Full Terms & Conditions of access and use can be found at
ISSN: 2044-2041 (Print) 2044-205X (Online) Journal homepage: https://www.tandfonline.com/loi/tinw20
Effects of agricultural landscapes and land uses
in highly biodiverse tropical streams of the
Andrés Morabowen, Verónica Crespo-Pérez & Blanca Ríos-Touma
To cite this article: Andrés Morabowen, Verónica Crespo-Pérez & Blanca Ríos-Touma (2019):
Effects of agricultural landscapes and land uses in highly biodiverse tropical streams of the
Ecuadorian Choco, Inland Waters, DOI: 10.1080/20442041.2018.1527597
To link to this article: https://doi.org/10.1080/20442041.2018.1527597
Published online: 24 May 2019.
Submit your article to this journal
Article views: 6
View Crossmark data
Eﬀects of agricultural landscapes and land uses in highly biodiverse tropical
streams of the Ecuadorian Choco
Verónica Crespo-Pérez ,
and Blanca Ríos-Touma
Centro de Investigación de la Biodiversidad y el Cambio Climático–BioCamb, Ingeniería en Biodiversidad y Recursos Genéticos, Facultad de
Ciencias de Medio Ambiente, Universidad Tecnológica Indoamérica, Quito, Ecuador;
Laboratorio de Entomología, Museo QCAZ-I, Pontiﬁcia
Universidad Católica del Ecuador, Escuela de Ciencias Biológicas, Quito, Ecuador;
Facultad de Ingenierías y Ciencias Aplicadas, Ingeniería
Ambiental, Grupo de Investigación en Biodiversidad, Medio Ambiente y Salud-BIOMAS-, Universidad de las Américas, Quito, Ecuador
The ecosystem-level consequences of agricultural land use in Neotropical forests have not been
fully studied. In areas like the Choco-Darien, conﬂict exists between the conservation of highly
diverse ecosystems and the use of economically important production areas. Current agricultural
practices involve complete deforestation, with consequent multiple eﬀects on stream
ecosystems. To address the issue of land use change in tropical rivers of Ecuador, we studied
streams draining 3 diﬀerent land use types in the Mashpi River drainage (Ecuadorian Choco): (1)
pristine montane cloud forest, (2) organic farms that included forest patches, and (3) palmito
(Bactris gasipaes) production land with extensive use of the insecticide endosulfan and the
herbicide glyphosate. We sampled macroinvertebrates (quantitative and qualitative samples) and
periphyton, and measured environmental variables during dry and wet seasons, and found a
direct relationship between the decline of certain macroinvertebrate groups (e.g., Anacroneuria,
Hyallela) and the type of land use. Furthermore, we found that species loss in streams draining
organic farms was negligible. Species richness of macroinvertebrates was considerably lower in
palmito monoculture farmlands than in the other 2 types of land use. Stream communities of
the Mashpi drainage area have been transformed by human agricultural disturbances, and
urgent changes to land management practices are necessary.
Received 16 December 2017
Accepted 7 September 2018
palmito (Bactris gasipaes)
The Ecuadorian Choco has a dense network of streams
draining the northern Andes to the Paciﬁc Ocean that
sustains a vast number of diﬀerent organisms. The
Choco is classiﬁed as a biodiversity hotspot (Myers
et al. 2000) but is threatened by human disturbance
(Chen et al. 2004, Townsend et al. 2008, Sundbäck
et al. 2010). Several studies have shown that diﬀerent
types of anthropogenic disturbances produce diﬀerent
eﬀects on the biodiversity supported by rivers in tropical
regions (Aratrakorn et al. 2006, Fitzherbert et al. 2008,
Wagner et al. 2010). The main anthropogenic impacts
to rivers and streams in this region are extensive defores-
tation for logging, conversion of forest to oil palm pro-
duction accompanied by chemical river pollution,
overexploitation of ﬁsheries, and channel modiﬁcations
such as dams (Dodds 2002, Liess and Von der Ohe
In recent decades, clearing forests for agriculture is the
most common disturbance impacting rivers in the Neo-
tropical region (Laurance 1999, Sierra 2013, González-
Jaramillo et al. 2016) and continues to destroy millions
of hectares annually (Achard et al.2002, Lorion and
Kennedy 2009). This activity has been especially preva-
lent in recent decades in impoverished tropical regions
like northwestern Ecuador (Laurance 1999, Ministerio
del Ambiente 2012, Sierra 2013, González-Jaramillo
et al. 2016). Deforestation can degrade stream habitats
by inﬂuencing runoﬀregimes and evapotranspiration
patterns (Iwata et al. 2003) as well as by changing
water temperature regimes and altering the abundance
and diversity of food resources (Henry et al. 1994, Ben-
stead et al. 2003, Bojsen and Jacobsen. 2003, Benstead
and Pringle 2004). These processes can signiﬁcantly
change benthic community structure and decrease mac-
roinvertebrate diversity (Benstead et al. 2003,Bojsen and
Jacobsen 2003, Iwata et al. 2003, Dudgeon et al. 2006,
Wantzen and Wagner 2006).
Evidence increasingly indicates that deforestation
caused by conversion to agriculture destructively
impacts benthic stream communities, but the eﬀects
on tropical stream ecosystems require further research
(Lorion and Kennedy 2009). Evidence to support the
conservation of intact riparian forests to maintain
© 2019 International Society of Limnology (SIL)
CONTACT Blanca Ríos-Touma firstname.lastname@example.org
diversity in tropical stream communities is limited
(Dudgeon 2000,Iwataetal.2003, Tomanova et al.
2006). The aim of this study was to increase knowledge
of the eﬀects of land use changes on aquatic communi-
ties in highly biodiverse Neotropical forests. We studied
the communities of the Mashpi watershed in a biodiver-
sity hotspot, the Ecuadorian Choco, a region with a
recent history of deforestation for the cultivation of
hearts of palm (Bactris gasipaes)monoculture.The
Mashpi drainage is a mosaic of protected forests,
organic agroforestry farms, and palmito (Bactris gasi-
paes) monoculture, creating an ideal environment to
test the eﬀects of diﬀerent land uses on aquatic biodi-
versity. This heterogeneity allowed us to determine
diﬀerences between the benthic communities in
streams draining these diﬀerent land uses. Our research
questions were: (1) Are there diﬀerences in the benthic
communities of streams draining diﬀerent types of land
use? (2) Which environmental factors can explain the
diﬀerences among macroinvertebrate assemblages
from streams draining diﬀerent land uses? and (3) Do
agroforestry farms produce fewer impacts on stream
biota than monoculture farms?
This contribution presents the ﬁrst evidence of com-
munity changes of aquatic benthic invertebrates in the
Choco hotspot region in the Mashpi drainage area due
to pollution and deforestation caused by monoculture
agricultural expansion. We also document how diﬀerent
agricultural practices, monoculture versus organic farm-
ing, can have diﬀerent levels of impact on benthic com-
munities of the streams draining them.
Materials and methods
The Mashpi River basin is located in the northwestern
Andean ridge of Ecuador (Fig. 1), which eventually
becomes part of the Esmeraldas River basin that drains
to the Paciﬁc Ocean. We studied 9 streams, all ﬁrst-
order tributaries of the Mashpi River, ∼500 m a.s.l.
The sites were divided into 3 categories based on land
use patterns within their drainage areas. The ﬁrst cate-
gory included 3 streams located within the intact
Mashpi forest, composed mostly of pristine montane
forests. The second category included 3 streams drain-
ing organic agroforestry farms that did not use pesti-
cides and had signiﬁcant riparian vegetation. The last
category included 3 streams that drained monoculture
palmito farmland with no natural vegetation cover
and little to no riparian vegetation. All streams experi-
enced the same macro-environmental conditions
because they are in close proximity (<2 km apart) and
at similar elevation.
The 9 streams were sampled 4 times: December 2014
(wet season), February 2015 (wet season), May 2015
(dry season), and March 2015 (dry season). Each stream
was divided into 3 stations along a 50 m length reach.
Sampling stations in each stream reach were located as
follows: one downstream (0 m), one in the middle
(25 m), and one upstream (50 m). In each station, we
took a qualitative 1 min kick sample using a D-net
(Hauer and Lamberti 2006), attempting to cover all the
microhabitats of the stream. Additionally, in each stream
we collected 3 quantitative samples, one at each station,
in riﬄe habitats using a Surber sampler (500 cm
size 200 µm; Surber 1937). Most macroinvertebrates
were identiﬁed to genus, except for Collembola, Haplo-
taxida, Isopoda, Trombidiformes, Unionoida, Basomma-
tophora, Gordioidea, and Tricladia, which were
identiﬁed only to order, and the Chironomidae, which
were identiﬁed to subfamily, using specialized keys to
Figure 1. Topography of the Mashpi River subbasin in Ecuador, with locations of the 9 studied streams, monoculture plantations of the
zone, and the protected rain forest area.
2A. MORABOWEN ET AL.
North and South American macroinvertebrates (Merritt
and Cummins 1996, Roldán 1988, Domínguez and Fer-
nández 2009, Hanson et al. 2010).
We used chlorophyll aconcentration as a measure of
periphyton biomass using the spectrophotometric
method (Steinman and Lamberti 1996, APHA 1998).
At each site, 3 cobbles were collected randomly at each
transect. A 4 cm
area of each cobble was scraped and
the removed material ﬁltered with a vacuum pump
using microﬁber glass ﬁlters, GF/F 47 mm diameter.
The ﬁlters were stored at −20 °C until 1 d before extrac-
tion and then placed in 96% ethanol at 4 °C for 24 h to
extract the chlorophyll. We centrifuged 15 mL of the
extract to sediment any impurity in the ﬁlter. We mea-
sured absorbance at 665 and 750 nm using a spectropho-
tometer (HACH DK3900, Loveland, CO, USA). To
correct for phaeophytin content, the extract was acidiﬁed
with 0.5 mL of 0.1N HCL, and absorbance was measured
again at 665 and 750 nm wavelengths.
At each sampling event we measured pH, dissolved oxy-
gen (mg/L), conductivity (µS/cm), temperature (°C), dis-
/s), and percentage of oxygen using YSI PRO
probes (Yellow Springs, OH, USA). For the discharge
measurements we used the salt dilution method (White
1978), in which a bucket with a known amount of dis-
solved salt (volume and conductivity) was added
upstream, and then conductivity was measured every
10 s at 15–25 m downstream. Mean current velocity
and discharge were calculated as the time elapsed for
half the salt to pass the stream reach divided by the
length of the reach. At each stream, 750 mL water sam-
ples were collected, frozen, and taken to the laboratory
for measurements of nitrate and phosphate concentra-
tions. We analysed phosphate concentration using the
soluble reactive phosphorus (SRP) method (Murphy
and Riley 1962) and nitrate concentration using the cad-
mium reduction method (Henriksen and Selmer-Olsen
1970) with a spectrophotometer (HACH DK3900, Love-
land CO, USA).
We characterized substrate using the pebble count
method by measuring the intermediate axis of 100 ran-
dom sediment particles in each stream habitat (Kondolf
and Li 1992). We measured benthic coarse particulate
organic matter (CPOM), the main energy source for
members of the basal trophic chain in forest streams
(Vannote et al. 1980, Cummins et al. 1989, Abelho and
Graça 1998), to account for potential diﬀerences in food
availability for macroinvertebrates. CPOM (>1 mm)
was collected from Surber samples after all macroinverte-
brates were removed. The material was dried at 90 °C for
24 h, weighed, and then combusted in a muﬄe furnace at
500 °C for 4 h and weighed again to obtain the ash-free
dry mass, calculated as the diﬀerence between the initial
and ﬁnal weight (dried CPOM minus combusted
CPOM; Steinman and Lamberti 1996).
We assessed the diﬀerences and similarities among and
between treatments using an analysis of similarities
(ANOSIM) in Primer v6 (Ivybridge, UK), a test widely
used to test spatial diﬀerences in community assemblage
(Chapman and Underwood 1999). This analysis was per-
formed with data from both the standardized D-net sam-
ples (relative abundance, without rare taxa) and Surber
samples (density, without rare taxa). The one-way ANO-
SIM is based on the statistical test R, which varies from
−1 to +1, with values closer to 1 representing the largest
diﬀerences between groups (Clarke and Warwick 2001).
To determine which macroinvertebrates were respon-
sible for diﬀerences found between and among stream
types, we performed a SIMPER analysis excluding com-
mon taxa (present in 95% of samples), allowing us to
pinpoint the taxa that characterized each stream cate-
gory. In addition, with this analysis we could detect
which groups were aﬀected or favored by the diﬀerent
types of land use associated with the stream drainage.
We performed SIMPER analyses with data from quanti-
tative (density) and qualitative (relative abundance) sam-
ples in Primer v6. (Ivybrige, UK).
To visualize how the community composition diﬀered
among streams from the same treatment and streams
from diﬀerent treatments, we performed a principal
coordinates analysis (PCO) and a cluster analysis with
a similarity proﬁle (SIMPROF) test to validate groups
(Clarke and Warwick 2001) for macroinvertebrate rela-
tive abundance (kick samples) and density (Surber sam-
ples), again excluding common taxa. The data matrix
was transformed using a square root transformation,
and the similarity matrix was calculated using the
Bray-Curtis similarity index (Clarke and Warwick
2001). Stress was calculated as a measure of the accuracy
of the similarity matrix; values <0.2 correspond to a rea-
sonable ﬁt (Clarke and Warwick 2001). The PCO was
performed using Primer v6 (Ivybridge, UK).
To test for signiﬁcant diﬀerences in the environmental
variables among treatments, we used a Kruskal-Wallis H
test, a nonmetric rank analysis of variance in STATIS-
TICA 8.0 (Weiß 2007). A principal component analysis
(PCA) using Primer v6 (Ivybridge UK) was then
INLAND WATERS 3
performed to spatially visualize diﬀerences in the physico-
chemical variables of streams within and between treat-
ments. All variables were normalized as a requirement
for the use of Euclidean distance as a measure of
To determine signiﬁcant variability in the communities
in diﬀerent categories of streams, we performed a full fac-
torial ANOVA in STATISTICA 8.0 (StatSoft Inc. 2007) to
account for the eﬀect of month of sampling and land use
on community metrics. Several community metrics were
used: richness (S), rareﬁed richness (SRar), abundance
(N), density, and the Shannon Wiener diversity index
(N1) in its exponential form to express the result as true
number of species (Jost 2006). These metrics were calcu-
lated using Primer v6 (Clarke and Gorley 2006).
Environmental changes among land uses
Water temperature (Table 1,Fig. 2a) was signiﬁcantly
higher in the streams draining human-inﬂuenced lands
(i.e., agroforestry and monoculture farmlands; Kruskal-
Wallis H (2 d.f., n= 9) = 7.26; p= 0.027), averaging 22.4
°C in agroforestry farmlands and 22 °C in monoculture
farmlands. The average water temperature of streams
crossing forest streams was 21.6 °C (Fig. 2a).
Stream pH (Table 1,Fig. 2b) also varied signiﬁcantly
between the studied streams (Kruskal-Wallis H (2 d.f.,
n= 9) = 6.16 p= 0.0459). Water was more acidic in
monoculture streams, with an average pH of 7.39, than
in agroforestry and forest streams, averaging 7.73 and
7.66, respectively (Table 1,Fig. 2b).
Diﬀerences in discharge between land uses were close
to signiﬁcant (Table 1,Fig. 2c) (Kruskal-Wallis H (2 d.f.,
n= 9) = 5.96; p= 0.0509). Agroforestry farmland streams
had higher and more variable discharge than the other 2
treatments, especially compared to monoculture streams,
where we found almost no variation in discharge
The amount of CPOM (Fig. 2d) measured from the
Surber samples varied dramatically between forest streams,
where we found the highest CPOM, and human-
inﬂuenced streams, where we found lower quantities of
CPOM, although diﬀerences were not signiﬁcant (Kruskal
Wallis H (d.f. 2, n= 9)= 5.42 p= 0.067). The average
amount of organic matter was almost the same in agrofor-
estry farmlands streams and monoculture streams (5.15 g)
but higher in forest-draining streams (8.41 g; Table 1,Fig. 2
d). Conductivity, substrate, oxygen, nitrate, and phosphate
showed no signiﬁcant diﬀerences among land uses, and the
averages were similar among all stream types.
The PCA (Fig. 3) of the environmental variables and
primary production showed that the ﬁrst 2 axes
explained 60% of the stream distribution on the plot.
Streams draining monoculture farmlands were the only
closely grouped streams and were negatively correlated
with discharge, temperature, and pH and positively
correlated with nitrate, phosphate, and chlorophyll a
concentrations. By contrast, the agroforestry farmland-
draining streams were positively correlated with temper-
ature and discharge. Finally, forest streams were posi-
tively correlated with oxygen and CPOM (Fig 3).
Primary production among land uses
No signiﬁcant diﬀerence was found in periphyton bio-
mass, estimated through chlorophyll aconcentration,
among land uses (p= 0.42). The highest average concen-
trations were found in streams draining monoculture
farmlands (mean [SD] = 0.18 [0.12] µg/cm
). The forest
streams had intermediate values (0.69 [0.06] µg/cm
and, streams draining agroforestry farms had the lowest
values (0.068 [0.02] µg/cm
Benthic communities across land uses
We found 23 305 invertebrate individuals at the 9 sites
during the 4 months of sampling; 10 837 were collected
with the kick D-net method and 12 468 with the Surber
sampling method. All specimens were classiﬁed in 140
taxa belonging to 18 diﬀerent orders, the majority of
which were Coleoptera, with 32 genera; others included
Table 1. Environmental variables measured across streams draining diﬀerent land uses in Mashpi River basin, Ecuador. CPOM = coarse
particulate organic matter; AFDM = ash-free dry mass.
Land use Site
Reference Boshungo 21.67 23.52 6.71 7.47 2.13 0.01 6.39 0.40 0.13
Chakra 21.67 48.92 7.31 7.78 3.37 0.02 6.84 0.40 0.16
MalTrib 21.57 43.63 7.45 7.74 3.10 0.01 12.02 0.43 0.14
Inés 22.67 36.90 7.12 7.80 3.02 0.08 4.95 0.30 0.12
Mashungo 22.52 33.10 6.49 7.58 2.57 0.01 5.43 0.40 0.13
Pamb 22.07 70.80 7.31 7.82 2.66 0.04 5.10 0.30 0.15
Monoculture MasPest1 22.05 44.15 6.99 7.39 3.37 0.00 4.98 0.40 0.12
MasPest2 21.97 48.10 7.08 7.39 2.00 0.00 5.68 0.40 0.13
Taipest 22.00 52.17 6.81 7.39 2.60 0.01 4.83 0.48 0.33
4A. MORABOWEN ET AL.
26 taxa of Diptera, 11 genera of Ephemeroptera, and 29
genera of Trichoptera.
No signiﬁcant diﬀerences were found (full factorial
ANOVA) in community diversity metrics, except for
Shannon N1 (equivalent number of species) between
the streams draining diﬀerent land use types (Fig. 4).
However, a reduction in richness was evident in streams
draining monoculture farmlands compared with forest
and agroforestry farms streams (Fig. 5).
The PCO (which explained 57.4% of the variance) with
average cluster showed that streams grouped by geo-
graphical proximity rather than land use (Fig. 6). How-
ever, the average of the 3 streams for each land use
showed a signiﬁcant clustering, diﬀerentiating monocul-
ture streams from forest and agroforestry streams (Fig. 7).
The ANOSIM using relative abundance data and
excluding ubiquitous taxa showed that the least similar
communities compared with forest streams were those
draining monoculture farmlands (R=0.404, p=0.001).
Diﬀerences between streams draining forest and agrofor-
estry were lower (R=0.3, p= 0.001). Finally, we found
higher similarity in the relative abundance of taxa in agro-
forestry draining streams and monoculture draining
streams (R=0.237,p= 0.001). The ANOSIM using density
data, excluding ubiquitous taxa, showed that the least sim-
ilar communities compared with forest streams were those
draining monoculture farmlands (R=0.11,p= 0.029), fol-
lowed by the agroforestry farmlandsdraining streams (R=
0.026, p= 0.029). Finally, the organic agroforestry stream
communities and monoculture draining streams were
the least similar (R=0.116,p=0.029).
The SIMPER analysis showed that the dissimilarity
percentages among land uses, using the relative abun-
dance metrics, were smaller when comparing monocul-
ture streams to forest streams (68.7%) than the
dissimilarity of streams draining agroforestry farms
compared with those draining forest streams (69.8%;
SIMPER analysis). The taxa that completely disappeared
from monoculture streams and caused most of the diﬀer-
ences with the other treatments were Anacroneuria sp.
(Perlidae; Plecoptera), Hyallela sp. (Hyallelidae; Crusta-
cea), and Corydalus sp. (Corydalidae; Megaloptera)
Figure 2. (a) Temperature across streams draining diﬀerent land uses in the Mashpi River basin, Ecuador; (b) pH across streams draining
diﬀerent land uses in the Mashpi River basin, Ecuador; (c) discharge (m
/s) across streams draining diﬀerent land uses in the Mashpi
River basin, Ecuador; (d) coarse particulate organic matter (CPOM, g ash free dry mass) across streams draining diﬀerent land uses in the
Mashpi River basin, Ecuador.
INLAND WATERS 5
(SIMPER Analysis). The taxa that slowly declined in rel-
ative abundance from forests to monoculture draining
streams were Campylocia sp. (Euthyplocidae, Ephemer-
optera), Nectopsyche sp.(Leptoceridae; Trichoptera),
Chimarra sp. (Philopotamidae; Trichoptera), and Zelu-
sia sp. (Baetidae, Ephemeroptera). Diptera (Chironomi-
dae; Orthocladiinae; Limoniidae) were the only taxa
found in higher abundance in streams draining mono-
culture farmlands and agroforestry farmlands.
The SIMPER analysis showed that the dissimilarity
percentages among land uses, using density metrics,
were higher when comparing monoculture streams to
forest streams (64.5%) than the dissimilarity of streams
draining agroforestry farms compared with those
draining forest streams (60.7%); the percentage of dis-
similarity among agroforestry farmlands and forest
Figure 3. Principal component analysis (PCA) on eﬀects of abiotic variables on streams draining diﬀerent land uses clustered by com-
munity similarities in the Mashpi River basin, Ecuador. 60 represents the percentage of similarity according to the average cluster using
Bray-Curtis similarity matrix.
Figure 4. Shannon diversity (N1) diﬀerence across streams drain-
ing diﬀerent land uses in the Mashpi River basin, Ecuador.
Figure 5. Macroinvertebrate rareﬁed richness (SRaR) diﬀerence
across streams draining diﬀerent land uses in the Mashpi River
6A. MORABOWEN ET AL.
streams was intermediate (62.8%; SIMPER analysis).
The diﬀerences in densities of some taxa were not as
marked as the diﬀerences in relative abundance
between treatments. Campylocia sp. (Euthyplocidae;
Ephemeroptera), Anacroneuria sp. (Perlidae; Plecop-
tera), and Chimarra sp. (Philopotamidae; Trichoptera)
were the major contributors to the diﬀerences among
forest streams and streams draining monoculture farm-
lands. Contributors to the diﬀerences between streams
draining agroforestry farms and monoculture streams
were caused by Nectopsyche sp. (Leptoceridae; Trichop-
tera), Palaemnema sp. (Platystictidae; Odonata),
Cylloepus sp. (Elmidae; Coleoptera), and Psephenus
sp. (Psephenidae; Coleoptera).
ent land uses on communities of aquatic invertebrates in
streams in the Choco biodiversity hotspot to understand
how these communities have changed along a gradient
of human impact, from pristine forests to monoculture
palmito (Bactris gasipaes) plantations. In general, we
found that the invertebrate communities and the environ-
mental characteristics of streams draining anthropogenic
Figure 6. Principal coordinates of streams draining diﬀerent land uses in the Mashpi River basin, Ecuador. Groups are deﬁned by an
average group (land use) clustering with 60 representing the percentage of similarity according to the average cluster using Bray-Curtis
Figure 7. Average group (land use) clustering of sites with SIMPROF test from macroinvertebrate assemblages of streams draining
diﬀerent land uses in the Mashpi River basin, Ecuador.
INLAND WATERS 7
land uses were diﬀerent from streams draining lands with
lower levels of anthropogenic disturbance. Although the
benthic invertebrate communities of streams draining
agroforestry farms did not diﬀer signiﬁcantly from intact
forest stream communities, we found important diﬀeren-
ces between communities in monoculture streams and
those in intact forest streams, and some taxa were lost in
streams draining monoculture lands.
Eﬀects of land use on benthic fauna
One of the most detrimental human practices torivers and
streams is the establishment of monocultures that elimi-
nate forests to maximize production, especially for small
streams, which are among the most threatened habitats
because of the extent of land converted to agriculture
(Harding et al. 1998). Complete forest clearing reduces
allochthonous inputs to streams, which modiﬁes their tro-
phic structure (Abelho and Graça 1998). Shredder abun-
dance is inﬂuenced by several factors, such as leaf input
reduction, leaf type, and microorganism leaf conditioning
(Golladay et al. 1983, Graça 2001). In our study, we found a
decline in the relative abundance and density of the shred-
der Nectopsyche sp. in streams draining monoculture
farmlands, probably caused by the lack of diversity in
leaf input, together with a reduced fungi community colo-
nizing the leaves, as found in other studies (Chergui and
Pattee 1991, Graça et al. 1993, Rosenberg and Resh 1993,
Kiran 1996, Scrimgeour and Kendall 2003, Encalada
et al. 2010). Nectopsyche sp. was an important component
of the forest streams; it declined in streams draining agro-
forestry farms and almost disappeared in streams draining
monoculture farmlands. The decline in abundance of Lep-
toceridae (Trichoptera) has been attributed to their sensi-
tivity to aquatic pollutants (Rios-Touma et al. 2014)but
also to their dependency on diﬀerent sources of allochth-
onous material (Wallace et al. 1997, Rios-Touma et al.
2011). In our study, the shredder functional feeding
group (e.g., the caddisﬂies Phylloicus sp. and Nectopsyche
sp.) was crucial in processing CPOM. These insects accel-
erate litter fragmentation for other taxa to feed on and pro-
duce fecal pellets that contribute to secondary production.
The absence of these taxa cascades through the food web,
eventually reducing stream production (Webster and
Benﬁeld 1986,Graça2001, Allan and Castillo 2007).
Other eﬀects of deforestation near streams are an
increase in superﬁcial runoﬀ, deposition of ﬁne sedi-
ment, increased pesticides, and increased nutrient
input, accompanied by higher water temperature (Prin-
gle and Bernstead 2001, Iwata et al. 2003, Kasangaki
et al. 2008). Several studies found that plecopterans are
susceptible to organic pollution and lack of dissolved
oxygen (Armitage et al. 1983, Lenat 1988, Ríos-Touma
et al. 2014). In our study, Anacroneuria sp. (Plecoptera)
was completely absent from monoculture streams, possi-
bly because of higher temperatures.
Another macroinvertebrate group absent in monocul-
ture streams was Campylocia sp., detritivorous mayﬂy
burrowers that ingest large amounts of ﬁne particles
deposited in sedimentation areas (Fenoglio et al. 2008).
This genus of mayﬂies uses its large mandibles to stay
ﬁxed in sediments under stones, a habitat preference
that might explain its absence in monoculture streams
because stream sediments are usually the main sinks of
pollutants that enter streams (Cameron et al. 2002, Mag-
bauna et al. 2013). Although we did not document the
presence of pesticides in water or sediments because no
local laboratories suitable for this analysis were available,
we know that large amounts of herbicides are used and
washed into these streams because we found several
used cans of glyphosate in the plantations, and local res-
idents told us they used it extensively to clear weeds.
The amphipod Hyallela sp. was also absent from the
streams draining monoculture lands, but it was found in
streams draining agroforestry farms. This crustacean is
particularly sensitive to the ingredients present in glypho-
sate herbicide (Tsui and Chu 2004). Although we did not
examine glyphosate content in running waters in our
study, the community reported widespread use of this her-
bicide on roadsides, near streambeds, and of course inside
palmito plantations. During our ﬁeld trips we saw empty
kegs of herbicide thrown carelessly inside monoculture
plantations, some close to the streams (AM, pers. observ.).
In addition, local workers of monoculture farmlands esti-
mated that 0.2 L of herbicide dissolved in 2 L of water
were spread per hectare each month, which is within the
limits permitted by the Instituto Nacional de Investigación
Agropecuaria (INIAP) in Ecuador (2–4L/ha).
Stream communities from monoculture lands were
not similar to those in streams draining the other land
uses, mainly because of the high numbers of midges,
Orthocladiinae, in streams draining monoculture lands.
Chironomids are a well-known group of insects tolerant
to environmental extremes and to chemical and organic
pollution. Their high recolonization rates in these habi-
tats are due to their short life cycles and good ﬂight
capacity (Armitage et al. 2012).
Variables responsible for community diﬀerences
Many studies have shown that intensive agricultural
lands have lower evapotranspiration rates than natural
vegetation (Eagleson 1978, Gardner 1983, Canadell
et al. 1996, Costa et al. 2003, Li et al. 2007, Raymond
et al. 2008, Coe et al. 2011), especially in annual crops
and perennial pastures with reduced root density and
8A. MORABOWEN ET AL.
depth (Coe et al. 2011). The root system in forests also
plays a key role in stabilizing stream banks and prevent-
ing erosion (e.g., Chamberlin et al. 1991, Tabacchi et al.
2000). Our results showed that discharge was higher in
streams draining both organic agroforestry farmlands
and palmito monoculture streams, likely because these
streams have a less complex root system in their banks,
causing water to enter streams at a higher rate. In our
study, discharge was highly inﬂuential in the distribution
patterns of macroinvertebrates and caused diﬀerences in
benthic invertebrate communities between human-
impacted and forested streams.
The streams draining agroforestry lands and mono-
culture farms were warmer than streams draining forests,
possibly related to the lack of streamside canopy that
exposes streams to high solar radiance, causing higher
runoﬀand making water in these streams warmer and
less oxygenated. The agroforestry lands of the area
were formerly used for cattle grazing (∼6 years ago,
according to land owners), and riparian forest recovery
has probably not yet occurred. The Mashpi drainage
basin has a high level of solar radiance when there is
no cloud cover, making forest cover important for miti-
gating high temperatures. Other studies have also found
that logging in drainage basins leads to an increase in the
average water temperatures of streams (Burton and Lik-
ens 1973, Holtby and Scrivener 1988, St-Hilaire et al.
2000). As stated earlier, the absence of Plecoptera in all
monoculture streams may be caused by these high tem-
peratures. In addition, the dissolved oxygen capacity
decreases with water temperature (Dodds 2002). This
decrease in oxygen could also be aﬀecting communities
in these warmer monoculture streams, although we did
not detect signiﬁcant changes in oxygen concentration
between our streams.
Impact diﬀerences between monoculture and
Less invasive and destructive agricultural practices (i.e.,
preservation of buﬀering vegetation near streams and
avoiding water pollution) have less impact on diversity
and richness of macroinvertebrates, although no signiﬁ-
cant diﬀerence was found for richness. We did ﬁnd
large diﬀerences in the abundance of certain taxa in
communities in streams draining monoculture farm-
lands compared with agroforestry farmlands. None of
the taxa in the main orders (Odonata, Plecoptera, Crus-
tacea, Trichoptera, and Ephemeroptera) disappeared in
taxa did disappear from monoculture streams (e.g., Epi-
gomphus sp., Hyallela sp., Erpetogomphus sp., Austro-
limnius sp., Alisotrichia sp., and Hydroptila sp.).
Interestingly, in streams draining agroforestry farms,
some taxa increased in abundance (e.g., Nectopsyche
sp., Tricorythodes sp., Campylocia sp., Anacroneuria
sp., and Leptonema sp.) compared to forest streams.
These genera are possibly favored by intermediate levels
of disturbance in agroforestry farms streams. Lepto-
nema sp. and Campylocia sp. are part of the collec-
tor–gatherer functional feeding group, and in these
streams may be favored by the amount of food available
and lower abundance of competitors than in mature
and stable communities of forest streams (Wiggins
1996,2004, Tomanova et al. 2006,Reynaga2009).
Finally, Anacroneuria are facultative predators (Merritt
and Cummins 1996, Tomanova et al.2006,Reynaga
and Rueda 2010), which in biologically diverse streams
with diverse substrates, such as the forest streams, are
able to exploit a variety of prey types where they
become more abundant.
Deforestation near streams aﬀects the presence of some
Trichoptera, some Plecoptera, and all Amphipoda, and
favors the colonization of tolerant Chironomidae.
Streams of the area were relatively well conserved, and
they have high diversity. The main threats to this diver-
sity are deforestation, sedimentation, and presumably
chemical and organic pollution. These streams are the
source of water for many people in lowlands and should
be a focus for conservation eﬀorts and research. We
found that even small changes in land use could lead
to local extinction of some groups of benthic
The lack of signiﬁcance in the community metrics
(diversity [S], abundance [N], and density) in streams
draining diﬀerent land use types may be because (1)
diversity of these rivers is high enough that the diﬀeren-
ces can only be seen in certain taxa, and (2) these com-
munities may be resilient to negative impacts because
high diversity and high density of macrobenthos
allow constant recolonization. Although nonsigniﬁcant,
macroinvertebrate richness was low in monoculture
streams compared to forest and agroforestry lands
(Fig. 5), and we did ﬁnd a signiﬁcant increase in Shannon
diversity (N1) of agroforestry land streams compared to
the other land uses. We think this ﬁnding might be a case
of intermediate disturbance hypothesis (Townsend et al.
1997), in which new niches are open in streams draining
To better understand the eﬀects of deforestation and
pesticides in these streams, mesocosmos experiments
should be performed that control emergence success
and drift with controlled exposures to herbicides and
INLAND WATERS 9
organic matter availability that mimic conditions seen in
the streams. Hyallela sp., Anacroneuria sp., Campylocia
sp., Corydallus sp., and other taxa rare or absent from
the monoculture streams would be interesting to study
in these controlled experiments to diﬀerentiate the
eﬀects of deforestation from pesticide exposure. The
eﬀect of the possible presence of pesticides, especially
herbicides, was not obvious in the community diversity
and abundance metrics. The eﬀects of chronic exposure
can only be seen clearly by studying the life history of
macroinvertebrates, including emergence success, espe-
cially when other studies link glyphosate exposure with
an increase in drift and emergence propensities of just
a few glyphosate-sensitive taxa. This herbicide can also
reduce the size and success rate of emerging adult insects
(Magbanua et al. 2013,2016).
We are thankful to Cliﬀord Kyle, Álvaro Barragán and the 2
anonymous reviewers for their insightful comments. To Choc-
olate Mashpi (Agustina Arcos and Alejandro Solano), Pambi-
liño Preserve (Oliver Torres), Mashpi Lodge, and Mashpi
Preserve (Carlos Morochz) for ﬁeld facilities. This work was
supported by Universidad Tecnológica Indoamérica (UTI)
Grant 121.068.2014 and an agreement between UTI and
Mashpi Preserve. BRT was partially supported by Universidad
de Las Américas, Ecuador (UDLA, Grant: AMB.BRT.17.01).
No potential conﬂict of interest was reported by the author(s).
Verónica Crespo-Pérez http://orcid.org/0000-0002-8811-
Blanca Ríos-Touma http://orcid.org/0000-0002-3921-0908
Abelho M, Graça MAS. 1998. Litter in a ﬁrst-order stream of a
temperate deciduous forest (Margaraça Forest, central
Portugal). Hydrobiologia. 386:147–152.
Achard F, Eva HD, Stibig HJ, Mayaux P, Gallego J, Richards T,
Malingreau JP. 2002. Determination of deforestation rates
of the world’s humid tropical forests. Science. 297:999–
Allan JD, Castillo MM. 2007. Stream ecology: structure and
function of running waters. Berlin/Heidelberg (Germany):
Springer Science & Business Media.
[APHA] American Public Health Association. 1998. Standard
methods for the examination of water and wastewater. 20th
ed. Washington (DC).
Aratrakorn S, Thunhikorn S, Donald PF. 2006. Changes in
bird communities following conversion of lowland forest
to oil palm and rubber plantations in southern Thailand.
Bird Conserv Int. 16:71–82.
Armitage PD, Moss D, Wright JF, Furse MT. 1983. The perfor-
mance of a new biological water quality score system based
on macroinvertebrates over a wide range of unpolluted run-
ning-water sites. Water Res. 173:333–347.
Armitage PD, Pinder LC, Cranston P. 2012. The
Chironomidae: biology and ecology of non-biting midges.
Berlin/Heidelberg (Germany): Springer Science & Business
Benstead JP, Douglas MM, Pringle CM. 2003. Relationships of
stream invertebrate communities to deforestation in eastern
Madagascar. Ecol Appl. 13:1473–1490.
Benstead JP, Pringle CM. 2004. Deforestation alters the
resource base and biomass of endemic stream insects in
eastern Madagascar. Freshwater Biol. 49:490–501.
Bojsen BH, Jacobsen D. 2003.Eﬀects of deforestation on macro-
invertebrate diversity and assemblage structure in
Ecuadorian Amazon streams. Arch Hydrobiol. 158:317–342.
Burton TM, Likens GE. 1973. The eﬀect of strip-cutting on
stream temperatures in the Hubbard Brook. Bioscience.
Cameron KC, Di HJ, Condron LM. 2002. Nutrient and pesti-
cide transfer from agricultural soils to water in New
Zealand. In: Haygarth PM, Jarvis SC, editors. Agriculture,
hydrology and water quality. Oxfordshire (UK): CAB
International; p. 373–393.
Canadell J, Jackson RB, Ehleringer JB, Mooney HA, Sala OE,
Schulze ED. 1996. Maximum rooting depth of vegetation
types at the global scale. Oecologia. 108:583–595.
Chamberlin TW, Harr RD, Everest FH. 1991. Timber harvest-
ing, silviculture, and watershed processes. In: Meehan WR,
editor. Inﬂuences of forest and rangelandmanagement on sal-
monid ﬁshes and their habitats. Bethesda (MD): American
Fisheries Society Special Publication 19; p. 181–205.
Chapman M, Underwood A. 1999. Ecological patterns in multi-
variate assemblages: information and interpretation of negative
values in ANOSIM tests. Mar Ecol Prog Ser. 180:257–265.
Chen CY, Hathaway KM, Folt CL. 2004. Multiple stress eﬀects
of Vision® herbicide, pH, and food on zooplankton and lar-
val amphibian species from forest wetlands. Environ
Toxicol Chem. 23:823–831.
Chergui H, Pattee E. 1991. An experimental study of the break-
down of submerged leaves by hyphomycetes and inverte-
brates in Morocco. Freshwater Biol. 26:97–110.
Clarke KR, Gorley RN. 2006. Primer v6: user manual/tutorial.
Plymouth (UK): Plymouth Marine Laboratory.
Clarke KR, Warwick RM. 2001. Change in marine communi-
ties: an approach to statistical analysis and interpretation.
2nd ed. Plymouth (UK): Primer-E.
Coe MT, Latrubesse EM, Ferreira ME, Amsler ML. 2011.The
eﬀects of deforestation and climate variability on the
streamﬂow of the Araguaia River, Brazil. Biogeochemistry.
Costa MH, Botta A, Cardille JA. 2003.Eﬀects of large-scale
changes in land cover on the discharge of the Tocantins
River, Southeastern Amazonia. J Hydrol. 283:206–217.
Cummins KW, Wilzbach MA, Gates DM, Perry JB, Taliaferro
WB. 1989. Shredders and riparian vegetation. BioScience.
Dodds WK. 2002. Freshwater ecology: concepts and environ-
mental applications. San Diego (CA): Academic Press.
10 A. MORABOWEN ET AL.
Domínguez E, Fernández HR. 2009. Macroinvertebrados
bentónicos sudamericanos, sistemática y biología [South
American benthic macroinvertebrates, systematics and biol-
ogy]. Tucumán (Argentina): Fundación Miguel Lillo.
Dudgeon D. 2000. Large-scale hydrological changes in tropical
Asia: prospects for riverine biodiversity. BioScience.
Dudgeon D, Arthington AH, Gessner MO, Kawabata ZI,
Knowler DJ, Lévêque C, Naiman RJ, Prieur-Richard AH,
Soto D, Stiassny ML, Sullivan CA. 2006. Freshwater biodi-
versity: importance, threats, status and conservation chal-
lenges. Biol Rev. 81:163–182.
Eagleson PS. 1978. Climate, soil, and vegetation:
1. Introduction to water balance dynamics. Water Resour
Encalada AC, Calles J, Ferreira V, Canhoto CM, Graça MA.
2010. Riparian land use and the relationship between the
benthos and litter decomposition in tropical montane
streams. Freshwater Biol. 55:1719–1733.
Fenoglio S, Bo T, Czekaj A. Rościszewska E. 2008. Feeding
habits, ﬁne structure and microhabitat preference of
Euthyplocia hecuba (Hagen, 1861) (Ephemeroptera:
Euthyplociidae) nymphs from Honduras. Folia Biol-
Fitzherbert EB, Struebig MJ, Morel A, Danielsen F, Brühl CA,
Donald PF, Phalan B. 2008. How will oil palm expansion
aﬀect biodiversity? Trends Ecol Evol. 23:538–545.
Gardner WR. 1983. Soil properties and eﬃcient water use: an
overview. Limitations to eﬃcient water use in crop produc-
tion, Ch. 2A. Madison (WI): American Society of
Agronomy, Crop Science Society of America, Soil Science
Society of America; p. 45–64.
Golladay SW, Webster JR, Benﬁeld EF. 1983. Factors aﬀecting
food utilization by a leaf shredding aquatic insect: leaf spe-
cies and conditioning time. Ecography. 6:157–162.
González-Jaramillo V, Fries A, Rollenbeck R, Paladines J,
Oñate-Valdivieso F, Bendix J. 2016. Assessment of defores-
tation during the last decades in Ecuador using NOAA-
AVHRR satellite DATA. Erdkunde.70:217–235.
Graça MAS. 2001. The role of invertebrates on leaf litter
decomposition in streams-a review. Int Rev Hydrobiol.
Graça MAS, Maltby L, Calow P. 1993 Importance of fungi in
the diet of Gammarus pulex and Asellus aquaticus.I:
Feeding strategies. Oecologia. 93:139–144.
Hanson P, Springer M, Ramirez A. 2010. Capítulo 1:
Introducción a los grupos de macroinvertebrados
acuáticos [Introduction to groups of aquatic macroinverte-
brates]. Rev Biol Trop. 58:3–37.
Harding JS, Benﬁeld EF, Bolstad PV, Helfman GS, Jones EBD.
1998. Stream biodiversity: the ghost of land use past. P Nat
Acad Sci USA. 95:14843–14847.
Hauer FR, Lamberti GA, editors. 2006. Methods in stream
ecology. Cambridge (MA): Academic Press.
Henriksen A, Selmer-Olsen AR. 1970. Automatic methods for
determining nitrate and nitrite in water and soil extracts.
Henry R, Uieda VS, Afonso de Oliveira AA, Kikuchi RM. 1994.
Input of allochthonous matter and structure of fauna in a
Brazilian headstream. Verh Int Ver Theor Angew Limnol.
Holtby B, Scrivener C. 1988. Observed and simulated eﬀects of
climatic variability, clear-cut logging and ﬁshing on the
numbers of chum salmon and coho salmon returning to
Carnation Creek, British Columbia. Can Spec Publ Fish
Aquat Sci. 105:68–81.
Iwata T, Nakano S, Inoue M. 2003. Impacts of past riparian
deforestation on stream communities in a tropical rainforest
in Borneo. Ecol Appl. 13:461–473.
Jost L. 2006. Entropy and diversity. Oikos. 113:363–375.
Kasangaki A, Chapman LJ, Balirwa J. 2008. Land use and the
ecology of benthic macroinvertebrate assemblages of high-
altitude rainforest streams in Uganda. Freshwater Biol.
Kiran U. 1996. Fungi and food preferences of aquatic inverte-
brates. National Academy Science Letters-India.
Kondolf GM, Li S. 1992. The pebble count technique for quan-
tifying surface bed material size in instream ﬂow studies.
Laurance WF. 1999.Reﬂections on the tropical deforestation
crisis. Biol Conserv. 91:109–117.
Lenat DR. 1988. Water quality assessment of streams using a
qualitative collection method for benthic macroinverte-
brates. J N Am Benthol Soc. 7:222–233.
Li KY, Coe MT, Ramankutty N, De Jong R. 2007. Modelling
the hydrological impact of land-use change in West
Africa. J Hydrol. 337:258–268.
Liess, M, Ohe, PC. 2005. Analyzing eﬀects of pesticides on
invertebrate communities in streams. Environ Toxicol
Lorion CM, Kennedy BP. 2009.Relationships between defor-
estation, riparian forest buﬀers and benthic macroinverte-
brates in Neotropical headwater streams. Freshwater Biol.
Magbanua FS, Townsend CR, Hageman KJ, Matthaei CD.
2013. Individual and combined eﬀects of ﬁne sediment
and the herbicide glyphosate on benthic macroinvertebrates
and stream ecosystem function. Freshwater Biol. 58:1729–
Magbanua FS, Townsend CR, Hageman KJ, Piggott JJ,
Matthaei CD. 2016. Individual and combined eﬀects of
ﬁne sediment and glyphosate herbicide on invertebrate
drift and insect emergence: a stream mesocosm experiment.
Freshwater Sci. 35:139–151.
Merritt RW, Cummins KW. 1996. An introduction to the
aquatic insects of North America. Dubuque (IA): Kendall/
Ministerio del Ambiente. 2012. Línea Base de Deforestación
del Ecuador Continental, Quito-Ecuador [Baseline of defor-
estation in continental Ecuador, Quito-Ecuador].
Murphy J, Riley J. 1962. A modiﬁed single solution method for
the determination of phosphate in natural waters. Anal
Chim Acta. 27:31–36.
Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GA,
Kent J. 2000. Biodiversity hotspots for conservation priori-
ties. Nature. 403:853–858.
Pringle CM, Benstead JP. 2001. The eﬀects of logging on trop-
ical river ecosystems. In: Fimbel RA, Grahal A, Robinson
JG, editors. The cutting edge: conserving wildlife in logged
tropical forests. New York (NY): Columbia University
Press; p. 305–325.
INLAND WATERS 11
Raymond PA, Oh NH, Turner RE, Broussard W. 2008.
Anthropogenically enhanced ﬂuxes of water and carbon
from the Mississippi River. Nature. 451:449–452.
Reynaga MC. 2009. Hábitos alimentarios de larvas de
Trichoptera (Insecta) de una cuenca subtropical [Food hab-
its of Trichoptera larvae (Insecta) from a subtropical basin].
Ecología Austral. 19:207–214.
Reynaga MC, Rueda MP. 2010. Trophic analysis of two species
of Atopsyche (Trichoptera: Hydrobiosidae). Limnologica.
Ríos-Touma B, Acosta R, Prat N. 2014. The Andean Biotic
Index (ABI): revised tolerance to pollution values for mac-
roinvertebrate families and index performance evaluation.
Rev Biol Trop. 62:249–273.
Ríos-Touma B, Encalada AC, Prat N. 2011. Macroinvertebrate
assemblages of an Andean high-altitude tropical stream: the
importance of season and ﬂow. I nt Rev Hydrobiol. 96:667–685.
Roldán G. 1988. Guía para el estudio de los macroinvertebrados
acuáticos del Departamento de Antioquia [Study guide for
the aquatic macroinvertebrates of the Department of
Antioquia]. Universidad de Antioquia, Fondo FEN, Medellín.
Rosenberg DM, Resh VH. 1993. Freshwater monitoring and
benthic macroinvertebrates. New York (NY): Chapman & Hall.
Scrimgeour GJ, Kendall S. 2003.Eﬀects of livestock grazing on
benthic invertebrates from a native grassland ecosystem.
Freshwater Biol. 48:347–362.
Sierra R. 2013. Patterns and factors of deforestation in conti-
nental Ecuador 1990–2010. And an approach to the next
10 years. Quito (Ecuador): Conservación Internacional
Ecuador and Forest Trends.
Steinman AD, Lamberti GA. 1996. Biomass and pigments of
benthic algae. In: Hauer FR, Lamberti GA, editors. Methods
in stream ecology. San Diego (CA): Academic Press; p. 297.
Sundbäck K, Alsterberg C, Larson F. 2010.Eﬀects of multiple
stressors on marine shallow-water sediments: response of
microalgae and meiofauna to nutrient–toxicant exposure.
J Exp Mar Biol Ecol. 388:39–50.
Surber EW. 1937. Rainbow trout and bottom fauna production
in one mile of stream. T Am Fish Soc. 66:193–202.
St-Hilaire A, Morin G, El-Jabi N, Caissie D. 2000. Water
temperature modelling in a small forested stream:
implication of forest canopy and soil temperature. Can J
Civil Eng. 27:1095–1108.
Tabacchi E, Lambs L, Guilloy H, Planty-Tabacchi AM,
Muller E, De’camps H. 2000. Impacts of riparian
vegetation on hydrological processes. Hydrol Process.
Tomanova S, Goitia E, Heledic J. 2006.Trophic levels and
functional feeding groups of macroinvertebrates in
Neotropical streams. Hydrobiologia. 556:251–264.
Townsend CR, Scarsbrook MR, Dolédec S. 1997. The interme-
diate disturbance hypothesis, refugia, and biodiversity in
streams. Limnol Oceanogr. 42:938–949.
Townsend CR, Uhlmann SS, Matthaei CD. 2008. Individual
and combined responses of stream ecosystems to multiple
stressors. J Appl Ecol. 45:1810–1819.
Tsui MTK, Chu LM. 2004. Comparative toxicity of glyphosate-
based herbicides: aqueous and sediment porewater expo-
sures. Arch Environ Con Tox. 46:316–323.
Vannote RL, Minshall GW, Cummins KW, Sedell JR, Cushing
CE. 1980. The river continuum concept. Can J Fish Aquat
Wallace JB, Eggert SL, Meyer JL, Webster JR. 1997. Multiple
trophic levels of a forest stream linked to terrestrial litter
inputs. Science. 277:102–104.
Wanger TC, Iskandar DT, Motzke I, Brook BW, Sodhi NS,
Clough Y, Tscharntke T. 2010.Eﬀects of land-use change
on community composition of tropical amphibians and rep-
tiles in Sulawesi, Indonesia. Conserv Biol. 24:795–802.
Wantzen KM, Wagner R. 2006. Detritus processing by inver-
tebrate shredders: a Neotropical-temperate comparison. J
N Am Benthol Soc. 25:216–232.
Webster JR, Benﬁeld EF. 1986. Vascular plant breakdown in
freshwater ecosystems. Annu Rev Ecol Syst. 17: 567–594.
Weiß CH. 2007. StatSoft, Inc., Tulsa, OK.: STATISTICA,
Version 8. Adv Stat Anal. 91:339.
White KE. 1978. Dilution methods. In: Herchy RW, editor.
Hydrometry. Hoboken (NJ): John Wiley and Sons; p. 47–55.
Wiggins GB. 1996. Larvae of the North American caddisﬂygen-
era (Trichoptera). Toronto(ON): University of Toronto Press.
Wiggins GB. 2004. Caddisﬂies: the underwater architects.
Toronto (ON): University of Toronto Press.
12 A. MORABOWEN ET AL.