Fundam. Appl. Limnol.  Vol.

182/1, 17–30Article
Stuttgart, January 2013

Macroinvertebrate trophic responses to nutrient

addition in a temperate stream in South America

Carolina Ocon* 1, María Vanesa Lopez-van Oosterom 1, María Isabel Muñoz2

and Alberto Rodrigues-Capítulo 1

With 6 figures and 3 tables

Abstract: The present continuous anthropogenic pressure has resulted in an enhancement of nutrient inputs into
rivers and streams. This situation has worsened, mainly in agricultural areas, causing an accelerated eutrophica-
tion. Macroinvertebrate feeding strategies reflect the species’ adaptations to environments. For a characterization
of the macroinvertebrate trophic response to eutrophication, we added nitrogen and phosphorus into the La Choza
stream and examined the gut contents of those taxa in order to analyze dietary alterations and assign each macro-
invertebrate to a functional feeding group (FFG). Complementary C and N stable isotopes analysis was carried
out. The gut contents of all taxa studied contained principally detritus, but statistical analyses indicated significant
differences before and after fertilization in some taxa. At the control site, and the treatment site before fertilization,
the FFGs maintained constant proportions, the gathering-collector species being dominant in all samples. After
fertilization the composition of the taxa at the treatment site varied. The δ13C isotopes values showed that most taxa
used detritus as a basal resource. δ15N values were generally coincident with gut content analyses. These results
provide a prediction of the functional responses of macroinvertebrates to the environmental consequences expected
from accelerated land use. The functional responses constitute a powerful tool to assess eutrophication and its con-
sequences in temperate plain streams.

Key words: nutrients, macroinvertebrates, gut contents, stable isotopes, functional feeding groups.

Introduction in South America (Motta & Uieda 2004, Tomanova

et al. 2006). Most of the work within the Neotropical
Ecosystem processes are key indicators of river health region (e.g., Poi de Neiff 1990, Bonetto & Wais de
and integrity (Bunn et al. 1999). Many studies have Badgen 1995, Callisto et al. 2001) has been modelled
emphasized the relevance of biotic interactions in on the studies of Cummins and co-workers (e.g., Cum-
community structure at both small and large spatial mins et al. 1966, Cummins 1973, Cummins & Klug
scales (e.g., Hildrew et al. 1985, Englund 1991). The 1979, Merritt & Cummins 1996) in the United States.
interest in studying the trophic relationships among However, King et al. (1988) have pointed out that the
the invertebrates of rivers and streams has recently in- transfer of information on functional feeding groups
creased. However, most investigations have concerned (FFGs) between continents requires extreme caution.
temperate species of the Northern Hemisphere and lit- Since the analysis of gut contents provides in-
tle is known about the feeding habits of aquatic fauna formation regarding trophic relationships among the

Authors’ addresses:
1  ILPLA-Instituto
de Limnología Dr. Raúl A. Ringuelet, CONICET La Plata -Universidad Nacional de La Plata, Bv. 120 e 61 y
62, La Plata, Argentina
2  Departamento de Ecología. Universidad de Barcelona, Av. Diagonal 645, Barcelona, Spain

* Corresponding author; [email protected]

© 2013 E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart, Germany www.schweizerbart.de

DOI: 10.1127/1863-9135/2013/0382 eschweizerbart_xxx

1863-9135/13/0382 $ 3.50
 18 Carolina Ocon et al.

species in a community, feeding strategies reflect the sponse to a long-term input of N and P into water in
adaptation of species to ongoing environmental condi- the form of fertilizer, thus simulating the increase in
tions. According to Fenoglio et al. (2005), improving the concentration of nutrients that would be a conse-
our knowledge of feeding behavior and trophic ecol- quence of the continued intensification of unregulated
ogy is indispensable to better understand applied and land use. We assume here that the addition of nutrients
basic elements of stream ecology. The same authors would increase the nutritional quality (in terms of a de-
consider that increased human influence in aquatic crease in the C:N and C:P ratios) and/or quantity (bio-
ecosystems lead to changes in feeding and growth of mass) of the basal resources for the primary consum-
aquatic invertebrates, altering composition and struc- ers in our study site. Slavik et al. (2004) and Sabater et
ture of benthic communities. al. (2011) demonstrated that algal biomass increased
The present continuous anthropogenic pressure and epilithon stoichiometry changed after nutrient ad-
has increased nutrient inputs into rivers and streams dition. An alteration in the diet of certain consumers,
(Rodrigues Capítulo et al. 2010), mainly in areas with mainly the scrapers and detritivores, would also be ex-
agricultural potential. According to Aizen (2009), in pected following a change in the quality of the food re-
Argentina the use of monocultivation has involved a sources. We predicted that the consumers would show
progressively widening area, and as a result lands for- preferences for chemically enriched algae and detri-
merly employed for both diversified agriculture and tus, and their diet would accordingly contain a larger
livestock pasturing have been displaced by the inten- proportion of those food items. We report here the re-
sive and exclusive culture of soybeans, a crop which sults of analyses of the gut contents and stable isotopes
has become the main agricultural product in much of of the macroinvertebrates in the Argentine temperate-
the country and especially in the pampean plain. This plains stream La Choza after the performance of an
author concludes that this shift has led to an increased artificial-fertilization experiment. A BACI (Before-
use of agrochemicals (pesticides and fertilizers). In After Control-Impact) design was applied with the
addition, traditional ranching involving free-range aim of determining significant differences attributable
grazing has been supplanted by the use of feedlots to nutrient addition. In the same way, the relationships
containing high densities of animals for the purpose with changes in basal resources (Feijóo et al. 2012)
of rapid fattening, which produces a high load of or- were taken into account.
ganic waste matter that eventually finds its way into
the environment. According to Rodrigues Capítulo et
al. (2010), these nutrient increases will favor autotro- Material and methods
phy, particularly by those species capable of strategies
for surviving in more turbid and enriched environ- Study area
ments. An elevation in the nutrient content of rivers The La Choza Stream (Fig. 1) is located in Luján, Buenos Aires
and streams can alter biological processes by stimu- province, Argentina (in the Río de la Plata basin). Pampean
lating the growth of primary producers and accelerat- grassland is the typical biome here, the region being a steppe
composed of gramineous grasses (Cabrera 1971). The main
ing the rates of decomposition (Elwood et al. 1981, previous anthropogenic activities were scattered agriculture,
Harvey et al. 1998), while enhanced eutrophication though now that form of land use has been replaced by inten-
will encourage herbivores and detritivores (Sabater et sive soybean farming, along with widespread cattle raising. At
al. 2005). The adverse effects of these environmental present, the cultivation of this crop has increased by 82 % in the
changes on habitats and their resident biota should Province of Buenos Aires compared to the 1970s and in addi-
tion it now occupies more than 50 % of the arable land through-
therefore be minimized through appropriate manage- out the entire country (Aizen et al. 2009). The climate of this
ment policies. For this step to be taken, as suggested region is humid and temperate with a mean temperature of
by Erwin (2009), we must first understand the nature 20 °C and a mean rainfall of 900 mm/year. The basin′s drainage
of the ecological changes that will most probably take area is 15,200 ha. The homogeneity of the streams in the region
made possible the selection of two segments of 100 m in length
place within each specific region before implementing
in the same lotic system, located 5 km away from each other.
the appropriate wetland management and restoration. The first reach was located upstream (control site: 34 º 39′ 14″ S
The present study is part of the GlobRio Project, – 59 º 10′ 00″ W), and the second downstream from the experi-
whose design consists of analyzing the effects of an mental input of nutrients (treatment site: 34 º 47′ 51″ S – 58 º
experimental addition of nutrients on biodiversity, 3′ 17″ W). The distance from the source is 2.5 km for control
site and 7.5 km for the treatment site.
food web, and fluvial-system functioning (Rodrigues The stream is characterized by a low slope (< 1 %) and the
Capítulo et al. 2010). The aim of the present study was width varies between 4 and 10 m in both reaches. This environ-
to quantify changes in the diets of invertebrates in re- ment has high turbidity, abundant organic matter and high sus-

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 19

Fig. 1. Map of the study area showing the La Choza Stream with the location of the sampling stations.

pended solids; furthermore the occurrence of high nutrient con- al. 2012). The water samples were collected and analyzed daily
centrations is an essential feature of these streams (Feijóo et al. by the staff of the Universidad Nacional de Luján, who deter-
2012). The bed comprises consolidated carbonates (“caliche”) mined values of dissolved inorganic nitrogen (DIN) and total
covered by fine sediments (mud and silt). Aquatic vegetation phosphorus (TP) (APHA 1998).
was represented by Ludwigia peploides and Bacopa monnieri.
Mean water velocity is 0.04 m s–1 in the control site and 0.08 m Macroinvertebrates sampling
s–1 in the treatment site. Mean water flow is 9.63 L s–1 in the
control and 14.83 L s–1 in the treatment. Mean dissolved oxygen Macroinvertebrates were collected following Rodrigues Capí-
concentration is 9.61 mg l–1 in the control site and 9.09 mg l–1 tulo et al. (2009). Three replicates of benthic macroinverte-
in treatment, conductivity is 1110 µS cm–1 and 1513 μS cm–1 brates were removed with an Ekman grab (100 cm2) for the
respectively. Mean water temperature is 17.6 °C in the control analysis of the FFGs. The macroinvertebrates associated with
site and 17.1 °C in the treatment site. the macrophytes were collected with sieves (250 µm) within the
area subsumed by a 1,300 cm2 Plexiglas square. In both reaches
Experimental design the same microhabitats were sampled. The samples were fixed
in situ with 5 % formaldehyde and the organisms sorted, identi-
Sampling was carried out bimonthly over a period of two years fied and counted under a stereomicroscope using the taxonomic
(from March 2007 to December 2008). keys of Fernández & Domínguez (2001).
From March 2007 through October 2007 both selected
reaches in La Choza stream were studied before nutrient ad- Analyses of gut contents and FFG
dition. From November 2007 until the end of the experiment determination
in December 2008, a fertilizer (Nitrofoska®) rich in nitrogen
(in the form of nitrates and ammonium salts) and phosphorus Only taxa found at both sites (except in the case of bivalves)
(as phosphoric anhydride) was added continuously upstream were selected for the analysis of gut contents: Perithemis sp. 1,
from the treatment site placing 12 bags in the water to give a Orthemis nodiplaga, Aeshna bonariensis, Phyllogomphoides
total of 750 g of the fertilizer and 250 g of urea, thus attaining a joaquini, Coenagrionidae sp. 1 (Odonata), Chironomidae (Dip-
concentration threefold higher than the average original levels tera), Diplodon delodontus (Mollusca, Bivalvia), and Pomacea
in the water. The bags were located along three transects sepa- canaliculata (Mollusca, Gastropoda) for the control site and all
rated by 20 m at the beginning of the treatment reach (4 bags of the above taxa but with Corbicula fluminea (Mollusca, Bi-
per transect). The proportion of added nutrients was calculated valvia) instead of D. delodontus for the treatment site (because
to maintain the N/P ratio in the water. The nutrient bags were the latter species was absent at this site). For each species, 10
replaced 2 or 3 times a week during the fertilization period to individuals (at the same instar) per site were collected on all
keep the added nutrients at a constant concentration (Feijóo et sampling dates and fixed in the field with 5 % formaldehyde.

eschweizerbart_xxx
 20 Carolina Ocon et al.

In the laboratory, the invertebrates were dissected under a ste- Statistical analysis
reoscopic microscope and the foregut separated and placed in
vials with Bengal’s rose colorant for 24 h to stain their contents. BACI (Before-After­ Control-Impact) design was applied fol-
These contents were removed, homogenized in distilled water, lowing Stewart-Oaten & Bence (2001) with the aim of estab-
and filtered through a 0.45-µm membrane. The filters were lishing whether significant differences between the sites had oc-
clarified by the addition of immersion oil and 15 randomly se- curred as a result of the experimental fertilization. With respect
lected fields per slide counted under an optical microscope at a to the FFGs this approach was performed through the use of
magnification of 100× to 400×. The area covered by each item the relative-abundance values. Differences in gut contents for
of gut contents was estimated with a graduated eyepiece. The each taxon were tested by means of the same method using per-
size of consumed particles was also measured using the same centages of each food item (arc sin transformation of values).
method. The quantification of a given ingested item was based The data analyses were performed using SPSS software Ver-
on the fraction its area covered relative to the area subsumed sion 12.0 (2003). An additional Student t test was applied for
by the total gut contents, expressed as the relative frequency each analysed taxa to compare the isotope data before and after
of each food item (following Winterbourn et al. 1984, Jaarsma fertilization at both sites. The same test was applied to analyse
et al. 1998, Diaz Villanueva & Albariño 1999). We classified the gut contents of P. joaquini before and after fertilization at
the gut contents according to five categories: detritus (uniden- both sites because this species showed apparent differences not
tifiable organic mater), diatoms, vascular plants, animal mat- demonstrated for BACI analyses.
ter and mineral material. The gut content of the predators was
furthermore identified whenever possible (identifiable animal
remains).
Invertebrates were partitioned into FFGs using the com-
Results
plete list of taxa and their densities (as relative-frequency val-
ues) obtained at each site and on each sampling date based on Water phosphorus and nitrogen concentrations
the literature (Merritt & Cummins 1996, Cummins et al. 2005). before and after fertilization
In the case of taxa whose gut contents were analysed the domi-
nant food was considered (more than 60 % of the gut contents). At the control reach, the respective nutrient concen-
In both cases the nomenclature established by Merritt and Cum- trations before and after October 2007 (date of addi-
mins (1996) was used. These FFGs were: (1) shredders (feed- tion of fertilizer in treatment site) were 205.6 ± 47.3
ing on coarse particulate organic matter > 1 mm in size), (2) and 93.4 ± 3 mg l–1 for soluble reactive phosphorus
filtering collectors (sifting fine particulates of 0.45 µm to 1 mm
from the flowing water column), (3) gathering collectors (gath-
and 601.6 ± 102 and 492.4 ± 110 mg l–1 for dissolved
ering fine particulates of organic matter from the debris and inorganic nitrogen (mean values ± SD). In contrast, in
sediments on the bed of the stream), (4) scrapers (scraping off the treatment stretch, after the addition of the fertilizer
and consuming the organic layer of algae, microorganisms, and the respective soluble-reactive-phosphorus levels in-
dead organic matter attached to stones and other substrates),
creased from 247.6 ± 37.7 to 431.6 ± 29 mg l–1, while
and (5) predators (feeding on other animals).
the dissolved-inorganic-nitrogen values decreased
Isotope analysis from 1331 ± 214 to 828 ± 199 mg l–1. In both reaches N
level decreased after the date of addition of nutrients.
In addition, an analysis of the stable N (δ15N) and C (δ13C) iso-
topes was carried out following Muñoz et al. (2009) to establish
This decrease is caused by denitrification processes,
the diet (in terms of assimilation of nutrients) and the trophic principally during the night (Acuña et al. 2011). The N
level of the species under study. Several individuals of each values, however, were higher in the treatment site (ap-
selected species were collected in numbers depending on their proximately twofold both before and after treatment).
weights (10 for Odonata, 3 for Mollusca) once before (Octo-
The mean ratios between the dissolved-inorganic-
ber 2007) and once after fertilization (December 2008). Next,
in the laboratory, the macroinvertebrates were stored for 24 h nitrogen and the soluble-reactive-phosphorus values
to let the gut contents evacuate and then, the organisms were were 5.4 and 2.3 for control and treatment reaches,
dried at 60 °C to constant weight and ground into a powder to respectively, during the enrichment period.
ensure their homogeneity. The resulting samples were analyzed
by a Carlo Erba elemental analyzer coupled to a mass spectro- Macroinvertebrate FFGs
photometer (University of Barcelona). The isotopic composi-
tions were quantified through the use of international standard The FFG was obtained for all taxa determined (Table
reference materials (Pee Dee Belemnite for carbon and atmos- 1). The control site exhibited consistent proportions
pheric N2 for nitrogen). The 13C/12C and 15N/14N ratios were
expressed as the relative difference in parts per thousand be- among the FFGs analyzed throughout the study period.
tween the sample and the conventional standard. The stable The gathering collectors were dominant on all sam-
isotopes and stoichiometry ratios of basal resources (detritus, pling dates, but an increase in predators was observed
periphyton, epipelon, macrophytes and seston) were obtained in May and December 2008 (Fig. 2a). Before fertiliza-
from Feijóo et al. (2012). The mixing model SIAR V4 (Stable
Isotope Analysis in R) was used to determine the contribution
tion, the treatment site likewise exhibited a high num-
of the different food items and basal resources to the diet of the ber of gathering collectors; but unlike the control site,
studied taxa. the filtering collectors were a more abundant group.

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 21

Table 1. FFG for all the taxa determined in the reaches studied in the La Choza stream.
Taxa Control reach Treatment reach Habitat FFG
Before After Before After Before After
Dugesiidae x x x x x x predator
Temnocephalla sp. x x x x x x parasites
Nematoda x x x x x x gathering-collector
Naididae Naidinae x x x x x x gathering-collector
Naididae Pristininae x x x x x x gathering-collector
Naididae Tubificinae x x x x x x gathering-collector
Naididae Rhyacodrilinae x x x x x x gathering-collector
Opistocystidae x x x gathering-collector
Enchytraeidae x x x x x x gathering-collector
Aelosomatidae x x gathering-collector
Hirudinea x x x x x x predator
Pomacea canaliculata x x x x x scrapers
Heleobia parchappei x x x x x x scrapers
Drepanotrema kermatoides x x x x x x scrapers
Biomphalaria peregrina x x x x x scrapers
Musculium sp. x x x x scrapers
Hebetancylus moricandi x x x x x x scrapers
Eupera sp. x x x scrapers
Chilina sp. x x scrapers
Pisidium vile x x x x x filtering-collectors
Pisidium sterkianum x x x x x x filtering-collectors
Corbicula fluminea x x x x x x filtering-collectors
Diplodon delodontus x x x x filtering-collectors
Tardigrada x x x x x x gathering-collector
Sminthuridae x x x x x gathering-collector
Caenis sp. x x x x x x gathering-collector
Americabaetis sp. x x x x x x gathering-collector
Callibaetis sp. x x gathering-collector
Campsurus sp. x x gathering-collector
Perithemis sp. x x x predator
Aeshna sp. x x predator
Phyllogomphoides joaquini x x x x x x predator
Coenagrionidae x x x x x x predator
Hydrophilidae larvae x x x x x x predator
Hydrophilidae adults x x x x x x gathering-collector
Elmidae larvae x x x x x x gathering-collector
Elmidae adults x x x x x x gathering-collector
Curculionidae larvae x x gathering-collector
Dytiscidae x x x x predator
Noteridae x x predator
Staphylinidae x x predator
Corixidae x x x predator
Belostoma elegans x x x x x predator
Veliidae x x predator
Chironomidae x x x x x x gathering-collector
Ceratopogonidae x x x x x x gathering-collector
Tipulidae x x gathering-collector
Ephydridae x x x gathering-collector
Empididae x x x x gathering-collector
Culicidae x x filtering-collectors
Tabanidae x x predator
Leptoceridae x x gathering-collector
Hidroptilidae x x gathering-collector
Hyallela curvispina x x x x x x gathering-collector
Palaeomonetes argentinus x x x x x gathering-collector
Macrobrachium borelii x x x x gathering-collector
Aegla bonariensis x x x x x gathering-collector
Ostracoda x x x x x x filtering-collectors
Daphnidae x x x x x x filtering-collectors
Macrothricidae x x x x x x filtering-collectors
Chidoridae x x x x x x filtering-collectors
Leydigia sp. x x filtering-collectors
Cyclopoida x x x x x x predator
Harpacticoida x x x x x x gathering-collector
Acari x x x x x x predator

eschweizerbart_xxx
 22 Carolina Ocon et al.

In the samples collected immediately after fertilization however, filterers (F = 7.29, p = 0.01) and gathering
(December 2007 through February 2008), a marked collectors (F = 11.79, p = 0.003) showed significant
increase in the proportion of filtering collectors was differences.
noted, though this abrupt change did not persist in
subsequent samples. In May and December 2008 the
Analyses of gut contents
predators became the dominant group, whereas in July
the prevalence of gathering collectors increased, and Detritus was an abundant component within the gut
by September the filtering collectors once again held contents of most of the taxa studied (Fig. 3). With the
sway (Fig. 2b). According to the BACI analysis these Coenagrionidae, the second most abundant item was
results were not statistically significant for predators animal remains, principally composed of the remains
(F = 2.73, p > 0.05) and scrapers (F = 2.69, p > 0.05), of Chironomidae and Copepoda. Animal remains were

Fig. 2. Percentage of functional feeding groups in the control and the treatment sites in the La Choza Stream. a) control, b) treat-
ment. Arrow shows fertilization date.

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 23

observed in the gut contents of P. joaquini at the con- animal matter was observed, whereas in the guts of
trol site but not in the treatment site. The Student t test, this same group at the treatment site the proportion of
however, showed that these differences between sites detritus increased at the expense of diatoms, and nei-
were not significant (t = – 0.01, p > 0.05). ther vascular plants nor animal matter were present.
In the gut of D. delodontus at the control site di- In the guts of P. canaliculata at the control site,
atoms were the second most abundant item (mean detritus (44 %), diatoms (32 %), and vascular plants
value, 25 %); the same was found for C.  fluminea at (in lesser amounts at 21 %) were principally observed,
the treatment site although here the diatoms were at whereas in the guts of this species analyzed at the treat-
times accompanied by the presence of animal remains ment site the vascular plants were dominant over the
(at 3 %). At the treatment site a significant decrease in diatoms, whose proportion had become significantly
the prevalence of diatoms in guts was noted follow- lower (F = 3.738, p = 0.03; Table 2).
ing fertilization along with an increase in the level of The analysis of interaction between sites (control
detritus (Table 2). vs. impact data) indicated that in some instances dif-
In the guts of the Chironomidae at the control site, ferences between the two reaches existed before the
a greater abundance of diatoms, vascular plants, and experimental addition of the nutrients. Nevertheless,

Fig. 3. Mean relative abundances of food items in gut contents for each of the taxa studied in the La Choza Stream.

eschweizerbart_xxx
 24 Carolina Ocon et al.

Table 2. Comparisons made by BACI method for each taxon studied and for each food item. Only statistically significant values
are shown. n.s.: not significant.
Comparisons Detritus Diatoms Vascular Animal Mineral
plants matter material
Chironomidae
Before-After n.s. n.s. n.s. n.s. n.s.
Control-Impact F = 302.3, p = 0.000 n.s. n.s. n.s. F = 22.0 p = 0.019
Before-After * Control-Impact n.s. n.s. n.s. n.s. n.s.
Times (Before-After) n.s. n.s. n.s. n.s. n.s.
Control-Impact * Times (Before-After) F = 9.7, p = 0.000 F = 16.9, p = 0.000 n.s. n.s. F = 7.4 p = 0.000
Bivalvia
Before-After n.s. n.s. n.s. n.s. n.s.
Control-Impact n.s. n.s. n.s. n.s. n.s.
Before-After * Control-Impact n.s. n.s. n.s. n.s. n.s.
Times (Before-After) n.s. n.s. n.s. n.s. n.s.
Control-Impact * Times (Before-After) F = 47.8, p = 0.000 F = 62.7, p = 0.000 – – n.s.
Odonata
Before-After n.s. n.s. n.s. n.s. n.s.
Control-Impact n.s. n.s. n.s. n.s. n.s.
Before-After * Control-Impact n.s. n.s. n.s. n.s. n.s.
Times (Before-After) n.s. n.s. n.s. n.s. n.s.
Control-Impact * Times (Before-After) n.s. n.s. n.s. n.s. n.s.
Pomacea
Before-After n.s. n.s. n.s. – F = 17.9, p = 0.014
Control-Impact F = 8.4, p = 0.006 n.s. n.s. – F = 9.3, p = 0.009
Before-After * Control-Impact F = 4.5, p = 0.037 n.s. n.s. – n.s.
Times (Before-After) F = 105.9, p = 0.009 n.s. n.s. – n.s.
Control-Impact * Times (Before-After) n.s. F = 3.7, p = 0.03 n.s. – n.s.

Fig. 4. Stable-isotope (δ15N and δ13C) values for basal resources and for each species analyzed. Symbols: Circles: control, Tri-
angles: treatment, white symbols: before, black symbols: after, Per – Perithemis sp., Pjo – P. joaquini, Pc – P. canaliculata, Cf –
C. fluminea, Dd – D. delodontus, Ses – seston, COPM – coarse organic particulate matter, FOPM – fine organic particulate matter,
Ep – epipelon, Eptn – epiphyton, Mac – macrophytes.

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 25

the changes recorded with the Chironomidae and the by a predominance of gathering collectors. This pat-
Bivalvia for the gut items detritus and diatoms and tern continued at the control site for the duration of
with P. canaliculata for diatoms after the fertilization the study period, whereas the treatment site exhibited
can be attributed only to that intervention (e.g., in the
BACI analysis: the interaction between the control
and impact data compared for the before and after
time points). With the Odonata, however, significant
differences were never observed between the two sites
(Table 2).
According to the quantitative analysis of the sta-
ble isotopes (Fig. 4), δ15N indicated that the Odonata
were always predators although the insects of that
order also had a tendency to decrease in δ15N signa-
tures in the treatment site after fertilization perhaps
the consumption of less enriched preys. The Student
t test showed that with respect to this shift a signifi-
cant difference had occurred with Perithemis sp. (t =
– 2.21, p = 0.03) at that site after nutrient input, though
not with P. joaquini (t = –1.16, k > 0.05). For bivalves,
which tend to be detritivores, we found evidence of
the consumption of animal matter (C. fluminea, t = 4,
p = 0.02). The mixing model results showed the im-
portance of detritus (CPOM mainly) for both species
of bivalves (Fig. 5). P. canaliculata, however, exhib-
ited a more herbivorous behavior, albeit with some
tendency to consume detritus. These differences, how-
ever, were not significant (t = 0.96, p > 0.05). The mix-
ing model results for this species showed that before
fertilization the more abundant item was epipelon in
the control site and detritus in the treatment site. After
nutrient addition in both sites the more abundant item
was macrophytes (Fig. 6). With respect to δ13C the
majority of the organisms studied were related with
detritus (coarse and fine particulate organic matter)
as basal resource, whereas the filterers D. delodontus
and C. fluminea (before fertilization) and the predator
Perithemis sp (control site, before fertilization) were
related with suspended ultrafine particulate organic
matter (seston). In a similar way P. canaliculata (con-
trol, after fertilization) was related with epiphyton.
For stoichiometric ratios of primary producers who
responded to treatment through an increase in their P
content and a decline of the stoichiometric ratios were
epiphyton (< C:N), epipelon (< C:N, C:P and N:P),
CPOM (< C:P and N:P) and FPOM (< C:N) (Table 3).

Discussion
Fig. 5. Boxplot of the proportions of different sources for Bi-
At both the experimental and control sites a similar valvia species. a) C. fluminea in treatment site, before fertili-
pattern with respect to the abundance of the FFGs zation, b) C.  fluminea in treatment site, after fertilization, c)
was maintained before fertilization, as characterized D. delodontus in control site, after fertilization.

eschweizerbart_xxx
 26 Carolina Ocon et al.

Table 3. Mean and standard deviation of stoichiometric ratios for basal resources in the La Choza stream. Decreases of the stoi-
chiometric ratios related with treatment are shown in bold.
C:N C:P N:P
Site Mean SD Mean SD Mean SD
Macrophyte
Before Control 17.588   4.875   557.683 184.267 32.879 11.692
Before Treatment  –  –   –  –  –  –
After Control 14.324   2.591   243.681   79.736 16.890   4.484
After Treatment 14.238   1.037   267.021   22.871 18.768   1.235
Epiphyton
Before Control 11.644  –   456.316  – 39.190  –
Before Treatment 13.349  –   402.849  – 30.178  –
After Control 12.473   1.423   132.607   51.448 10.878   4.648
After Treatment  9.200   0.898    93.726   10.368 10.182   0.353
Epipelon
Before Control 14.742 14.742   489.256    0.000 33.187   0.000
Before Treatment 14.538   1.961  714.120 385.709 50.076 27.483
After Control 50.346  – 3044.407  – 60.470  –
After Treatment 12.041  1.111  433.142 321.772 36.841 29.733
Seston
Before Control   7.912   0.609   215.185   20.533 27.175   0.708
Before Treatment   8.094   0.486   232.932   24.253 28.741   1.270
After Control   8.294   0.344   124.462   20.714 14.958   1.854
After Treatment 11.217   1.330   155.482   56.764 14.206   6.087
CPOM
Before Control 16.220   4.883     2.544    1.484   0.150   0.074
Before Treatment   9.751   3.855    4.070    0.694  0.459   0.171
After Control 23.586   0.568     5.872    0.219   0.249   0.008
After Treatment 17.772   0.210    2.374    0.101  0.134   0.004
FPOM
Before Control 12.282   0.428     2.467    1.043   0.199   0.079
Before Treatment 11.769  1.531    1.579    0.075  0.135   0.011
After Control 11.104   1.809     1.125    0.129   0.102   0.008
After Treatment  8.266   0.870    1.966    0.422  0.238   0.044

an increase in the proportions of other groups, prin- source for insects. Thus, organic matter may be a sig-
cipally the filtering collectors; with these organisms nificant source of nutrients – and one of limitless avail-
feeding mainly on fine or ultrafine organic matter and ability – for the various communities of tropical and
on diatoms that had been enriched by the addition of subtropical streams. Shieh et al. (2002) asserted that,
the nutrients. with the exception of predators, all the invertebrate
The experimental addition of nutrients to this pam- taxa consume a higher amount of amorphous detritus
pean stream caused slight changes in the diet of some than any other type of food. This food resource would
invertebrates. Detritus was an abundant component of be a dietary supplement – and one that would be very
the gut contents in most of the taxa studied. Several advantageous in nutritionally unstable environments.
reports have established that detritus is a key source The detritus from the treatment site had a higher nutri-
of nutrition for lotic food webs (e.g., Cummins et al. tional quality after the nutrient addition (Feijóo et al.
1966, Hildrew et al. 1985, Closs & Lake 1994, Acuña 2012) as had been previously observed by Sabater et
et al. 2005) in agreement with our results in the La al. (2011). In the studied stream the detritus is a mix-
Choza Stream. Motta & Uieda (2004) concluded that ture of epipelon, epiphyton and macrophytes, with the
the stability of the trophic structure of the insect com- largest contribution from algae.
munity in Ribeirão do Atalho (Brazil) may be related Shieh et al. (2003) demonstrated that the relative
to the great prevalence of organic matter as a food re- prevalence of animal prey in the diet of macroinverte-

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 27

Fig. 6. Boxplot of the proportions of different sources for P. canaliculata. a) control site, before fertilization, b) control site, after
fertilization, c) treatment site, before fertilization, d) treatment site, after fertilization.

brates decreased with increasing environmental stress in environmental conditions when specific resources
(increases of nutrients concentration), whereas under may become unavailable. Because of this, the stable
the same circumstances the occurrence of diatoms in- isotopes analyses are a useful tool to determine that
creased. In our study, there were no significant differ- the organisms are really assimilating.
ences in predator gut contents with respect to animal C.  fluminea is capable of both filter-feeding and
matter between the two experimental situations (BACI pedal-feeding – i.e., drawing from both the water col-
analyses). Measurement of δ15N, however, indicated umn and the bottom sediments, respectively, for nutri-
differences in the general diet of some predatory spe- tion (Wittman et al. 2008). In our study, the dominant
cies. Furthermore, our results indicated that – after the item was detritus, with a particle size ranging from 10
enrichment – detritus, rather than diatoms, became to 350 µm and corresponding to fine-to-ultrafine mate-
the most abundant basal food-web resource of all the rial, with various diatoms being well represented. In
FFGs. Nevertheless, the gut diatom content did exhib- the example of D. delodontus, the gut contained prin-
it differences, but only for P. canaliculata (a scraper) cipally ultrafine detritus and minerals at the control
and C. fluminea (a filterer), after the fertilization. Flex- site. The statistical analyses (BACI) revealed signifi-
ibility in the mode of feeding no doubt facilitates the cant differences in the proportions of detritus and dia-
adaptation of aquatic insects to changes in food avail- toms within the gut contents of the Bivalvia, with this
ability within a given environment (Motta & Uieda difference being attributable to the fertilization. Since,
2004). Accordingly, Mihuc & Mihuc (1995) stressed however, C. fluminea and D. delodontus are two dif-
that observed food habits do not necessarily reflect an ferent bivalve species; the results do not necessarily
accurate assessment of fundamental functional trophic allow a definitive conclusion.
relationships in benthic invertebrates since most taxa In the present study, the gut contents of the Chi-
may function as generalists, if necessary, even when ronomidae larvae contained detritus, minerals, and di-
reported as specialists. Consequently, Albariño & Díaz atoms, and although fertilization resulted in a general
Villanueva (2006) indicated that a generalist strategy preference for the ingestion of detritus in this group,
would be better adapted during unpredictable changes the size of the ingested material was always between

eschweizerbart_xxx
 28 Carolina Ocon et al.

0.5 µm and 1 mm – in agreement with the findings of course principally at night (Acuña et al. 2011). This
Chessman (1986) and Henríquez de Oliveira et al. phosphorus enrichment affected the responses of the
(2003). These organisms are usually classified as gath- organism present in terms of both the feeding prefer-
ering collectors, as has been seen in our study, but the ences and the proportions of the food items ingested.
FFGs of scrapers, predators, and shredders are also A general increase in detritus ingestion was observed
represented within this family. This confirms that the in the scrapers and the collectors. The trophic structure
FFG approach is not always a better metric for evalu- also exhibited changes, in terms of the proportions of
ating the feeding habit of macroinvertebrates. the different FFGs undergoing notable variations. Sla-
P. canaliculata are extremely polyphagous snails, vik et al. (2004) found that densities of invertebrate
feeding on vegetal, detrital, and animal matter. This taxa such as certain Chironomidae and Ephemerop-
species is primarily macrophytophagous, preferring tera increased in a fertilized reach while filterer den-
floating or submerged plants (Estebenet & Martín sities were higher in the reference reach. Alterations
2002) and our observations also show that their gut in resource supply have been shown to dramatically
contents contained conspicuous vegetal matter and change trophic dynamics of streams in arctic tundra
periphytic organisms. Nutrient addition furthermore (Peterson et al. 1993) and those in forested streams
changed the proportions of the food items in the diet (Wallace et al. 1999). According to Cross et al. (2007)
of this species resulting in a significant decrease in a longer period of fertilization would be needed to
diatom content at the treatment site. confirm this conclusion statistically since effects on
Bottom-up effects of nutrient enrichment can stim- community composition require at least three years of
ulate consumer biomass as a result of increased pri- continuous fertilization, as was previously observed.
mary production, but also through detrital-microbial However, effects may be seen in the short term when
pathways (Cross et al. 2005, Cross et al. 2007). Be- it comes to sensitive taxa or assemblages of primary
cause bacteria are organisms with a high affinity for producers even in eutrophic environments (Gómez
phosphorus (Coveney & Wetzel 1992), these microbes et al. 2009), but especially in environments with low
may react to nutrient enrichment. Accordingly, Sabat- nutrient concentration (Slavik et al. 2004, Sabater et
er et al. (2011) found an increase in bacterial density al. 2011).We can therefore conclude that the premise
and activity after nutrient addition in their comparable on which the study was based – i.e., that the artificial
experimental study. This increase can mean a more fertilization would result in dietary changes in the
balanced elemental content (e.g., C, N, and P) of the scrapers and detritivores – proved valid since the ef-
detritus that would alter the nutritional relationship fects on those biota were, in fact, observed, and can
between the producers and the consumers (Sterner even be extended to the predators. These results are
& Elser 2001) in favor of the consumption of detri- meaningful in constituting a first approximation of the
tus by the latter. In our study the artificial fertilization functional responses of macroinvertebrates to the en-
caused increases in the quantity and quality (decreases vironmental alterations expected within the context of
in stoichiometric ratios) of some primary resources global land-use change in temperate plain streams of
so as to produce a bottom-up effect in this manner. South America.
Our results are consistent with those of Gómez et al.
(2009), who found that the microalgae associated with Acknowledgements
the sediment responded significantly to the addition of This study was supported by BBVA Foundation’s GlobRio
nutrient by doubling their biomass. Those authors also Project (Global changes in fluvial systems: effects on biodiver-
found changes in the structure of the diatom taxoceno- sity, food web and system functions) and CONICET PIP Nº
5305. We extend thanks to Dr. Donald F. Haggerty and Mónica
sis along with an increase in their density in response
Caviglia for editing the final version of the manuscript and are
to the fertilization of the La Choza Stream. especially grateful to Joaquín Cochero and Nora Gómez for
The statistical analyses indicated significant differ- providing chemical data. Scientific Contribution N° 904 from
ences between the status of the stream before and after Instituto de Limnología ″Dr. Raúl A. Ringuelet″ (CONICET La
fertilization with respect to some of the observed items Plata -UNLP).
in the gut contents of certain taxa. Although, contrary
to expectations, the nitrogen levels in the water did not References
become elevated after fertilization during the experi-
Acuña, V., Muñoz, I., Giorgi, A., Omella, M., Sabater, F. &
ment, the phosphorus loading of this lotic system dou- Sabater, S., 2005: Drought and postdrought recovery cycles
bled. The decreases in nitrogen concentrations were in an intermittent Mediterranean stream: structural and func-
due to denitrification processes that occur in this water tional aspects. – J. N. Am. Benthol. Soc. 24: 919 – 933.

eschweizerbart_xxx
 Macroinvertebrate trophic responses to nutrient addition 29

Acuña, V., Vilches, C. & Giorgi, A., 2011: As productive and chiometry in a nutrient-rich Pampean stream. – XVI Con-
slow as a stream can be – the metabolism of a Pampean greso de la Asociación Ibérica de Limnología. Guimaraes,
stream. – J. N. Am. Benthol. Soc. 30: 71–83. Minho, Portugal.
Aizen, M., Garibaldi, L. & Dondo, M., 2009: Expansión de la Fenoglio, S., Bo, T. & Cucco, M., 2005: Winter prey preference
soja y diversidad de la agricultura argentina. – Ecologia Aus- of Perlodes microcephalus (Pictet, 1833: Plecoptera, Perlo-
tral [online] 19: 45–54. didae) nymphs in an Apenninic creek, northwestern Italy. –
Albariño, R. & Díaz Villanueva, V., 2006: Feeding ecology of Entomological News 116: 245 – 252.
two plecopterans in low order Andean-Patagonian streams. – Fernández, H. & Domínguez, E., 2001: Guía para la determi-
Int. Rev. Hydrobiol. 91: 122 –135. nación de los artrópodos bentónicos sudamericanos. – (Eu-
APHA, 1998: Standard methods for the examination of water. deT – Serie: Investigaciones de la UNT). pp. 1–282.
20th Ed. – American Public Health Association, Washington Gómez, N., Sierra, M., Cochero, J., Licursi, M. & Bauer,
DC, pp. 1–1134. D., 2009: Epipelic biofilms as indicators of environmental
Bonetto, A. & Wais de Badgen, I. R., 1995: Southern South changes in lowland fluvial systems. – In: Bailey, W. (ed.):
American streams and rivers. – In: Cushing, C., Cummins, Biofilms: Formation, Development and Properties. – Nova
K. & Minshall, G. (eds): Ecosystems of the world: rivers and Science Publishers, Hauppauge NY.
stream ecosystems. – Elsevier, pp. 257– 293. Harvey, C., Peterson, B., Bowden, W., Hershey, A., Miller, M.,
Bunn, S., 1995: Biological monitoring of water quality in Aus- Deegan, L. & Finlay, J., 1998: Biological responses to ferti-
tralian: workshop summary and future directions. – Austral. lization of Oksrukuyik Creek, a tundra stream. – J. N. Am.
J. Ecol. 20: 220 – 227. Benthol. Soc. 17: 190 – 209.
Cabrera, A., 1971: Fitogeografía de la República Argentina. – Henríquez de Oliveira, A., Nessimian, J. & Dorvillé, F., 2003:
Boletin Sociedad Argentina de Botánica 14: 1– 42. Feeding habits of chironomid larvae (Insecta: Diptera) from
Callisto, M., Moreno, C. & Barbosa, F., 2001: Habitat diver- a stream in the Floresta da Tijuca, Rio de Janeiro, Brazil. –
sity and benthic functional feeding groups at Cerra do Cipo, Braz. J. Biol. 63: 269 – 281.
southeast Brazil. – Rev. Brasil. Biol. 61: 259 – 266. Hildrew, A., Townsend, C. & Hasham, A., 1985: The predatory
Chessman, B., 1986: Dietary studies of aquatic insects from chironomidae of an iron-rich stream: feeding ecology and
two Victorian Rivers. – Austral. J.  Mar. Freshw. Res. 37: food web structure. – Ecol. Entomol. 10: 403 – 413
129 –146. IBM SPSS Statistics version 19.0. IBM Corporation, New York.
Closs, G. & Lake, P., 1994: Spatial and temporal variation in Jaarsma, N., De Boer, S., Townsend, C., Thompson, R. & Ed-
the structure of an intermittent-stream food web. – Ecol. wards, E., 1998: Characterising food-webs in two New Zea-
Monogr. 64: 1– 21. land streams. – N. Z. J. Mar. Freshw. Res. 32: 271– 286.
Coveney, M. & Wetzel, R., 1992: Effects of nutrients on specif- King, J., Day, J., Hurley, P., Henshallhoward, M. & Davies, B.,
ic growth rate of bacterioplankton in oligotrophic lake water 1988: Macroinvertebrate communities and environment in a
cultures. – Appl. Environ. Microbiol. 58: 150 –156. southern African mountain stream. – Can. J. Fisher. Aquat.
Cross, W., Johnson, B., Wallace, J. & Rosemond, A., 2005: Sci. 45: 2168 – 2181.
Contrasting response of stream detritivores to long-term nu- Merritt, R. & Cummins, K., 1996: Trophic relations of mac-
trient enrichment. – Limnol. Oceanogr. 50: 1730 –1739. roinvertebrates. – In: Merritt, R. & Cummins, K. (eds): An
Cross, W., Wallace, J. & Rosemond, A., 2007: Nutrient enrich- introduction to the aquatic insects of North America, 3rd edi-
ment reduces constraints on material flows in a detritus- tion. – Kendall/Hunt, Dubuque, Iowa, pp. 453 – 474.
based food web. – Ecology 88: 2563 – 2575. Mihuc, T., 1997: The functional trophic role of lotic primary
Cummins, K., 1973: Trophic relations of aquatic insects. – consumers: generalist versus specialist strategies. – Freshw.
Annu. Rev. Entomol. 18: 183 – 206. Biol. 37: 455 – 462.
Cummins, K., Coffman, W. & Roff, P., 1966: Trophic relation- Mihuc, T. & Mihuc, J., 1995: Trophic ecology of five shredders
ships in a small woodland stream. – Verh. Internat. Verein. in a Rocky Mountain stream. – J. Freshw. Ecol. 10: 209 – 216.
Limnol. 16: 627– 638. Motta, R. & Uieda, V., 2004: Diet and trophic groups of an
Cummins, K. & Klug, M., 1979: Feeding ecology of stream aquatic insect community in a tropical stream. – Braz. J. Biol.
invertebrates. – Annu. Rev. Ecol. Syst. 10: 147–172. 64: 809 – 817.
Elwood, J., Newbold, J., Trimble, A. & Stark, R., 1981: The Muñoz, I., Romaní, A., Rodrígues-Capítulo, A., González Es-
limiting role of phosphorus in a woodland stream ecosystem: teban, J. & García-Berthou, E., 2009: Relaciones tróficas en
effects of P enrichment on leaf decomposition and primary el ecosistema fluvial. – In: Elosegi, A. & Sabater, S. (eds):
producers. – Ecology 62: 146 –158. Conceptos y técnicas para el estudio de la ecología de ríos. –
Englund, G., 1991: Asymmetric resource competition in a fil- Fundación BBVA, pp. 315 – 335.
ter-feeding stream insect (Hydropsyche siltalai: Trichoptera). Poi de Neiff, A., 1990: Dry weight loss and colonization by in-
– Freshw. Biol. 26: 425 – 432. vertebrates of Eichornia crassipes under aerobic conditions.
Erwin, K., 2009: Wetlands and global climate change: the role – Trop. Ecol. 30: 175 –182.
of wetland restoration in a changing world. – Wetlands Ecol. Rodrigues-Capítulo, A., Muñoz, I., Bonada, N., Gaudes A. &
Manage. 17: 71– 84. Tomanova S., 2009: La biota de los ríos: los invertebrados.
Estebenet, A. & Martín, P., 2002: Pomacea canaliculata – In: Elosegi, A. & Sabater, S. (eds): Conceptos y técnicas
(Gastropoda: Ampullariidae): Life-history traits and their para el estudio de la ecología de ríos. – Fundación BBVA,
plasticity. – Biocell 26: 83 – 89. pp. 248 – 264.
Feijóo, C., Leggieri, L., Ocon, C., Rodrigues Capítulo, A., Rodrigues Capítulo, A., Gómez, N., Giorgi, A. & Feijoó, C.,
Giorgi, A., Colautti, D., Muñoz, I., Ferreiro, N., Licursi, M., 2010: Global changes in Pampean lowland streams (Argen-
Gómez, N. & Sabater, S., 2012: Moderate phosphorus en- tina): implications for biodiversity and functioning. – Hyd-
richment maintains homeostatic response of food web stoi- robiologia 657: 53 –70. DOI: 10.1007/s10750-010-0319-3.

eschweizerbart_xxx
 30 Carolina Ocon et al.

Sabater, S., Acuña, V., Giorgi, A., Guerra, E., Muñoz, I. & Ro- to the Biosphere. – Princeton University Press, Princeton,
maní A., 2005: Effects of nutrient inputs in a forested Medi- pp. 80 –133.
terranean stream under moderate light availability. – Arch. Stewart-Oaten, A. & Bence, J., 2001: Temporal and spatial vari-
Hydrobiol. 163: 479 – 496. ation in environmental impact assessment. – Ecol. Monogr.
Sabater, S., Artigas, J., Gaudes, A., Muñoz, I., Urrea, G. & 71: 305 – 339.
Romani, A., 2011: Long-term moderate nutrient inputs Tomanova, S., Goitia, E. & Helesie, J., 2006: Trophic levels
enhance autotrophy in a forested Mediterranean stream. – and functional feeding groups of macroinvertebrates in neo-
Freshw. Biol. 56: 1266 –1280. tropical streams. – Hydrobiologia 556: 251– 264.
Shieh, S., Ward, J. & Kondratieff, B., 2002: Energy flow Winterbourn, M., Cowie, B. & Rounick, J., 1984: Food re-
through macroinvertebrates in a polluted plains stream. – sources and ingestion patterns of insects along a West Coast,
J. N. Am. Benthol. Soc. 21: 660 – 675. South Island, river system. – N. Z. J. Mar. Freshw. Res. 18:
SIAR V4 (Stable Isotope Analysis in R): http://cran.r-project. 43 – 51.
org/web/packages/siar/siar.pdf Wittmann, M., Reuter, J., Schladow, G., Hackley, S., Allen,
Sterner, R. & Elser, J., 2002: The Stoichiometry of Autotroph B., Chandra, S. & Caires, A., 2008: Asian clam (Corbicula
Growth: Variation at the Base of Food Webs. – In: Ecologi- fluminea) of Lake Tahoe: Preliminary scientific findings in
cal Stoichiometry: the Biology of Elements from Molecules support of a management plan. – UC Davis Tahoe Environ-
mental Research Center, pp. 1– 47.

Submitted: 15 June 2012; accepted: 17 January 2013.

eschweizerbart_xxx