- Email: [email protected]
Habitat suitability of the invasive water hyacinth and its relation to water quality and macroinvertebrate diversity in a tropical reservoir
Accepted Manuscript Title: Habitat suitability of the invasive water hyacinth and its relation to water quality and macroinvertebrate diversity in a tropical reservoir Author: Tien Hanh Thi Nguyen id="aut0010" orcid="0000-0001-8183-328X"> Pieter Boets Koen Lock Minar Naomi Damanik Ambarita Marie Ane Eurie Forio Peace Sasha Luis Elvin Dominguez Granda Thu Huong Thi Hoang Gert Everaert Peter L.M. Goethals PII: DOI: Reference:
S0075-9511(15)00040-7 http://dx.doi.org/doi:10.1016/j.limno.2015.03.006 LIMNO 25452
To appear in: Received date: Revised date: Accepted date:
4-12-2014 26-3-2015 31-3-2015
Please cite this article as: Nguyen, T.H.T., Boets, P., Lock, K., Ambarita, M.N.D., Forio, M.A.E., Sasha, P., Granda, L.E.D., Hoang, T.H.T., Everaert, G., Goethals, P.L.M.,Habitat suitability of the invasive water hyacinth and its relation to water quality and macroinvertebrate diversity in a tropical reservoir, Limnologica (2015), http://dx.doi.org/10.1016/j.limno.2015.03.006 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Habitat suitability of the invasive water hyacinth and its relation to water quality
2
and macroinvertebrate diversity in a tropical reservoir
3
Tien Hanh Thi Nguyen
4
Ambarita1, Marie Ane Eurie Forio1, Peace Sasha1, Luis Elvin Dominguez Granda3, Thu
5
Huong Thi Hoang 4, Gert Everaert1, Peter L.M. Goethals1
6
1: Ghent University, Laboratory of Environmental Toxicology and Aquatic Ecology, J.
7
Plateaustraat 22, B-9000 Ghent, Belgium
8
2: Research Institute for Aquaculture No1, Dinh Bang, Tu Son, Bac Ninh, Vietnam
9
3: Escuela Superior Politécnica del Litoral (ESPOL), Water and Sustainable
10
Development Center (CADS), Campus Gustavo Galindo, Km. 30.5 Vía Perimetral,
11
Guayaquil - Ecuador
12
4: Hanoi University of Technology, Institute for Environmental Science and Technology,
13
No 1 Dai Co Viet, Hanoi, Vietnam
cr
ip t
, Pieter Boets1,*, Koen Lock1, Minar Naomi Damanik
14
te
d
M
an
us
1,2
*Author for correspondence: Pieter Boets, [email protected], Tel: +329 264 3776,
16
Fax: +329 264 4199
17 18
Ac ce p
15
Page 1 of 28
Abstract
20
In this study, we assessed the relationship between the occurrence of the invasive water
21
hyacinth (Eichhornia crassipes) and water quality properties as well as macroinvertebrate
22
diversity in a tropical reservoir, situated in western Ecuador. Macroinvertebrates and
23
physico-chemical water quality variables were sampled at 32 locations (during the dry
24
season of 2013) in both sites covered and non-covered by water hyacinth in the Daule-
25
Peripa reservoir. The results indicated that, in terms of water quality, only turbidity was
26
significantly different between sampling sites with and without water hyacinth (Mann-
27
Whitney U-test, p<0.01). The habitat suitability model showed that water hyacinth was
28
present at sites with a low turbidity. The percentage water hyacinth cover increased with
29
decreasing turbidity. The Biological Monitoring Working Party-Colombia score and the
30
Margalef diversity index were significantly higher (Mann-Whitney U-test, p<0.01) at
31
sampling sites where water hyacinth was present compared to water hyacinth absent sites.
32
However, there were no significant differences in the Shannon–Wiener index, Evenness
33
index and Simpson index between the sampling sites with and without water hyacinth.
34
Our results suggest that water hyacinth cover was an important variable affecting the
35
diversity of macroinvertebrates in the Daule-Peripa reservoir, with intermediate levels of
36
water hyacinth cover having a positive effect on the diversity of macroinvertebrates.
37
Information on the habitat suitability of water hyacinth and its effect on the physico-
38
chemical water quality and the macroinvertebrate community are essential to develop
39
conservation and management programs for large tropical reservoirs such as the Daule-
40
Peripa reservoir and the Guayas river basin, where water resources are being at high risk
41
due to expansion of agricultural and industrial development activities.
cr
us
an
M
d
te
Ac ce p
42
ip t
19
43
Keywords: Daule-Peripa reservoir, invasive species, water hyacinth, Eichhornia
44
crassipes, macroinvertebrates, water quality
45 46
Page 2 of 28
1. Introduction
48
Alien invasive plant species, mainly of terrestrial habitats, have recently received much
49
attention due to their direct or indirect impact on ecosystem structure and functioning
50
(Vilà et al., 2011). The impact of plant invasions in aquatic ecosystems has received less
51
attention. Several authors reported that invasive plant species can cause serious
52
ecological and economic impacts such as loss of native species diversity, hybridization
53
with native species, changes in ecosystem processes and functioning, and an increase of
54
pests and diseases (Rodriguez, 2006; Villamagna and Murphy, 2010; Stiers et al., 2011).
55
Alien invasive plant species may alter the available structure in aquatic habitats by
56
creating a shift to a homogeneous habitat, thereby negatively affecting biological
57
communities (Theel et al., 2008; Schultz and Dibble, 2012). Indeed, simplification of the
58
macrophyte growth form seems to negatively affect the abundance of biotic communities
59
such as macroinvertebrates (Walker et al., 2013). Schultz and Dibble (2012) found that
60
mechanisms underlying the impact of aquatic invasive plants are not very different from
61
native aquatic plants species. They identified three invasive traits largely responsible for
62
negative effects on biological communities: increased growth rate, allelopathic chemical
63
production and phenotypic plasticity. Despite the negative impact that is often observed,
64
positive relationships between invasion by alien macrophytes and macroinvertebrate
65
diversity have been observed as well (e.g. Brendonck et al., 2003; Villamagna, 2009).
66
Water hyacinth Eichhornia crassipes (Mart.) Solms is one of the most widely spread
67
invasive aquatic macrophytes in the world (Villamagna and Murphy, 2010). The species
68
has
69
Pantanal in western Brazil (Parolin et al., 2012). Water hyacinth is a free floating aquatic
70
plant, that can spread fast and has a strong growth (Malik, 2007). According to Malik
71
(2007), water hyacinth can double its size (area covered) in five days and a mat of
72
medium sized plants may contain two million plants per hectare that weigh 270 to 400
73
tons. Under favorable conditions the biomass of water hyacinth can be doubled within
74
only 12 days (Parolin et al., 2012). Nowadays, water hyacinth is distributed worldwide
75
(Parolin et al., 2012) and the International Union for Conservation of Nature listed the
76
species as one of the 100 most harmful invasive species (Lowe et al., 2000).
Ac ce p
te
d
M
an
us
cr
ip t
47
its
origin
in
the
Amazon
basin
and
in
the
vast
wetlands
of
the
Page 3 of 28
Water hyacinth tolerates a wide range of nutrients, temperature and pH levels and grows
78
in a wide variety of ecosystem types (Malik, 2007). Environmental factors such as
79
temperature, pH, solar radiation, and salinity of the water can influence the growth and
80
performance of water hyacinth (Gupta et al., 2012). The optimum conditions for growth
81
of water hyacinth are a pH between 6 to 8 (Malik, 2007) and a temperature that ranges
82
between 28 and 30°C (Gupta et al., 2012). A study by Wilson et al. (2005) showed that
83
nutrient concentrations and temperature are two of the most important factors
84
determining water hyacinth growth and reproduction. Salinity on the other hand can
85
cause major constraints for water hyacinth growth, since water hyacinth can not survive
86
at a salinity higher than 2‰ (Olivares and Colonnello, 2000).
87
Water hyacinth can have an effect on the physical and chemical composition of water.
88
The introduction of water hyacinth can cause a change in water clarity, hydrological
89
regime, dissolved oxygen concentration, nutrient concentrations and other pollutants in
90
the water body (Villamagna and Murphy, 2010). Negative as well as positive effects have
91
been recorded. Water hyacinth is known to reduce dissolved oxygen concentrations
92
because the mats prevent the transfer of oxygen from the air to the water surface, while
93
the plant does not release oxygen into the water (Meerhoff et al., 2003). Water hyacinth
94
has the potential to stabilize pH and temperature and prevent stratification in lotic
95
systems (Giraldo and Garzon, 2002) and in this way change ecosystem structure and
96
functioning. Rommens et al. (2003) and Sooknah and Wilkie (2004) reported that water
97
hyacinth has the capacity to absorb nutrients (e.g. nitrate, ammonium, phosphate) from
98
the water column affecting phyto- and zooplankton abundance.
99
Water hyacinth plays a important role for phytoplankton, zooplankton and fish in
100
freshwater ecosystems by providing habitat complexity, shelter and feeding grounds
101
(Brendonck et al., 2003; Meerhoff et al., 2003; Villamagna and Murphy, 2010).
102
According to de Marco et al. (2001) and Brendonck et al. (2003) the roots and the leaves
103
of water hyacinth offer an important substratum and habitat for macroinvertebrate
104
colonisation. Indeed, a positive relationship between macroinvertebrate diversity and the
105
cover of water hyacinth has been found (Schramm et al., 1987; Kouame et al., 2011).
106
Kouame et al. (2011) found a high diversity and density of macroinvertebrate
Ac ce p
te
d
M
an
us
cr
ip t
77
Page 4 of 28
107
assemblages associated with root masses of water hyacinth in Lake Taabo (Ivory Coast).
108
They ascribed this to improvements in physico-chemical properties (e.g. conductivity,
109
nutrients,
110
macroinvertebrate community. Moreover, Bailey and Litterick (1993) also found a
111
positive effect of water hyacinth on dissolved oxygen and a variety of potential food
112
resources for aquatic invertebrates in water hyacinth root-mats resulting in higher
113
abundances of macroinvertebrates. However, water hyacinth limits the development of
114
phytoplankton via preventing light penetration and absorbing nutrients which reduces
115
phytoplankton and leads to a decrease in zooplankton abundance (Villamagna and
116
Murphy, 2010). In addition, water hyacinth causes significant ecological alterations in the
117
invaded community by modifying the habitat, disrupting the food chain and nutrient
118
cycling, and consequently changing invertebrate and fish assemblage structure and finally
119
the entire food web (Brendonck et al., 2003; Toft et al., 2003; Coetze et al., 2014). Hence,
120
water hyacinth influences species richness, diversity and composition of invaded
121
communities and causes huge impacts to ecosystems structure and functioning
122
(Villamagna and Murphy, 2010).
123
Next to its ecological impact, water hyacinth has attracted global attention due to serious
124
problems caused to power plants, navigation, irrigation and recreation (Epstein, 1998; Lu
125
et al., 2007; Malik, 2007). According to Malik (2007), many large hydropower plants
126
have to spend considerable time and money in clearing water hyacinth in order to prevent
127
it from entering in the turbines and causing damage and power interruptions. In China,
128
water hyacinth is reported to clog lakes and rivers, impede water flows, obstruct
129
navigation, and damage irrigation and hydroelectricity facilities, thus the total cost for
130
water hyacinth control is estimated to be more than $12.35 million each year (Lu et al.,
131
2007). Indeed, the invasion of water hyacinth causes huge ecological and economic
132
consequences worldwide (Lu et al., 2007). However, their effects to human society vary
133
according to the magnitude of invasion, the ecosystem services of the water body and the
134
response of water hyacinth to management efforts (Villamagna and Murphy, 2010).
135
There have been several studies carried out on the effect of water hyacinth on water
136
quality and aquatic diversity, the invasion mechanism and the utilization in temperate
turbidity),
which
positively
influenced
the
Ac ce p
te
d
M
an
us
cr
ip t
temperature,
Page 5 of 28
regions (Villamagna and Murphy, 2010). However, until now the habitat suitability of
138
water hyacinth and the relationship between water hyacinth and ecological communities
139
largely remains understudied in tropical reservoirs. Therefore, the aim of this study was
140
to determine the habitat suitability of water hyacinth and to assess the effect of water
141
hyacinth cover on the chemical water quality and on the macroinvertebrate community in
142
a tropical reservoir. This information is essential to develop conservation and
143
management programs for tropical reservoirs and specifically the Guayas river basin,
144
situated in western Ecuador, where water resources are being at high risk due to
145
expansion of agricultural and industrial development activities.
146
2. Materials and Methods
147
2.1 Study area
148
The Daule-Peripa reservoir is located in the Guayas river basin in the western part of
149
Ecuador (Fig. 1). Hydropower and agriculture (banana, rice, maize, African oil palm
150
(Elaeis guineensis) and cacao) are the main economic sectors of the Guayas basin
151
(Alvarez-Mieles et al., 2013). The reservoir which was constructed in 1987 receives
152
water from the Daule and Peripa rivers (Gelati et al., 2011), and provides hydropower
153
generation, irrigation, flood protection and drinking water supply (Arriaga, 1989). The
154
reservoir has a water storage capacity of 6,000 million m³, a maximum surface area of
155
approximately 30,000 ha and a strongly fluctuating water level between 70-85 m in
156
difference due to the operation of the dam (CELEC, 2013).
cr
us an
M
d
te
Ac ce p
157
ip t
137
158
2.2 Sampling
159
The physio-chemical water variables and macroinvertebrate samples were collected in the
160
Daule-Peripa reservoir during the dry season (November) of 2013. In total, 32 samples
161
were taken. Eleven samples were taken at non-vegetated sites and 21 sites were
162
characterised by vegetation (the presence of water hyacinth) (Fig. 1 and Fig. 2). At the
163
vegetated sites, water hyacinth was the only floating macrophyte occurring. Only at one
164
vegetated site a native emergent plant species (Sagittaria spec.) occurred at very low
Page 6 of 28
abundance, at all other sites no native plants species were recorded. The percentage of
166
water hyacinth cover was visually estimated from the bank along a transect of 100 m
167
where the site where macroinvertebrates and water samples were taken was considered as
168
the centre. The vegetation cover classes were divided as follows (based on the Braun-
169
Blanquet cover/abundance scale): 0 = non-vegetated/absent, 1 = 1-5% (rare), 2 = 5-25%
170
(occasional), 3 = 25-50% (frequent), 4 = 50-75% (common) and 5 = 75-100%
171
(abundant). Classes rather than exact values for cover of water hyacinth were used as this
172
yields more reliable measures (Ellenberg and Mueller-Dombois, 1974). The 21 vegetated
173
sites were classified, based on the vegetation cover, as follows: class 0: 11 sites, class 1: 5
174
sites, class 2: 0 sites, class 3: 2 sites, class 4: 3 sites and class 5: 11 sites. Values of pH,
175
Dissolved Oxygen (DO) (mg.L-1), Chlorophyll (µg.L-1), Chloride (mg.L-1), Turbidity
176
(FTU), Conductivity (µS.cm-1) and Total Dissolved Solids (TDS) (mg.L-1) were
177
measured at the surface water layer with a multiprobe (model YSI 6600 V2 and YSI 6600
178
V1, YSI manufacturer). Moreover, water samples from each sampling location were
179
collected and stored in plastic bottles (1l), kept cool and dark immediately after collection
180
and transported to the laboratory for further analysis. At the laboratory, the water samples
181
were kept in a refrigerator (cool and dark) for nutrient analysis. The Hach Lange GMBH
182
photometric method was used to determine chemical oxygen demand (COD) (mg L-1),
183
total phosphorus (TP) (mg P.L-1), ammonium (mg N.L-1), nitrate (mg N.L-1), nitrite (mg
184
N.L-1) and total nitrogen (TN) (mg N.L-1). However, the values of all variables analysed
185
at the laboratory except for ammonium were below the detection limit of the Hach Lange
186
GMBH. Consequently, those variables were not included in the analysis. The lowest
187
detection limit of COD was 5 mg/L, TN was 1 mg/L, TP was 0.5 mg/L, NO3- -N was
188
0.23 mg/L, NO2- -N was 0.015 mg/L and NH4+-N was 0.015 mg/L.
cr
us
an
M
d
te
Ac ce p
189
ip t
165
190
Macroinvertebrates were collected by standard hand net consisting of a metal frame
191
holding a conical net (mesh-size 300 μm) at the same sites where water quality was
192
measured. Macroinvertebrates were collected during 10 minutes sampling including all
193
different microhabitats present at the sampling site (Gabriels et al., 2010). At the non-
194
vegetated sites, samples were collected by vigorously sweeping the net along the
195
reservoir margins and by disturbing the bank substratum at wading depth. For those sites
Page 7 of 28
that contained macrophytes, macroinvertebrates were obtained by submerging the
197
handnet under the roots of the macrophytes and by disturbing the stalks and leaves.
198
Afterwards, the net was quick and carefully lifted out of the water to prevent the escape
199
of mobile organisms. Macroinvertebrates attached to the roots of water hyacinth were
200
manually collected. Similarly macroinvertebrates were collected from stones and other
201
substrates at both vegetated and non-vegetated sites (Gabriels et al., 2010). Samples
202
collected were sieved (mesh size of 500µm) in the laboratory and sorted in white trays.
203
Samples from each location were placed in separate small plastic vials containing 80%
204
ethanol for preservation. After sorting, organisms were counted and identified under a
205
stereomicroscope. Macroinvertebrates were identified to family level using the
206
identification keys developed by Domínguez and Fernández (2009). The biotic
207
macroinvertebrate index BMWP-Colombia (Biological Monitoring Working Party-
208
Colombia) was calculated according to the modified method proposed by Zúñiga and
209
Cardona (2009). The BMWP-Colombia was calculated per site based on a summation of
210
all macroinvertebrate taxon scores. Each macroinvertebrate taxon receives a score that
211
reflects its susceptibility to pollution, where pollution-intolerant taxa receive high scores,
212
whereas pollution-tolerant taxa are given low scores (Zúñiga and Cardona, 2009). The
213
total score for each site indicates water quality categories ranging from very bad (0-15),
214
bad (16-35), poor (36-60), moderate (61-100) to good (>100). Moreover, the Shannon-
215
Wiener Diversity Index (Shannon and Wiener, 1949), Evenness index (Menhinick, 1964),
216
Simpson’s diversity index (Simpson, 1949) and Margalef index (Margalef, 1968) were
217
calculated to assess the taxa evenness and taxa richness for each sampling site. A
218
combination of several diversity and biotic indices was calculated in order to take
219
advantages of the strengths of each and develop a more complete understanding of
220
community structure (Hooper et al., 2005).
cr
us
an
M
d
te
Ac ce p
221
ip t
196
222
2.3 Data analysis
223
Generalized linear models were developed in R (version 3.0.3, The R Project for
224
Statistical
225
determined the occurrence of water hyacinth. Prior to developing the GLM, correlations
226
and variance inflation factors (VIFs) between the predictor variables were examined, in
Computing,
http://www.r-project.org) to
investigate which variables
Page 8 of 28
order to avoid problems of collinearity (Zuur et al., 2009). All variables with a correlation
228
of 0.7 or higher were removed. The presence or absence of water hyacinth was the
229
predicted variable, whereas temperature, conductivity, chlorophyll, oxygen saturation and
230
turbidity were retained as predictor variables. We started fitting the full model, i.e.
231
including all candidate predictor variables. Next, a stepwise backward selection
232
procedure was followed based on the Akaike Information Criterion (AIC), where the
233
model with the lowest AIC value was retained as the final model. Relations between the
234
residuals (the differences between observations and predictions by the retained model)
235
and predictor variables were evaluated and the normality of the residuals was tested using
236
a QQ-plot (probability plot). Retained models were only considered reliable if no
237
relations between the residuals and the predictor variables were visually observed and
238
residuals were normally distributed (Zuur et al., 2009); the retained models were rejected
239
otherwise.
240
Non-parametric tests were performed, since the data were non-normal distributed (tested
241
with Shapiro-Wilk Normality test). A Mann-Whitney U-test was used to compare
242
physico-chemical variables between sampling sites with and without water hyacinth.
243
Spearman’s rank correlation was used to explore the relationships between percentage
244
water hyacinth cover and macroinvertebrate metrics. A Kruskal-Wallis ANalysis Of
245
VAriance (ANOVA) followed by post-hoc multiple comparisons was performed to test
246
whether significant differences in ecological indices existed between different classes of
247
vegetation cover. Species data were log(x+1) transformed before multivariate analysis to
248
ensure data normality (Clarke, 1993). Non-metric Multidimensional Scaling (NMDS)
249
was used to visualize the similarity of macroinvertebrate communities between habitats
250
types onto two-dimensional charts. ANOSIM (ANalysis Of SIMilarities) was used for
251
testing the similarities (based on the Bray-Curtis similarity) of macroinvertebrate
252
community composition between sites with and without water hyacinth. A SIMPER
253
(SIMilarity PERcentages) analysis was applied for identifying which species primarily
254
contributed to the observed differences in macroinvertebrate assemblages between habitat
255
types. All multivariate analyses were performed using the PRIMER software package
256
(Clarke, 1993).
Ac ce p
te
d
M
an
us
cr
ip t
227
Page 9 of 28
3. Results
258
3.1 Physico-chemical water quality
259
Table 1 shows the water quality results measured at the Daule-Peripa reservoir. Based on
260
a Mann-Whitney U-test it was found that only turbidity differed significantly between
261
sampling sites with and without water hyacinth (Table 1). Based on the low conductivity
262
and high oxygen levels, a relatively good water quality was observed for most sampling
263
sites, regardless of the presence or absence of water hyacinth.
us
264
cr
ip t
257
3.2. Habitat suitability of water hyacinth
266
After model selection only turbidity was retained in the final model (Appendix A). Based
267
on the generalized linear model we found that water hyacinth is present at sites with a
268
low turbidity (Fig. 3). The percentage water hyacinth cover increased with decreasing
269
turbidity (Appendix A, Fig. A1).
M
d te
270
an
265
3.3 Relationship between the macroinvertebrate community and habitat characteristics
272
Based on the BMWP-Colombia scores the sampling sites of the Daule-Peripa reservoir
273
were categorized in four water quality classes: moderate, poor, bad and very bad
274
(Appendix B, Fig. B1). The BMWP-Colombia score and Margalef index were
275
significantly higher (Mann-Whitney U-test, p<0.01) for sites containing water hyacinth
276
compared to sites containing none. However, there were no distinct differences in the
277
Shannon–Wiener index, Evenness index and Simpson index between the sampling sites
278
with and without water hyacinth (all p > 0.05).
279
A Spearman rank-order correlation was run to determine the relationship between
280
percentage of cover by water hyacinth and the ecological indices. There was a significant,
281
positive correlation between percentage cover by water hyacinth and the BMWP-
282
Colombia and Margalef index (r = 0.64, r = 0.54 respectively; p < 0.05). Based on a
283
Kruskal-Wallis ANOVA a significant difference was detected in BMWP-Colombia
Ac ce p
271
Page 10 of 28
scores (chi-squared = 13.8, df = 4, p = 0.008) and in Margalef index between different
285
vegetation classes (chi-squared = 12.6, df = 4, p = 0.01). Based on post-hoc tests we
286
found a significant difference for the BMWP-Colombia index between vegetation class 0
287
and class 5 and for the Margalef index between vegetation class 3 and all other classes (p
288
< 0.05) (Fig. 4). When turbidity was plotted in function of the BMWP-Colombia,
289
samples with a high percentage water hyacinth cover were clearly separated from sites
290
without water hyacinth or sites with a low cover (Fig. 5). Sites with a lower turbidity had
291
a higher BMWP-Colombia score.
292
In total, 32 different macroinvertebrate taxa were found at the different sites with a
293
maximum of 19 different taxa for sites without water hyacinth and a maximum of 25 taxa
294
for sites covered by water hyacinth (Appendix C). Sites with an intermediate vegetation
295
cover (25-50%) had a relatively high abundance and diversity of macroinvertebrate taxa.
296
SIMPER (SIMilarity PERcentages) analysis indicated that the similarity in species
297
composition between samples with water hyacinth was 49.4%, while similarity between
298
samples without water hyacinth was 29.6%. Dissimilarity in species composition between
299
samples with and without water hyacinth was 67.6%. Densities of Chironomidae,
300
Hyallelidae,
301
Coenagrionidae, Tubificidae, Mesoveliidae and Caenidae reached higher densities at
302
sampling sites with water hyacinth, while Glossiphoniidae, Thiaridae, Gerridae and
303
Corixidae reached higher densities at sampling sites without water hyacinth (Appendix
304
C). ANOSIM (ANalysis Of SIMilarity) indicated that there was a significant difference
305
in macroinvertebrate community composition between sites with and without water
306
hyacinth (R=0.47; p<0.01). Non-metric MultiDimensional Scaling (NMDS) also
307
indicated that samples with water hyacinth were more similar to each other than samples
308
without water hyacinth (Fig. 6).
Dugesiidae,
Libellulidae,
Baetidae,
Notonectidae,
Ac ce p
te
Prostigmata,
d
M
an
us
cr
ip t
284
309 310
4. Discussion
311
This study is one of the few studies investigating the relationship between the occurrence
312
of the invasive water hyacinth and physico-chemical water quality conditions and
Page 11 of 28
macroinvertebrate diversity in a tropical reservoir. In terms of physico-chemical
314
characteristics, our results show that only turbidity differed significantly between
315
sampling sites with and without water hyacinth. Water hyacinth reduces the effect of
316
waves caused by the wind and motorboats, which explains the lower average turbidities
317
that were observed at stations containing water hyacinth in the Daule-Peripa reservoir.
318
On the other hand, water hyacinth will also more easily establish at sites where the wave
319
action is not too strong, where there is a low flow velocity and where consequently
320
turbidity is low as well (Opande et al., 2004). Our habitat suitability model, indicating
321
that water hyacinth is present at sites with a low turbidity, supports this finding. The
322
lower average turbidity measured at sites where water hyacinth was presented is also
323
supported by Jeppesen et al. (1997) because it was found that water hyacinth has the
324
capacity to stabilise the sediment and thus reduce turbidity. There was no significant
325
difference in average chlorophyll concentrations between covered and non-covered sites,
326
however, the maximum concentration measured at water hyacinth absent sites was more
327
than two times higher than that at sites where water hyacinth was present. This may be
328
attributed to water hyacinth absorbing nutrients from the water column and reducing light
329
penetration when forming dense vegetation mats (McVea and Boyd, 1975), thus limiting
330
the productivity of phytoplankton at sites where water hyacinth is present (Malik, 2007;
331
Villamagna and Murphy, 2010). Dissolved oxygen concentration, although not
332
significantly different, was lower at sites where water hyacinth is present compared to
333
sites where the species is absent. The lower oxygen levels measured, can be ascribed to
334
the water hyacinth mats that prevent the transfer of oxygen from the air to the water and
335
the reduced effects of mixing due to wind (Meerhoff et al., 2003; Villamagna, 2009).
336
A positive relationship between the occurrence of water hyacinth and macroinvertebrate
337
diversity was observed. Higher BMWP-Colombia and Margalef scores were reported for
338
sites containing water hyacinth compared to locations without water hyacinth. The
339
Margalef index reveals that sites covered for 25-50% with water hyacinth have a higher
340
species diversity compared to water hyacinth absent sites. Similarly, Kouame et al.
341
(2011) found that the vegetation cover by water hyacinth positively altered the
342
biodiversity of benthic invertebrate assemblages. Indeed, macrophytes can provide
343
excellent
Ac ce p
te
d
M
an
us
cr
ip t
313
microhabitats
that
promote
the
establishment
and
colonization
of
Page 12 of 28
macroinvertebrates. Furthermore, macroinvertebrates use macrophytes for refuge and
345
shelter against predation (Walker et al., 2013). In addition, the root mat of water hyacinth
346
can provide new habitats for colonization by macroinvertebrates and in this way increase
347
the diversity of aquatic macroinvertebrates (Masifwa et al., 2001; Rocha-Ramirez et al.,
348
2007; Barker et al., 2014). When the percentage of water hyacinth cover was higher than
349
50% we did not observe an increase in positive effect, because the cover might be too
350
dense, which negatively affects the physico-chemical water quality conditions and
351
consequently also the diversity of macroinvertebrates. This may be explained by the
352
intermediate cover hypothesis that was put forward by Villamagna (2009) to explain the
353
observed patterns in invertebrate abundance and diversity of Lake Chapala in Mexico.
354
This hypothesis suggests that the biotic community (e.g. fish) increases with increasing
355
prey abundance and provision of refuge as a consequence of increased vegetation.
356
However, a too high vegetation density may reduce oxygen availability and increase the
357
competition between organisms thereby negatively affecting the density of the biotic
358
community.
359
Our study showed that the dissimilarity in species composition between samples with and
360
without water hyacinth was 67.6%, indicating differences in the taxa associated with
361
water hyacinth. At sampling sites where water hyacinth was present, both pollution
362
tolerant (e.g. Tubificidae and Chironomidae) and pollution sensitive (e.g. Hyallelidae,
363
Baetidae and Coenagrionidae) families reached higher abundances compared to sites
364
where water hyacinth was absent. According to Rocha-Ramirez et al. (2007), the density
365
of invertebrates is not only affected by the presence of water hyacinth, but also by
366
environmental variables such as temperature, salinity, dissolved oxygen, and turbidity. In
367
addition, particular groups of invertebrates are correlated with specific variables (e.g.
368
mayflies prefer a lower conductivity) (Rocha-Ramirez et al., 2007) and can respond
369
differently to water hyacinth (Villamagna and Murphy, 2010). In our study, it seems that
370
the measured environmental conditions in combination with the presence of water
371
hyacinth positively influenced the diversity and abundance of macroinvertebrates.
372
Water hyacinth is considered one of the world’s worst weeds (Lowe et al., 2000). The
373
explosive growth rate of water hyacinth has caused serious socio-economic consequences
Ac ce p
te
d
M
an
us
cr
ip t
344
Page 13 of 28
such as obstruction of water ways, reducing hydropower production, etc. which often
375
requires immediate action (Masifwa et al., 2001). Our study showed that water hyacinth
376
is positively related to macroinvertebrate diversity and water quality. This benefit should
377
be weighed and compared to other impacts on ecosystem services before management
378
actions are initiated. It was reported that about one third of the Daule-Peripa reservoir is
379
covered by water hyacinth and that the presence of water hyacinth causes blockage of
380
navigation between different communities in the reservoir and increases the incidence of
381
infectious diseases (e.g. malaria, dengue and mosquito born infectious diseases)
382
(Gerebizza, 2009). However, studies of water hyacinth effects on local economy (e.g.
383
cost for removal of water hyacinth, cost of local people for transportation by boats,
384
medical treatment and the problems with hydropower generation) in the Daule-Peripa
385
reservoir have not been assessed. Small and isolated mats of water hyacinth can probably
386
provide a unique habitat that contributes to overall biotic diversity and ecosystem
387
functioning (Villamagna and Murphy, 2010). In case of the Daule-Peripa reservoir, we
388
found that an intermediate vegetation cover of water hyacinth was positively related with
389
the diversity of macroinvertebrates, whereas the impact of water hyacinth on hydropower
390
generation is expected to be limited. However, further investigation is needed to quantify
391
these findings and to assess the cost and benefits related to the presence of water
392
hyacinth. Integrating the ecological knowledge obtained from this study with economic
393
assessments and public perception could help decision makers to identify priority habitats
394
to be targeted for the control of water hyacinth and to prioritize conservation actions in an
395
operative way (Caplat and Coutts, 2011).
396
In conclusion, the findings of this study revealed that only turbidity was significantly
397
different between sampling sites with and without water hyacinth. The presence of water
398
hyacinth was an important variable affecting the diversity of macroinvertebrates in the
399
Daule-Peripa
400
macroinvertebrate diversity and the occurrence of water hyacinth it is important to
401
consider the socio-economic costs related to the management of water hyacinth. Work
402
conducted by Villamagna (2009) and Villamagna et al. (2012) on the effect of water
403
hyacinth cover on bird diversity and abundance found evidence for a more indirect
404
influence on waterbirds via changes in trophic structure, prey community composition,
Ac ce p
te
d
M
an
us
cr
ip t
374
reservoir.
Despite
the
overall
positive
relationship
between
Page 14 of 28
and energy flows throughout the system. This work is believed to serve as baseline data
406
for further studies on tropical reservoirs and as an incentive to assess the indirect effect of
407
water hyacinth on other communities (e.g. fish, phytoplankton and zooplankton).
408
Acknowledgments
409
The first author is grateful for the financial support for this work from Belgian Technical
410
Cooperation (BTC) in the framework of her PhD grant entitled “Ecological impact
411
assessment of hydropower generation on river systems”. Pieter Boets is supported by a
412
postdoctoral fellowship from the Fund for Scientific Research (FWO Vlaanderen,
413
Belgium). We would like to thank for the support provided by the VLIR Ecuador
414
Biodiversity Network project and the assistance of CADS-ESPOL for the help during the
415
fieldwork.
an
us
cr
ip t
405
416
References
418
Alvarez-Mieles, G., Irvine, K., Griensven, A.V., Arias-Hidalgo, M., Torres, A., Mynett,
419
A.E., 2013. Relationships between aquatic biotic communities and water quality
420
in a tropical river-wetland system (Ecuador). Environ. Sci. Pol. 34, 115-127.
421
Arriaga, L., 1989. The Daule- Peripe Dam Project, Urban Development of Guayaquil and
422
Their Impact on Shrimp Mariculture. In: Olsen, S. and Arriaga, L., editors. A
424
d
te
Ac ce p
423
M
417
Sustainable Shrimp Mariculture Industry for Ecuador. Narragansett, RI: Coastal Resources Center, University of Rhode Island.
425
Bailey, R.G., Litterick, M.R., 1993. The Macroinvertebrate Fauna of Water Hyacinth
426
Fringes in the Sudd Swamps (River Nile, Southern Sudan). Hydrobiologia 250,
427 428 429
97-103.
Barker, J.E., Hutchens, J.J., Luken, J.O., 2014. Macroinvertebrates Associated with Water Hyacinth Roots and a Root Analog. Freshw. Sci. 33, 159-167.
430
Brendonck, L., Maes, J., Rommens, W., Dekeza, N., Nhiwatiwa, T., Barson, M.,
431
Callebaut, V., Phiri, C., Moreau, K., Gratwicke, B., Stevens, M., Alyn, N.,
432
Holsters, E., Ollevier, F., Marshall, B., 2003. The impact of water hyacinth
433
(Eichhornia crassipes) in a eutrophic subtropical impoundment (Lake Chivero,
434
Zimbabwe). II. Species diversity. Arch. Hydrobiol. 158, 389-405.
Page 15 of 28
435
Caplat, P., Coutts, S., 2011. Integrating Ecological Knowledge, Public Perception and
436
Urgency of Action into Invasive Species Management. Environ. Manage. 48, 878-
437
881.
440 441
accesed on17/10/2014. In Spanish.
ip t
439
CELEC, 2013. Revista 25 Anos de la presa Daule-peripa. http://www.hidronacion.org,
Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community structure. Australian J. Ecol. 18, 117-143.
cr
438
Coetzee, J., Jones, R., Hill, M., 2014. Water hyacinth, Eichhornia crassipes
443
(Pontederiaceae), reduces benthic macroinvertebrate diversity in a protected
444
subtropical lake in South Africa. Biodivers. Conserv. 23, 1319-1330.
us
442
de Marco, P., Araujo, M.A.R., Barcelos, M.K., dos Santos, M.B.L., 2001. Aquatic
446
invertebrates associated with the water-hyacinth (Eichhornia crassipes) in an
447
eutrophic reservoir in tropical brazil. Stud. Neotrop. Fauna Environ. 36, 73-80.
451 452 453
M
Epstein, P., 1998. Nairobi - Weeds bring disease to the east African waterways. Lancet
d
450
sistemática y Biología. Tucumán, Fundación Miguel Lillo, 656p.
351, 577-577.
Ellenberg, D., Mueller-Dombois, D., 1974. Aims and methods of vegetation ecology.
te
449
Domínguez, E., Fernández, H.R., 2009. Macroinvertebrados bentónicos sudamericanos:
New York, NY: Wiley
Ac ce p
448
an
445
454
Gabriels, W., Lock, K., De Pauw, N., Goethals, P.L.M., 2010. Multimetric
455
Macroinvertebrate Index Flanders (MMIF) for biological assessment of rivers
456
and lakes in Flanders (Belgium). Limnologica 40,199-207.
457
Gelati, E., Madsen, H., Rosbjerg, D., 2011. Stochastic reservoir optimization using El
458
Niño information: case study of Daule Peripa, Ecuador. Hydrology Research,
459 460
42(5), 413-431.
Gerebizza,
E.,
2009.
The
Daule
Peripa
project.
Accessed
from:
461
http://eurodad.org/uploadedfiles/whats_new/reports/the%20daule%20peripa%20p
462
roject_crbm.pdf (Editor).
463
Giraldo, E., Garzon, A., 2002. The potential for water hyacinth to improve the quality of
464
Bogota River water in the Muna Reservoir: comparison with the performance of
465
waste stabilization ponds. Water Sci. Technol. 45, 103-110.
Page 16 of 28
466 467
Gupta, P., Roy, S., Mahindrakar, A.B., 2012. Treatment of Water Using Water Hyacinth, Water Lettuce and Vetiver Grass - A Review. Res. Environ. 2, 202-215. Hooper, D.U., Chapin III, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton,
469
J.H., Lodge, D.M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A.J.,
470
Van-dermeer, J., Wardle, D.A., 2005. Effects of biodiversity on ecosystem
471
functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35.
ip t
468
Jeppesen, E., Jensen, J., Sandergaard, M., Lauridsen, T., Pedersen, L., Jensen, L., 1997.
473
Top-down control in freshwater lakes: the role of nutrient state, submerged
474
macrophytes and water depth. Hydrobiologia 342-343, 151-164.
us
cr
472
Kouame, M.K., Dietoa, M.Y., Edia, E.O., Da Costa, S.K., Ouattara, A., Gourene, G.,
476
2011. Macroinvertebrate communities associated with macrophyte habitats in a
477
tropical man-made lake (Lake Taabo, Cote d'Ivoire). Knowl. Manage. Aquat.
478
Ecosys. 18.
an
475
Lowe, S., Browne, M., Boudjelas, S., De Poorter, M., 2000. 100 of the World’s Worst
480
Invasive Alien Species. A selection from the Global Invasive Species Database.
481
The Invasive Species Specialist Group (ISSG) a specialist group of the Species
482
Survival Commission (SSC) of the World Conservation Union (IUCN).
485 486 487 488 489 490 491 492 493 494
d
te
484
Lu, J.B., Wu, J.G., Fu, Z.H., Zhu, L., 2007. Water hyacinth in China: A sustainability science-based management framework. Environ. Manag. 40, 823-830.
Ac ce p
483
M
479
Malik, A., 2007. Environmental challenge vis a vis opportunity: The case of water hyacinth. Environ. Int. 33,122-138.
Margalef, R., 1968. Perspectives in Ecological Theory. University of Chicago Press, Chicago, IL, p.111.
Masifwa, W.F., Twongo, T., Denny, P., 2001. The impact of water hyacinth, Eichhornia crassipes
(Mart)
Solms
on
the
abundance
and
diversity of
aquatic
macroinvertebrates along the shores of northern Lake Victoria, Uganda. Hydrobiologia 452, 79-88.
McVea, C., Boyd, C.E., 1975. Effects of Waterhyacinth Cover on Water Chemistry, Phytoplankton and Fish in Ponds. pp. 375-378.
495
Meerhoff, M., Mazzeo, N., Moss, B., Rodriguez-Gallego, L., 2003. The structuring role
496
of free-floating versus submerged plants in a subtropical shallow lake. Aquat.
Page 17 of 28
497 498 499
Ecol. 37, 377-391. Menhinick, E., 1964. A comparison of some species-individuals diversity indices applied to samples of field insects. Ecology 45, 859-861. Olivares, E., Colonnello, G., 2000. Salinity gradient in the Ma´namo River, a dammed
501
distributary of the Orinoco Delta, and its influence on the presence of Eichhornia
502
crassipes and Paspalum repens. Interciencia, 25:242–248.
ip t
500
Opande, G.O., Onyango, J.C., Wagai, S.O., 2004. Lake victoria: The water hyacinth
504
(Eichhornia crassipes [Mart.] Solms), its socio-economic effects, control
505
measures and resurgence in the Winam gulf. Limnologica 34, 105-109.
us
cr
503
Parolin, P., Rudolph, B., Bartel, S., Bresch, C., Poncet, C., 2012. Worldwide Invasion
507
Pathways of the South American Eichhornia crassipes. Xxviii International
508
Horticultural Congress on Science and Horticulture for People (Ihc2010):
509
International Symposium on Advances in Ornamentals, Landscape and Urban
510
Horticulture 937, 1133-1140.
M
an
506
Rocha-Ramirez, A., Ramirez-Rojas, A., Chavez-Lopez, R., Alcocer, J., 2007.
512
Invertebrate assemblages associated with root masses of Eichhornia crassipes
513
(Mart.) Solms-Laubach 1883 in the Alvarado Lagoonal System, Veracruz,
514
Mexico. Aquat. Ecol. 41, 319-333.
516
te
Rodriguez, L.F., 2006. Can invasive species facilitate native species? Evidence of how,
Ac ce p
515
d
511
when, and why these impacts occur. Biol. Invasions 8, 927-939.
517
Rommens, W., Maes, J., Dekeza, N., Inghelbrecht, P., Nhiwatiwa, T., Holsters, E.,
518
Ollevier, F., Marshall, B., Brendonck, L., 2003. The impact of water hyacinth
519 520 521 522
(Eichhornia crassipes) in a eutrophic subtropical impoundment (Lake Chivero, Zimbabwe). I. Water quality. Arch. Hydrobiol. 158, 373-388.
Schramm, H.L., Jirka, K.J., Hoyer, M.V., 1987. Epiphytic macroinvertebrates on dominant macrophyes in two central Florida lakes. J. Freshw. Ecol. 4, 151-161.
523
Schultz, R., Dibble, E., 2012. Effects of invasive macrophytes on freshwater fish and
524
macroinvertebrate communities: the role of invasive plant traits. Hydrobiologia
525
684(1), 1-14.
526 527
Shannon, C., Wiener, W., 1949. The Mathematical Theory of Communication. Urbana University of Illinois Press, 177.
Page 18 of 28
528
Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688.
529
Sooknah, R.D., Wilkie, A.C., 2004. Nutrient removal by floating aquatic macrophytes
530
cultured in anaerobically digested flushed dairy manure wastewater. Ecol. Eng.
531
22, 27-42. Stiers, I., Crohain, N., Josens, G., Triest, L., 2011. Impact of three aquatic invasive
533
species on native plants and macroinvertebrates in temperate ponds. Biol.
534
Invasions 13, 2715-2726.
cr
ip t
532
Theel, H.J., Dibble, E.D., Madsen, J.D., 2008. Differential influence of a monotypic and
536
diverse native aquatic plant bed on a macroinvertebrate assemblage; an
537
experimental implication of exotic plant induced habitat. Hydrobiologia 600(1),
538
77-87.
an
us
535
Toft, J.D., Simenstad, C.A., Cordell, J.R., Grimaldo, L.F., 2003. The effects of
540
introduced water hyacinth on habitat structure, invertebrate assemblages, and fish
541
diets. Estuaries 26, 746-758.
M
539
Vilà, M., Espinar, J.L., Hejda, M., Hulme, P.E., Jarošík, V., Maron, J.L., Pergl, J.,
543
Schaffner, U., Sun, Y., Pyšek, P., 2011. Ecological impacts of invasive alien
544
plants: a meta-analysis of their effects on species, communities and ecosystems.
545
Ecol. Lett. 14, 702-708.
547 548
te
Villamagna, A.M., 2009. Ecological effects of water hyacinth (Eichhornia crassipes) on
Ac ce p
546
d
542
Lake Chapala, Mexico (Doctoral dissertation, Virginia Polytechnic Institute and State University).
549
Villamagna, A.M., Murphy, B.R., 2010. Ecological and socio-economic impacts of
550
invasive water hyacinth (Eichhornia crassipes): a review. Freshw. Biol. 55, 282-
551
298.
552
Villamagna, A.M., Murphy, B.R., Karpanty, S.M., 2012. Community-Level Waterbird
553
Responses to Water Hyacinth (Eichhornia crassipes). Invasive Plant Sci. Manage.
554 555 556 557 558
5, 353-362. Walker, P.D., Wijnhoven, S., van der Velde, G., 2013. Macrophyte presence and growth form influence macroinvertebrate community structure. Aquat. Bot. 104, 80-87. Wilson, J.R., Holst, N., Rees, M., 2005. Determinants and patterns of population growth in water hyacinth. Aquat. Bot. 81, 51-67.
Page 19 of 28
559
Zúñiga, M.C., Cardona, W., 2009. Water quality and environmental flow bioindicators,
560
(in Spanish). In: Cantera, J., Carvajal, Y., L., Castro, L. (Eds.) Environmental
561
flow: Concepts, experiences and challenges, Ed. Del Valle University, Cali,
562
Colombia, 167-198.
564
Zuur, A.F., Ieno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed effects
ip t
563
models and extensions in ecology with R. New York, Springer.
cr
565 566
us
567
Ac ce p
te
d
M
an
568
Page 20 of 28
Table legends
570
Table 1. Median, minimum and maximum values of the measured environmental
571
variables of the sampling sites with and without water hyacinth, statistical significance
572
levels between both are indicated (Mann-Whitney U-test).
ip t
569
Water hyacinth absent Water hyacinth present p-value Temperature (°C)
109
Total Dissolved Solids (mg.L-1)
0.05
0.05
0.07
pH
7.35
7.01
7.76
Chlorophyll (µg.L-1)
7.00
5.46
25.43
Chlorides (mg.L-1)
2.15
1.71
Dissolved Oxygen (mg.L-1)
8.04
6.47
Oxygen Saturation (%)
102
Turbidity (FTU)
5.20
0.63
73.50 36.50
89
0.07
us
70
27.37 25.62 29.74
0.05
0.06
0.20
7.11
6.69
7.90
0.06
6.92
4.70
10.10
0.61
2.52
2.48
1.26
2.90
0.10
10.69
7.22
4.14
9.09
0.22
M
an
0.04
81.33 130.60 89.80 51.77 117.82
0.22
3.96
0.00
9.52
3.66
1.13
5.32
te Ac ce p
575
79
Max
d
Conductivity (µS.cm )
574
Median Min
27.86 26.54 28.91 -1
573
Max
cr
Median Min
Page 21 of 28
576
Figure legends
577
Figure 1. Map of study area with indication of the sampling sites and the presence (blue)
579
or absence (red) of water hyacinth in the Daule-Peripa reservoir, Ecuador .
580
Figure 2. Picture taken at the Daule-Peripa dam of a site completely covered by water
581
hyacinth.
582
Figure 3. Generalized linear model for the presence/absence of water hyacinth in
583
function of turbidity.
584
Figure 4. Values of BMWP-Colombia (a), Margalef index (b), Shannon-Wiener index
585
(c), Simpson index (d) and Evenness (e) for different percentages of water hyacinth cover
586
(0=absent, 1=1-5%, 3= 25-50%, 4 = 50-75%, 5 = 75-100%).
587
Figure 5. Scatter plot of the Biological Monitoring Working Party-Colombia (BMWP-
588
Colombia) and the turbidity, with indication of the different vegetation cover classes of
589
water hyacinth (0=absent, 1=1-5%, 3= 25-50%, 4 = 50-75%, 5 = 75-100%).
590
Figure 6. Non-metric Multidimensional Scaling (NMDS) plot with indication of samples
591
with (big symbols) and without (small symbols) water hyacinth and the ecological water
592
quality class according to the Biological Monitoring Working Party (BMWP) Colombia.
594 595 596
cr
us
an
M
d
te
Ac ce p
593
ip t
578
Page 22 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 23 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 24 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 25 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 4
Page 26 of 28
Ac
ce
pt
ed
M
Figure 5
Page 27 of 28
Figure 6
e
Resemblance: S17 Bray Curtis similarity 2D Stress: 0.19
pt
ce Ac
Moderate
BMWP Colombia Poor Bad Very bad
Page 28 of 28
S0075-9511(15)00040-7 http://dx.doi.org/doi:10.1016/j.limno.2015.03.006 LIMNO 25452
To appear in: Received date: Revised date: Accepted date:
4-12-2014 26-3-2015 31-3-2015
Please cite this article as: Nguyen, T.H.T., Boets, P., Lock, K., Ambarita, M.N.D., Forio, M.A.E., Sasha, P., Granda, L.E.D., Hoang, T.H.T., Everaert, G., Goethals, P.L.M.,Habitat suitability of the invasive water hyacinth and its relation to water quality and macroinvertebrate diversity in a tropical reservoir, Limnologica (2015), http://dx.doi.org/10.1016/j.limno.2015.03.006 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Habitat suitability of the invasive water hyacinth and its relation to water quality
2
and macroinvertebrate diversity in a tropical reservoir
3
Tien Hanh Thi Nguyen
4
Ambarita1, Marie Ane Eurie Forio1, Peace Sasha1, Luis Elvin Dominguez Granda3, Thu
5
Huong Thi Hoang 4, Gert Everaert1, Peter L.M. Goethals1
6
1: Ghent University, Laboratory of Environmental Toxicology and Aquatic Ecology, J.
7
Plateaustraat 22, B-9000 Ghent, Belgium
8
2: Research Institute for Aquaculture No1, Dinh Bang, Tu Son, Bac Ninh, Vietnam
9
3: Escuela Superior Politécnica del Litoral (ESPOL), Water and Sustainable
10
Development Center (CADS), Campus Gustavo Galindo, Km. 30.5 Vía Perimetral,
11
Guayaquil - Ecuador
12
4: Hanoi University of Technology, Institute for Environmental Science and Technology,
13
No 1 Dai Co Viet, Hanoi, Vietnam
cr
ip t
, Pieter Boets1,*, Koen Lock1, Minar Naomi Damanik
14
te
d
M
an
us
1,2
*Author for correspondence: Pieter Boets, [email protected], Tel: +329 264 3776,
16
Fax: +329 264 4199
17 18
Ac ce p
15
Page 1 of 28
Abstract
20
In this study, we assessed the relationship between the occurrence of the invasive water
21
hyacinth (Eichhornia crassipes) and water quality properties as well as macroinvertebrate
22
diversity in a tropical reservoir, situated in western Ecuador. Macroinvertebrates and
23
physico-chemical water quality variables were sampled at 32 locations (during the dry
24
season of 2013) in both sites covered and non-covered by water hyacinth in the Daule-
25
Peripa reservoir. The results indicated that, in terms of water quality, only turbidity was
26
significantly different between sampling sites with and without water hyacinth (Mann-
27
Whitney U-test, p<0.01). The habitat suitability model showed that water hyacinth was
28
present at sites with a low turbidity. The percentage water hyacinth cover increased with
29
decreasing turbidity. The Biological Monitoring Working Party-Colombia score and the
30
Margalef diversity index were significantly higher (Mann-Whitney U-test, p<0.01) at
31
sampling sites where water hyacinth was present compared to water hyacinth absent sites.
32
However, there were no significant differences in the Shannon–Wiener index, Evenness
33
index and Simpson index between the sampling sites with and without water hyacinth.
34
Our results suggest that water hyacinth cover was an important variable affecting the
35
diversity of macroinvertebrates in the Daule-Peripa reservoir, with intermediate levels of
36
water hyacinth cover having a positive effect on the diversity of macroinvertebrates.
37
Information on the habitat suitability of water hyacinth and its effect on the physico-
38
chemical water quality and the macroinvertebrate community are essential to develop
39
conservation and management programs for large tropical reservoirs such as the Daule-
40
Peripa reservoir and the Guayas river basin, where water resources are being at high risk
41
due to expansion of agricultural and industrial development activities.
cr
us
an
M
d
te
Ac ce p
42
ip t
19
43
Keywords: Daule-Peripa reservoir, invasive species, water hyacinth, Eichhornia
44
crassipes, macroinvertebrates, water quality
45 46
Page 2 of 28
1. Introduction
48
Alien invasive plant species, mainly of terrestrial habitats, have recently received much
49
attention due to their direct or indirect impact on ecosystem structure and functioning
50
(Vilà et al., 2011). The impact of plant invasions in aquatic ecosystems has received less
51
attention. Several authors reported that invasive plant species can cause serious
52
ecological and economic impacts such as loss of native species diversity, hybridization
53
with native species, changes in ecosystem processes and functioning, and an increase of
54
pests and diseases (Rodriguez, 2006; Villamagna and Murphy, 2010; Stiers et al., 2011).
55
Alien invasive plant species may alter the available structure in aquatic habitats by
56
creating a shift to a homogeneous habitat, thereby negatively affecting biological
57
communities (Theel et al., 2008; Schultz and Dibble, 2012). Indeed, simplification of the
58
macrophyte growth form seems to negatively affect the abundance of biotic communities
59
such as macroinvertebrates (Walker et al., 2013). Schultz and Dibble (2012) found that
60
mechanisms underlying the impact of aquatic invasive plants are not very different from
61
native aquatic plants species. They identified three invasive traits largely responsible for
62
negative effects on biological communities: increased growth rate, allelopathic chemical
63
production and phenotypic plasticity. Despite the negative impact that is often observed,
64
positive relationships between invasion by alien macrophytes and macroinvertebrate
65
diversity have been observed as well (e.g. Brendonck et al., 2003; Villamagna, 2009).
66
Water hyacinth Eichhornia crassipes (Mart.) Solms is one of the most widely spread
67
invasive aquatic macrophytes in the world (Villamagna and Murphy, 2010). The species
68
has
69
Pantanal in western Brazil (Parolin et al., 2012). Water hyacinth is a free floating aquatic
70
plant, that can spread fast and has a strong growth (Malik, 2007). According to Malik
71
(2007), water hyacinth can double its size (area covered) in five days and a mat of
72
medium sized plants may contain two million plants per hectare that weigh 270 to 400
73
tons. Under favorable conditions the biomass of water hyacinth can be doubled within
74
only 12 days (Parolin et al., 2012). Nowadays, water hyacinth is distributed worldwide
75
(Parolin et al., 2012) and the International Union for Conservation of Nature listed the
76
species as one of the 100 most harmful invasive species (Lowe et al., 2000).
Ac ce p
te
d
M
an
us
cr
ip t
47
its
origin
in
the
Amazon
basin
and
in
the
vast
wetlands
of
the
Page 3 of 28
Water hyacinth tolerates a wide range of nutrients, temperature and pH levels and grows
78
in a wide variety of ecosystem types (Malik, 2007). Environmental factors such as
79
temperature, pH, solar radiation, and salinity of the water can influence the growth and
80
performance of water hyacinth (Gupta et al., 2012). The optimum conditions for growth
81
of water hyacinth are a pH between 6 to 8 (Malik, 2007) and a temperature that ranges
82
between 28 and 30°C (Gupta et al., 2012). A study by Wilson et al. (2005) showed that
83
nutrient concentrations and temperature are two of the most important factors
84
determining water hyacinth growth and reproduction. Salinity on the other hand can
85
cause major constraints for water hyacinth growth, since water hyacinth can not survive
86
at a salinity higher than 2‰ (Olivares and Colonnello, 2000).
87
Water hyacinth can have an effect on the physical and chemical composition of water.
88
The introduction of water hyacinth can cause a change in water clarity, hydrological
89
regime, dissolved oxygen concentration, nutrient concentrations and other pollutants in
90
the water body (Villamagna and Murphy, 2010). Negative as well as positive effects have
91
been recorded. Water hyacinth is known to reduce dissolved oxygen concentrations
92
because the mats prevent the transfer of oxygen from the air to the water surface, while
93
the plant does not release oxygen into the water (Meerhoff et al., 2003). Water hyacinth
94
has the potential to stabilize pH and temperature and prevent stratification in lotic
95
systems (Giraldo and Garzon, 2002) and in this way change ecosystem structure and
96
functioning. Rommens et al. (2003) and Sooknah and Wilkie (2004) reported that water
97
hyacinth has the capacity to absorb nutrients (e.g. nitrate, ammonium, phosphate) from
98
the water column affecting phyto- and zooplankton abundance.
99
Water hyacinth plays a important role for phytoplankton, zooplankton and fish in
100
freshwater ecosystems by providing habitat complexity, shelter and feeding grounds
101
(Brendonck et al., 2003; Meerhoff et al., 2003; Villamagna and Murphy, 2010).
102
According to de Marco et al. (2001) and Brendonck et al. (2003) the roots and the leaves
103
of water hyacinth offer an important substratum and habitat for macroinvertebrate
104
colonisation. Indeed, a positive relationship between macroinvertebrate diversity and the
105
cover of water hyacinth has been found (Schramm et al., 1987; Kouame et al., 2011).
106
Kouame et al. (2011) found a high diversity and density of macroinvertebrate
Ac ce p
te
d
M
an
us
cr
ip t
77
Page 4 of 28
107
assemblages associated with root masses of water hyacinth in Lake Taabo (Ivory Coast).
108
They ascribed this to improvements in physico-chemical properties (e.g. conductivity,
109
nutrients,
110
macroinvertebrate community. Moreover, Bailey and Litterick (1993) also found a
111
positive effect of water hyacinth on dissolved oxygen and a variety of potential food
112
resources for aquatic invertebrates in water hyacinth root-mats resulting in higher
113
abundances of macroinvertebrates. However, water hyacinth limits the development of
114
phytoplankton via preventing light penetration and absorbing nutrients which reduces
115
phytoplankton and leads to a decrease in zooplankton abundance (Villamagna and
116
Murphy, 2010). In addition, water hyacinth causes significant ecological alterations in the
117
invaded community by modifying the habitat, disrupting the food chain and nutrient
118
cycling, and consequently changing invertebrate and fish assemblage structure and finally
119
the entire food web (Brendonck et al., 2003; Toft et al., 2003; Coetze et al., 2014). Hence,
120
water hyacinth influences species richness, diversity and composition of invaded
121
communities and causes huge impacts to ecosystems structure and functioning
122
(Villamagna and Murphy, 2010).
123
Next to its ecological impact, water hyacinth has attracted global attention due to serious
124
problems caused to power plants, navigation, irrigation and recreation (Epstein, 1998; Lu
125
et al., 2007; Malik, 2007). According to Malik (2007), many large hydropower plants
126
have to spend considerable time and money in clearing water hyacinth in order to prevent
127
it from entering in the turbines and causing damage and power interruptions. In China,
128
water hyacinth is reported to clog lakes and rivers, impede water flows, obstruct
129
navigation, and damage irrigation and hydroelectricity facilities, thus the total cost for
130
water hyacinth control is estimated to be more than $12.35 million each year (Lu et al.,
131
2007). Indeed, the invasion of water hyacinth causes huge ecological and economic
132
consequences worldwide (Lu et al., 2007). However, their effects to human society vary
133
according to the magnitude of invasion, the ecosystem services of the water body and the
134
response of water hyacinth to management efforts (Villamagna and Murphy, 2010).
135
There have been several studies carried out on the effect of water hyacinth on water
136
quality and aquatic diversity, the invasion mechanism and the utilization in temperate
turbidity),
which
positively
influenced
the
Ac ce p
te
d
M
an
us
cr
ip t
temperature,
Page 5 of 28
regions (Villamagna and Murphy, 2010). However, until now the habitat suitability of
138
water hyacinth and the relationship between water hyacinth and ecological communities
139
largely remains understudied in tropical reservoirs. Therefore, the aim of this study was
140
to determine the habitat suitability of water hyacinth and to assess the effect of water
141
hyacinth cover on the chemical water quality and on the macroinvertebrate community in
142
a tropical reservoir. This information is essential to develop conservation and
143
management programs for tropical reservoirs and specifically the Guayas river basin,
144
situated in western Ecuador, where water resources are being at high risk due to
145
expansion of agricultural and industrial development activities.
146
2. Materials and Methods
147
2.1 Study area
148
The Daule-Peripa reservoir is located in the Guayas river basin in the western part of
149
Ecuador (Fig. 1). Hydropower and agriculture (banana, rice, maize, African oil palm
150
(Elaeis guineensis) and cacao) are the main economic sectors of the Guayas basin
151
(Alvarez-Mieles et al., 2013). The reservoir which was constructed in 1987 receives
152
water from the Daule and Peripa rivers (Gelati et al., 2011), and provides hydropower
153
generation, irrigation, flood protection and drinking water supply (Arriaga, 1989). The
154
reservoir has a water storage capacity of 6,000 million m³, a maximum surface area of
155
approximately 30,000 ha and a strongly fluctuating water level between 70-85 m in
156
difference due to the operation of the dam (CELEC, 2013).
cr
us an
M
d
te
Ac ce p
157
ip t
137
158
2.2 Sampling
159
The physio-chemical water variables and macroinvertebrate samples were collected in the
160
Daule-Peripa reservoir during the dry season (November) of 2013. In total, 32 samples
161
were taken. Eleven samples were taken at non-vegetated sites and 21 sites were
162
characterised by vegetation (the presence of water hyacinth) (Fig. 1 and Fig. 2). At the
163
vegetated sites, water hyacinth was the only floating macrophyte occurring. Only at one
164
vegetated site a native emergent plant species (Sagittaria spec.) occurred at very low
Page 6 of 28
abundance, at all other sites no native plants species were recorded. The percentage of
166
water hyacinth cover was visually estimated from the bank along a transect of 100 m
167
where the site where macroinvertebrates and water samples were taken was considered as
168
the centre. The vegetation cover classes were divided as follows (based on the Braun-
169
Blanquet cover/abundance scale): 0 = non-vegetated/absent, 1 = 1-5% (rare), 2 = 5-25%
170
(occasional), 3 = 25-50% (frequent), 4 = 50-75% (common) and 5 = 75-100%
171
(abundant). Classes rather than exact values for cover of water hyacinth were used as this
172
yields more reliable measures (Ellenberg and Mueller-Dombois, 1974). The 21 vegetated
173
sites were classified, based on the vegetation cover, as follows: class 0: 11 sites, class 1: 5
174
sites, class 2: 0 sites, class 3: 2 sites, class 4: 3 sites and class 5: 11 sites. Values of pH,
175
Dissolved Oxygen (DO) (mg.L-1), Chlorophyll (µg.L-1), Chloride (mg.L-1), Turbidity
176
(FTU), Conductivity (µS.cm-1) and Total Dissolved Solids (TDS) (mg.L-1) were
177
measured at the surface water layer with a multiprobe (model YSI 6600 V2 and YSI 6600
178
V1, YSI manufacturer). Moreover, water samples from each sampling location were
179
collected and stored in plastic bottles (1l), kept cool and dark immediately after collection
180
and transported to the laboratory for further analysis. At the laboratory, the water samples
181
were kept in a refrigerator (cool and dark) for nutrient analysis. The Hach Lange GMBH
182
photometric method was used to determine chemical oxygen demand (COD) (mg L-1),
183
total phosphorus (TP) (mg P.L-1), ammonium (mg N.L-1), nitrate (mg N.L-1), nitrite (mg
184
N.L-1) and total nitrogen (TN) (mg N.L-1). However, the values of all variables analysed
185
at the laboratory except for ammonium were below the detection limit of the Hach Lange
186
GMBH. Consequently, those variables were not included in the analysis. The lowest
187
detection limit of COD was 5 mg/L, TN was 1 mg/L, TP was 0.5 mg/L, NO3- -N was
188
0.23 mg/L, NO2- -N was 0.015 mg/L and NH4+-N was 0.015 mg/L.
cr
us
an
M
d
te
Ac ce p
189
ip t
165
190
Macroinvertebrates were collected by standard hand net consisting of a metal frame
191
holding a conical net (mesh-size 300 μm) at the same sites where water quality was
192
measured. Macroinvertebrates were collected during 10 minutes sampling including all
193
different microhabitats present at the sampling site (Gabriels et al., 2010). At the non-
194
vegetated sites, samples were collected by vigorously sweeping the net along the
195
reservoir margins and by disturbing the bank substratum at wading depth. For those sites
Page 7 of 28
that contained macrophytes, macroinvertebrates were obtained by submerging the
197
handnet under the roots of the macrophytes and by disturbing the stalks and leaves.
198
Afterwards, the net was quick and carefully lifted out of the water to prevent the escape
199
of mobile organisms. Macroinvertebrates attached to the roots of water hyacinth were
200
manually collected. Similarly macroinvertebrates were collected from stones and other
201
substrates at both vegetated and non-vegetated sites (Gabriels et al., 2010). Samples
202
collected were sieved (mesh size of 500µm) in the laboratory and sorted in white trays.
203
Samples from each location were placed in separate small plastic vials containing 80%
204
ethanol for preservation. After sorting, organisms were counted and identified under a
205
stereomicroscope. Macroinvertebrates were identified to family level using the
206
identification keys developed by Domínguez and Fernández (2009). The biotic
207
macroinvertebrate index BMWP-Colombia (Biological Monitoring Working Party-
208
Colombia) was calculated according to the modified method proposed by Zúñiga and
209
Cardona (2009). The BMWP-Colombia was calculated per site based on a summation of
210
all macroinvertebrate taxon scores. Each macroinvertebrate taxon receives a score that
211
reflects its susceptibility to pollution, where pollution-intolerant taxa receive high scores,
212
whereas pollution-tolerant taxa are given low scores (Zúñiga and Cardona, 2009). The
213
total score for each site indicates water quality categories ranging from very bad (0-15),
214
bad (16-35), poor (36-60), moderate (61-100) to good (>100). Moreover, the Shannon-
215
Wiener Diversity Index (Shannon and Wiener, 1949), Evenness index (Menhinick, 1964),
216
Simpson’s diversity index (Simpson, 1949) and Margalef index (Margalef, 1968) were
217
calculated to assess the taxa evenness and taxa richness for each sampling site. A
218
combination of several diversity and biotic indices was calculated in order to take
219
advantages of the strengths of each and develop a more complete understanding of
220
community structure (Hooper et al., 2005).
cr
us
an
M
d
te
Ac ce p
221
ip t
196
222
2.3 Data analysis
223
Generalized linear models were developed in R (version 3.0.3, The R Project for
224
Statistical
225
determined the occurrence of water hyacinth. Prior to developing the GLM, correlations
226
and variance inflation factors (VIFs) between the predictor variables were examined, in
Computing,
http://www.r-project.org) to
investigate which variables
Page 8 of 28
order to avoid problems of collinearity (Zuur et al., 2009). All variables with a correlation
228
of 0.7 or higher were removed. The presence or absence of water hyacinth was the
229
predicted variable, whereas temperature, conductivity, chlorophyll, oxygen saturation and
230
turbidity were retained as predictor variables. We started fitting the full model, i.e.
231
including all candidate predictor variables. Next, a stepwise backward selection
232
procedure was followed based on the Akaike Information Criterion (AIC), where the
233
model with the lowest AIC value was retained as the final model. Relations between the
234
residuals (the differences between observations and predictions by the retained model)
235
and predictor variables were evaluated and the normality of the residuals was tested using
236
a QQ-plot (probability plot). Retained models were only considered reliable if no
237
relations between the residuals and the predictor variables were visually observed and
238
residuals were normally distributed (Zuur et al., 2009); the retained models were rejected
239
otherwise.
240
Non-parametric tests were performed, since the data were non-normal distributed (tested
241
with Shapiro-Wilk Normality test). A Mann-Whitney U-test was used to compare
242
physico-chemical variables between sampling sites with and without water hyacinth.
243
Spearman’s rank correlation was used to explore the relationships between percentage
244
water hyacinth cover and macroinvertebrate metrics. A Kruskal-Wallis ANalysis Of
245
VAriance (ANOVA) followed by post-hoc multiple comparisons was performed to test
246
whether significant differences in ecological indices existed between different classes of
247
vegetation cover. Species data were log(x+1) transformed before multivariate analysis to
248
ensure data normality (Clarke, 1993). Non-metric Multidimensional Scaling (NMDS)
249
was used to visualize the similarity of macroinvertebrate communities between habitats
250
types onto two-dimensional charts. ANOSIM (ANalysis Of SIMilarities) was used for
251
testing the similarities (based on the Bray-Curtis similarity) of macroinvertebrate
252
community composition between sites with and without water hyacinth. A SIMPER
253
(SIMilarity PERcentages) analysis was applied for identifying which species primarily
254
contributed to the observed differences in macroinvertebrate assemblages between habitat
255
types. All multivariate analyses were performed using the PRIMER software package
256
(Clarke, 1993).
Ac ce p
te
d
M
an
us
cr
ip t
227
Page 9 of 28
3. Results
258
3.1 Physico-chemical water quality
259
Table 1 shows the water quality results measured at the Daule-Peripa reservoir. Based on
260
a Mann-Whitney U-test it was found that only turbidity differed significantly between
261
sampling sites with and without water hyacinth (Table 1). Based on the low conductivity
262
and high oxygen levels, a relatively good water quality was observed for most sampling
263
sites, regardless of the presence or absence of water hyacinth.
us
264
cr
ip t
257
3.2. Habitat suitability of water hyacinth
266
After model selection only turbidity was retained in the final model (Appendix A). Based
267
on the generalized linear model we found that water hyacinth is present at sites with a
268
low turbidity (Fig. 3). The percentage water hyacinth cover increased with decreasing
269
turbidity (Appendix A, Fig. A1).
M
d te
270
an
265
3.3 Relationship between the macroinvertebrate community and habitat characteristics
272
Based on the BMWP-Colombia scores the sampling sites of the Daule-Peripa reservoir
273
were categorized in four water quality classes: moderate, poor, bad and very bad
274
(Appendix B, Fig. B1). The BMWP-Colombia score and Margalef index were
275
significantly higher (Mann-Whitney U-test, p<0.01) for sites containing water hyacinth
276
compared to sites containing none. However, there were no distinct differences in the
277
Shannon–Wiener index, Evenness index and Simpson index between the sampling sites
278
with and without water hyacinth (all p > 0.05).
279
A Spearman rank-order correlation was run to determine the relationship between
280
percentage of cover by water hyacinth and the ecological indices. There was a significant,
281
positive correlation between percentage cover by water hyacinth and the BMWP-
282
Colombia and Margalef index (r = 0.64, r = 0.54 respectively; p < 0.05). Based on a
283
Kruskal-Wallis ANOVA a significant difference was detected in BMWP-Colombia
Ac ce p
271
Page 10 of 28
scores (chi-squared = 13.8, df = 4, p = 0.008) and in Margalef index between different
285
vegetation classes (chi-squared = 12.6, df = 4, p = 0.01). Based on post-hoc tests we
286
found a significant difference for the BMWP-Colombia index between vegetation class 0
287
and class 5 and for the Margalef index between vegetation class 3 and all other classes (p
288
< 0.05) (Fig. 4). When turbidity was plotted in function of the BMWP-Colombia,
289
samples with a high percentage water hyacinth cover were clearly separated from sites
290
without water hyacinth or sites with a low cover (Fig. 5). Sites with a lower turbidity had
291
a higher BMWP-Colombia score.
292
In total, 32 different macroinvertebrate taxa were found at the different sites with a
293
maximum of 19 different taxa for sites without water hyacinth and a maximum of 25 taxa
294
for sites covered by water hyacinth (Appendix C). Sites with an intermediate vegetation
295
cover (25-50%) had a relatively high abundance and diversity of macroinvertebrate taxa.
296
SIMPER (SIMilarity PERcentages) analysis indicated that the similarity in species
297
composition between samples with water hyacinth was 49.4%, while similarity between
298
samples without water hyacinth was 29.6%. Dissimilarity in species composition between
299
samples with and without water hyacinth was 67.6%. Densities of Chironomidae,
300
Hyallelidae,
301
Coenagrionidae, Tubificidae, Mesoveliidae and Caenidae reached higher densities at
302
sampling sites with water hyacinth, while Glossiphoniidae, Thiaridae, Gerridae and
303
Corixidae reached higher densities at sampling sites without water hyacinth (Appendix
304
C). ANOSIM (ANalysis Of SIMilarity) indicated that there was a significant difference
305
in macroinvertebrate community composition between sites with and without water
306
hyacinth (R=0.47; p<0.01). Non-metric MultiDimensional Scaling (NMDS) also
307
indicated that samples with water hyacinth were more similar to each other than samples
308
without water hyacinth (Fig. 6).
Dugesiidae,
Libellulidae,
Baetidae,
Notonectidae,
Ac ce p
te
Prostigmata,
d
M
an
us
cr
ip t
284
309 310
4. Discussion
311
This study is one of the few studies investigating the relationship between the occurrence
312
of the invasive water hyacinth and physico-chemical water quality conditions and
Page 11 of 28
macroinvertebrate diversity in a tropical reservoir. In terms of physico-chemical
314
characteristics, our results show that only turbidity differed significantly between
315
sampling sites with and without water hyacinth. Water hyacinth reduces the effect of
316
waves caused by the wind and motorboats, which explains the lower average turbidities
317
that were observed at stations containing water hyacinth in the Daule-Peripa reservoir.
318
On the other hand, water hyacinth will also more easily establish at sites where the wave
319
action is not too strong, where there is a low flow velocity and where consequently
320
turbidity is low as well (Opande et al., 2004). Our habitat suitability model, indicating
321
that water hyacinth is present at sites with a low turbidity, supports this finding. The
322
lower average turbidity measured at sites where water hyacinth was presented is also
323
supported by Jeppesen et al. (1997) because it was found that water hyacinth has the
324
capacity to stabilise the sediment and thus reduce turbidity. There was no significant
325
difference in average chlorophyll concentrations between covered and non-covered sites,
326
however, the maximum concentration measured at water hyacinth absent sites was more
327
than two times higher than that at sites where water hyacinth was present. This may be
328
attributed to water hyacinth absorbing nutrients from the water column and reducing light
329
penetration when forming dense vegetation mats (McVea and Boyd, 1975), thus limiting
330
the productivity of phytoplankton at sites where water hyacinth is present (Malik, 2007;
331
Villamagna and Murphy, 2010). Dissolved oxygen concentration, although not
332
significantly different, was lower at sites where water hyacinth is present compared to
333
sites where the species is absent. The lower oxygen levels measured, can be ascribed to
334
the water hyacinth mats that prevent the transfer of oxygen from the air to the water and
335
the reduced effects of mixing due to wind (Meerhoff et al., 2003; Villamagna, 2009).
336
A positive relationship between the occurrence of water hyacinth and macroinvertebrate
337
diversity was observed. Higher BMWP-Colombia and Margalef scores were reported for
338
sites containing water hyacinth compared to locations without water hyacinth. The
339
Margalef index reveals that sites covered for 25-50% with water hyacinth have a higher
340
species diversity compared to water hyacinth absent sites. Similarly, Kouame et al.
341
(2011) found that the vegetation cover by water hyacinth positively altered the
342
biodiversity of benthic invertebrate assemblages. Indeed, macrophytes can provide
343
excellent
Ac ce p
te
d
M
an
us
cr
ip t
313
microhabitats
that
promote
the
establishment
and
colonization
of
Page 12 of 28
macroinvertebrates. Furthermore, macroinvertebrates use macrophytes for refuge and
345
shelter against predation (Walker et al., 2013). In addition, the root mat of water hyacinth
346
can provide new habitats for colonization by macroinvertebrates and in this way increase
347
the diversity of aquatic macroinvertebrates (Masifwa et al., 2001; Rocha-Ramirez et al.,
348
2007; Barker et al., 2014). When the percentage of water hyacinth cover was higher than
349
50% we did not observe an increase in positive effect, because the cover might be too
350
dense, which negatively affects the physico-chemical water quality conditions and
351
consequently also the diversity of macroinvertebrates. This may be explained by the
352
intermediate cover hypothesis that was put forward by Villamagna (2009) to explain the
353
observed patterns in invertebrate abundance and diversity of Lake Chapala in Mexico.
354
This hypothesis suggests that the biotic community (e.g. fish) increases with increasing
355
prey abundance and provision of refuge as a consequence of increased vegetation.
356
However, a too high vegetation density may reduce oxygen availability and increase the
357
competition between organisms thereby negatively affecting the density of the biotic
358
community.
359
Our study showed that the dissimilarity in species composition between samples with and
360
without water hyacinth was 67.6%, indicating differences in the taxa associated with
361
water hyacinth. At sampling sites where water hyacinth was present, both pollution
362
tolerant (e.g. Tubificidae and Chironomidae) and pollution sensitive (e.g. Hyallelidae,
363
Baetidae and Coenagrionidae) families reached higher abundances compared to sites
364
where water hyacinth was absent. According to Rocha-Ramirez et al. (2007), the density
365
of invertebrates is not only affected by the presence of water hyacinth, but also by
366
environmental variables such as temperature, salinity, dissolved oxygen, and turbidity. In
367
addition, particular groups of invertebrates are correlated with specific variables (e.g.
368
mayflies prefer a lower conductivity) (Rocha-Ramirez et al., 2007) and can respond
369
differently to water hyacinth (Villamagna and Murphy, 2010). In our study, it seems that
370
the measured environmental conditions in combination with the presence of water
371
hyacinth positively influenced the diversity and abundance of macroinvertebrates.
372
Water hyacinth is considered one of the world’s worst weeds (Lowe et al., 2000). The
373
explosive growth rate of water hyacinth has caused serious socio-economic consequences
Ac ce p
te
d
M
an
us
cr
ip t
344
Page 13 of 28
such as obstruction of water ways, reducing hydropower production, etc. which often
375
requires immediate action (Masifwa et al., 2001). Our study showed that water hyacinth
376
is positively related to macroinvertebrate diversity and water quality. This benefit should
377
be weighed and compared to other impacts on ecosystem services before management
378
actions are initiated. It was reported that about one third of the Daule-Peripa reservoir is
379
covered by water hyacinth and that the presence of water hyacinth causes blockage of
380
navigation between different communities in the reservoir and increases the incidence of
381
infectious diseases (e.g. malaria, dengue and mosquito born infectious diseases)
382
(Gerebizza, 2009). However, studies of water hyacinth effects on local economy (e.g.
383
cost for removal of water hyacinth, cost of local people for transportation by boats,
384
medical treatment and the problems with hydropower generation) in the Daule-Peripa
385
reservoir have not been assessed. Small and isolated mats of water hyacinth can probably
386
provide a unique habitat that contributes to overall biotic diversity and ecosystem
387
functioning (Villamagna and Murphy, 2010). In case of the Daule-Peripa reservoir, we
388
found that an intermediate vegetation cover of water hyacinth was positively related with
389
the diversity of macroinvertebrates, whereas the impact of water hyacinth on hydropower
390
generation is expected to be limited. However, further investigation is needed to quantify
391
these findings and to assess the cost and benefits related to the presence of water
392
hyacinth. Integrating the ecological knowledge obtained from this study with economic
393
assessments and public perception could help decision makers to identify priority habitats
394
to be targeted for the control of water hyacinth and to prioritize conservation actions in an
395
operative way (Caplat and Coutts, 2011).
396
In conclusion, the findings of this study revealed that only turbidity was significantly
397
different between sampling sites with and without water hyacinth. The presence of water
398
hyacinth was an important variable affecting the diversity of macroinvertebrates in the
399
Daule-Peripa
400
macroinvertebrate diversity and the occurrence of water hyacinth it is important to
401
consider the socio-economic costs related to the management of water hyacinth. Work
402
conducted by Villamagna (2009) and Villamagna et al. (2012) on the effect of water
403
hyacinth cover on bird diversity and abundance found evidence for a more indirect
404
influence on waterbirds via changes in trophic structure, prey community composition,
Ac ce p
te
d
M
an
us
cr
ip t
374
reservoir.
Despite
the
overall
positive
relationship
between
Page 14 of 28
and energy flows throughout the system. This work is believed to serve as baseline data
406
for further studies on tropical reservoirs and as an incentive to assess the indirect effect of
407
water hyacinth on other communities (e.g. fish, phytoplankton and zooplankton).
408
Acknowledgments
409
The first author is grateful for the financial support for this work from Belgian Technical
410
Cooperation (BTC) in the framework of her PhD grant entitled “Ecological impact
411
assessment of hydropower generation on river systems”. Pieter Boets is supported by a
412
postdoctoral fellowship from the Fund for Scientific Research (FWO Vlaanderen,
413
Belgium). We would like to thank for the support provided by the VLIR Ecuador
414
Biodiversity Network project and the assistance of CADS-ESPOL for the help during the
415
fieldwork.
an
us
cr
ip t
405
416
References
418
Alvarez-Mieles, G., Irvine, K., Griensven, A.V., Arias-Hidalgo, M., Torres, A., Mynett,
419
A.E., 2013. Relationships between aquatic biotic communities and water quality
420
in a tropical river-wetland system (Ecuador). Environ. Sci. Pol. 34, 115-127.
421
Arriaga, L., 1989. The Daule- Peripe Dam Project, Urban Development of Guayaquil and
422
Their Impact on Shrimp Mariculture. In: Olsen, S. and Arriaga, L., editors. A
424
d
te
Ac ce p
423
M
417
Sustainable Shrimp Mariculture Industry for Ecuador. Narragansett, RI: Coastal Resources Center, University of Rhode Island.
425
Bailey, R.G., Litterick, M.R., 1993. The Macroinvertebrate Fauna of Water Hyacinth
426
Fringes in the Sudd Swamps (River Nile, Southern Sudan). Hydrobiologia 250,
427 428 429
97-103.
Barker, J.E., Hutchens, J.J., Luken, J.O., 2014. Macroinvertebrates Associated with Water Hyacinth Roots and a Root Analog. Freshw. Sci. 33, 159-167.
430
Brendonck, L., Maes, J., Rommens, W., Dekeza, N., Nhiwatiwa, T., Barson, M.,
431
Callebaut, V., Phiri, C., Moreau, K., Gratwicke, B., Stevens, M., Alyn, N.,
432
Holsters, E., Ollevier, F., Marshall, B., 2003. The impact of water hyacinth
433
(Eichhornia crassipes) in a eutrophic subtropical impoundment (Lake Chivero,
434
Zimbabwe). II. Species diversity. Arch. Hydrobiol. 158, 389-405.
Page 15 of 28
435
Caplat, P., Coutts, S., 2011. Integrating Ecological Knowledge, Public Perception and
436
Urgency of Action into Invasive Species Management. Environ. Manage. 48, 878-
437
881.
440 441
accesed on17/10/2014. In Spanish.
ip t
439
CELEC, 2013. Revista 25 Anos de la presa Daule-peripa. http://www.hidronacion.org,
Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community structure. Australian J. Ecol. 18, 117-143.
cr
438
Coetzee, J., Jones, R., Hill, M., 2014. Water hyacinth, Eichhornia crassipes
443
(Pontederiaceae), reduces benthic macroinvertebrate diversity in a protected
444
subtropical lake in South Africa. Biodivers. Conserv. 23, 1319-1330.
us
442
de Marco, P., Araujo, M.A.R., Barcelos, M.K., dos Santos, M.B.L., 2001. Aquatic
446
invertebrates associated with the water-hyacinth (Eichhornia crassipes) in an
447
eutrophic reservoir in tropical brazil. Stud. Neotrop. Fauna Environ. 36, 73-80.
451 452 453
M
Epstein, P., 1998. Nairobi - Weeds bring disease to the east African waterways. Lancet
d
450
sistemática y Biología. Tucumán, Fundación Miguel Lillo, 656p.
351, 577-577.
Ellenberg, D., Mueller-Dombois, D., 1974. Aims and methods of vegetation ecology.
te
449
Domínguez, E., Fernández, H.R., 2009. Macroinvertebrados bentónicos sudamericanos:
New York, NY: Wiley
Ac ce p
448
an
445
454
Gabriels, W., Lock, K., De Pauw, N., Goethals, P.L.M., 2010. Multimetric
455
Macroinvertebrate Index Flanders (MMIF) for biological assessment of rivers
456
and lakes in Flanders (Belgium). Limnologica 40,199-207.
457
Gelati, E., Madsen, H., Rosbjerg, D., 2011. Stochastic reservoir optimization using El
458
Niño information: case study of Daule Peripa, Ecuador. Hydrology Research,
459 460
42(5), 413-431.
Gerebizza,
E.,
2009.
The
Daule
Peripa
project.
Accessed
from:
461
http://eurodad.org/uploadedfiles/whats_new/reports/the%20daule%20peripa%20p
462
roject_crbm.pdf (Editor).
463
Giraldo, E., Garzon, A., 2002. The potential for water hyacinth to improve the quality of
464
Bogota River water in the Muna Reservoir: comparison with the performance of
465
waste stabilization ponds. Water Sci. Technol. 45, 103-110.
Page 16 of 28
466 467
Gupta, P., Roy, S., Mahindrakar, A.B., 2012. Treatment of Water Using Water Hyacinth, Water Lettuce and Vetiver Grass - A Review. Res. Environ. 2, 202-215. Hooper, D.U., Chapin III, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton,
469
J.H., Lodge, D.M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A.J.,
470
Van-dermeer, J., Wardle, D.A., 2005. Effects of biodiversity on ecosystem
471
functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35.
ip t
468
Jeppesen, E., Jensen, J., Sandergaard, M., Lauridsen, T., Pedersen, L., Jensen, L., 1997.
473
Top-down control in freshwater lakes: the role of nutrient state, submerged
474
macrophytes and water depth. Hydrobiologia 342-343, 151-164.
us
cr
472
Kouame, M.K., Dietoa, M.Y., Edia, E.O., Da Costa, S.K., Ouattara, A., Gourene, G.,
476
2011. Macroinvertebrate communities associated with macrophyte habitats in a
477
tropical man-made lake (Lake Taabo, Cote d'Ivoire). Knowl. Manage. Aquat.
478
Ecosys. 18.
an
475
Lowe, S., Browne, M., Boudjelas, S., De Poorter, M., 2000. 100 of the World’s Worst
480
Invasive Alien Species. A selection from the Global Invasive Species Database.
481
The Invasive Species Specialist Group (ISSG) a specialist group of the Species
482
Survival Commission (SSC) of the World Conservation Union (IUCN).
485 486 487 488 489 490 491 492 493 494
d
te
484
Lu, J.B., Wu, J.G., Fu, Z.H., Zhu, L., 2007. Water hyacinth in China: A sustainability science-based management framework. Environ. Manag. 40, 823-830.
Ac ce p
483
M
479
Malik, A., 2007. Environmental challenge vis a vis opportunity: The case of water hyacinth. Environ. Int. 33,122-138.
Margalef, R., 1968. Perspectives in Ecological Theory. University of Chicago Press, Chicago, IL, p.111.
Masifwa, W.F., Twongo, T., Denny, P., 2001. The impact of water hyacinth, Eichhornia crassipes
(Mart)
Solms
on
the
abundance
and
diversity of
aquatic
macroinvertebrates along the shores of northern Lake Victoria, Uganda. Hydrobiologia 452, 79-88.
McVea, C., Boyd, C.E., 1975. Effects of Waterhyacinth Cover on Water Chemistry, Phytoplankton and Fish in Ponds. pp. 375-378.
495
Meerhoff, M., Mazzeo, N., Moss, B., Rodriguez-Gallego, L., 2003. The structuring role
496
of free-floating versus submerged plants in a subtropical shallow lake. Aquat.
Page 17 of 28
497 498 499
Ecol. 37, 377-391. Menhinick, E., 1964. A comparison of some species-individuals diversity indices applied to samples of field insects. Ecology 45, 859-861. Olivares, E., Colonnello, G., 2000. Salinity gradient in the Ma´namo River, a dammed
501
distributary of the Orinoco Delta, and its influence on the presence of Eichhornia
502
crassipes and Paspalum repens. Interciencia, 25:242–248.
ip t
500
Opande, G.O., Onyango, J.C., Wagai, S.O., 2004. Lake victoria: The water hyacinth
504
(Eichhornia crassipes [Mart.] Solms), its socio-economic effects, control
505
measures and resurgence in the Winam gulf. Limnologica 34, 105-109.
us
cr
503
Parolin, P., Rudolph, B., Bartel, S., Bresch, C., Poncet, C., 2012. Worldwide Invasion
507
Pathways of the South American Eichhornia crassipes. Xxviii International
508
Horticultural Congress on Science and Horticulture for People (Ihc2010):
509
International Symposium on Advances in Ornamentals, Landscape and Urban
510
Horticulture 937, 1133-1140.
M
an
506
Rocha-Ramirez, A., Ramirez-Rojas, A., Chavez-Lopez, R., Alcocer, J., 2007.
512
Invertebrate assemblages associated with root masses of Eichhornia crassipes
513
(Mart.) Solms-Laubach 1883 in the Alvarado Lagoonal System, Veracruz,
514
Mexico. Aquat. Ecol. 41, 319-333.
516
te
Rodriguez, L.F., 2006. Can invasive species facilitate native species? Evidence of how,
Ac ce p
515
d
511
when, and why these impacts occur. Biol. Invasions 8, 927-939.
517
Rommens, W., Maes, J., Dekeza, N., Inghelbrecht, P., Nhiwatiwa, T., Holsters, E.,
518
Ollevier, F., Marshall, B., Brendonck, L., 2003. The impact of water hyacinth
519 520 521 522
(Eichhornia crassipes) in a eutrophic subtropical impoundment (Lake Chivero, Zimbabwe). I. Water quality. Arch. Hydrobiol. 158, 373-388.
Schramm, H.L., Jirka, K.J., Hoyer, M.V., 1987. Epiphytic macroinvertebrates on dominant macrophyes in two central Florida lakes. J. Freshw. Ecol. 4, 151-161.
523
Schultz, R., Dibble, E., 2012. Effects of invasive macrophytes on freshwater fish and
524
macroinvertebrate communities: the role of invasive plant traits. Hydrobiologia
525
684(1), 1-14.
526 527
Shannon, C., Wiener, W., 1949. The Mathematical Theory of Communication. Urbana University of Illinois Press, 177.
Page 18 of 28
528
Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688.
529
Sooknah, R.D., Wilkie, A.C., 2004. Nutrient removal by floating aquatic macrophytes
530
cultured in anaerobically digested flushed dairy manure wastewater. Ecol. Eng.
531
22, 27-42. Stiers, I., Crohain, N., Josens, G., Triest, L., 2011. Impact of three aquatic invasive
533
species on native plants and macroinvertebrates in temperate ponds. Biol.
534
Invasions 13, 2715-2726.
cr
ip t
532
Theel, H.J., Dibble, E.D., Madsen, J.D., 2008. Differential influence of a monotypic and
536
diverse native aquatic plant bed on a macroinvertebrate assemblage; an
537
experimental implication of exotic plant induced habitat. Hydrobiologia 600(1),
538
77-87.
an
us
535
Toft, J.D., Simenstad, C.A., Cordell, J.R., Grimaldo, L.F., 2003. The effects of
540
introduced water hyacinth on habitat structure, invertebrate assemblages, and fish
541
diets. Estuaries 26, 746-758.
M
539
Vilà, M., Espinar, J.L., Hejda, M., Hulme, P.E., Jarošík, V., Maron, J.L., Pergl, J.,
543
Schaffner, U., Sun, Y., Pyšek, P., 2011. Ecological impacts of invasive alien
544
plants: a meta-analysis of their effects on species, communities and ecosystems.
545
Ecol. Lett. 14, 702-708.
547 548
te
Villamagna, A.M., 2009. Ecological effects of water hyacinth (Eichhornia crassipes) on
Ac ce p
546
d
542
Lake Chapala, Mexico (Doctoral dissertation, Virginia Polytechnic Institute and State University).
549
Villamagna, A.M., Murphy, B.R., 2010. Ecological and socio-economic impacts of
550
invasive water hyacinth (Eichhornia crassipes): a review. Freshw. Biol. 55, 282-
551
298.
552
Villamagna, A.M., Murphy, B.R., Karpanty, S.M., 2012. Community-Level Waterbird
553
Responses to Water Hyacinth (Eichhornia crassipes). Invasive Plant Sci. Manage.
554 555 556 557 558
5, 353-362. Walker, P.D., Wijnhoven, S., van der Velde, G., 2013. Macrophyte presence and growth form influence macroinvertebrate community structure. Aquat. Bot. 104, 80-87. Wilson, J.R., Holst, N., Rees, M., 2005. Determinants and patterns of population growth in water hyacinth. Aquat. Bot. 81, 51-67.
Page 19 of 28
559
Zúñiga, M.C., Cardona, W., 2009. Water quality and environmental flow bioindicators,
560
(in Spanish). In: Cantera, J., Carvajal, Y., L., Castro, L. (Eds.) Environmental
561
flow: Concepts, experiences and challenges, Ed. Del Valle University, Cali,
562
Colombia, 167-198.
564
Zuur, A.F., Ieno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed effects
ip t
563
models and extensions in ecology with R. New York, Springer.
cr
565 566
us
567
Ac ce p
te
d
M
an
568
Page 20 of 28
Table legends
570
Table 1. Median, minimum and maximum values of the measured environmental
571
variables of the sampling sites with and without water hyacinth, statistical significance
572
levels between both are indicated (Mann-Whitney U-test).
ip t
569
Water hyacinth absent Water hyacinth present p-value Temperature (°C)
109
Total Dissolved Solids (mg.L-1)
0.05
0.05
0.07
pH
7.35
7.01
7.76
Chlorophyll (µg.L-1)
7.00
5.46
25.43
Chlorides (mg.L-1)
2.15
1.71
Dissolved Oxygen (mg.L-1)
8.04
6.47
Oxygen Saturation (%)
102
Turbidity (FTU)
5.20
0.63
73.50 36.50
89
0.07
us
70
27.37 25.62 29.74
0.05
0.06
0.20
7.11
6.69
7.90
0.06
6.92
4.70
10.10
0.61
2.52
2.48
1.26
2.90
0.10
10.69
7.22
4.14
9.09
0.22
M
an
0.04
81.33 130.60 89.80 51.77 117.82
0.22
3.96
0.00
9.52
3.66
1.13
5.32
te Ac ce p
575
79
Max
d
Conductivity (µS.cm )
574
Median Min
27.86 26.54 28.91 -1
573
Max
cr
Median Min
Page 21 of 28
576
Figure legends
577
Figure 1. Map of study area with indication of the sampling sites and the presence (blue)
579
or absence (red) of water hyacinth in the Daule-Peripa reservoir, Ecuador .
580
Figure 2. Picture taken at the Daule-Peripa dam of a site completely covered by water
581
hyacinth.
582
Figure 3. Generalized linear model for the presence/absence of water hyacinth in
583
function of turbidity.
584
Figure 4. Values of BMWP-Colombia (a), Margalef index (b), Shannon-Wiener index
585
(c), Simpson index (d) and Evenness (e) for different percentages of water hyacinth cover
586
(0=absent, 1=1-5%, 3= 25-50%, 4 = 50-75%, 5 = 75-100%).
587
Figure 5. Scatter plot of the Biological Monitoring Working Party-Colombia (BMWP-
588
Colombia) and the turbidity, with indication of the different vegetation cover classes of
589
water hyacinth (0=absent, 1=1-5%, 3= 25-50%, 4 = 50-75%, 5 = 75-100%).
590
Figure 6. Non-metric Multidimensional Scaling (NMDS) plot with indication of samples
591
with (big symbols) and without (small symbols) water hyacinth and the ecological water
592
quality class according to the Biological Monitoring Working Party (BMWP) Colombia.
594 595 596
cr
us
an
M
d
te
Ac ce p
593
ip t
578
Page 22 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 23 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 24 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 25 of 28
Ac
ce
pt
ed
M
an
us
cr
i
Figure 4
Page 26 of 28
Ac
ce
pt
ed
M
Figure 5
Page 27 of 28
Figure 6
e
Resemblance: S17 Bray Curtis similarity 2D Stress: 0.19
pt
ce Ac
Moderate
BMWP Colombia Poor Bad Very bad
Page 28 of 28