ArticlePDF Available

Species richness increases income in agroforestry systems of eastern Amazonia

  • Maranhão State University, São Luís, Brazil


Biodiversity is believed to reduce risks (resistance and resilience against perturbations), to increase productivity via niche expansion, and possibly also to improve resource efficiency via mutually benefic species interactions. Agroforestry has been postulated as an ideal pathway of maintaining or restoring biodiversity in a socioeconomically sustainable manner. This study tests the relevance of agroforestry species diversity and richness on socioeconomic performance in a wide range of agroforestry systems in 38 farms aggregated in four clusters of sites in eastern Amazonia. We cover both commercial and subsistence agroforestry, ranging from simply structured plantations to diverse systems (enriched fallows, multi-strata home gardens), as well as pastures and shifting cultivation for comparisons. We quantify (i) all cultivated species, classifying them economically into species with commercial value, primarily subsistence purpose species or ‘non-productive’ species, and (ii) socioeconomic system variables (costs, monetary/non-monetary income, degree of satisfaction). Land-use intensity (per-hectare costs and income) was highest in commercial agroforestry and subsistence home gardens, and lowest in enriched fallows and pastures. All agroforestry systems resulted in higher income:cost ratios and greater satisfaction than pastures and shifting cultivation. Net income, non-monetary income and income:cost ratio were maximum in home gardens. Total species richness was negatively related with costs and monetary income, but not with non-monetary income, due to occupation of space by ‘non-productive’ species (juveniles or species providing ecosystem services). By contrast, productive (combining commercial and subsistence) species richness was positively related with (mainly non-monetary) income, net income and income:cost ratio. According to GLM, both productive species richness and Shannon–Wiener diversity positively affected net income. Future efforts for food security and poverty reduction need to focus more on species-rich agroforestry systems, both in terms of applied research and of extension service programs. Notably, the ubiquitous and successful home gardens merit far more attention.
Species richness increases income in agroforestry systems
of eastern Amazonia
Ernesto Go
´mez Cardozo .Henry Mavisoy Muchavisoy .Hulda Rocha Silva .
Marcelo Luı
´s Corre
ˆa Zelaraya
´n.Marcio Fernandes Alves Leite .
Guillaume Xavier Rousseau .Christoph Gehring
Received: 26 August 2014 / Accepted: 17 June 2015
ÓSpringer Science+Business Media Dordrecht 2015
Abstract Biodiversity is believed to reduce risks
(resistance and resilience against perturbations), to
increase productivity via niche expansion, and possi-
bly also to improve resource efficiency via mutually
benefic species interactions. Agroforestry has been
postulated as an ideal pathway of maintaining or
restoring biodiversity in a socioeconomically sustain-
able manner. This study tests the relevance of
agroforestry species diversity and richness on socioe-
conomic performance in a wide range of agroforestry
systems in 38 farms aggregated in four clusters of sites
in eastern Amazonia. We cover both commercial and
subsistence agroforestry, ranging from simply struc-
tured plantations to diverse systems (enriched fallows,
multi-strata home gardens), as well as pastures and
shifting cultivation for comparisons. We quantify
(i) all cultivated species, classifying them economi-
cally into species with commercial value, primarily
subsistence purpose species or ‘non-productive’
species, and (ii) socioeconomic system variables
(costs, monetary/non-monetary income, degree of
satisfaction). Land-use intensity (per-hectare costs
and income) was highest in commercial agroforestry
and subsistence home gardens, and lowest in enriched
fallows and pastures. All agroforestry systems resulted
in higher income:cost ratios and greater satisfaction
than pastures and shifting cultivation. Net income,
non-monetary income and income:cost ratio were
maximum in home gardens. Total species richness was
negatively related with costs and monetary income,
but not with non-monetary income, due to occupation
of space by ‘non-productive’ species (juveniles or
species providing ecosystem services). By contrast,
productive (combining commercial and subsistence)
species richness was positively related with (mainly
non-monetary) income, net income and income:cost
ratio. According to GLM, both productive species
richness and Shannon–Wiener diversity positively
affected net income. Future efforts for food security
and poverty reduction need to focus more on species-
rich agroforestry systems, both in terms of applied
research and of extension service programs. Notably,
the ubiquitous and successful home gardens merit far
more attention.
Keywords Babassu pasture Commercial
agroforestry Fallow enrichment Home gardens
Non-monetary income Subsistence farming
Agroforestry exists on all continents and most biomes
(Zomer et al. 2009), in a multitude of systems which
E. G. Cardozo H. M. Muchavisoy
H. R. Silva M. L. C. Zelaraya
´nM. F. A. Leite
G. X. Rousseau C. Gehring (&)
Agroecology Program of Maranha
˜o State University,
˜o Luı
´s, Brazil
Agroforest Syst
DOI 10.1007/s10457-015-9823-9
vary widely both in terms of their socioeconomic
origins and purposes and their botanical composition
(Atangana et al. 2014). Systems origins range from
traditional indigenous (Diemont and Martin 2009)to
‘modern’ (Yamada and Gholz 2002), purposes range
from subsistence to commercial (Nair 1985), and
species richness and diversity from low/simplistic
(alley cropping, tree plantation) (Atangana et al. 2014;
Nair 2013) to high/complex (Schroth and da Mota
2014; Nair 2013). Consequently, ecological and
socioeconomic systems sustainability will likewise
vary widely (Atangana et al. 2014).
The multiple relationships between species diver-
sity and agroecosystem functioning are fundamental
issues of ecological theory. Expected biodiversity
benefits are higher system stability (via increased
resistance and resilience i.e., against pests or diseases
(Ratnadass et al. 2012) and weather/climate instabil-
ities (Lin 2011), increased resource exploration and
productivity via niche expansions (Weiher and Keddy
2001; Garcı
´a-Barrios and Ong 2004), and possibly
also increased efficiencies caused by the stimulation of
positive species interactions (Cardinale et al. 2002;
´a-Barrios and Ong 2004). Forms of such rela-
tionships, redundancies, and possible diversity thresh-
olds have been hotly debated (Schulze and Mooney
1993; Loreau 2000; Schleuning et al. 2015), and
functional diversity can be more relevant than taxo-
nomic diversity (Dı
´az and Cabido 2001).
From an economic perspective, portfolio diversity
bears both advantages and disadvantages (Godsey
2010). Diversity can reduce risks, but also increase
management complexity (Altieri et al. 2011). Low
diversity and concentration of investments in a few
components permits the development of scale effects,
both in systems management and in the sale of
products (Rosa et al. 2009). Sociologically, diversity
of cultures and of species use increase resilience, serve
as insurance against unexpected or disruptive events
and provide components that facilitate renewal after
disturbances of ‘socioecological systems’ (Berkes
et al. 2003; Cabell and Oelofse 2012).
This study tests the hypothesis that species richness
and diversity can drive socioeconomic performance of
agroforestry systems. We investigate possible rela-
tionships of commercial, subsistence and ‘non-pro-
ductive’ species on key socioeconomic parameters
over a wide range of agroforestry systems in eastern
Amazonia. Specifically, we (i) compare simple or
complex subsistence or commercial agroforestry sys-
tems—among another and with extensive pastures
with babassu palms (dominating in area) and slash-
and-burn shifting cultivation (sustaining most of rural
population); (ii) establish specific relationships
between commercial, subsistence, and ‘non-produc-
tive’ species diversity and costs, benefits, profitability,
and satisfaction; and (iii) give practical recommenda-
tions for agroforestry research and development.
Study region and study sites
Research was conducted in the eastern periphery of
Amazonia in four clusters of sites, in the Brazilian
states of Maranha
˜o3°280S/44°530W and Para
48°220W. Three of the four clusters are located in
Maranhao State, the Tome
´-Ac¸u cluster of Para
´State is
approximately 400 km further westward (Fig. 1).
Climate classification according to Ko
¨ppen is Aw
and Ami (Alvares et al. 2013), varying slightly
between clusters (2100 mm annual rainfall in the
˜o clusters and 2300 mm in Tome
´-Ac¸u, 5 vs.
4 months hydric deficit). All soils are acid and
nutrient-poor upland soils, classified as sandy loam
Oxisols. Vegetation is predominantly extensive pas-
tures or secondary forests. Frequent slash-and-burning
has drastically increased dominance of the babassu
palm (Attalea speciosa Mart., Arecaceae), covering an
estimated 100,000 ha in Maranha
˜o alone (Porro
Agroforestry systems
Research was conducted with 38 families (owners and
careholders of farms, consisting of farmer, wife, and
the children living there): 27 families/sites with
agroforestry systems, three with babassu pastures,
and eight with slash-and-burn shifting cultivation. We
classify land-use systems according to their socioeco-
nomic finalities in three classes (‘commercial’,
‘mixed’, and ‘subsistence’), as proposed by Nair
(1993), and we distinguish three types of subsistence
agroforestry. All land-use systems have three or more
site replications, but not all systems occurred in all
Agroforest Syst
clusters (Fig. 2). At site selection, we strived to
maintain a high level of within-system structural
homogeneity, though species composition did vary
regionally and between clusters of sites.
Commercial agroforestry plantations
The main products of these regularly spaced
agroforestry plantations are coconut (Cocos nucifera),
Fig. 1 Clusters of study sites
Fig. 2 Agroforestry systems, classifications, and their aggregation in site clusters. Values in brackets give the number of sites
Agroforest Syst
acai (Euterpe oleracea), cacao (Theobroma cacao),
cupuassu (Theobroma grandiflorum), and black pep-
per (Piper nigrum). We directly compare two types of
commercial agroforestry plantations that are ecolog-
ically similar but differ socioeconomically by distin-
guishing into:
(a) Purely commercial operations: Commercial agro-
forestry enterprise (CAE) Large-scale plantations
mainly by the pioneer Japanese immigrants (Yamada
and Gholz 2002); and
(b) Mixed systems: Commercial agroforestry by
smallholder farmers (CASF) Inspired by the Japa-
nese agroforestry plantations, but owned by small-
holder farmers. These plantations differ from the
commercial large-scale operations in the inclusion
of additional subsistence species such as acai,
banana (Musa spp.) and pupunha (Bactris gasipaes).
Subsistence agroforestry
Home gardens Multistrata agroforestry systems
surround the houses, providing vital shade and
omnipresent throughout our region. Dominant
overstory trees are cupuassu, acai, cacao, banana,
mango (Mangifera indica), and jackfruit (Artocarpus
integrifolia), understory fruticulture and small
domestic animals are further important components.
We distinguish home gardens according to their
size into:
(a) Small home garden (SH): home garden \1ha.
(b) Medium-sized home garden (MH): home garden
[1 ha.
Enriched fallow (EF) This agroforestry system is
established by enrichment plantings of fruit and timber
in the understory of old secondary forests. The degree
of human control over vegetation structure and species
composition is far lower than in all other systems
covered by our study. Fallow enrichment developed
independently in various site clusters by isolated
initiatives of innovative farmers partly was initiated
(and thereafter abandoned) by a local NGO.
Enrichment plantings are an ecologically valuable
option for income generation in otherwise
‘unproductive’ forest reserves required by law. In
our study, secondary forest age ranged from 6 to
[30 years and enrichment planting age was
3–12 years. Banana, acai, and bacuri (Platonia
insignis) are the most important enrichment species.
We compare these agroforestry systems with the
region’s two predominant land-use systems, which
likewise constitute agroforestry in the wider sense
(silvopastoral pastures with babassu and sequential
agroforestry shifting cultivation):
Pasture with babassu (PB) Extensive pastures
predominate throughout most of the Amazonian arc
of deforestation, with brachiaria grass and stocking
rates of typically \1 Nelore cattle ha
et al. 2010). A conspicuous feature is the babassu
palms growing within these pastures, providing shade
for the cattle, and a source for babassu nut extractivism
(not quantified in this study).
Slash-and-burn shifting cultivation (SB) In terms of
area, this age-old production system is far less important
than the pastures (i.e., low percentages of active fields),
but this system sustains 74.4 % of the region’s rural
population (MDA 2011). We investigate the
socioeconomic variables of the cultivation phase, and
do not attempt to quantify the biodiversity nor the
extraction of timber and non-timber products during the
region typically is young and strongly degraded
(consequence of repeated slash-and-burn cycles and the
shortening fallow phases), extractivism (some medicinal
plants, nectar and pollen for bees, sometimes charcoal
production) is likely to be only of minor relevance.
We identified and quantified all planted agroforestry
species and all spontaneously occurring species
C5 cm diameter at breast height (i.e., dbh at
1.30 m), in the case of cacao and cupuassu C5cm
diameter at 30 cm height, and also, because of its
prominent economical relevance, black pepper. The
inventory was conducted in one circular sampling unit
per site with 50 m diameter (1963 m
) in all agro-
forestry systems without regular spacing (Richards
1996; Brown 2002; Soto-Pinto et al. 2009). Given
their regular spacing and low mid-scale spatial vari-
ability, we preferred a differing sampling scheme in
the regularly spaced commercial agroforestry planta-
tions (both CAE and CASF) with three 25 925-m
quadrants per site (1875 m
). This scheme has
Agroforest Syst
previously been successfully used by Kato (2009) and
Somarriba et al. (2013) in similar plantations. We
subsequently corrected for the slightly different sam-
pling sizes. We discarded large boarder zones to
neighboring vegetation.
In a first step, the identification of agroforestry
species counted on the help of the farmers who
indicated the cultivated plants and gave them associ-
ating local names, which were subsequently trans-
formed into scientific nomenclature. In all doubtful
cases, taxonomic classification was based on subse-
quent analysis in the herbarium of Maranha
˜o State
University, following the Angiosperm Phylogeny
Group III classification system (Angiosperm Phy-
logeny Group III 2009).
We calculate (i) species richness (number of
species in sampling area), and (ii) diversity indices
of Simpson and Shannon–Wiener and equitability
index of Pielou (Magurran 1988), based on species’
abundance and frequency shares and using FITOPAC
software (Shepherd and Fitopac shell 2009).
We classified all agroforestry species into (1)
commercial species (market production), (2) subsis-
tence species (mainly auto-consumption, small quan-
tities are also sold), and (3) ‘non-productive’ (species
without immediate productive value but often exerting
important ecosystem-services such as shade, organic-
matter cycling/soil-cover or nectar/pollen for bees,
medicinal), as well as juvenile plants.
Socioeconomic variables
We collected all socioeconomic variables via semi-
structured interviews with open questions (Sibelet and
Smektala 1999). We classified costs and income as
Monetary income
Value of commercialized production (on-farm prices).
We do not include timber as future income source.
Non-monetary income
Value of non-commercialized production that would
have been obtained if the farmer had sold production
instead of consuming it (on-farm prices).
Total costs Costs for maintenance and harvest
operations, comprising external inputs and (own-
family or hired) labor. Original installation costs are
not considered.
Net income
Sum of monetary and non-monetary income minus
total costs.
Income:cost ratio
An indicator of economic efficiency (das Chagas
Oliveira et al. 2013).
Estimates of costs, wages, and income were
obtained using the prices reigning in each county in
2012. We subsequently extrapolated the values of our
sampling sites to per-hectare estimates.
We estimated the degree of satisfaction via auto-
evaluation of life quality of the farmers and wives
(Veenhoven 1994,2007), combining aspects of work
routine in the different land-use systems, comfort,
food availability and quality, and leisure aspects.
Farmers classified their satisfaction utilizing the
following numerical scale:
8–10: Very satisfied
6–8: Satisfied
4–6: Indifferent
2–4: Dissatisfied
0–2: Very dissatisfied
We verified normality of distribution of all data via
Shapiro–Wilk and Lilliefor’s tests against normality,
and checked for homogeneity of variance with
Levene’s test (Crawley 2007). We compared the
different agroforestry systems via one-way ANOVA
and post hoc Spjøtfoll–Stoline test (Tukey for unequal
replication numbers; Spjøtvoll and Stoline 1973), and
analyzed relationships between biodiversity and
socioeconomic variables via linear and logarithmic
regressions. We identified and eliminated two outlier
values. Our experimental setup is unbalanced, due to
the non-occurrence of some systems in some clusters
of sites. We believe this is not a serious problem,
because of (i) small regional edaphic differences
(chap. 2.1), and (ii) lacking statistical differences of all
biodiversity and socioeconomic variables between the
three Maranha
˜o and the Tome
´-Ac¸u clusters of sites,
Agroforest Syst
both in home gardens (small and medium-sized
combined) and in slash-and-burn shifting cultivation
systems. We also investigate the impact of total and
productive species richness and diversity and of the
type of agroforestry system on costs and income via
generalized linear modeling (GLM). All analyses
involving species diversity exclude our two ‘control’
systems (pastures with babassu containing only one
species, and slash-and-burn shifting cultivation with-
out data on the fallow phase). Statistical analyses were
conducted with STATISTICA 8.0 (StatSoft 2007). In
order to gain a better overview of all variables and
agroforestry systems under investigation, we also
conducted a principal component analysis using
INFOSTAT software (Di Rienzo et al. 2012).
Species composition and systems management
We identified, in the 27 agroforestry sites, a total of 83
species, distributed in 73 genera and 34 plant families.
Details on taxonomy, use, and origin of the overall ten
most abundant species of this study are given in Annex 1.
Figure 3explores between-system differences in the
relative abundances of the principal agroforestry species,
and large and systematic differences turn apparent.
Commercial agroforestry enterprise
Species-poor plantations, cocoa, banana and black
pepper combine 90 % of all plants. Contrary to the
other two, the main finality of banana is in its
ecosystem services (rapid shade, organic matter).
The main management activities are understory
clearing and periodic prunings, fertilizer and pesticide
applications, and cocoa processing. This system relies
exclusively on hired labor and amply applies external
In direct comparison, commercial agroforestry by
smallholder farmers was slightly more diverse, five
species combine 93 % of all plants. Production of
cocoa and black pepper is completely commercial-
ized, whereas ac¸ai, banana, and cupuassu serve both
for auto-consumption and the market. Next to under-
story clearing and periodic prunings, cocoa and
cupuassu fruit processing is important, relying almost
exclusively on family labor, though additional labor is
hired for black pepper harvesting. Contrary to
Fig. 3 The most abundant species in eastern Amazonian agroforestry systems
Agroforest Syst
commercial agroforestry enterprises, there is no use of
external inputs (fertilizers or pesticides).
Small (\1 ha) home gardens
Surrounding the houses and typically inherited from
the preceding generation(s). Five species (cupuassu,
ac¸ai, banana, jackfruit, and cocoa) combine 76 % of
all plants. Almost all production is for home con-
sumption, though excess cupuassu and ac¸ai also are
sold to local markets. The main management activities
are periodic pruning conducted exclusively by family
labor; no use of external inputs. Female labor
predominates in cupuassu fruit processing.
Medium-sized ([1 ha) home gardens
Often developed around initial natural clusters of ac¸ai
palms close to springs, and subsequently systemati-
cally enriched and enlarged. Ac¸ai and cupuassu
combine 79 % of all plants. The main management
activities are periodic prunings conducted exclusively
by family labor; no use of external inputs.
Enriched fallow
The most important enrichment species are banana
ac¸ai and bacurı
´, combining 55 % of all plants. Most of
the remainder are spontaneously occurring species of
the secondary forest overstory, which furnishes
ecosystem services such as shade, organic matter/
litter, wildlife feed (Cecropia sp.), or nectar and pollen
for bees (Andira sp.). The multi-use babassu palm
provides charcoal, palm oil, construction material, and
feed for (rodent) wildlife. The main management
activities are periodic understory clearing and shade
regulation, using exclusively family labor and no
external inputs.
Species richness and diversity
Table 1shows the key biodiversity indicators of our
agroforestry systems. We exclude pastures with
babassu (with merely one woody species, the babassu
palm) and slash-and-burn shifting cultivation (no data
on fallow phase) from this analysis. Within agro-
forestry systems, species richness and Shannon diver-
sity are highest in enriched fallows and small home
gardens, and lowest in commercial agroforestry
enterprises. Dominance (Simpson) is lowest in small
home gardens, equitability (Pielou) does not differ
between systems.
Socioeconomic profile of agroforestry species
Of the total of 83 species, 4 % were exclusively
commercial, 23 % served both auto-consumption and
commercial purposes, and 73 % ‘non-productive’
species were maintained because of the ecosystem
services they provided, were juveniles or medicinal
plants. Species composition differed markedly
between systems (Table 2).
Total species number was similar in enriched
fallows and small home gardens, but home gardens
had a much higher quantity of productive species.
Costs and benefits, net income
The only two systems with significant costs caused by
external inputs (fertilizers and pesticides) and by hired
labor were the two exclusively commercial systems,
commercial agroforestry enterprises and pastures with
Table 1 Total species richness and biodiversity indicators in agroforestry systems of eastern Amazonia
Agroforestry system NNumber of species per
plot (1963 m
Shannon–Wiener Simpson Pielou
Commercial agroforestry enterprise 5 4.40 ±0.69b 0.73 ±0.11b 0.66 ±0.05a 0.54 ±0.04a
Commercial agroforestry by smallholder farmers 4 7.50 ±1.19ab 0.98 ±0.13ab 0.50 ±0.09ab 0.51 ±0.08a
Small home garden 7 12.14 ±1.78a 1.70 ±0.02a 0.26 ±0.04b 0.68 ±0.05a
Medium-sized home garden 4 10.00 ±3.10ab 1.08 ±0.10ab 0.49 ±0.04ab 0.51 ±0.05a
Enriched fallow 7 12.57 ±1.60a 1.54 ±0.25a 0.37 ±0.10ab 0.60 ±0.08a
Means ±SE, absence of common letters within the same column indicates significant between—system differences (monospecific
pastures with babassu not included in statistical analysis), nnumber of sites
Values obtained in regularly spaced agroforestry plantations corrected for their 4.5 % smaller sampling size
Agroforest Syst
babassu. In the commercial agroforestry enterprise
plantations, external inputs summed 36 % and hired
labor 64 % of total costs. In all other agroforestry
systems, costs were caused exclusively by proper
(family) labor.
Figure 4exhibits large differences between systems
in monetary and non-monetary benefits and net income.
As to be expected, monetary income and costs are
highest in commercial agroforestry enterprises. Costs
are also high in slash-and-burn shifting cultivation. Both
costs and returns are far lower (i.e., more extensive land
use) both in enriched fallows and in pastures with
babassu. Net per-hectare income is highest in home
gardens (due to low costs and high non-monetary
income), and lowest in the extensive pastures with
babassu and in slash-and-burn shifting cultivation.
Based on the minimum wage of R$622.00 in 2012,
small and medium-sized home gardens annually
generated 7.47 and 6.77 minimum wages per hectare
respectively, whereas pastures with babassu and slash-
and-burn shifting cultivation generated only 0.77 and
1.85 minimum wages per hectare respectively, the
other systems were intermediate between these two
extremes (CAE, 5.02, CASF, 5.48, EF, 2.43).
Figure 5compares the income:cost ratios as a
measure of the ‘socioeconomic efficiency.’ Efficiency
is highest in home gardens (especially in the small
ones), and lowest in commercial agroforestry planta-
tions (both enterprise and smallholder farmer ven-
tures), the enriched fallows, and especially in the
predominating pastures with babassu and slash-and-
burn shifting cultivation land-use systems.
Table 2 Species number and their socioeconomic profile in agroforestry systems of eastern Amazonia (means ±SE)
Species profile Agroforestry system
Commercial 2.2 (±0.2) 1.3 (±0.5) 0 0 0
Auto-consumption and commercial 0.0 3.0 (±0.9) 7.7 (±0.9) 4.8 (±0.3) 3.0 (±0.3)
Non-productive (juveniles and ecosystem services) 1.6 (±0.6) 2.5 (±1.0) 4.3 (±1.2) 5.3 (±3.1) 9.6 (±1.7)
All species 4.20 ±0.66 7.50 ±1.19 12.14 ±1.78 10.00 ±3.10 12.57 ±1.60
CAE Commercial agroforestry enterprise, CASF commercial agroforestry by smallholder farmers, SH small home garden, MH
medium-sized home garden, EF enriched fallow
Fig. 4 Total costs, monetary and non-monetary benefits, and
net income generated in agroforestry and predominating land-
use systems of eastern Amazonia (means ±SE). CAE Com-
mercial agroforestry enterprise, CASF commercial agroforestry
by smallholder farmers, SH small home garden, MH medium-
sized home garden, EF enriched fallow, PB pasture with
babassu, SB slash-and-burn shifting cultivation. Absence of
common letters indicates significant difference between systems
Agroforest Syst
Impacts of species richness and diversity on costs,
benefits, and income
Figure 6explores the relationships between total
agroforestry species richness and costs and income
(top), and between productive (i.e., commercial and
auto-consumption) species richness and non-monetary
income, net-income and income:cost ratio (bottom).
Relationships were non-linear in some cases. Rela-
tionships with Shannon–Wiener diversity were similar
though less expressive and significant, whereas there
were no relationships with Simpson dominance or
Pielou equitability (data not shown).
Total species richness was negatively related both
with costs and with monetary (but not with non-
monetary) income, presumably a consequence of
Fig. 5 Income-to-cost ratio
of agroforestry and of
predominating land-use
systems of eastern
Amazonia (means ±SE).
CAE Commercial
agroforestry enterprise,
CASF commercial
agroforestry by smallholder
farmers, SH small home
garden, MH medium-sized
home garden, EF enriched
fallow, PB pasture with
babassu, SB slash-and-burn
shifting cultivation.
Absence of common letters
indicates significant
difference between systems
Fig. 6 Negative impact of total agroforestry species richness
on total costs, monetary and (excluding purely commercial
agroforestry enterprises) non-monetary income (top), and
positive impact of productive species richness on non-monetary
income, net income, and income:cost ratio (bottom) in eastern
Amazonian agroforestry systems
Agroforest Syst
occupying space with non-productive ‘other’ species.
By contrast, non-monetary income, net income and the
income:cost ratio were positively related with pro-
ductive species richness.
Table 3depicts the joint impacts of total and of
productive (commercial and auto-consumption) spe-
cies richness and diversity and of the categorical
variable ‘agroforestry system’ on net income, as
identified by generalized linear modeling (GLM).
When considering all species, neither species richness
or diversity affected net income, whereas limiting our
analysis to the productive species, both species richness
and diversity turn important predictors of net income.
GLM based on total species consistently indicated
‘agroforestry system’ as predictors of costs, monetary
and non-monetary income, and failed to do so for
species richness and diversity. The only exception to
this is in the significant effects both of species richness
and Shannon–Wiener diversity on profitability (the
income-cost ratio), presumably the outcome of the
higher profitability of home gardens relative to
plantation agroforests (data not shown). GLM limited
to productive species likewise consistently identified
significant effects of ‘agroforestry system’ on costs,
monetary and non-monetary income, but failed to do
so for species richness or diversity (data not shown).
Multivariate correlations between biodiversity
and monetary variables
The first two axes of principal component analysis
accounted for 88.0 % of total variation (Fig. 7). Axis-
1 (70.0 % of variation) identifies systems with high
and low diversity, and axis-2 (18.0 % of variation) is
largely income-related. Whereas commercial agro-
forestry enterprises constitute the species-poor
extreme of axis-1, small home gardens and enriched
fallows are on the diverse other end. The main
difference in axis-2 coincides with the latter two
systems, reflecting economic differences caused by
the contrasting predominance of ‘productive’ versus
‘non-productive’ species.
The degree of satisfaction was systematically higher in
all agroforestry systems than in the two predominating
land-use systems, pasture with babassu and slash-and-
burn shifting cultivation (Fig. 8). All farmer families
of all types of agroforestry systems were ‘very
satisfied’, satisfaction was maximum in commercial
agroforestry enterprises and near-maximum in the
home gardens.
Taxonomic and/or functional diversity are known to
be key for the functioning and stability of (agro)
ecosystems (Schulze and Mooney 1993; Loreau 2000;
Schleuning et al. 2015). Relative to tree or crop
monocultures, multi-species agroforests could have
higher productivity (niche expansion, improved
resource exploitation; Cannell et al. 1996; Schroth
Table 3 Impacts of species
richness and diversity and
of agroforestry system on
net income, for all species
(top) and for productive
(commercial and auto-
consumption) species
Significant or near-
significant pvalues are
highlighted in bold
SS df MS Fpvalue
Intercept 2,207,159 1 2,207,159 1.210 0.285
Species richness 138,380 1 138,380 0.076 0.786
Shannon diversity 1,248,133 1 1,248,133 0.685 0.418
Simpson diversity 2404 1 2404 0.001 0.971
Agroforestry system 46,476,381 4 11,619,095 6.374 0.002
Error 34,636,260 19 1,822,961
SS df MS Fpvalue
Intercept 13,773,557 1 13,773,557 6.526 0.019
Species richness 9,856,415 1 9,856,415 4.670 0.045
Shannon diversity 9,220,406 1 9,220,406 4.368 0.050
Simpson diversity 3,918,752 1 3,918,752 1.857 0.189
Agroforestry system 22,022,129 4 5,505,532 2.608 0.068
Error 40,102,858 19 2,110,677
Agroforest Syst
et al. 2001), and higher stability (resistance against and
resilience following disturbances), though species’
identity and functional traits strongly affect such
relationships (Dı
´az and Cabido 2001). From a market
perspective, product diversity reduces market risks
(vulnerability against price fluctuations) (Faye et al.
2011; Vallejo et al. 2015), but likewise reduces scale
effects and increases marketing costs (Souza et al.
2011). Our study establishes positive linear relation-
ships between productive species richness and both
monetary and non-monetary income and income:cost
ratios over a wide range of agroforestry systems (see
Fig. 6). According to GLM, productive (but not total)
species richness strongly affected net income.
Species-poor commercial agroforestry plantations
have both the highest costs and the highest (monetary)
returns. On the other end are low-intensity and (total)
species-rich enriched fallows. Lower intensity is
caused by the occupation of space and light capture
by ‘non-productive’ species (Steffan-Dewenter et al.
2007; Godsey 2010; Clough et al. 2011), which reduce
both the costs and monetary (but not non-monetary)
income. Next to the ‘species of the future’ (still
unproductive juveniles), this category is composed of
plants renown for the ecosystem services they provide
(shade, N
fixation, organic matter cycling) (Jose
2009; Godsey 2010). Taking only the productive
species (which generate commercial and/or non-
monetary auto-consumption benefits), agroforestry
species richness and diversity are positively related
Fig. 7 Principal component analysis of tree diversity and
income variables over five agroforestry systems (27 sites) in
eastern Amazonia (CAE commercial agroforestry enterprise,
CASF commercial agroforestry by smallholder farmers, SH
small home garden, MH medium-sized home garden, EF
enriched fallow). The first two axes of PCA accounted for
88 % of total variation
Fig. 8 Degree of satisfaction of farmers in the different land-
use systems of eastern Amazonia. CAE Commercial agro-
forestry enterprise, CASF commercial agroforestry by small-
holder farmers, SH small home garden, MH medium-sized home
garden, EF enriched fallow, PB pasture with babassu, SB slash-
and-burn shifting cultivation
Agroforest Syst
with non-monetary income, net income, and, most
expressed of all, the income:cost ratio (i.e., efficiency).
The strong relationships with net income and income:-
cost ratio point to efficiency increase, possibly gener-
ated by synergies between agroforestry species
´a-Barrios and Ong 2004).
Next to the above-mentioned positive ecological
biodiversity effects, two further—socioeconomic—
benefits of agroforestry species diversity are impor-
tant: (i) a better temporal distribution of labor demands
and income generation, and (ii) reduced (financial and
non-financial) risks. Whereas overall labor availability
as the main input factor in subsistence smallholder
systems is high, peak labor demands during planting
operations and subsequently during weeding and
harvest are limiting factors in traditional slash-and-
burn shifting cultivation (Metzger 2002). Income in
food and other products is likewise very unevenly
distributed, with critically low-income phases during
dry season and at onset of rainy season (Huss-
Ashmore and Goodman 1988). Tree and palm crops
in agroforestry can reduce seasonality (a crucial issue
in poverty alleviation), both in labor requirements and
in income generation, this effect could increase with
agroforestry species diversity.
Even though profitability (i.e., the income:cost
ratio) is lowest in commercial agroforestry enterprise
plantations, the average ratio of almost 2 (1.97)
nevertheless is a sound investment. Yamada and Gholz
(2002) confirm the efficiency of this system developed
by Japanese immigrants in southern Para
´State, with
returns generated by 10–20 ha agroforestry plantations
equivalent to those of 400–1200 ha pasture. The
income:cost ratio is 36.7 % higher (2.70) in commer-
cial agroforestry by smallholder farmers, which in turn
contains 87.5 % more productive species relative to
pure commercial operations. Future financial returns in
both commercial agroforestry systems are bound to
increase in both systems, due to the timber species
overstory (not considered in our study).
We underestimate total income generation in
pastures with babassu as we omit babassu nut extrac-
tivism for palm oil and charcoal production. We do
this from the farmer’s perspective, as this income is
generated by non-farm actors (the babassu nut-cracker
women), and is sporadic and geographically unpre-
dictable. This is because of a legal singularity in
˜o State, which guarantees free access and
exploitation of babassu nuts on all private lands (the
‘Free Babassu’ law). Manual babassu nut cracking still
provides a vital income source for the rural poor
especially during dry season (Porro et al. 2004), for
some 300,000 families in Maranha
˜o State alone
(Almeida et al. 2001). However, (almost exclusively
female) labor is tough, involves health risks and
generates very marginal income, well below the
poverty line. Thanks to the advent of new income
opportunities (migration to urban areas, government
programs), babassu extractivism is declining, with a
9.7 % reduction of oil production between 2006 and
2010 (IBGE 2010). The ‘silvopastoral’ babassu pas-
tures not only are by far the least productive system,
they also have the lowest biodiversity (one single
‘forestry’ species). Carbon stocks of these sites are
likewise much lower (Muchavisoy 2013). This con-
firms the low socioeconomic and ecological efficiency
of extensive pastures, which only are profitable
because of the unequal land distribution/low land
prices, connected scale effects and high labor effi-
ciency, and (mainly in the past) direct and indirect
government subsidies (Porro et al. 2004).
Our other control, slash-and-burn shifting cultiva-
tion, is stricken by a sustainability crisis. Estimates of
the socioeconomic importance of slash-and-burn
shifting cultivation are scarce, old, and insecure. This
form of land use is believed to sustain some 300–500
million people worldwide (Brady 1996). Whereas
non-degraded fallows can maintain considerable bio-
diversity (Padoch and Pinedo-Vasquez 2010), this is
not the case in our study region, where increasing
land-use pressure and reduced fallow periods (in our
study region only 2–5 years) cause continuous degra-
dation, reduced productive potential and yields, lower
resilience, and a widespread increase in rural poverty,
confirming trends of many other parts of the humid
tropics (Styger et al. 2007; Lawrence et al. 2010).
There is a large intensification potential via techno-
logical efficiency increases (Pascual 2005), next to
understory enrichment plantings in older fallows (as
investigated in this study), main intensification path-
ways are in ecological enrichment of young regrowth
with fast-growing legume trees (Koutika et al. 2005)
and/or fire-free land preparation (Denich et al. 2004),
or alternatively slash-and-char land preparation
(McHenry 2009).
We investigate only the agricultural phase of
shifting cultivation, which explains the higher per-
hectare income generation (intensity) relative to
Agroforest Syst
extensive pastures with babassu. We did not quantify
income-generation via extractivism of useful species
during the fallow phase. As our secondary forests are
severely degraded and species-poor, this income
source is likely very small. Our results indicate a
low profitability of shifting cultivation, as both per-
hectare net income and income:cost ratio are signif-
icantly higher in all agroforestry systems.
Our study identifies home-garden agroforestry as
the most efficient and promising of all land-use
systems, top performing both in net benefits per area
(averages of small and mid-sized home gardens
generate R$ 4532 ha
or 7.3 minimum monthly
wages (of R$ R$ 622 in census year 2012) and in
profitability (income:cost ratio of 5.5). Coincidentally,
we encountered maximum agroforestry biodiversity in
the small home gardens (1.7 ±0.02 Shannon diver-
sity index and 12.1 ±1.8 species richness/plot) which
is intermediate relative to Shannon diversity reported
in other studies ranging from 1.00 (home gardens in
the northern periphery of Amazonia; da Semedo and
Barbosa 2007) to 2.21 and 2.30 in home gardens of
neighboring Para
´State and in urban home gardens in
´State of southern Brazil (Gomes 2010; Vieira
et al. 2012). Home gardens developed independently
in many cultures throughout the world, especially in
the humid tropics. Even though they still are definitely
under-researched (Nair 2001) and have never been
part of systematic agronomic improvement efforts
(Kumar 2006), these systems persist to the modern day
and constitute a remarkable success story both in
ecological and socioeconomic terms (Peyre et al.
2006; Galluzzi et al. 2010). Our results point to three
key socioeconomic features for rural poverty reduc-
tion efforts destined to those with no money and little
land: low costs/zero external inputs, maximum prof-
itability (income:cost ratio) and high (mainly non-
monetary) per-hectare income generation. Home gar-
dens generate high-quality products for auto-con-
sumption, the degree of satisfaction of their farmers is
near maximum.
(i) Agroforestry species diversity reduces costs
and increases (especially non-monetary)
income and profitability. This is important
both from theoretical and practical
(ii) All agroforestry systems are more sustainable
than extensive pastures with bab assu and slash-
and-burn shifting cultivation, the two predom-
inating land-use systems throughout Amazo-
nia. This reiterates the notion that agroforestry
can provide a solution to the socioecological
sustainability crisis in the tropics.
(iii) Farmer’s satisfaction is maximum both in
commercial agroforestry plantations and in
the home gardens, which form the two most
intense forms of agroforestry, and which at
the same time are the extreme ends in terms
of their economic and biodiversity settings.
(iv) Traditional species-rich home gardens out-
perform all other systems in terms of net
income per area, profitability, and non-mon-
etary income, providing the explanation for
their pan-tropical success and persistence to
date. Home gardens have so far been left on
the sidelines of applied agronomic research
and extension efforts in Brazil and world-
wide. Future investments into home-garden
agroforestry would be highly efficient for
poverty reduction and nutritional security in
eastern Amazonia and throughout the tropics.
Investment is needed both in bidirectional
research (e.g., selection/introduction of pro-
ductive species, optimized management of
organic matter and nutrient cycling in joint
trials with local communities), and in subse-
quent extension efforts (formation of home
garden expertise in extension services and
training programs, specialized rural credit
lines, certification efforts etc.).
Acknowledgments This research was partially financed by
the Research Fund of Maranha
˜o State (FAPEMA), and two
research fellowships were financed by CAPES (Brazilian
Council of Higher Education). We would also like to thank
INCRA (Federal Colonization and Land Reform Agency),
Embrapa Eastern-Amazonia, MST (Movement of the Landless)
and the Agricultural Cooperative of Tome
´-Ac¸ u (CAMTA) for
their practical and infrastructure support.
Annex 1
See Table 4.
Agroforest Syst
Almeida AWB, Shiraishi J, Mesquita BA, Abreu Arau
´jo HF,
Martins CC, da Silva MHP (2001) Economia do babac¸u.
Interstate Movement of Babassu Nutcrackers (MIQCB),
˜o Luis
Altieri MA, Funes-Monzote FR, Petersen P (2011) Agroeco-
logically efficient agricultural systems for smallholder
farmers: contributions to food sovereignty. Agron Sustain
Dev 32:1–13. doi:10.1007/s13593-011-0065-6
Alvares CA, Stape JL, Sentelhas PC, de Moraes G, Leonardo J,
Sparovek G (2013) Ko
¨ppen’s climate classification map for
Brazil. Meteorol Zeitschrift 22:711–728. doi:10.1127/
Angiosperm Phylogeny Group III (2009) An update of the
Angiosperm Phylogeny Group classification for the orders
and families of flowering plants: APG III. Bot J Linn Soc
Atangana A, Khasa D, Chang S, Degrande A (2014) Definitions
and classification of agroforestry systems. In: Atangana A,
Khasa D, Chang S, Degrande A (eds) Tropical agro-
forestry. Springer, Dordrecht, pp 35–47
Berkes F, Colding J, Folke C (2003) Navigating social-eco-
logical systems, building resilience for complexity and
change. Cambridge University Press, New York. doi:10.
Brady NC (1996) Alternatives to slash-and-burn: a global
imperative. Agric Ecosyst Environ 58:3–11
Brown S (2002) Measuring, monitoring, and verification of
carbon benefits for forest-based projects. Philos Trans A
Math Phys Eng Sci 360:1669–1683. doi:10.1098/rsta.2002.
Cabell JF, Oelofse M (2012) An indicator framework for
assessing agroecosystem resilience. Ecol Soc 17:1–18.
Cannell MGR, Noordwijk M, Ong CK (1996) The central
agroforestry hypothesis: the trees must acquire resources
that the crop would not otherwise acquire. Agrofor Syst
Cardinale BJ, Palmer MA, Collins SL (2002) Species diversity
enhances ecosystem functioning through interspecific
facilitation. Nature 415:426–429. doi:10.1038/415426a
Clough Y, Barkmann J, Juhrbandt J et al (2011) Combining high
biodiversity with high yields in tropical agroforests. Proc
Natl Acad Sci USA 108:8311–8316. doi:10.1073/pnas.
Crawley MJ (2007) The R book. Wiley, New York
da Semedo RJCG, Barbosa RI (2007) Fruit trees in urban home
gardens of Boa Vista, Roraima, Brazilian Amazonia. Acta
Amaz 37:497–504. doi:10.1590/S0044-59672007000400003
das Chagas Oliveira F, Collado A
´C, Leite LFC (2013) Autonomy
and sustainability: an integrated analysis of the development
of new approaches to agrosystem management in family-
based farming in Carnaubais Territory, Piauı
´, Brazil. Agric
Syst 115:1–9. doi:10.1016/j.agsy.2012.09.005
Denich M, Vielhauer K, Kato MA, Block A, Kato OR, Sa
¨cke W, Vlek PLG (2004) Mechanized land preparation
in forest-based fallow systems: the experience from East-
ern Amazonia. Agrofor Syst 6:91–106. doi:10.1023/B:
Di Rienzo JA, Casanoves F, Balzarini MG, Gonzalez L, Tablada
M, Robledo CW, InfoStat versio
´n (2012) Grupo InfoStat,
FCA. Universidad Nacional de Co
´rdoba, Argentina. http://
Table 4 The ten most abundant agroforestry species: shares in total abundance (over all sites), uses and origins
Common name Scientific name Family Abundance
Main uses Origin
´Euterpe oleracea Mart. Arecaceae 25.9 AC Food Native
Banana Musa spp. Musaceae 20.2 AC Food, shade,
organic matter
Cocoa Theobroma cacao L. Malvaceae 16.4 CO Food, organic matter Native
Cupuassu Theobroma grandiflorum (Wild.
ex Spreng.) K. Schum.
11.6 AC Food, organic matter Native
Black pepper Piper nigrum L. Piperaceae 4.9 CO Food Exotic
Bacuri Platonia insignis Mart Clusiaceae 2.3 AC Food Native
Jaca Artocarpus integrifolia L. Moraceae 1.4 AC Food timber Exotic
Cecropia Cecropia sp. Urticaceae 1.3 NP Food, fauna
shade medicinal
Pati Syagrus cocoides Mart. Arecaceae 1.2 NP Food, fauna Native
Mango Mangifera indica L. Anacardiaceae 1 AC Food Exotic
Others 13.7 – –
CO Commercial, AC mainly auto-consumption, NP non-productive (providing ecosystem services, juveniles or medicinal use)
We consider as ‘native’ species all species originating in Amazonia or the Cerrado forests, or introduced into the region before the
year 1500
Agroforest Syst
´az S, Cabido M (2001) Vive la diffe
´rence: plant functional
diversity matters to ecosystem processes. Trends Ecol Evol
16:646–655. doi:10.1016/S0169-5347(01)02283-2
Diemont SAW, Martin JF (2009) Lacandon Maya ecosystem
management: sustainable design for subsistence and envi-
ronmental restoration. Ecol Appl 19:254–266. doi:10.
Faye MD, Weber JC, Abasse TA, Boureima M, Larwanou M,
Bationo AB, Diallo BO, Sigue
´H, Dakouo JM, Samake
Sonogo Diaite
´D (2011) Farmers’ preferences for tree
functions and species in the West African Sahel. For Trees
Livelihoods 20:113–136. doi:10.1080/14728028.2011.
Galluzzi G, Eyzaguirre P, Negri V (2010) Home gardens:
neglected hotspots of agro-biodiversity and cultural
diversity. Biodivers Conserv 19:3635–3654. doi:10.1007/
´a-Barrios L, Ong CK (2004) Ecological interactions,
management lessons and design tools in tropical agro-
forestry systems. Agrofor Syst 61:221–236. doi:10.1023/
Godsey LD (2010) Economic budgeting for agroforestry prac-
tices. University of Missouri Center for Agroforestry,
Gomes GS (2010) Quintais agroflorestais no municı
´pio de Irati-
´, Brasil: agrobiodiversidade e sustentabilidade
ˆmica e ambiental. PhD thesis, Federal Univer-
sity of Parana
´, Curitiba, Brazil
Huss-Ashmore R, Goodman JL (1988) Seasonality of work,
weight, and body composition for women in highland
Lesotho. Res Pap Sci Archaeol 5:29–44
IBGE (Brazilian Institute of Geography and Statistics) (2010)
Diretoria de pesquisas, coordenac¸a
˜o de agropecua
˜o da extrac¸a
˜o vegetal e da silvicultura, v.25,
Accessed July 16 2013
Jose S (2009) Agroforestry for ecosystem services and envi-
ronmental benefits: an overview. Agrofor Syst 76:1–10.
Kato OR (2009) Projeto dende
ˆ: Sistemas agroflorestais na
agricultura familiar (The oil palm project: agroforestry
systems in smallholder agriculture). Embrapa Eastern
Amazonia, Bele
Koutika L-S, Nolte C, Yemefack M et al (2005) Leguminous
fallows improve soil quality in south-central Cameroon as
evidenced by the particulate organic matter status. Geo-
derma 125:343–354. doi:10.1016/j.geoderma.2004.09.009
Kumar BM (2006) Carbon sequestration potential of tropical
homegardens. In: Kumar BM, Nair PKR (eds) Tropical
homegardens: a time-tested example of sustainable agro-
forestry., Advances in agroforestry 3Springer, Dordrecht,
pp 185–204
Lawrence D, Radel C, Tully K, Schmook B, Schneider L (2010)
Untangling a decline in tropical forest resilience: con-
straints on the sustainability of shifting cultivation across
the globe. Biotropica 42:21–30. doi:10.1111/j.1744-7429.
Lin BB (2011) Resilience in agriculture through crop diversi-
fication: adaptive management for environmental change.
Bioscience 61:183–193. doi:10.1525/bio.2011.61.3.4
Loreau M (2000) Biodiversity and ecosystem functioning:
recent theoretical advances. Oikos 91:3–17. doi:10.1034/j.
Magurran AE (1988) Ecological diversity and its management.
Princeton University Press, Princeton
McHenry MP (2009) Agricultural bio-char production, renew-
able energy generation and farm carbon sequestration in
Western Australia: certainty, uncertainty and risk. Agric
Ecosyst Environ 129:1–7. doi:10.1016/j.agee.2008.08.006
MDA (Ministe
´rio do Desenvolvimento Agra
´rio) (2011)
´sticas do meio rural 2010–2011, 4th edn. Dieese, Sa
Metzger JP (2002) Landscape dynamics and equilibrium in
areas of slash-and-burn agriculture with short and long
fallow period (Bragantina region, NE Brazilian Amazon).
Landsc Ecol 17:419–431
Muchavisoy HM (2013) Carbon stocks in rainforests, secondary
regrowth and agroforestry systems of eastern Amazonia.
M.Sc. thesis, Maranha
˜o State University, Sa
˜o Luı
´s, Brazil
Nair PKR (1985) Classification of agroforestry systems. Agro-
for Syst 3:97–128. doi:10.1007/BF00122638
Nair PKR (1993) An introduction to agroforestry. Kluwer
Academic Publishers, Dordrecht
Nair PKR (2001) Do tropical homegardens elude science, or is it
the other way around? Agrofor Syst 53:239–245. doi:10.
Nair PKR (2013) Agroforestry: trees in support of sustainable
agriculture. In: Reference module in earth systems and
environmental sciences. Elsevier, Amsterdam. doi:10.
Padoch C, Pinedo-Vasquez M (2010) Saving slash-and-burn to
save biodiversity. Biotropica 42:550–552. doi:10.1111/j.
Pascual U (2005) Land use intensification potential in slash-and-
burn farming through improvements in technical effi-
ciency. Ecol Econ 52:497–511. doi:10.1016/j.ecolecon.
Peyre A, Guidal A, Wiersum KF, Bongers F (2006) Dynamics of
homegarden structure and function in Kerala, India. Agro-
for Syst 66:101–115. doi:10.1007/s10457-005-2919-x
Porro R (2005) Palms, Pastures, and Swidden Fields: the
grounded political ecology of agro-extractive/shifting-
cultivator peasants in Maranha
˜o, Brazil. Hum Ecol
33:17–56. doi:10.1007/s10745-005-1654-2
Porro R, Mesquita BA, Santos I (2004) Expansa
˜o e trajeto
´ria da
´ria na Amazo
ˆnia: vales dos rios Mearim e Pindare
˜o. Editora Universidade de Brası
Acre, Brası
´lia, p 183
Ratnadass A, Fernandes P, Avelino J, Habib R (2012) Plant
species diversity for sustainable management of crop pests
and diseases in agroecosystems: a review. Agron Sustain
Dev 32:273–303. doi:10.1007/s13593-011-0022-4
Richards PW (1996) The tropical rain forest, 2nd edn. Cam-
bridge University Press, London
Rosa LS, Vieira TA, Santos APA, Menezes AAS, Rodrigue
AF, Perote JRS, Lopez CVC (2009) Limites e oportu-
nidades para a adoc¸a
˜o de sistemas agroflorestais pelos
agricultores familiares da microrregia
˜o Bragantina, PA. In:
Porro R (ed) Alternativa agroflorestal na Amazo
ˆnia em
˜o. EMBRAPA, Brası
´lia, pp 645–670
Agroforest Syst
Sarmento CMB, Veiga JB, Rischkowsky B, Kato OR, Sieg-
mund-Schultze M (2010) Caracterizac¸a
˜o e avaliac¸a
pastagem do rebanho de agricultores familiares do nordeste
paraense. Acta Amaz 40:415–423. doi:10.1590/S0044-
Schleuning M, Fru
¨nd J, Garcı
´a D (2015) Predicting ecosystem
functions from biodiversity and mutualistic networks: an
extension of trait-based concepts to plant-animal interac-
tions. Ecography 38:380–392. doi:10.1111/ecog.00983
Schroth G, da Mota MSS (2014) Agroforestry: complex multi-
strata agriculture. In: Encyclopedia of agriculture and
food systems. 195–207. doi:10.1016/B978-0-444-52512-3.
Schroth G, Lehmann J, Rodrigues MRL, Barros E, Mace
ˆdo JLV
(2001) Plant-soil interactions in multistrata agroforestry in
the humid tropics. Agrofor Syst 53:85–102
Schulze ED, Mooney HA (1993) Biodiversity and ecosystem
function. Springer, Berlin
Shepherd GJ, Fitopac-shell IGP (2009) Fitopac user
´s manual.
Federal University of Campinas, Brazil
Sibelet MFN, Smektala G (1999) Guide me
´thodologique pour la
conduite d’une e
´tude en milieu rural. Cours de l’ENGREF,
Somarriba E, Cerda R, Orozco L et al (2013) Carbon stocks and
cocoa yields in agroforestry systems of Central America.
Agric Ecosyst Environ 173:46–57. doi:10.1016/j.agee.
Soto-Pinto L, Anzueto M, Mendoza J et al (2009) Carbon
sequestration through agroforestry in indigenous commu-
nities of Chiapas, Mexico. Agrofor Syst 78:39–51. doi:10.
Souza HN, Graaff J, Pulleman MM (2011) Strategies and eco-
nomics of farming systems with coffee in the Atlantic
Rainforest Biome. Agrofor Syst 84:227–242. doi:10.1007/
Spjøtvoll E, Stoline MR (1973) An extension of the t-method of
multiple comparison to include the cases with unequal
sample sizes. J Am Stat Assoc 68:975–978
StatSoft, Inc (2007) STATISTICA (Data Analysis Software
System) Version 8.0.
Steffan-Dewenter I, Kessler M, Barkmann J et al (2007)
Tradeoffs between income, biodiversity, and ecosystem
functioning during tropical rainforest conversion and
agroforestry intensification. Proc Natl Acad Sci USA
104:4973–4978. doi:10.1073/pnas.0608409104
Styger E, Rakodondramasy HM, Pfeffer MJ, Fernandes ECM,
Bates DM (2007) Influence of slash-and-burn farming
practices on fallow succession and land degradation in the
rainforest region of Madagascar. Agric Ecosyst Environ
119:257–269. doi:10.1016/j.agee.2006.07.012
Vallejo M, Casas A, Perez-Negron E, Moreno-Calles A, Her-
nandez-Ordonez O, Tellez O, Davila P (2015) Agroforestry
systems of the lowland alluvial valleys of the Tehuacan-
Cuicatlan Biosphere Reserve: an evaluation of their bio-
cultural capacity. J Ethnobiol Ethnomed 11:8. doi:10.1186/
Veenhoven R (1994) El estudio de la satisfaccio
´n con la vida.
Psychosoc Interv 3:87–116
Veenhoven R (2007) Healthy happiness: effects of happiness on
physical health and the consequences for preventive health
care. J Happiness Stud 9:449–469. doi:10.1007/s10902-
Vieira TA, Rosa SRL, Santos MMLS (2012) Agrobiodiversi-
dade de quintais agroflorestais no municı
´pio de Bonito,
Estado do Para
´. Rev Cie
ˆncias Agra
´rias (Amazonian J Agric
Environ Sci) 55:159–166
Weiher E, Keddy P (2001) Ecological assembly rules: per-
spectives, advances and retreats. Cambridge University
Press, Cambridge
Yamada M, Gholz HL (2002) An evaluation of agroforestry
systems land-use change on soil nutrient dynamics in
Amazonia. Agrofor Syst 55:81–87. doi:10.1023/A:
Zomer RJ, Trabucco A, Coe R, Place F (2009) Trees on farm: an
analysis of global extent and geographical patterns of agro-
forestry. ICRAF Working Paper 89, World Agroforestry
Centre, Nairobi, Kenya. http://www.worldagroforestry.
Agroforest Syst
... Os Sistemas Agroflorestais ou simplesmente SAF, na Amazônia, são práticas desenvolvidas desde o início da agricultura, em que práticas indígenas tiveram grande contribuição na transferência de conhecimentos e adaptação dos sistemas hoje conhecidos (Miller & Nair, 2006 Nos SAF, produtos agrícolas são cultivados juntamente a essências florestais frutíferas e/ou madeiráveis, a depender do tipo e arranjo que se adota; em muitos casos, utiliza-se do mesmo espaço e tempo, às vezes com a presença de animais domésticos (Nair 1985;Paludo & Costabeber, 2012) que podem prestar serviços ecossistêmicos (Vasconcellos & Beltrão, 2018) capaz de garantir a sustentabilidade socioeconômica, agroecológico e/ou ambiental (Cardozo et al., 2015;Schembergue, Cunha, Carlos, Pires, & Faria, 2017;Viswanath & Lubina, 2017). ...
... A classificação dos sistemas dependerá da finalidade a que os SAF se destinam (Nair, 1987) e podem influenciar diretamente na definição do plano de ação de implantação do sistema de produção (Cardozo et al., 2015). Por exemplo, o nível socioeconômico, escala de produção e o nível de gerenciamento do sistema podem ser utilizados como critérios atribuídos aos sistemas em: comerciais; intermediários ou de subsistência (Nair, 1985;1987). ...
... São operações puramente comerciais, geralmente produzidas por empresa florestal. Nair (1985;1987); Barros et al. (2009);Cardozo et al. (2015). ...
Embora existam estudos que apontem a importância de modelos com alternativas de diversificação da produção nas propriedades rurais, diversos fatores influenciam sua implementação e sustentabilidade, principalmente, quando se trata de propriedade da agricultura familiar. A presente pesquisa, objetivou avaliar a contribuição do Projeto Rural Sustentável (PRS), que incentivou a Recuperação de Áreas Degradadas (RAD) nas pequenas e médias propriedades por meio de fomento e assistência técnica a produtores da agricultura familiar, com adoção de tecnologias, principalmente com base nos Sistemas Agroflorestais (SAF). Os SAF assumem papel social, econômico e ambiental estratégico no grupo estudado, além das características, fatores técnicos/biofísicos e políticas públicas do PRS que favoreceram a sua implantação. Para tanto, as propriedades rurais dependem primeiramente de uma fonte de financiamento, campanhas de educação e conscientização para que os sistemas se apresentam como alternativas na RAD em propriedades da Agricultura Familiar.
... Agroforestry has been proposed a sustainable alternative to slash-and-burn shifting cultivation in the tropics. The core principle of agroforestry systems (AFS) lies in combining trees with crops, and/ or animals in the same plot of land (a multistrata system) (Atangana et al., 2014) to mimic plant succession in the spontaneous forest (Cezar et al., 2015;Young, 2017), while including crop production (Cardozo et al., 2015. When appropriately managed, agroforestry practices improve the topsoil physico-chemical properties by increasing phosphorus and potassium contents (Pinho et al., 2012), maintain soil organic matter content (Leite et al., 2014), and promote nutrient cycling via nutrient pumping and safety net mechanisms (Seneviratne et al., 2006), which all strictly depend on ecosystem services delivered by the soil microbes (Wagg et al., 2014). ...
... Site selection and classification was based on the work of Cardozo et al. (2015) and Leite et al. (2016), as follows: ...
... The subplots and transects were sampled as above. Further details about the sampling scheme are presented in previous studies (Cardozo et al., 2015;Leite et al., 2016) and can be found in Figure S1. ...
An alarming and increasing deforestation rate threatens Amazon tropical ecosystems and subsequent degradation due to frequent fires. Agroforestry systems (AFS) may offer a sustainable alternative, reportedly mimicking the plant‐soil interactions of the natural mature forest. However, the role of microbial community in tropical AFS remains largely unknown. This knowledge is crucial for evaluating the sustainability of AFS and practices given the key role of microbes in the aboveground‐belowground interactions. The current study, by comparing different AFS and successions of secondary and mature forests, showed that AFS fostered distinct groups of bacterial community, diverging from the mature forests, likely a result of management practices while secondary forests converged to the same soil microbiome found in the mature forest, by favoring the same groups of fungi. Model simulations reveal that AFS would require profound changes in aboveground biomass and in soil factors to reach the same microbiome found in mature forests. In summary AFS practices did not result in ecosystems mimicking natural forest plant‐soil interactions but rather reshaped the ecosystem to a completely different relation between aboveground biomass, soil abiotic properties, and the soil microbiome.
... Het is reeds bekend dat vooral volwassen exemplaren van de pinapalm in boslandbouw systemen droge tijden kunnen tolereren (De Sousa et al., 2021). Onder andere in het oosten van de Amazone worden dergelijke producten geteeld in verschillende boslandbouwsystemen (Cardozo et al., 2015). Secundaire bossen worden daar soms verrijkt met banaan (Musa spp.), podosiri, en geelhart (Platonia insignis). ...
... Boslandbouw aan huis is het efficiëntste systeem en levert netto per hectare het hoogste inkomen. De kosten bleken het hoogst te zijn bij de zwerflandbouw, maar per hectare zijn weilanden het minste productief en bevatten de minste biodiversiteit (Cardozo et al., 2015). Verder blijkt de productie van podosiri sterk afhankelijk te zijn van bestuiving door verschillende insecten (bijen, kevers, torren, vliegen, wespen) die elkaar aanvullen of compenseren. ...
Technical Report
Full-text available
This report is an outcome of a land-use planning project. The villages included in this pilot project have been affected by several flooding events over the past decade. Partially because of climate change, these villages will be exposed to serious flooding more often. There are several ways in which the risk and the effects can be mitigated. This document briefly describes the historical, social and natural context. Subsequently, it broadly indicates which measures are the most appropriate according to the acquired information and which follow-up steps will be necessary to achieve a good implementation. This report serves to support managers and other involved stakeholders in conducting sustainable landscape management and land-use planning, in order to control the problem of flooding and reduce the risks. Some of the insights are also applicable to other climate- and environment-related problems, and also to other areas.
... Para Gómez Cardozo et al. (2015) los SAFs pueden ser utilizados para restaurar ecosistemas (por la diversidad que alberga dentro de sí) y que pueden ser mucho más sostenibles en el tiempo por su capacidad de resiliencia y estabilidad que le da al agroecosistema (Nair et al., 2008). ...
Full-text available
En el presente estudio exploratorio efectuado en el distrito de Inambari del departamento de Madre de Dios en Perú, se evaluó la población de macro artrópodos en 20 parcelas con diferentes tipos de sistemas agroforestales y con diferentes edades, los cuales fueron clasificadas en tres grupos por edad y por tipo de sistema agroforestal. En cada parcela se colectaron los macro artropodos del suelo de seis monolitos de suelo que tuvieron las siguientes dimensiones: 25 cm x 25 cm x 30 cm, preservándose los individuos en alcohol al 95%, para posteriormente identificarlas y reportarlas en cantidad de individuos/m2. Se colectaron 3046 individuos, encontrándose que las órdenes más importantes a nivel de toda la zona de estudio fueron: Haplotaxida, Coleoptera, Hymenoptera, Dictyoptera e Isoptera con más de 100 individuos por metro cuadrado. Las lombrices de tierra fueron las especies más dominantes en los sistemas agroforestales 1 y 3, mientras que en el sistema agroforestal 3 dominaron los Isopteros. En los sistemas agroforestales más antiguos dominaron los Isópteros mientras que en los sistemas más jóvenes los Haplotaxidos
... The community forestry program (HKm) raises new hope for the community around the forest area for a better future [1][2][3]. The community has the opportunity to plant seasonal crops among forest so that it will increase the income of farmers' households [4][5][6]. With the existence of HKm, it is hoped that the welfare of local communities can increase through the optimal, fair and sustainable use of forest resources while maintaining the sustainability of forest and environmental functions [7][8][9][10][11]. ...
This study aims to: (1) identify the types of existing MPTS plants in the HKm area, and (2) determine the featured MPTS plants in HKm areas. This research was carried out in the HKm area of North Batukliang District, Central Lombok Regency. Data and information collection was carried out through: interviews with 30 samples, in-depth interviews, document tracing, and observations. Data analysis was performed with Process Hierarchy Analysis (AHP). The results of the study are: (1) The dominant types of MPTS planted by the community in the HKm area of North Batukliang, Central Lombok are: durian, banana, coffee, cocoa, jackfruit, rambutan, avocado, mangosteen; and (2) The featured MPTS plants that can be developed in the HKm North Batukliang area of Central Lombok from in order as follows: durian, banana, coffee, cocoa, avocado, jackfruit, rambutan, and mangosteen. It is recommended that the government or other parties support the provision of quality plant seeds, especially for the five priority MPTS plant types.
... They are structurally and functionally more intricate than silvicultural and agricultural monoculture systems (Rathore et al., 2021). AFSs allow the improvement of biodiversity (Bradley et al., 2012, Cardozo et al., 2015 nutrient cycling, and soil health build-up (Nair et al., 2010), without compromising farm productivity . Nevertheless, tree-crop combinations can escalate the proficiency of land use Yadav et al., 2021a) and boost economic and ecological advantages (David et al., 2013). ...
Full-text available
Environmental crises, land degradation, and frequent crop failure threaten the livelihoods of millions of the populace in the semi-arid agroecosystems. Therefore, different combinations of annual crops with fruit trees were assessed to restore the soil carbon, and enhance farm productivity and profitability in a semi-arid climate. The study hypothesized that the integration of fruit trees with seasonal crops may enhance farm productivity, economic returns, and environmental sustainability. Integration of phalsa (Grewia asiatica) with mung bean (Vigna radiata)-potato (Solanum tuberosum) system recorded the highest system productivity (25.9 Mg/ha) followed by phalsa-cowpea (Vigna unguiculata)-mustard (Brassica juncea) systems (21.2 Mg/ha). However, Karonda (Carissa sp.)-mung bean-potato system recorded maximum net return (3529.1 US$/ha), and water use efficiency (33.0 kg/ha-mm). Concerning the benefit-cost (B:C) ratio, among the agroforestry systems, the karonda+cowpea-mustard system registered a maximum BC ratio (3.85). However, SOC density remained higher (9.10 Mg/ha) under the phalsa + cowpea-mustard and Moringa + mung bean-potato system (9.16 Mg/ha) over other systems. Similarly, phalsa-mung bean-potato system had the highest C sustainability index (27.6), carbon sequestration potential (0.6-0.67 Mg/ha/year), and water use efficiency (33.0 kg/ha-mm). Hence, the study suggested that the integration of fruit trees with short-duration leguminous and oilseeds offer a myriad of benefits and an efficient system for restoring soil C without compromising the food and livelihood security of the rural populace in semiarid regions.
... In regularly-spaced agroforestry plantations, the sampling scheme was adapted to three quadrangular plots of 625 m 2 and six subplots of 15m 2 . Further details about the sampling scheme are presented in Leite et al. (2016), and Cardozo et al. (2015). ...
Full-text available
The potential of agroforestry systems (AFS) for atmospheric carbon sequestration in degraded tropical lands is of key interest for climate change and rural development policies. This study evaluated aboveground and soil (0-20 cm) carbon stocks of AFS, secondary forests (SF), conserved and logged mature forests, on 88 sites in the eastern Brazilian Amazon. Tree carbon stock was higher in young (<10 years) and advanced (>30 years) AFS (10.2 ± 2.0 and 47.2 ± 8.1 Mg ha-1, respectively) when compared to the same age SF (5.8 ± 2.5 and 26.5 ± 19.5 Mg ha-1). However, aboveground and total carbon stocks were statistically similar within the same age categories of AFS and SF, because shrub pool were higher in SF. Conserved mature forests had the highest carbon stocks (190.2 ± 11.0 Mg ha-1), and carbon stocks in logged mature forests (119.4 ± 5.1 Mg ha-1) were similar to the advanced stages of AFS (108.6 ± 7.5 Mg ha-1). Litter and soil organic carbon (SOC) did not differ significantly between land-use systems nor along succession. At 30 years, aboveground carbon recovery was 46% (±16) in AFS and 35% (±21) in SF. Vegetation structural diversity (measured by DBH and height variation) was a good predictor of aboveground carbon stocks. Our results show the potential of AFS for carbon recovery, especially in the tree pool at late stages of development. Structurally more complex AFS provide an alternative to recover degraded lands and to develop synergies between climate change mitigation, adaptation, and goods production in Amazon
... The income:cost ratio is an indicator of economic efficiency, defined as the ratio of the income generated after the SWC practices implementation to the practices' investment of each scenario. It has been used in many studies [52] and is similar to the concept of the input:output ratio or return on investment (ROI), which means the ratio of project input capital to output capital, that is, how many units of capital can be produced by investing one unit of capital. The greater the N, the better the economic efficiency. ...
Full-text available
Soil and water conservation (SWC) practices on agricultural watersheds have been the most effective practices for preventing soil erosion for several decades. The ecosystem services (ES) protected or enhanced by SWC practices include the comprehensive effects of protecting and conserving water sources, protecting and improving soil, carbon fixation, increasing agricultural production, and so on. Due to the lack of ES evaluation indicators and unified calculation methods in line with regional characteristics, this study proposes a framework of scenario analysis by using ES mapping, ES scoring, and economic analysis technology for ES and economic-benefit trade-offs under different scenarios. The study area was the Xiaoyang catchment located in Ningdu County, Jiangxi Province, which is a typically hilly red-soil region of southern China. From the results of scenario analysis, an obvious phenomenon is that some SWC practices can affect the value of some ES indicators, while some have no clear trend. By computing the ES scores for the four scenarios, the ranking was S3 (balanced), S1 (conservation), S2 (economic), and S0 (baseline). S3 ranks second in net income (with CNY 4.73 million), preceded only by S2 (CNY 6.36 million). Based on the above rankings, S3 is the relatively optimal scenario in this study. The contributions of this study are the method innovation with the localization or customized selection of ES indicators, and scenario analysis with ES scores and economic-benefit trade-offs in different scenarios.
... Management of timber and non-timber forest product from restored areas represent another opportunity for sustainable rural development, as well as the Payment for Environmental Services. Restoration through agroforestry also has been demonstrated to improve local livelihoods in the tropics, with income, sustainable production, and food security (Cardozo et al., 2015;Padovan et al., 2022). ...
Forest restoration has attracted the attention of different organizations, investors, and donors with the launch of the UN Decade for Ecosystems Restoration (2021–2030), along with climate and biodiversity commitments. Restoration can address many of mankind's challenges, such as biodiversity loss, climate change, water security, and poverty. In the Brazilian Amazon, the ~28 million inhabitants are among the most vulnerable of the country, and this has only worsened during the COVID-19 pandemic. Meanwhile, millions of hectares are suitable for forest restoration. The growing demand for large-scale forest restoration projects have been prioritizing biophysical objectives (e.g., number of trees, hectares of land, and carbon) while it should be prioritizing the local people's well-being and a fair transition to a sustainable economy based on forest services' recovery. Nonetheless, many challenges need to be overcome to realize this potential. Amazonian states need to control illegality, enforce the existing policies and promote innovative ones to halt deforestation and enable large-scale restoration. Better governance and social engagement are urgently needed but depend upon, recognition of indigenous peoples and local communities' rights, needs, and knowledge. Forest restoration represents an opportunity for the emergence of a more inclusive development paradigm, much needed in the Amazon region, especially in the post COVID-19 world.
... Studies have shown that out of the diversity of species produced, such as those highlighted by Cardozo et al. (2015), the cocoa (T. security with regard to market changes, in addition to allowing them to sell their products directly at fairs, which, according to Roces and Montiel (2010), strengthens their contact with consumers who value agroecological production. ...
Full-text available
Agroecological practices are alternatives for valuing and conserving the biodiversity of the Amazon as well as for integrating family farmers who survive from the exploitation of natural resources in the face of the expansion of large agricultural projects that have advanced in the region. In this respect, the objective of this study was to evaluate the scientific production related to agroecology in the Brazilian Amazon, in order to identify the main research gaps and set the foundations for new studies that will strengthen the debate on and policies to encourage sustainable agriculture. Our methodology was based on a systematic review of the literature using four academic research databases. The studies were published from 2000 to 2019. Through our database research, we obtained 36 articles that focused on agroecology in the Brazilian Amazon. Our results showed a certain level of diversity in terms of the geographic distribution of the municipalities included in the studies analyzed. Additionally, we detected diversity in terms of the dimensions addressed. For instance, reference was made to sustainable production systems based on agroecological principles, with emphasis on agroforestry systems. Additionally, socioeconomic and cultural aspects, such as the valuation of traditional knowledge and the importance of women in rural areas, were analyzed. Thus, the realization of in-depth studies that will analyze the process of agroecological transition with the use of different techniques, such as the statistical treatment of data and geoprocessing, can qualify agricultural production based on the agroecological practices in the Amazon.
Full-text available
The cocoa tree (Theobroma cacao L.) is cultivated typically in agroforestry systems in close association with a rich list of tree species and other useful plants on the same plot. Cocoa based agroforestry systems are credited for stocking significant amounts of carbon and hence have the potential to mitigate climate change. Since cocoa yields decrease non-linearly with increasing shade, a need is to design optimal cocoa agroforestry systems with high yields and high carbon stocks. We estimated the carbon stocked in a network of 229 permanent sample plots in cacao-based agroforestry systems and natural forests in five Central American countries. Carbon stocks were fractioned by both system compartments (aboveground, roots, soil, litter, dead wood – fine and coarse, and total) and tree use/form (cocoa, timber, fruit, bananas, shade and ornamentals, and palms). Cocoa plantations were assigned to a five-class typology and tested for independence with growing region using contingency analysis. Most Central American cocoa plantations had mixed or productive shade canopies. Only 4% of cocoa plantations were full sun or rustic (cocoa under thinned natural forest). Cocoa tree density was low (548 ± 192 trees ha−1). Total carbon (soil + biomass + dead biomass) was 117 ± 47 Mg ha−1, with 51 Mg ha−1 in the soil and 49 Mg ha−1 (42% of total carbon) in aboveground biomass (cocoa and canopy trees). Cocoa trees accumulated 9 Mg C ha−1 (18% of carbon in aboveground biomass). Timber and fruit trees stored 65% of aboveground carbon. The annual rate of accumulation of carbon in aboveground biomass ranged between 1.3 and 2.6 Mg C ha−1 y−1. Trade-offs between carbon levels and yields were explored qualitatively using functional relationships documented in the scientific and technical literature, and expert knowledge. We argue that it is possible to design cocoa-based AFS with good yields (cocoa and shade canopy) and high carbon stock levels. The botanical composition of the shade canopy provides a large set of morphological and functional traits that can be used to optimize shade canopy design. Our results offer Central American cocoa producers a rigorous estimate of carbon stocks in their cocoa plantations. This knowledge may help them to certify and sell their cocoa, timber, fruits and other goods to niche markets with good prices. Our results will also assist governments and the private sector in (i) designing better legal, institutional and policy frameworks, local and national, promoting an agriculture with trees and (ii) contributing to the development of the national monitoring, reporting and verification systems required by the international community to access funding and payment for ecosystem services.
Full-text available
Agroforestry, the inclusion of woody perennials within farming systems, has been both a traditional landuse approach developed by subsistence farmers throughout the tropics, and a livelihood option promoted by landuse managers and international development efforts. Agroforestry systems range from subsistence livestock and pastoral systems to home gardens, alley intercropping, and biomass plantations with a wide diversity of biophysical conditions and socio-ecological characteristics. The extent of its practice has never been quantified leading to widely varied estimates about its importance. This paper is the first attempt to quantify the extent of agroforestry at the global level. A geospatial analysis of remote sensing derived global datasets investigated the correspondence and relationship of tree cover, population density and climatic conditions within agricultural land at 1 km resolution. Among the key results are that agroforestry is a significant feature of agriculture in all regions, that its extent varies significantly across different regions (e.g. more significant in Central America and less in East Asia), that tree cover is strongly positively related to humidity, and that there are mixed relationships between tree cover and population density depending on the region. This first analysis suggests that patterns of tree cover are influenced by a range of factors we were not able to examine at the global scale and a number of follow up analyses are recommended.
Several definitions have been proposed to agroforestry, of which the most commonly used are those of Lundgreen and Raintree (Agricultural research for development: potentials and challenges in Asia, 1982, pp 37–49) and Leakey (Agroforest Today 8:1, 1996). Agroforestry is any land-use system, practice or technology, where woody perennials are integrated with agricultural crops and/or animals in the same land management unit, in some form of spatial arrangement or temporal sequence. Agroforestry is also a dynamic and ecologically -based natural resource management system. Agroforestry refers to the deliberate introduction or retention of trees on farms to increase, diversify, and sustain production for increased social, economic, and environmental benefits. Agroforestry system classification can be based on vegetation structure, function of woody perennials in the system, levels of management input, and environmental conditions and ecological suitability of the system. Agroforestry practices rather than systems are also used as the unit of an ecologically -based classification that is rooted in the role of trees in agricultural landscape.
Agroforestry refers to the growing of trees and crops, sometime with animals, in interacting combinations on the same unit of land. Although such practices were prevalent for many centuries in different parts of the world, scientific efforts to understand and utilize their sustainability attributes and production benefits started only in the late 1970s. Today, integration and utilization of trees and shrubs that provide multiple products and services in land-use systems offer sustainable solutions to several serious land management and environmental issues such as food security, environmental protection, and climate-change mitigation, in both developing countries and the industrialized world.
The biota of the earth is being altered at an unprecedented rate. We are witnessing wholesale exchanges of organisms among geographic areas that were once totally biologically isolated. We are seeing massive changes in landscape use that are creating even more abundant succes­ sional patches, reductions in population sizes, and in the worst cases, losses of species. There are many reasons for concern about these trends. One is that we unfortunately do not know in detail the conse­ quences of these massive alterations in terms of how the biosphere as a whole operates or even, for that matter, the functioning of localized ecosystems. We do know that the biosphere interacts strongly with the atmospheric composition, contributing to potential climate change. We also know that changes in vegetative cover greatly influence the hydrology and biochemistry ofa site or region. Our knowledge is weak in important details, however. How are the many services that ecosystems provide to humanity altered by modifications of ecosystem composition? Stated in another way, what is the role of individual species in ecosystem function? We are observing the selective as well as wholesale alteration in the composition of ecosystems. Do these alterations matter in respect to how ecosystems operate and provide services? This book represents the initial probing of this central ques­ tion. It will be followed by other volumes in this series examining in depth the functional role of biodiversity in various ecosystems of the world.