the different sets of variables, evaluated with co-
inertia analysis, highlight the hierarchical relevance of plant cover and woody
diversity on ES. The biological complexity associated with intact forest cover
appears to generate “bundles” of co-occurring ES, with this land use
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
© 2022 The Authors. Ecosphere published by Wiley Periodicals LLC on behalf of The Ecological Society of America.
Ecosphere. 2022;13:e3979. https://onlinelibrary.wiley.com/r/ecs2 1 of 19

https://doi.org/10.1002/ecs2.3979
21508925, 2022, 3, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.3979 by CAPES, Wiley Online Library on [30/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 19 GALINDO ET AL.
demonstrating the highest infiltration, and low runoff and sediment losses.
Our findings demonstrate that forests and tree-based agricultural systems may
better contribute to the provision of multiple ES, including biodiversity conser-
vation and hydrologic regulation.
KEYWORDS
ecosystem services, hydrologic function, soil aggregation, soil macrofauna, woody and
herbaceous vegetation
INTRODUCTION impact and the formation of stable biogenic structures by

edaphic fauna (Lavelle, 2002). These biological soil aggre-
In the 20th century, large-scale transformation of natural gates have greater resistance to degradation and support
ecosystems toward intensified agriculture and livestock multiple ES such as soil C storage, infiltration, and nutri-
operations has contributed to greater food production, ent retention (Brussaard, 2012; Elliott & Coleman, 1988;
but this has occurred at the expense of key regulating Lavelle et al., 2020; Zhang et al., 2013). Although the
and supporting ecosystem services (ES) (Millennium Eco- importance of forests for the conservation of soil biodi-
system Assessment, 2005; Tallis & Polasky, 2009). This versity and the maintenance of soil structure is well rec-
decline in the provision of ES at a global scale highlights ognized (Brussaard, 2012; Lavelle, 2002), impacts of land
the need to develop alternative agroecosystems that can use conversion on both plants and soil macrofauna, and
maintain productivity, but also generate multiple ES, the implications for diverse soil processes remain poorly
especially in areas of high biodiversity and ecological fra- understood; this is especially true for aboveground–
gility (Bennett et al., 2009). In the Andean region, the belowground biological interactions that influence hydro-
intensification of these agricultural landscapes can have logical regulation and related ecological functions.
considerable effects on ES with regard to soil processes, The development of effective public policies is often
leading to decreased fertility (Caulfield et al., 2020; de limited by a lack of knowledge in priority conservation
Valença et al., 2017), reduced water infiltration potential, areas. The region under consideration here, the
lower water holding capacity, and decreased soil stability Farallones de Cali National Park (in western Colombia)
and erosion control (Guzman et al., 2019; Otero and its buffer areas, represents a key biodiversity hot spot
et al., 2011; Zimmerer, 1993). Increased use of fertilizers (Myers et al., 2000), but it is representative of many tropi-
with agricultural intensification can increase food and fiber cal areas and remains poorly studied with regard to biodi-
production (Rolando et al., 2018), but often diminishes the versity and associated ES provision. This region forms
provision of other ES related to hydrological function and part of the Cali River watershed and comprises conserva-
water quality (Bennett et al., 2009; Rolando et al., 2017). tion and special management areas with great impor-
Conversion of natural vegetation to agriculture and tance for the provision of water to the city of Cali, with
other land uses can also affect local biodiversity, including nearly 2.2 million inhabitants. Over the last several
belowground organisms (Grimaldi et al., 2014; Lavelle decades, much of this mountainous landscape surround-
et al., 2014; Negrete-Yankelevich et al., 2007) with impor- ing the city has gone through a process of transformation,
tant implications for hydrological regulation (Bruijnzeel, from native forest to agricultural production. The land
2004). In some highland regions of the Andes, the transfor- uses now occupying formerly forested areas include cof-
mation of natural cover can lead to irreversible changes in fee plantations, annual crops (mainly vegetables), cattle
soil structure and function. For example, conversion of par- grazing, and abandoned areas dominated by shrubs that
amo to pine plantations in Ecuador resulted in a 50% have not yet returned to forest (UNIVALLE CVC, 2007).
decrease in water yield (Buytaert et al., 2007). Also, it has Although local legislation aims to regulate land manage-
been shown that conversion of natural ecosystems to culti- ment impacts and this region is located in a protected
vated and intensively grazed land can lead to increased zone for providing water to the city, knowledge regarding
peaks of water yield and severe reduction during low flows the impacts of land use intensification on biodiversity
(Crespo et al., 2010). and the provision of multiple ES remains limited. The
In contrast to more intensive agricultural practices, present study sought to address this knowledge gap by
agroecosystems that conserve surface residues and examining the effects of common land uses on biodiver-
healthy soil macrofauna communities (e.g., agroforestry) sity (of plant and soil macroinvertebrates), soil physico-
help to improve soil aggregation, due to reduced raindrop chemical properties, and hydrological properties. We
ECOSPHERE 3 of 19
hypothesized that (1) the conversion from forest to agri- are spinach (Spinacia oleracea), chayote (Sechium
cultural land uses decreases soil chemical fertility and edule), squash (Cucurbita pepo), and herbs, including
water infiltration, while increasing runoff and sediment rosemary Rosmarinus officinalis, thyme Thymus
loss via erosion, and (2) the provision of soil-based ES vulgaris, and Allium fistulosum. Weeds in this system
decreases with decreasing diversity of plant and soil were controlled regularly by hand. The areas sampled
macroinvertebrate communities along a gradient of for pasture were selected from small, extensively man-
intensifying land use. aged cattle operations on hilly terrain with low stocking
rates of 1.2 0.9 AU ha1 (Gonzalez-Quintero
et al., 2020). Ten representative plots were selected for
MATERIALS AND METHODS each land use with a randomized selection of plots for
each category selected across the landscape. In each of
Study area and sampling design the 50 plots selected, a sampling area of 30 20 m was
oriented with the long axis along the contour of the
The study took place in a forest reserve area of the Cali slope. Within each sampling area, we conducted a
River watershed (12,526 ha), in the southwestern Andean detailed characterization of the woody and herbaceous
range (Cordillera Occidental; between 3 270 2800 N and vegetation as well as thorough evaluation of soil biologi-
76 360 3400 W), in the municipality of Cali, Valle del Cauca cal and physicochemical properties. The sampling was
Department, Colombia. The study area ranged from 1000 conducted during the last rainy season of 2016
to 2300 m in elevation and is comprised largely of pre- (October–December) and during the first rainy season of
montane moist forest (bh-PM) and wet forest (bmh-PM) 2017 (April).
Holdridge ecological zones. Mountainous relief and steep
terrain dominate the landscape with shallow valleys and
a dense network of first- and second-order streams. Soils Ecosystem services framework
are predominantly of volcanic origin and have a fine silt-
clay texture that is highly susceptible to erosion The ES categories and groups defined by the Common
(UNIVALLE CVC, 2007). With a mean annual tempera- International Classification of Ecosystem Services
ture of 24 C in the lower course of the Cali River, mean (Haines-Young & Potschin, 2018) were used to develop
annual precipitation ranges from 1300 to 2400 mm in the “proxy” measures for key regulating ES. For hydrological
highest elevation, largely distributed between two rainy regulation, we evaluated infiltration, runoff, and sedi-
periods (March–May and September–November). In the ment production. Soil fertility and structure were
upper part, the reserve area is dominated by secondary assessed by evaluating soil chemical fertility (based on
forest (47%) and agricultural land (35%) that includes a standard laboratory measures) following standardized
mosaic of relatively intensive cropping systems and analytical procedures (Motsara & Roy, 2008), and the
extensively grazed grasslands, as well as shrubland (13%), morphology of soil aggregates was separated visually
and rural population centers (5%). The farming systems (Velasquez et al., 2007). Aboveground biodiversity was
are comprised of both perennial cropping systems also evaluated through the structure and diversity of veg-
(mainly coffee, with some bananas and other fruit tree etation, while belowground diversity and activity were
plantations) and annual crops of herbs and vegetables evaluated via soil macrofauna communities and their
(Fondo Patrimonio Natural et al., 2014; UNIVALLE function.
CVC, 2007) (Figure 1).
In the present study, the most representative land
uses were selected for evaluation: (1) farms with annual Soil fertility
crops, (2) coffee plantations, (3) pasture areas with low-
intensity cattle grazing, (4) abandoned pastures with Soil sampling was conducted during the last rainy season
shrub regrowth (following 2–3 years of succession), and of 2016 (December). Five equidistant soil subsamples
(5) secondary forests (with no evidence of human distur- (0–15 cm) were taken along a diagonal in each 20 30 m
bance for at least 15 years) (Figure 1). The density in sampling plot using a soil corer to obtain a 1 kg com-
coffee plantations ranged from 2500 to 4000 plants ha1, posite sample for analyses of soil chemical fertility
with less than 20 shade trees ha1, while agrochemical parameters. Samples were generally collected in between
inputs (herbicides, insecticides, and fertilizer) were used existing plants, to minimize the effect of any particular
regularly. Annual cropping systems were mostly fertil- species. Analyses were conducted following methods out-
ized using poultry manure (8–10 Mg ha1) and relied on lined by Motsara and Roy (2008). In each sampling
green mulching. The most common crops in this system area, pH was measured using potentiometric titration
F I G U R E 1 View of (a) multiple land uses within a transformed pre-montane moist forest landscape, with examples of study sites in
(b) secondary forest, (c) pasture, (d) abandoned lands, (e) annual crops, and (f) coffee plantations, in the southwestern Andean range, near
Cali, Valle del Cauca, Colombia
(1:1 [v/v] soil to water suspension) and total C (C) and method at pH 7, and cation exchange capacity was deter-
total N (TN) were determined by dry combustion using mined through the sum of interchangeable bases and the
an elemental analyzer. Exchangeable Ca, K, Mg, and Na resulting acidity at pH 7. KCl-extractable Al was deter-
were determined with the ammonium acetate extraction mined by titration with 0.01 N NaOH to endpoint pH 7.8.
ECOSPHERE 5 of 19
Available P (P-Bray II) was evaluated with ammonium earthworms and other soil macrofauna. Root aggregates
fluoride and hydrochloric acid. Ammonium (NH4) and are generally observed as small and rounded forms that are
nitrate (NO3) were determined using KCl extraction attached to roots. Aggregates of physical origin have angu-
(Motsara & Roy, 2008). Soil texture was determined by lar forms and plane surfaces (Lavelle & Spain, 2005). We
hydrometer method with sodium hexametaphosphate as evaluated for each sample the dry mass of biogenic aggre-
a dispersant (Bouyoucos, 1962). The above analyses were gates, root aggregates, physical aggregates, litter, roots, and
conducted at the Universidad Nacional de Colombia, in rocks. The different fractions were dried in an oven at
Bogota. Soil bulk density was estimated on five subsam- 105 C and their dry mass evaluated.
ples per plot by inserting a metal cylinder (5 cm
diameter 5 cm in length) to a depth of 7.5 cm. Soils
were carefully excavated and dried in an oven at 105 C Plant biodiversity and cover
before weighing.
Two vegetation categories were considered: woody and
herbaceous plants. Other woody plants as palms
Infiltration, runoff, and erosion control (Arecaceae), bamboos (subfamily Bambusoideae), and
arboreal ferns were also recorded. All trees and woody
In order to estimate infiltration, water retention, and ero- individuals with a diameter ≥1 cm at breast height (1.3 m
sion, a portable rainfall simulator was used and adjusted above the soil surface) were considered. Woody plants
to the average precipitation of the region (Otero were evaluated along six transects of 30 m 2 m, rep-
et al., 2011). The device was installed 1.3 m above the resenting 60% of the plot area. In the same study plots,
soil, with 240 drippers of 0.6 mm in diameter distributed the herbaceous plants were evaluated at five different
over an area of 0.14 m2. The simulator has an intensity sampling points using a 1 1 m quadrat oriented along
range of 80–150 mm h1, a raindrop velocity of 4.5 m s1, the plot diagonal. Plant species composition was com-
and a range of kinetic energy between 14.05 and pared among land uses considering species richness and
26.34 J mm1 m1 (Cobo & Amézquita, 1999; Otero diversity (Shannon diversity index H0 ).
et al., 2011). The rain simulator was used for periods of Overall vegetative cover was estimated using a verti-
30 min and conducted at three subsampling points in cal densiometer (GRS Densiometer) (Stumpf, 1993) at
each plot. The volume of runoff was measured every 100 points (2 m spacing) in each 20 30 m plot. Plant
5 min during each 30-min period using a canal at the bot- cover was recorded both, above and below 1.7 m, at each
tom of each sampling area that permitted the collection point including trees, shrubs, vines, epiphytes, herba-
of water in plastic containers. In each study plot, the soil ceous plants, litter, seedlings, grasses, coarse wood debris,
moisture content (0–5 cm) was determined before and bare soil, and rocks. To investigate their association with
after each test. Infiltration was estimated by measuring soil characteristics and ES, the general abundance, indi-
the difference between the volume of water applied and ces of diversity, and the percent coverage of growth types
that collected as runoff during the sampling period. In of woody and herbaceous were considered.
order to estimate erosion, the runoff samples were dried
in an oven at 105 C and the mass of sediment produced
per unit area was calculated. The values reported for run- Biodiversity of edaphic macrofauna
off, infiltration, and sediments correspond to the average
of three subsamples in each plot. Average slope (%) was Sampling of macrofauna was conducted using five soil
also evaluated for each plot using the elevation difference monoliths (25 25 10 cm) following the ISO/TSBF
between two terrain data points. methodology (ISO, 2011). Soils were excavated, and sam-
ples were hand-sorted for removal of all visible inverte-
brates. Specimens were identified and separated
Morphology of soil aggregates manually and then preserved in a mix of 70% alcohol or
4% formaldehyde. The individuals collected were counted
Aggregate fractions were separated according to Velasquez and separated into broad taxonomic groups, generally at
et al. (2007). Five soil monoliths (10 10 10 cm) were the level of order in the class Insecta: Blattodea (includ-
collected adjacent to the samples for soil fertility analysis. ing suborder Blattoidea, infraorder Isoptera), Isopoda,
Soils within each monolith were separated by hand Dermaptera, Hemiptera, Coleoptera (adults and larvae),
according to their size and visual inspection to determine Orthoptera, Lepidoptera (larvae), and Diptera (larvae).
biological or physical origin. Biogenic aggregates have The family Formicidae in the Hymenoptera was treated
rounded forms generally associated with the activity of as a major group. Classes Arachnida, Chilopoda,
Diplopoda, Gastropoda, and Oligochaeta were also con- conducted in the R environment using vegan, MASS, and
sidered. Samples were taken adjacent to the sampling ade4 packages (R Development Core Team, 2018).
points for aggregate morphology and soil fertility.
RESULTS
Statistical analysis
Soil physical and chemical properties
Variables were grouped into four databases at the plot
level (n = 50): (1) data for hydrological ES In the 50 farms evaluated, clay loam soils predominated
(e.g., infiltration, runoff, and sediment and aggregate with no significant texture differences between land use
morphology); (2) soil physical and chemical properties; types (Table 1). Bulk density was significantly different
(3) plant cover, abundance, species richness, and Shan- between land uses, with low values in forests and higher
non diversity of woody and herbaceous plants, as well as values in abandoned lands and coffee plantations.
vegetative cover; and (4) soil macrofauna variables, In general, the soils were less acid in the annual
including mean abundance, richness, and Shannon diver- cropping systems (5.7) than in abandoned lands (5.1), pas-
sity of the principal taxa. tures (5.0), coffee systems (4.9), and forests (4.7). Al satura-
Land uses were compared for the different soil and tion was relatively high in forests (1.9 meq 100 g1), but
vegetation parameters using one-way analysis of variance not within the toxic range for the region (Ortíz-Escobar
(ANOVA) with Tukey tests for mean comparisons. In et al., 2004), and low in other land uses (0.8–
each case, data for each group of soil and plant variables 1.0 meq 100 g1). Total C and N contents exhibited highest
were checked for assumptions of normality and homoge- values in annual cropping plots and forests (6% C and
neity of variance and transformed as needed to meet the 0.3% TN), while the lowest values were found in aban-
assumptions of ANOVA. For the comparison of variables doned pastures (2.3% C and 0.1% N; Table 2). Available K,
expressed as percentages, such as plant community com- P, and N (as NO3) were significantly higher in soils with
position and macrofauna, generalized linear models were annual crops than in other land uses (Table 2).
used. For the data with Poisson distributions (count PCA of soil physical and chemical parameters con-
data), the deviation was approximated to a χ 2 distribution firmed these trends. Principal component (PC) 1
with the goal of realizing a significance test from compar- explained 29.1% of the variation across sites and was
ing all models with and without the principal effect (land largely associated with indicators of soil fertility. This axis
use types). For each database, exploratory analyses were separated annual crops (with high nutrient availability
implemented to visualize the relationships between vari- and pH) from other land uses (Appendix S1: Figure S1).
ables without issues of collinearity using the principal PC 2 (19.6% of the variance explained) opposed those
component analysis (PCA). A Monte Carlo test of permu- sites with low soil pH and high organic matter accumula-
tation was also conducted to test multivariate differences tion in forest sites with respect to sites with active or
among land uses. Additionally, we used co-inertia analy- abandoned cropping systems. Multivariate separation of
sis (Dolédec & Chessel, 1994; Dray et al., 2003) to explore the land uses by physiochemical variables was highly sig-
covariation among the four datasets. All analyses were nificant and explained 27.5% of the variance.
T A B L E 1 Soil physical characteristics across five land use types in the Cali River watershed (mean SE and p value of each parameter)
sampled in December 2016
Property Forest Abandoned Pasture Coffee Annual crops F df p

BD 0.82 0.07b 1.21 0.08a 1.11 0.11ab 1.20 0.05a 0.91 0.08ab 4.8 4, 45 0.002
VWC (%) 34.7 3.8 ns 44.8 2.7 ns 45.4 3.6 ns 35.38 2.8 ns 40.6 3.7 ns 2.2 4, 45 0.079
Clay (%) 19.8 1.9 ns 28.2 2.8 ns 20.6 2.8 ns 26.6 2.8 ns 22 3.0 ns 1.8 a
4, 49 0.134
Silt (%) 32.0 3.4 ns 35.0 3.0 ns 32.8 3.0 ns 36.6 2.8 ns 32.4 2.7 ns 0.4 a
4, 49 0.789
Sand (%) 48.2 5.0 ns 36.8 5.4 ns 46.6 5.6 ns 36.8 4.5 ns 45.6 5.5 ns 1.1a 4, 49 0.347
Slope (%) 41.5 2.9a 40.7 3.1a 31.3 2.6ab 38.0 4.3ab 27.2 2.3b 3.4 4, 45 0.016
Note: Values with different letters indicate statistical differences between different land uses, according to Tukey tests.
Abbreviations: BD, bulk density; ns, not significant (p > 0.05); VWC, volumetric water content.
a
Generalized linear model.
ECOSPHERE 7 of 19
TABLE 2 Chemical fertility parameters across five land use types in the Cali River watershed (mean SE and p value of each
parameter)
Property Forest Abandoned Pasture Coffee Annual crops F df p

1
Ca (meq 100 g ) 12.2 3.6 ns 15.5 3.9 ns 13.6 3.8 ns 10.9 2.4 ns 15.6 1.9 ns 0.4 4, 45 0.804
1
K (meq 100 g ) 0.4 0.0b 0.1 0.0b 0.3 0.1b 0.4 0.1b 1.4 0.2a 17.1 4, 45 0.001
1
Mg (meq 100 g ) 3.5 0.8 ns 6.0 1.8 ns 3.3 1.1 ns 2.7 0.5 ns 3.3 0.5 ns 5.09 4, 45 0.729
Na (meq 100 g1) 0.1 0.0b 0.1 0.0b 0.1 0.0b 0.1 0.0b 0.2 0.0a 10.5 4, 45 0.001
1
Al (meq 100 g ) 1.9 0.6a 1.0 0.4a 0.8 0.4a 1.0 0.4a 0.1 0.0b 2.7 4, 45 0.042
1
CICE (meq 100 g ) 17.9 4 ns 22.7 5.2 ns 18.2 4.6 ns 15.1 2.5 ns 20.7 2.4 ns 0.5 4, 45 0.701
pH 4.7 0.2b 5.1 0.2ab 5.0 1.2ab 4.9 0.2b 5.7 0.1a 3.5 4, 45 0.014
C (%) 6.2 1.1a 2.3 0.3c 5.5 0.7ab 3.4 0.2bc 6.0 1.0a 5.92a 49 0.001
TN (%) 0.29 0.04a 0.12 0.02c 0.25 0.03ab 0.17 0.01bc 0.29 0.03a 8.5a 49 0.001
1
Avail P (mg kg ) 15.7 11.2b 4.8 2.2b 18.2 11.4b 6.2 2.1b 56.2 11.8a 7.6 4, 45 0.001
1
NO3 (mg kg ) 1328 172.8a 397 126.3bc 834 230.5ab 810 124.4ab 1482 267.1a 5.1 4, 45 0.002
NH4 (mg kg1) 24.1 10.2 ns 29.7 5.7 ns 30.6 9.2 ns 23.5 6.5 ns 24.6 4.2 ns 0.2 4, 45 0.934
Abbreviations: Al, aluminum; Ca, calcium; CICE, cationic interchange capacity; CO, total carbon; K, potassium; Mg, magnesium; Na, sodium; NH4, ammonia;
NO3, nitrates; NT, total nitrogen; P, available phosphorous; pH, acidity.
a
Hydrological function and aggregate pastures, and coffee cultivations, there was a higher pro-
morphology portion of surface roots (1.7%–0.7% dry mass), compared
to annual crops (0.3% dry mass; Table 3).
Infiltration rate was highest in forests and coffee planta- PC 1 (26.7% of the variance) was associated with infil-
tions (21.0 and 16.5 cm min1, respectively) and lowest tration rates, sediment production, runoff, and presence of
in abandoned lands, pasture, and annual crops (between roots (Figure 3). This axis separated forests (with high
12.7 and 14.6 cm min1). Runoff was higher in the areas water infiltration rates and fine superficial roots and root
of abandoned land and pastures (4.6 and 3.6 cm min1, associated aggregates) from the other land uses that had
respectively) and lowest in forests (0.7 cm min1). Sedi- higher runoff and sediment production. PC 2 (18.4% of var-
ment production was greater in agricultural land uses iance explained) opposed annual cropping, with a large
(7.2–13.6 Mg ha1), especially in sites with bare soils proportion of non-macroaggregated soil, from land uses
(i.e., abandoned land, pastures, and annual crops), and with higher proportions of root aggregates (Figure 3).
lowest in forest areas (1.1 Mg ha1) (Figure 2).
In general, biogenic and root aggregate were domi-
nant and together comprised 77% of the soil mass. In cof- Soil macrofauna
fee plantations, abandoned land, and annual crops,
biogenic aggregates represented the largest proportion In total, 10,768 individuals (ind) were collected, which
(ranging from 57% to 66% of the dry mass), followed by were separated into 95 taxonomic groups, (at the level of
forests and pastures (34%–43%; Table 3). Root aggregates orders, families, and genera), and further grouped into
predominated in pastures and forests (47% and 31% dry 16 coarser taxonomic groups for analysis. The highest
mass, respectively), followed by abandoned land and cof- abundance was found in annual cropping systems
fee plantations (23% and 22% dry mass, respectively). In (1317 485 ind m2), and the lowest in abandoned
annual cropping systems, the presence of root aggregates lands (400 58 ind m2; Figure 4). Mean richness (con-
was notably low (0.2% dry mass). Non-macroaggregated sidering the 16 taxonomic groups selected) was highest in
soil was highest under annual cropping (38% dry mass) forests (10 1 taxa plot1) (Figure 4), followed by coffee
and forests (23% dry mass), followed by pastures, aban- plantations, pastures, annual croplands, and abandoned
doned land, and coffee plantations (Table 3). Litter mass lands (6 1 taxa plot1). Similarly, Shannon diversity
was similar among forests, coffee plantations, annual was highest in forest soils (1.8) and considerably lower in
crops, and abandoned land (0.3%–0.9% dry mass) but the agricultural systems (ranging from 1.1 to 1.4)
lowest in pastures (0.1% dry mass). In forests, abandoned, (Figure 4). The most abundant group was Formicidae
Infiltration Runoff Sediments

a 6 a a
a
20 ab 15 a
ab
b b
15 b 4
–1
Mg ha–1
a
cm min –1
10
cm min
bc bc
10
2
5
5 c
b
0 0 0
A AC CP F P A AC CP F P A AC CP F P
F I G U R E 2 Values (mean SE) for infiltration, runoff, and sediment production (erosion) in the five dominant land use types
(secondary forest [F], abandoned land [A], pasture [P], coffee plantations [CP], and annual crops [AC]) of the Cali River watershed. Values
with different letters indicate statistical differences between different land uses, according to multiple Tukey tests
T A B L E 3 Percent dry mass of aggregates, litter, and roots (based on hand-sorting according to their biological or physical origin;
mean SE) in the five dominant land use types of the Cali River watershed
Parameter Forest Abandoned Pasture Coffee Annual crops Fa df p

Litter mass (% dry mass) 0.9 0.3a 0.6 0.1ab 0.1 0.1b 0.3 0.2ab 0.7 0.4ab 2.6 4, 44 0.05
Roots (% dry mass) 1.7 0.5a 0.7 0.1abc 1.0 0.2ab 0.7 0.1abc 0.3 0.1c 6.3 4, 44 0.001
Soil aggregation
Biogenic aggregates 42.7 3.3bc 63.2 7.8a 33.9 4.1c 65.7 3.2a 57.1 7.2ab 6.8 4, 45 0.001
(% dry mass)
Root aggregates (% dry mass) 30.9 3.0ab 23.2 7.4b 47.5 5.0a 22.1 3.1b 0.2 0.1 4.6 3, 36 0.007
Physical aggregates 0.2 0.1 ns 2.4 1.0 ns 0.0 0.0 ns 0.7 0.5 ns 0.7 0.3 ns 4.0 3, 36 0.014
(% dry mass)
Non-macroaggregated 23.0 4.2a 8.4 1.7b 15.0 2.8ab 9.9 1.1b 37.8 6.5a 12.5 3, 45 0.001
soil (% dry mass)
Rocks (%) 1.0 0.4ab 0.4 1.3b 3.0 1.8ab 0.6 0.2b 3.1 1.0a 4.6 4, 40 0.003
Note: Values with different letters indicate statistical differences between different land uses, according to multiple Tukey tests.
a
(47.4% of total individuals collected), and this trend was and the abundance of litter dwelling taxa, especially
not significantly different across land uses. The second- Araneae, Diplopoda, and Diplura. PC 2 (14.6% of vari-
most abundant group, Oligochaeta (23.5% of total), was ance explained; Figure 5) separates annual crops, with
most common under annual cropping and lowest in highest overall densities and Coleoptera, and Oligochaeta
abandoned lands, pastures, and forests. Coleoptera com- abundances from the other types of land use.
prised 5.0% of total individuals collected without appar-
ent differences between land use types. The Diplopoda
comprised 4.3% of total abundance and was observed Woody and herbaceous vegetation
mostly in forest. Isopoda and Coleoptera larvae com-
prised 3.6% of total abundance and were not different Forests and coffee plantations presented the highest
between land uses. Other taxa (Chilopoda, Diplura, abundance of woody vegetation (251 18 and
Aracnida, and Hemiptera) comprised between 2.2% and 242 26 stems ha1, respectively), notably decreased in
1.1% of total abundance, while Chilopoda and Araneae abandoned land, pastures, and annual crops. The mean
were principally observed in forests (Table 4). richness of woody plants was highest in forests
PC 1 (25.2% of variance explained) separate forests (42 3 species plot1), and lowest in coffee plantations,
from the other sites according to their taxonomic richness pastures, and annual crops (1–10 species plot1). A
ECOSPHERE 9 of 19
(a) (b)
1.0
NoAg d=2
Rock
0.5 Litt AC
BiAg
PC 1 (26.7%) Infi
0.0 PhAg
F
Sedi Root
CP
Runo A P
−0.5
RoAg
−1.0
−1.0 −0.5 0.0 0.5 1.0

PC 2 (18.4%)
F I G U R E 3 Principal component analysis biplot with the variables of regulation of soil hydrological function and soil aggregation
associated with the predominant land use types of the Cali River watershed (a). Between-class analysis of the principal types of coverage of
the Cali River watershed (b); secondary forest (F), abandoned land (A), pasture (P), coffee plantations (CP), and annual crops
(AC) according to the variables that describe the services that regulate the hydrological function of the soil, the decrease of erosion, and the
aggregation of soil (p = 0.001 according to the Monte Carlo permutation test, 37.0% of variance explained). Infiltration (Infi), runoff (RO),
sediments (Sedi), litter (Lit), roots (Root), and soil aggregation; biogenic aggregates (BiAg), physical aggregates (PhAg), root aggregates
(RoAg), non-macroaggregated soil (NAg), rocks (Rock), associated with the principal coverage types of the Cali River watershed. PC
1, principal component 1; PC 2, principal component 2
Richness (No. taxa plot –1)
4000
Abundance (ind m )
15
–2
3000 a
ab
10
2000
a b b
b
1000 ab ab ab
b 5
0
A AC CP F P A AC CP F P
F I G U R E 4 Mean abundance and richness of soil macrofauna in five land use types (secondary forest [F], abandoned land [A], pasture
[P], coffee plantations [CP], and annual crops [AC]) of the Cali River watershed. Boxplot lines represent median, 50 and 75 percentiles, and
dots represent data outside these percentiles. Values with different letters indicate statistical differences between different land uses,
according to Tukey tests
similar tendency was found when comparing Shannon more abundant in pastures and coffee plantations (161–
diversity, with a marked decrease for land uses with a 165 ind m2), with intermediate values in abandoned
higher level of intervention. Herbaceous plants were land and annual crops, and low values in forests
T A B L E 4 Abundance (number of individuals per square metre) of principal soil macrofauna taxa in five land use types of the Cali River
watershed (mean SE)
Parameter Forest Abandoned Pasture Coffee Annual crops Fa df p

Formicidae 228.6 47.9 ns 138.6 43.7 ns 577.1 331.1 ns 274.2 49.9 ns 620.4 425.97 ns 1.4 4, 44 0.23
Oligochaeta 113.4 42.6b 145.3 51.2b 113.1 32.1b 165.1 33.4ab 369.6 72.5a 4.2 4, 45 0.005
Coleoptera 99.6 22.7 ns 89.6 44.8 ns 220.4 77.3 ns 172.8 48.2 ns 378.0 217.5 ns 1.7 4, 45 0.15
Diplopoda 96.4 20.0a 16.6 5.5b 11.7 4.1b 17.3 6.6b 22.4 12.6b 8.8 4, 45 0.001
Isopoda 6.4 2.5 ns 10.6 3.5 ns 23.5 9.6 ns 32.0 10.2 ns 72.0 38.2 ns 3.7 4, 45 0.011
Coleoptera (larvae) 40.9 6.7 ns 17.3 5.2 ns 27.0 7.7 ns 18.6 5.7 ns 36.0 9.7 ns 2.1 4, 41 0.09
Chilopoda 39.1 6.2a 10.2 2.6b 6.4 2.1b 19.2 5.0ab 8.0 2.7b 10.8 4, 41 0.001
Diplura 33.4 8.1 ns 9.6 4.3 ns 1.1 0.5 ns 23.0 11.7 ns 8.8 4.5 ns 4.8 4, 45 0.003
Araneae 39.5 9.6a 11.5 3.5b 8.5 2.6b 12.8 5.2b 1.2 1.0b 9.4 4, 45 0.001
Hemiptera 6.8 1.7 ns 4.2 2.0 ns 11.7 4.2 ns 10.2 2.9 ns 7.2 3.7 ns 1.0 4, 45 0.41
Dictyoptera 16.4 8.6 ns 1.6 0.9 ns 3.2 2.9 ns 0.64 0.64 ns … 4.2 4, 36 0.012
Dermaptera … … 2.1 0.8 ns 12.8 5.7 ns 1.2 1.0 ns 6.1 4, 27 0.006
Myriapoda 1.1 0.7 ns 4.8 1.8 ns 5.7 3.2 ns 1.0 1.0 ns 2.0 1.0 ns 2.0 4, 45 0.114
Diptera (larvae) 3.2 0.9 ns 1.0 0.7 ns 1.4 0.7 ns 2.6 1.9 ns 3.6 1.2 ns 1.0 4, 45 0.396
Lepidoptera (larvae) 2.1 1.1 0.3 0.3 1.1 0.7 0.3 0.3 0.4 0.3 … …
Gastropoda 0.7 0.6 … … 0.6 0.6 2.4 1.6 … …
a
(a) (b)
d=2
1.0
F A
0.5
CP
ARAC P
CHILO
DYCT
PC 1 (25.2%) DIPLO
LEPIL MYRI
DIPLU
0.0
DERM
AC
Mac_RT
HEMI
ISOP
DIPTL GAST
−0.5 FORM
COLAL OLIG
COLA
TOT
−1.0
−1.0 −0.5 0.0 0.5 1.0
PC 2 (14.6%)
F I G U R E 5 Principal component analysis biplot of principal macrofauna taxa in five common land uses of the Cali River watershed (a).
Between-class analysis of principal types of coverage of the Cali River watershed (b); secondary forest (F), abandoned land (A), pasture (P),
coffee plantation (CP), and annual crops (AC), according to variation in the principal taxa of edaphic macrofauna (p = 0.001 according to
the Monte Carlo permutation test, 21.6% of variance explained). ARAC, Arachnida; CHILO, Chilopoda; COLA, Coleoptera; COLAL,
Coleoptera larvae; DERM, Dermaptera; DIPLO, Diplopoda; DIPLU, Diplura; DIPTL, Diptera larvae; DYCT, Dictyoptera; HEMI, Hemiptera;
HORM, Hormigas; ISOP, Isoptera; LEPIL, Lepidoptera larvae; Mac_RT, macrofauna richness; MYRI, Myriapoda; OLIG, Oligochaeta; PC
1, principal component 1; PC 2, principal component 2; TOT, total macrofauna density
ECOSPHERE 11 of 19
(16 ind m2). The Shannon diversity and richness of her- complex relationships between land uses and key ecologi-
baceous plants showed a similar tendency to abundance, cal processes. We first confirm that the simplification of
with relatively high values in pastures, coffee, and a ecosystems drives changes in soil macrofauna communi-
lower value in forests (Figure 6). ties and vegetation. These changes, in turn, strongly and
In terms of plant cover and growth forms, trees pre- negatively influence erosion, which is likely to affect the
dominated in forests (52%), and coffee plantation (19%) long-term productivity of these ecosystems and impact
in contrast to other land uses (Appendix S1: Table S1). At water quality within the Cali River watershed, with con-
the soil surface, litter cover was greatest in forests (48.7%) siderable implications for downstream inhabitants
and coffee plantations (18.5%). Bare soil predominated in (e.g., eutrophication, contamination from pesticides,
annual cropping systems (20%). Grasses were the most salts, and pathogens).
common cover type in pastures (55.3%) and in aban-
doned lands (37.7%; Appendix S1: Table S1).
PC 1 (32.5% of variance explained) separated forested Physicochemical properties of the soil
from the rest of land uses based on richness and abun-
dance of lianas, trees, understory shrubs, and litter The highest chemical fertility was observed with annual
(Appendix S1: Figure S2). PC 2 (15.3% of variance cropping systems, which is likely explained by the greater
explained) separated annual cropping systems from application of soil fertility amendments and/or incorpora-
purely herbaceous pastures and coffee plantations with tion of ashes from burned forest. Annual cropping sys-
more woody vegetation. tems in the region commonly produce aromatic and
medicinal plants and vegetables that receive frequent
applications of organic fertilizers, mainly in the form of
Covariation among groups of response chicken manure. This likely explains the higher quantity
variables of soil C and N and the abundance of generalist soil
invertebrate groups (mainly detritivores) in this land use.
Soil physiochemical properties, plant communities, and A similar pattern has been shown in other Andean
macrofauna presented significant co-inertia among them- regions where organic fertilizers are widely used
selves and with soil macroaggregation and infiltration (de Valença et al., 2017). Higher abundance of
and runoff patterns. Highest values of the matrix correla- detritivores, such as earthworms, can result in biogenic
tion index were recorded between vegetation cover vari- aggregates that accumulate C and N and inhibit micro-
ables and those associated with soil aggregation and bial decay due to their compact structure (Bossuyt
proxies of the hydrological services (42% of variance et al., 2005). The notable increase in earthworm abun-
explained; p = 0.001). Significant covariation was also dance is due to the multiplication of the pantropical inva-
observed between soil macrofauna and plant cover (40% sive species Pontoscolex corethrurus (Marichal
of variance explained; p = 0.001) and hydrological ES et al., 2010). This species can have complex effects on
parameters (26%). The physicochemical properties were water infiltration and runoff, with an accumulation of
most related to soil aggregation and vegetation variables compact macroaggregates in the soil sometimes offset by
(31% and 27% of variance explained, respectively; deposition at the soil surface of casts that are highly dis-
Figure 7). Tree cover, shrubs, seedlings, presence of persible until they dry (Chauvel et al., 1999). In many
trunks and litter, and woody plant diversity exhibited the agricultural regions, high N additions can increase plant
greatest relationship with infiltration rates. In contrast, growth and the activity of soil microorganisms, but a sig-
grass cover, richness and abundance of herbaceous nificant fraction is often lost to leaching and other delete-
plants, and absence of vegetation are most related to run- rious loss pathways. As a result, the excess N from
off and sedimentation (figures with specific variables not intensive agriculture can lead to widespread decline in
shown). water quality (Ren et al., 2014; Song et al., 2009; Wang
et al., 2019). The use of a continuous, permanent plant
cover or mulch may reduce the potential loss of C and N
DISC USS I ON by decreasing water erosion and N loss in sediment
(Pacheco & Sanches-Fernandes, 2016).
In this study, we hypothesized that changes in vegetation In contrast to the relatively high fertility observed in
cover and soil resulting from the conversion from forest the annual cropping systems, abandoned shrubland areas
to agricultural land uses promote changes in soil biodi- had much lower contents of soil C and nutrients. This is
versity that, in turn, negatively impact ES related to congruent with the notion that this land use is often
hydrological regulation. Our findings demonstrate highly degraded and likely explains its abandonment
Woody vegetation
400 60
Abundance (ind 0.1 ha )
4
–1
Richness (Sp. plot )

–1
a a
a
300 3
a 40
Shannon
b
200 2
20
100 1 c
b b
c c
c cc
c c c
c
0 0 0
Herbaceous vegetation
25
Richness (Sp. m )
Abundance (ind m )
–2
a
–2
1500 a a 3 a
20 a
a ab
Shannon
a
15 b 2
1000
b
10 bc
500 1
b c
5
bc c
0 0
0
F I G U R E 6 Abundance, richness, and Shannon diversity of woody and herbaceous vegetation in the five land use types (secondary
forest [F], abandoned land [A], pasture [P], coffee plantations [CP], and annual crops [AC]) of the Cali River watershed. Boxplot lines
represent median, 50 and 75 percentiles, and dots represent data outside these percentiles. Values with different letters indicate statistical
differences between different land uses, according to Tukey tests
Plant cover
0.2
0 *** Plant diversity
7*
*
0.4
0.17 ns
Soil macrofauna Chemical variables
Decomposer diversity Soil chemical fertility
*
0.
0.42*** **
21
. 31
0
**
0.
Soil
*
27
**
33
aggregation
0.
**
*
0.
38
***
0.13 *
Infiltration
runoff
Hydrological
regulation
F I G U R E 7 Schematic summary of the co-inertia analysis between the databases of ecosystem services of regulation of hydrological
function, vegetation, soil macrofauna, and physiochemical properties of the soil. The coefficient of correlation Rv and the probability
associated between groups of variables for databases is indicated on the relationship arrows. P-values for co-inertia analyses are indicated as
*, p < 0.05; **, p < 0.01; ***, p < 0.001
ECOSPHERE 13 of 19
from agricultural production. At the same time, we noted in the Colombian Andes (Farfan, 2014), and it remains
that forest soils, while relatively rich in C and N, had very unclear if these species will maintain similar levels of
low contents of available P and this likely has to do with productivity in the context of climate change. With the
the lack of P inputs and the relatively low pH of the forest growing demand for specialty coffees, traditional varieties
soils. tolerant to shade are becoming more popular again, and
farmers can opt to plant more trees. In this type of coffee
plantation, a higher price paid for quality may compen-
Ecosystem services of hydrological sate the reduction in production due to competition for
regulation light (Vaast et al., 2005) and likely contribute more to
long-term soil conservation efforts.
Slower water infiltration along with greater soil compac-
tion and erosion was generally observed in the more dis-
turbed land uses, especially in grazed pastures and Soil aggregation
abandoned land. This tendency is accentuated where
trees are scarce or absent (Abdelkadir & Yimer, 2011; Soils presented high levels of macroaggregation with
Lavelle et al., 2014; Velasquez et al., 2007). Compaction maximum values (>80%) observed in coffee plantations,
due to overgrazing and poor soil fertility management shrublands, and pastures. High proportions of biogenic
was still apparent in the shrublands left to fallow aggregates in shrublands and coffee cultivations are
2–5 years earlier and likely slowed ecological succession clearly associated with activities of the earthworm
and the restoration process (Schweiger et al., 2018). Shal- P. corethrurus. In disturbed tropical areas, this peregrine
low root systems and limited leaf litter production of the endogeic earthworm has been shown to replace native
pioneer vegetation are not able to generate a rapid species and produce large amounts of casts feeding on
improvement of the soil structure and associated C accu- organic stocks of the original ecosystem (Chauvel
mulation. Conversely, high infiltration rates in coffee et al., 1999; Marichal et al., 2010), thus increasing
plantations, similar to that observed in the forest, high- macroaggregation. Termites, ants, and other macro-
light the importance of an arboreal component to reduce invertebrate ecosystem engineers may also contribute to
raindrop impact and associated soil crusting, and thus biogenic soil aggregation in coffee and shrubland plant
ensure soil permeability (Gaitan et al., 2016; Hanson covers.
et al., 2004). Organic fertilization favors the activity of macrofauna
The highest rates of runoff and sedimentation were in agricultural soils and enhances the formation of aggre-
associated with agricultural uses, especially in areas of gates (Grimaldi et al., 2014; Lavelle et al., 2014;
shrubland and pasture with few trees and a low deposi- Velasquez et al., 2007). A significant relationship has
tion of leaf litter. This reduction in organic matter inputs been found between the amount of organic matter, the
and higher rates of erosion are likely associated with density and diversity of the macroinvertebrate commu-
nutrient deficiencies in many Andean soils (Fonte nity, and formation of aggregates in different
et al., 2012; Molina et al., 2007, 2012). Lower runoff in agroecosystems of Colombia, Nicaragua, and the Amazon
annual cropping systems may occur because these sys- region (Grimaldi et al., 2014; Rousseau et al., 2013). In
tems are located on relatively moderate slopes; addition- coffee plantations, the crown architecture and dense
ally, the use of low impact tillage practices may also foliage allow for a continuous input of litter, often com-
support water infiltration. However, sediment loss was parable to that observed in forests. Annual crops exhibit
relatively high under annual crops due to a continuous high proportions of biogenic aggregates (57%), but also
movement and exposure of soil during planting and the highest proportions of non-macroaggregated soil
weeding activities. (38%). In annual crops, the predominance of both aggre-
In Andean soils, the integration of different strata of gated and non-aggregated soils may result from periods
vegetation in agroforestry systems constitutes an alterna- where the soil lays fallow and untilled. Formation of bio-
tive to improve surface soils with low structure, low genic aggregates likely occurs during the rotation with
organic matter contents, deficient drainage, and steep cover plants such as chayote (S. edule (Jacq.) Sw.) and the
slopes (Beer et al., 2003). Coffee, as a perennial crop, pro- incorporation of harvest waste products, which are then
vides the best support for hydrological function, even destroyed during tillage and preparation of the soil for
with a limited association of shade tree species. The planting.
increased adoption of varieties that demand high The high proportion of biogenic aggregates in aban-
amounts of fertilization, with low tolerance to shade con- doned lands is probably inherited from the previous
ditions, has created issues of soil conservation elsewhere cropping phases and characteristics of the leaf litter and
root inputs that sustain activities of macroaggregating Ants are significant actors in the structuring of the soil
fauna. The litter quality of pioneer species in abandoned due to the large movement of soil and organic material
lands can promote the high decomposition rates by the operated by such genera as Atta, Acromyrmex, and
low C:N ratio in leaves, a high proportion of labile C, and Myrmicinae in transformed environments (Leal et al.,
the redistribution of resources during senescence 2014), among others. Oligochaeta, the second-most abun-
(Eviner & Chapin, 2003; Laughlin et al., 2010). In pas- dant group, were abundant everywhere, with maximum
tures, the aggregation can be driven by the root and densities in soils of the annual cropping systems, where
mycorrhizal systems (Furrazola et al., 2016) and earth- they found adequate organic material and N sources
worms that benefit from deposition of cattle manure (Feijoo et al., 1999; Jouquet et al., 2014). They had rela-
(Bacher et al., 2018; Guggenberger et al., 1996). In this tively low densities in pastures possibly a result of com-
type of plant cover, a dense root mat increases the cohe- paction due to livestock grazing (Barros et al., 2003;
sion of shallow particulates (Materechera et al., 1992). In Radford et al., 2001).
forests, as in pastures, aggregate stability is favored by the
cohesion of particles stuck together by root exudates or
entangled in mycorrhizal hyphae networks (Singh Vegetation
et al., 2013). The high prevalence of mycorrhizae in soils
in tropical grasslands suggests that they play an impor- In the Cali River watershed, a simplification in the diver-
tant role in the stabilization of macroaggregates, probably sity of woody vegetation (number of species, genera, and
in association with earthworms and/or roots. families) can be observed along an intensification gradi-
ent from forests to agricultural systems. Due to their
proximity to the city, the forest remnants are notably less
Soil biodiversity diverse and with fewer trees than forests of the surround-
ing intact watersheds, where Shannon indices of 4.8–6.5
Macroinvertebrate communities generally had high den- (on 0.1 ha Gentry transects) have been measured
sities and taxonomic richness in comparison with similar (Gamboa & Ramos, 1995) instead of 3.0 in the present
data sets collected worldwide (Lavelle et al., submitted). study. In coffee systems, the relatively low plant cover
The highest abundance was observed within the annual and basal area indicate a semiwooded model and conven-
cropping sites, where ants and earthworms exhibited tional management with low arboreal association
especially high densities. This result contrasts with a (Ehrenbergerova et al., 2016). Relatively more dispersed
majority of other studies that indicated depletion of these arboreal cover in coffee plantations and pastures favors
communities in agricultural land (Lavelle et al., 2014; the occurrence of generalist vegetation and the arrival of
Lavelle & Pashanasi, 1989; Marichal et al., 2014; Sanabria exotic species. In pastures and abandoned lands, the sub-
et al., 2014). However, these findings agree with other stitution of herbaceous families can be observed during
studies, showing that decomposer communities can be the process of succession, such as the replacement of
quite abundant in agricultural soils receiving sufficient Asteraceae, Conmmelinaceae, and Poaceae by families
organic inputs and with minimal disturbance (de Valença with greater cover and woody structure, such as
et al., 2017; Feijoo et al., 1999; Kelly et al., 2021; Mathieu Actinidaceae, Clusiaceae, and Melastomataceae. Among
et al., 2005; Ramírez et al., 2012; Rousseau et al., 2013). the most common species, the introduced grasses
Another factor that could play a role is that annual crops Cynodon plectostachyus and Brachiaria sp. stand out, as
are generally located in the flatter and more fertile areas they are both exotic, and form homogenous covers that
of the landscape. The similarity of communities observed frequently limit the regeneration of other shrub and arbo-
in secondary forests and agroforestry systems, such as real species. Inadequate pasture management likely
shade coffee plantations that have a continuous litter results in low pasture persistence on these tropical hill-
inputs, has been indicated by several authors (Barros sides (Etter & Villa, 2000), and pastures often shift to
et al., 2003; Montoya-Molina et al., 2016; Rousseau abandoned land as soil fertility declines. In abandoned
et al., 2013; Velasquez et al., 2007). lands, Pteridium aquilinum and Dicranopteris spp. are
Surprisingly, the density of ant communities, the common invasive ferns that can form monodominant
most abundant group, did not show significant differ- covers in many degraded lands of the tropics and sub-
ences among the land uses investigated—at species or tropics (Suazo-Ortuño et al., 2014). These abandoned
genera levels. Differences in the quantity and quality of lands are usually targeted for restoration projects with
key resources (e.g., food and habitat) between land uses active management of succession, and the planting of
have been shown to selectively favor specific taxa in other shrubs and pioneer trees that can eventually shade out
studies (Armbrecht et al., 2005; Sanabria et al., 2014). invasive ferns or grasses (Galindo et al., 2017; Holl, 2002).
ECOSPHERE 15 of 19
Covariation among ES probably influenced by the application of organic fertilizers

in annual crops. Although coffee systems have infiltration
This study confirmed the covariation of soil-based ES rates similar to those in forests, sediment production was
observed in other studies from Colombia (Figure 5; Lavelle higher, suggesting that the disturbance of the soil during
et al., 2014, Grimaldi et al., 2014) and elsewhere weed control and other management interventions may
(De Leijster et al., 2019; Rey Benayas & Bullock, 2012). Most contribute to erosional processes. We note that in this
of the suites of variables used as proxies for ES significantly study only fine roots in the first 10 cm were analyzed; how-
covaried: aboveground with belowground biodiversity, plant ever, the role of deep roots has also been shown to signifi-
production with soil chemical fertility component, water cantly impact soil structure and infiltration (Benegas
services with infiltration, and runoff components. et al., 2014; Ilstedt et al., 2007).
In general, the highest generation of multiple ES, or The co-inertia between all ES proxies and macrofauna
“bundles,” occurs in forests and land use covers with was significant, with a notable positive relationship
greater arboreal structure or most limited agricultural between species richness and infiltration. The significant
intensification (i.e., shade coffee). In these forest associa- co-inertia between plant and macrofauna communities
tions, variables of hydrological regulation stand out due highlights the interactions between functional diversity
to their decrease in runoff and high infiltration, their above and below the soil, as well as its effects on the reg-
control of erosion, and their services that support and ulation of ES. Even in annual crops, the soil manage-
maintain soil fertility. The significant co-inertia among ment, use of green manures, and litter inputs influence
plant cover, soil macrofauna communities, and soil phys- the provision of ES by increasing detritivore activity, soil
icochemical properties indicates a strong connection macroaggregation, and infiltration potential (Craswell &
among the biophysical components that determine eco- Lefroy, 2001; Franzluebbers, 2002; Kearney et al., 2019).
system function. Similar relationships were found when Our findings suggest that the promotion of biodiver-
comparing different sets of variables in tropical savannah sity and above- and belowground interactions should be
agroecosystems (Lavelle et al., 2014), the Amazon region a key conservation objective in protected areas and forest
(Grimaldi et al., 2014), and the Peruvian and Ecuadorian reserves due to the clear relationships with ES of hydro-
Andes, where land use drives biodiversity and the provi- logical regulation. Policy makers and land managers
sion of multiple ES (Caulfield et al., 2020; de Valença should work toward increasing vegetative coverage and
et al., 2017). Such generalized correlations support the diversity at the watershed scale in their land manage-
hypothesis proposed that transforming plant communi- ment plans. For example, in order to decrease the nega-
ties leads to indirect management of soil fauna communi- tive effects of extensive cattle ranching (e.g., erosion and
ties and ES that depend on their activities. Roots, soil compaction), alternatives such as silvopastoral sys-
earthworms, and other ecosystem engineers have a nota- tems, which increase tree density in pastures, should be
ble effect on soil aggregation and porosity, an important utilized (Giraldo et al., 2011; Montoya-Molina
driver of hydrological services. et al., 2016). With their inputs of organic material and
The greatest association with ES was observed with their favorable microsite conditions, silvopastoral systems
plant cover, the different growth forms, and indices of can promote diverse and abundant macrofauna commu-
plant diversity. The diversity of woody vegetation was nities, with positive effects on the soil-based ES bundles
positively correlated with water infiltration rate and neg- (Barros et al., 2003; Giraldo et al., 2011; Montoya-Molina
atively with runoff and sedimentation. Among the plant et al., 2016; Webster et al., 2019). In coffee systems, in
covers that permit the greatest infiltration, forests and addition to increasing tree diversity, the planting of living
coffee plantations stand out. While Benegas et al. (2014) mulches between rows and more selective control of
indicated that coffee root systems and trees in pastures weeds may diminish the loss of soils due to runoff caused
support soil permeability, an increase in the density of by excessive soil disturbance (Afandi et al., 2002). Finally,
trees has a greater effect on infiltration in pastures than it changes in the physical and hydraulic properties of the
does in coffee plantations. soil are generated as a result of the activity of ecosystem
Covariation between ES and physicochemical proper- engineers. The annual crops can generate rapid revenue
ties of soil highlights the importance of certain physico- to farmers but could generate negative ecosystem trade-
chemical properties as modulators of hydrological offs related to high erosion and increased N loss (via
regulation services (Grimaldi et al., 2014). Soil bulk density leaching and erosion) that degrades downstream water
increased in the presence of physical macroaggregates and quality. The introduction of filter strips, drainage terraces
directly influences runoff. In contrast, other properties, (Ramos et al., 2015), or radical terraces (De La Paix
such as soil C and NO3, show positive relationships with Mupenzi et al., 2012) can help to reduce these impacts in
soil aggregation in agricultural systems, an effect that is mountainous agricultural landscapes. A larger study of
the functional groups of vegetation and soil macrofauna ACKNOWLEDGMENTS

may show the capacity for resistance and adaptive Funding was provided by Colciencias and the Patrimonio
change inherent in the predominant productive models, Autonomo Fondo Nacional de Financiamiento para
especially in high-impact agricultural systems such as la Ciencia Francisco José de Caldas, Colombia (FPA44842-
extensive cattle ranching and transitory cultivations. In 042-2016) (80740-424-2019) and the Fundacio n CIPAV. We
the context of the study area being a buffer between the appreciate the assistance of the technical and administra-
city of Cali and Farallones National Natural Park, it is of tive staff of CIPAV, especially to J. Chara for their valuable
great importance to protect the forest patches that have comments on earlier version of the project. We thank the
not been converted to agriculture, to restore the degraded members of the Group on Ecology of Agroecosystems and
lands, and to manage the already converted lands with Natural Habitats (GEAHNA), and the Entomology staff of
more sustainable production practices in order to main- the Universidad del Valle. Particularly grateful to
tain ES that have downstream impacts on the watershed. A. Arenas for macrofauna identification and to J. A. Vargas
for plant identification, and C. Figueroa for GIS support.
Field assistants involved in this study: I. Nichols, E.
C O N C L U S IO N S Valderrama, E. Alzate, C. Londoño, H. Calero, V. Giraldo,
D. Giraldo, S. Orejuela, J. Galindo y A. Perdomo. Owner(s)
Our findings indicate that land use transformation of farms: E. Alvear, F. Sanchez, O. Jovel, M. Ruiz, A. Qui-
(intensification of agriculture) can have marked impacts ñones, A. Mosquera, C. Buitro n, D. Madriñan, D. Molano,
on ES related to hydrological regulation and other key F. Pérez, F. Buenaventura, F. Vallecilla, G. Polanía, H.
ecosystem functions, but that key trade-offs exist between Chica, H. Cuellar, H. Hidalgo, J. Certuche, J. Pérez, J.
land uses. For example, while annual cropping systems Gomez, J. Ordoñez, J. Osorio, J. Cortez, J. Sanchez, J.
demonstrate high soil C storage and macrofauna abun-
Arango, J. Mesa, L. Calvert, O. Mosquera, P. Alvarez, P.
dance, they have decreased macroaggregation, with nega- Lazo, P. Muños, R. Higuera, T. Herrera, W. Sanchez and
tive implications for downstream water quality due to W. Cruz.
relatively high erosion losses and potential for N
leaching. At the same time, some of the annual cropping CONFLICT OF INTEREST
management practices favor the activity of macrofauna The authors declare no conflict of interest.
and achieve positive effects in the provision of
ES. Meanwhile, coffee production systems contribute to DA TA AVAI LA BI LI TY S T ATE ME NT
improved aggregation and reduced runoff, but do not Data (Galindo et al., 2022) are available from Mountain
support soil C storage as well as forest or annual cropping Scholar: https://doi.org/10.25675/10217/234038.
systems. There is a notable impact of extensive cattle
grazing on steep hillsides, with negative effects on all ORCID
ES. Despite having low animal loads, this land use is Víctor Galindo https://orcid.org/0000-0003-0112-2432
associated with low infiltration and higher potential for Carolina Giraldo https://orcid.org/0000-0003-3028-
erosion. It is important to identify the optimal proportion 6436
and spatial distribution of more natural versus managed Patrick Lavelle https://orcid.org/0000-0002-2127-1067
land uses across the landscape so that soil macrofauna Inge Armbrecht https://orcid.org/0000-0003-0574-2559
and related drivers can be best managed for reducing sed- Steven J. Fonte https://orcid.org/0000-0002-3727-2304
iment loads and erosion at a watershed scale. The impor-
tance of the diversity above and belowground is notable RE FER EN CES
in the interactive relationships shown in the co-inertia Abdelkadir, A., and F. Yimer. 2011. “Soil Water Property Variations
analyses, indicating a hierarchical character to ecosystem in Three Adjacent Land Use Types in the Rift Valley Area of
organization. Thus, when there are changes in the diver- Ethiopia.” Journal of Arid Environments 75: 1067–71.
sity and structure of the vegetation, changes are gener- Afandi, T., K. Manik, B. Rosadi, M. Utomo, M. Senge, T. Adachi,
and Y. Oki. 2002. “Soil Erosion under Coffee Trees with Differ-
ated in the availability and quality of resources in other
ent Weed Managements in Humid Tropical Hilly Area of
trophic levels, in which edaphic macrofauna intervene. Lampung, South Sumatra, Indonesia.” Journal of the Japanese
In conclusion, creating landscapes that preserve and Society of Soil Physics 91: 3–14.
restore forests, rehabilitate productive areas with sil- Armbrecht, I., L. Rivera, and I. Perfecto. 2005. “Reduced Diversity
vopastoral and agroforestry systems, and increase struc- and Complexity in the Leaf-Litter Ant Assemblage of Colombian
tural diversity of vegetation in watershed areas is a Coffee Plantations.” Conservation Biology 19: 897–907.
recommend strategy that can help recover and promote Bacher, M. G., O. Fenton, G. Bondi, R. E. Creamer, M. Karmarkar,
and O. Schmidt. 2018. “The Impact of Cattle Dung Pats on
the ES mediated by soil biodiversity.
ECOSPHERE 17 of 19
Earthworm Distribution in Grazed Pastures.” BMC Ecology Dray, S., D. Chessel, and J. Thioulouse. 2003. “Co-Inertia Analysis
18: 59. and the Linking of Ecological Data Tables.” Ecology 84:
Barros, E., A. Neves, E. Blanchart, E. C. M. Fernandes, E. Wandelli, 3078–89.
and P. Lavelle. 2003. “Development of the Soil Macrofauna Ehrenbergerova, L., E. Cienciala, A. Kučera, L. Guy, and H.
Community under Silvopastoral and Agrosilvicultural Systems Habrova. 2016. “Carbon Stock in Agroforestry Coffee Planta-
in Amazonia.” Pedobiologia 47: 273–80. tions with Different Shade Trees in Villa Rica, Peru.” Agrofor-
Beer, J., C. Harvey, M. Ibrahim, J. M. Harmand, E. Somarriba, and estry Systems 90: 433–45.
F. Jiménez. 2003. “Servicios ambientales de los sistemas Elliott, E. T., and D. C. Coleman. 1988. “Let the Soil Work for Us.”
agroforestales.” Agroforestería en las Américas 10: 80–7. Ecological Bulletins 39: 23–32.
Benegas, L., U. Ilstedt, O. Roupsard, J. Jones, and A. Malmer. 2014. Etter, A., and L. A. Villa. 2000. “Andean Forests and Farming Sys-
“Effects of Trees on Infiltrability and Preferential Flow in Two tems in Part of the Eastern Cordillera (Colombia).” Mountain
Contrasting Agroecosystems in Central America.” Agriculture, Research and Development 20: 236–45.
Ecosystems and Environment 183: 185–96. Eviner, T. V., and F. S. Chapin. 2003. “Functional Matrix: A Con-
Bennett, E. M., G. D. Peterson, and L. J. Gordon. 2009. “Under- ceptual Framework for Predicting Multiple Plant Effects on
standing Relationships among Multiple Ecosystem Services.” Ecosystem Processes.” Annual Review of Ecology, Evolution,
Ecology Letters 12: 1394–404. and Systematics 34: 455–85.
Bossuyt, H., J. Six, and P. F. Hendrix. 2005. “Protection of Soil Car- Farfan, V. F. 2014. Agroforestería y sistemas agroforestales con café.
bon by Microaggregates within Earthworm Casts.” Soil Biology Manizales: FNC-Cenicafé.
and Biochemistry 37: 251–8. Feijoo, A. M., E. B. Knapp, P. Lavelle, and A. G. Moreno. 1999.
Bouyoucos, G. J. 1962. “Hydrometer Method Improved for Making “Quantifying Soil Macrofauna in a Colombian Watershed.”
Particle Size Analysis of Soils.” Agronomy Journal 54: 464–5. Pedobiologia 43: 513–7.
Bruijnzeel, L. A. 2004. “Hydrological Functions of Tropical Forests: Fondo Patrimonio Natural, CIPAV, CVC, DAGMA, PNN
Not Seeing the Soil for the Trees?” Agriculture, Ecosystems and Farallones de Cali, EMCALI. [Fondo Patrimonio Natural, Cen-
Environment 104: 185–228. tro para la Investigacion en Sistemas Sostenibles de Produc-
Brussaard, L. 2012. “Ecosystem Services Provided by the Soil Biota.” ci
on Agropecuaria, Corporaci on Autonoma Regional del Valle
In Soil Ecology and Ecosystem Services, edited by D.H. Wall, del Cauca, Parque Nacional Natural Farallones de Cali, and
45–59. Oxford: Oxford University Press. Empresas Municipales de Cali]. 2014. “Compensaci on por Ser-
Buytaert, W., V. Iñiguez, and B. de Bièvre. 2007. “The Effects of vicios Ambientales Hídricos en la cuenca del río Cali, Valle del
Afforestation and Cultivation on Water Yield in the Andean Cauca. Tomo 1.2.” In Coleccion los incentivos a la conservacion:
Paramo.” Forest Ecology and Management 251: 22–30. una mirada desde la pr actica, edited by Z. Calle. Bogota:
Caulfield, M., S. J. Fonte, J. Groot, S. Vanek, S. Sherwood, P. Fondo Patrimonio Natural.
Oyarzun, R. Borja, S. Dumble, and P. Tittonell. 2020. Fonte, S. J., S. J. Vanek, P. Oyarzun, S. Parsa, D. C. Quintero, I. M.
“Agroecosystem Patterns and Land Management Co-Develop Rao, and P. Lavelle. 2012. “Pathways to Agroecological Inten-
through Environment, Management and Land-Use Interac- sification of Soil Fertility Management by Smallholder
tions.” Ecosphere 11: e03113. Farmers in the Andean Highlands.” Advances in Agronomy
Chauvel, A., M. Grimaldi, E. Barros, E. Blanchart, T. Desjardins, M. 116: 125–84.
Sarrazin, and P. Lavelle. 1999. “Pasture Damage by an Amazo- Franzluebbers, A. J. 2002. “Water Infiltration and Soil Structure
nian Earthworm.” Nature 398: 32–3. Related to Organic Matter and its Stratification with Depth.”
Cobo, L., and E. Amézquita. 1999. “Diseño, construcci on y evaluaci
on Soil and Tillage Research 66: 197–205.
de un minisimulador de lluvia para estudios de susceptibilidad a Furrazola, E., L. Ojeda, and C. Hernandez. 2016. “Mycorrhizal Col-
erosion en laderas.” Revista Suelos Ecuatoriales 29: 66–70. onization and Species of Arbuscular Mycorrhizal Fungi in
Craswell, E. T., and R. D. B. Lefroy. 2001. “The Role and Function Grasses from the Cuenca Pecuaria “El Tabl on”, Cienfuegos,
of Organic Matter in Tropical Soils.” Nutrient Cycling in Cuba.” Cuban Journal of Agricultural Sciences 50: 321–31.
Agroecosystems 61: 7–8. Gaitan, L., I. Armbrecht, and S. Graefe. 2016. “Throughfall and Soil
Crespo, P., R. Célleri, W. Buytaert, J. Feyen, V. Iñiguez, P. Borja, Properties in Shaded and Unshaded Coffee Plantations and a
and B. De Bievre. 2010. “Land Use Change Impacts on the Secondary Forest: A Case Study from Southern Colombia.”
Hydrology of Wet Andean paramo Ecosystems.” IAHS-AISH Journal of Agriculture and Rural Development in the Tropics
Publication 336: 71–6. and Subtropics 117: 309–21.
De La Paix Mupenzi, J., L. Li, J. Ge, J. de Dieu, G. Habiyaremye, J. Galindo, V., Z. Calle, J. Chara, and I. Armbrecht. 2017. “Facilitation
Ngamije, and I. Baragahoranye. 2012. “Radical Terraces in by Pioneer Shrubs for the Ecological Restoration of Riparian
Rwanda.” East African Journal of Science and Technology 1: Forests in the Central Andes of Colombia.” Restoration Ecology
53–8. 25: 731–7.
De Leijster, V., M. J. Santos, M. J. Wassen, M. E. Ramos-Font, A. B. Galindo, V., C. Giraldo, P. Lavelle, I. Armbrecht, and S.J. Fonte.
Robles, M. Díaz, M. Staal, and P. A. Verweij. 2019. “Agroeco- 2022. “Data Associated with “Land Use Conversion to Agricul-
logical Management Improves Ecosystem Services in Almond ture Impacts Biodiversity, Erosion Control, and Key Soil Prop-
Orchards within One Year.” Ecosystem Services 38: 100948. erties in an Andean watershed”.” Data. Mountain Scholar.
Dolédec, S., and D. Chessel. 1994. “Co-Inertia Analysis: An Alterna- https://doi.org/10.25675/10217/234038
tive Method for Studying Species–Environment Relation- Gamboa, M. A., and J. E. Ramos. 1995. “Composici on florística y
ships.” Freshwater Biology 31: 277–94. diversidad vegetal de un bosque premontano en los Farallones
de Cali.” In Memorias del Primer Congreso Nacional sobre Lavelle, P. A. 2002. “Functional Domains in Soils.” Ecological
Biodiversidad. 71–6. Cali: Instituto de Estudios Del Pacífico, Research 17: 441–50.
Universidad del Valle. Lavelle, P. A., and B. Pashanasi. 1989. “Soil Macrofauna and Land
Giraldo, C., F. Escobar, J. D. Chara, and Z. Calle. 2011. “The Adop- Management in Peruvian Amazonia (Yurimaguas, Loreto).”
tion of Silvopastoral Systems Promotes the Recovery of Ecolog- Pedobiologia 33: 283–91.
ical Processes Regulated by Dung Beetles in the Colombian Lavelle, P. A., and A. Spain. 2005. Soil Ecology. Dordrecht: Springer.
Andes.” Insect Conservation and Diversity 4: 115–22. Lavelle, P. A., N. Rodríguez, O. Arguello, J. Bernal, C. Botero, P.
Gonzalez-Quintero, R., M. S. Sanchez-Pinz on, D. M. Bolívar- Chaparro, Y. Gomez, et al. 2014. “Soil Ecosystem Services and
Vergara, N. Chirinda, J. Arango, H. A. Pentévez, G. Correa- Land Use in the Rapidly Changing Orinoco River Basin of
Londoño, and R. Barahona-Rosales. 2020. “Technical and Colombia.” Agriculture, Ecosystems & Environment 185: 106–17.
Environmental Characterization of Very Small, Small, Lavelle, P. A., S. Spain, J. C. Fonte, E. Bedano, V. Blanchart, M.
Medium, and Large Cow-Calf Operations in Colombia.” Galindo, J. J. Grimaldi, E. Velasquez Jimenez, and A.
Revista Mexicana de Ciencias Pecuarias 11: 183–204. Zangerlé. 2020. “Soil Aggregation, Ecosystem Engineers and
Grimaldi, M., J. Oszwald, S. Dolédec, M. D. P. Hurtado, I. de Souza the C Cycle.” Acta Oecologica 105: 103561.
Miranda, X. Arnauld de Sartre, W. S. De Assis, et al. 2014. Leal, I. R., R. Wirth, and M. Tabarelli. 2014. “The Multiple Impacts
“Ecosystem Services of Regulation and Support in Amazonian of Leaf-Cutting Ants and their Novel Ecological Role in
Pioneer Fronts: Searching for Landscape Drivers.” Landscape Human-Modified Neotropical Forests.” Biotropica 46: 516–28.
Ecology 29: 311–28. Marichal, R., A. F. Martinez, C. Praxedes, D. Ruiz, A. F. Carvajal, J.
Guggenberger, G., R. J. Thomas, and W. Zech. 1996. “Soil Organic Oszwald, M. P. Hurtado, et al. 2010. “Invasion of Pontoscolex
Matter within Earthworm Cast of an Anecic-Endogeic Tropi- corethrurus (Glossoscolecidae, Oligochaeta) in Landscapes of the
cal Pasture Community, Colombia.” Applied Soil Ecology 3: Amazonian Deforestation Arc.” Applied Soil Ecology 46: 443–9.
263–74. Marichal, R., M. Grimaldi, M. A. Feijoo, J. Oszwald, C. Praxedes, D. H.
Guzman, C. D., F. Hoyos-Villada, M. Da Silva, F. A. Zimale, N. Ruiz Cobo, M. P. Hurtado, et al. 2014. “Soil Macroinvertebrate
Chirinda, C. Botero, A. Morales, B. Rivera, P. Moreno, and Communities and Ecosystem Services in Deforested Landscapes of
T. S. Steenhuis. 2019. “Variability of Soil Surface Characteris- Amazonia.” Applied Soil Ecology 83: 177–85.
tics in a Mountainous Watershed in Valle del Cauca, Materechera, S. A., A. R. Dexter, and A. M. Alston. 1992. “Forma-
Colombia: Implications for Runoff, Erosion, and Conserva- tion of Aggregates by Plant-Roots in Homogenized Soils.”
tion.” Journal of Hydrology 576: 273–86. Plant and Soil 142: 69–79.
Haines-Young, R., and M. B. Potschin. 2018. Common International Mathieu, J., J. P. Rossi, P. Mora, P. Lavelle, P. F. Da. S. Martins, C.
Classification of Ecosystem Services (CICES) V5.1 and Guidance Rouland, and M. Grimaldi. 2005. “Recovery of Soil
on the Application of the Revised Structure. Nottingham: Macrofauna Communities after Forest Clearance in Eastern
CICES. Amazonia, Brazil.” Conservation Biology 19: 1598–605.
Hanson, D. L., T. S. Steenhuis, M. F. Walter, and J. Boll. 2004. Millennium Ecosystem Assessment. 2005. Ecosystem and Human
“Effects of Soil Degradation and Management Practices on the Well-Being. Synthesis. Washington, DC: Island Press.
Surface Water Dynamics in the Talgua River Watershed in Molina, A., G. Govers, V. Vanacker, J. Poesen, E. Zeelmaekers, and
Honduras.” Land Degradation and Development 15: 367–81. F. Cisneros. 2007. “Runoff Generation in a Degraded Andean
Holl, K. D. 2002. “Effect of Shrubs on Tree Seedling Establishment Ecosystem: Interaction of Vegetation Cover and Land Use.”
in an Abandoned Tropical Pasture.” Journal of Ecology 90: Catena 71: 357–70.
179–87. Molina, A., V. Vanacker, V. Balthazar, D. Mora, and G. Govers.
Ilstedt, U., A. Malmer, E. Verbeeten, and D. Murdiyarso. 2007. “The 2012. “Complex Land Cover Change, Water and Sediment
Effect of Afforestation on Water Infiltration in the Tropics: A Yield in a Degraded Andean Environment.” Journal of Hydrol-
Systematic Review and Meta-Analysis.” Forest Ecology and ogy 472–473: 25–35.
Management 251: 45–51. Montoya-Molina, S., C. Giraldo-Echeverri, J. Montoya-Lerma, J.
International Organization for Standardization. 2011. Soil Quality- Chara, F. Escobar, and Z. Calle. 2016. “Land Sharing Vs. Land
Sampling of Soil Invertebrates. Part 5. Sampling and Extraction Sparing in the Dry Caribbean Lowlands: A Dung Beetles’ Per-
of Soil Macro-Invertebrates. ISO 23611-5, Geneva, Switzerland. spective.” Applied Soil Ecology 98: 204–12.
Jouquet, P., E. Blanchart, and Y. Capowiez. 2014. “Utilization of Motsara, M. R., and R. N. Roy. 2008. Guide to Laboratory Establish-
Earthworms and Termites for the Restoration of Ecosystem ment for Plant Nutrient Analysis. FAO Fertilizer and Plant
Functioning.” Applied Soil Ecology 73: 34–40. Nutrition Bulletin 19. Roma: Food and Agriculture Organiza-
Kearney, S. P., S. J. Fonte, E. García, P. Siles, K. M. A. Chan, and tion of the United Nations.
S. M. Smukler. 2019. “Evaluating Ecosystem Service Trade- Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da
Offs and Synergies from Slash-and-Mulch Agroforestry Sys- Fonseca, and J. Kent. 2000. “Biodiversity Hotspots for Conser-
tems in El Salvador.” Ecological Indicators 105: 264–78. vation Priorities.” Nature 403: 853–258.
Kelly, C., S. J. Fonte, A. Shrestha, K. M. Daane, and J. P. Mitchell. Negrete-Yankelevich, S., C. Fragoso, A. C. Newton, and O. W. Heal.
2021. “Winter Cover Crops and No-Till Promote Soil 2007. “Successional Changes in Soil, Litter and
Macrofauna Communities in Irrigated, Mediterranean Crop- Macroinvertebrate Parameters Following Selective Logging in
land of California.” Applied Soil Ecology 166: 104068. a Mexican Cloud Forest.” Applied Soil Ecology 35: 340–55.
Laughlin, D. C., J. J. Leppert, M. M. Moore, and C. H. Sieg. 2010. Ortíz-Escobar, M. E., R. D. Zapata-Hernandez, S. S. Khalajabadi,
“A Multi-Trait Test of the Leaf-Height-Seed Plant Strategy and H. F. Franco. 2004. “Aluminio intercambiable en suelos
Scheme with 133 Species from a Pine Forest Flora.” Functional con propiedades andicas y su relaci on con la toxicidad.”
Ecology 24: 493–501. Cenicafé 55: 101–10.
ECOSPHERE 19 of 19
Otero, J. D., A. Figueroa, F. A. Muñoz, and M. R. Peña. 2011. “Loss of Soil Suazo-Ortuño, I., L. Lopez-Toledo, J. Alvarado-Díaz, and M.
and Nutrients by Surface Runoff in Two Agro-Ecosystems within an Martínez-Ramos. 2014. “Land-Use Change Dynamics, Soil
Andean Paramo Area.” Ecological Engineering 37: 2035–43. Type and Species Forming Mono-Dominant Patches: The Case
Pacheco, F. A. L., and L. F. Sanches-Fernandes. 2016. “Environ- of Pteridium aquilinum in a Neotropical Rain Forest Region.”
mental Land Use Conflicts in Catchments: A Major Cause of Biotropica 47: 18–26.
Amplified Nitrate in River Water.” Science of the Total Envi- Tallis, H., and S. Polasky. 2009. “Mapping and Valuing Ecosystem
ronment 548–549: 173–88. Services as an Approach for Conservation and Natural-
R Core Team. 2018. R: A Language and Environment for Statistical Resource Management.” Annals of the New York Academy of
Computing. Vienna: R Foundation for Statistical Computing. Sciences 1162: 265–83.
Radford, B. J., A. C. Wilson-Rummenie, G. B. Simpson, K. L. Bell, UNIVALLE, and CVC [Universidad del Valle and Corporaci on
and M. A. Ferguson. 2001. “Compacted Soil Affects Soil Autonoma Regional del Valle del Cauca]. 2007. Plan de
Macrofauna Populations in a Semi-Arid Environment in Cen- ordenamiento y manejo ambiental de la cuenca hidrogr afica del
tral Queensland.” Soil Biology and Biochemistry 33: 1869–72. río Cali. Cali: Corporaci on Aut onoma Regional del Valle del
Ramírez, M., J. Chara, L. C. Pardo-Locarno, J. Montoya-Lerma, I. Cauca.
Armbrecht, C. H. Molina, and E. J. Molina. 2012. Vaast, P., R. van Kanten, P. Siles, B. Dzib, N. Franck, J. M.
“Biodiversidad de hormigas hip ogeas (Hymenoptera: Harmand, and M. Genard. 2005. “Shade: A Key Factor for Cof-
Formicidae) en agroecosistemas del Cerrito, Valle del Cauca.” fee Sustainability and Quality.” ASIC 2004. 20th International
Livestock Research for Rural Development 24: 15. Conference on Coffee Science, Bangalore, India, October
Ramos, M. C., C. Benito, and J. A. Martínez-Casasnovas. 2015. 11–15, 2004, Association Scientifique Internationale du Café
“Simulating Soil Conservation Measures to Control Soil and (ASIC), Paris, France, 887–896.
Nutrient Losses in a Small, Vineyard Dominated, Basin.” Agri- de Valença, A. W., S. J. Vanek, K. Meza, R. Canto, E. Olivera, M. Scurrah,
culture, Ecosystems and Environment 213: 194–208. E. A. Lantinga, and S. J. Fonte. 2017. “Land Use as a Driver of Soil
Ren, T., J. Wang, Q. Chen, F. Zhang, and S. Lu. 2014. “The Effects Fertility and Biodiversity across an Agricultural Landscape in the
of Manure and Nitrogen Fertilizer Applications on Soil Central Peruvian Andes.” Ecological Applications 27: 1138–54.
Organic Carbon and Nitrogen in a High-Input Cropping Sys- Velasquez, E., P. Lavelle, and M. Andrade. 2007. “GISQ, A
tem.” PLoS One 9: e97732. Multifunctional Indicator of Soil Quality.” Soil Biology and
Rey Benayas, J. M., and J. M. Bullock. 2012. “Restoration of Biodi- Biochemistry 39: 3066–80.
versity and Ecosystem Services on Agricultural Land.” Ecosys- Wang, D., L. Guo, L. Zheng, Y. Zhang, R. Yang, M. Li, F. Ma, X.
tems 15: 883–99. Zhang, and Y. Li. 2019. “Effects of Nitrogen Fertilizer and
Rolando, J. L., C. Turin, D. A. Ramírez, V. Mares, J. Monerris, and Water Management Practices on Nitrogen Leaching from a
R. Quiroz. 2017. “Key Ecosystem Services and Ecological Typical Open Field Used for Vegetable Planting in Northern
Intensification of Agriculture in the Tropical High-Andean China.” Agricultural Water Management 213: 913–21.
Puna as Affected by Land-Use and Climate Changes.” Agricul- Webster, E., A. C. M. Gaudin, M. Pulleman, P. Siles, and S. J.
ture, Ecosystems and Environment 236: 221–33. Fonte. 2019. “Improved Astures Support Early Indicators of
Rolando, J. L., J. C. B. Dubeux, D. A. Ramirez, M. Ruiz-Moreno, C. Soil Restoration in Low-Input Agroecosystems of Nicaragua.”
Turin, V. Mares, L. Sollenbert, and R. Quiroz. 2018. “Land Use Environmental Management 64: 201–12.
Effects on Soil Fertility and Nutrient Cycling in the Peruvian Zhang, S., Q. Li, Y. Lü, X. Zhang, and W. Liang. 2013. “Contribu-
High-Andean Puna Grasslands.” Soil Science Society of Amer- tions of Soil Biota to C Sequestration Varied with Aggregate
ica Journal 82: 463–74. Fractions under Different Tillage Systems.” Soil Biology & Bio-
Rousseau, L., S. J. Fonte, O. Téllez, R. Van Der Hoek, and P. chemistry 62: 147–56.
Lavelle. 2013. “Soil Macrofauna as Indicators of Soil Quality Zimmerer, K. 1993. “Soil Erosion and Labor Shortages in the Andes
and Land Use Impacts in Smallholder Agroecosystems of with Special Reference to Bolivia, 1953-91: Implications for
Western Nicaragua.” Ecological Indicators 27: 71–82. “Conservation-With-Development”.” World Development 21:
Sanabria, C., P. Lavelle, and S. J. Fonte. 2014. “Ants as Indicators of 1659–75.
Soil-Based Ecosystem Services in Agroecosystems of the
Colombian Llanos.” Applied Soil Ecology 84: 24–30.
Schweiger, A. H., I. Boulangeat, T. Conradi, M. Davis, and J. -C. SU PP O R TI N G I N F O RMA TI O N
Svenning. 2018. “The Importance of Ecological Memory for Additional supporting information may be found in the
Trophic Rewilding as an Ecosystem Restoration Approach.” online version of the article at the publisher’s website.
Biological Reviews 94: 1–15.
Singh, P. K., M. Singh, and B. N. Tripathi. 2013. “Glomalin: An
Arbuscular Mycorrhizal Fungal Soil Protein.” Protoplasma How to cite this article: Galindo, Víctor,
250: 663–9. Carolina Giraldo, Patrick Lavelle, Inge Armbrecht,
Song, X. Z., C. X. Zhao, X. L. Wang, and J. Li. 2009. “Study of and Steven J. Fonte. 2022. “Land Use Conversion
Nitrate Leaching and Nitrogen Fate under Intensive Vegetable to Agriculture Impacts Biodiversity, Erosion
Production Pattern in Northern China.” Comptes Rendus Biol-
Control, and Key Soil Properties in an Andean
ogies 332: 385–92.
Stumpf, K. A. 1993. “The Estimation of Forest Vegetation Cover
Watershed.” Ecosphere 13(3): e3979. https://doi.
Descriptions Using a Vertical Densitometer.” Geographic org/10.1002/ecs2.3979
Resource Solution 97: 1–10.

Ecosphere - 2022 - Galindo - Land Use Conversion To Agriculture Impacts Biodiversity Erosion Control and Key Soil

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ecosphere - 2022 - Galindo - Land Use Conversion To Agriculture Impacts Biodiversity Erosion Control and Key Soil

Uploaded by

Copyright:

Available Formats

Received: 1 June 2021 Revised: 30 September 2021 Accepted: 8 October 2021

Land use conversion to agriculture impacts biodiversity,

Víctor Galindo1,2 | Carolina Giraldo1 | Patrick Lavelle3 | Inge Armbrecht2 |

Ecosphere. 2022;13:e3979. https://onlinelibrary.wiley.com/r/ecs2 1 of 19

INTRODUCTION impact and the formation of stable biogenic structures by

Property Forest Abandoned Pasture Coffee Annual crops F df p

Property Forest Abandoned Pasture Coffee Annual crops F df p

Infiltration Runoff Sediments

Parameter Forest Abandoned Pasture Coffee Annual crops Fa df p

−1.0 −0.5 0.0 0.5 1.0

Parameter Forest Abandoned Pasture Coffee Annual crops Fa df p

−1.0 −0.5 0.0 0.5 1.0

Richness (Sp. plot )

Covariation among ES probably influenced by the application of organic fertilizers

the functional groups of vegetation and soil macrofauna ACKNOWLEDGMENTS

You might also like