Terrestrial Contributions to the Aquatic Food Web in the Middle Yangtze River
Citation: Wang J, Gu B, Huang J, Han X, Lin G, et al. (
Terrestrial Contributions to the Aquatic Food Web in the Middle Yangtze River
Jianzhu Wang 0
Binhe Gu 0
Jianhui Huang 0
Xingguo Han 0
Guanghui Lin 0
Fawen Zheng 0
Yuncong Li 0
Syuhei Ban, University of Shiga Prefecture, Japan
0 1 Collaborative Innovation Center for Geo-hazards and Eco-environment in Three Gorges Area, Hubei Province, The Three Gorges University , Yichang , China , 2 Soil and Water Science Department, University of Florida, Gainesville, Florida, United States of America, 3 Institute of Botany, the Chinese Academy of Sciences , Beijing , China , 4 Institute of Applied Ecology, the Chinese Academy of Sciences , Shenyang , China , 5 Center for Earth System Science, Tsinghua University , Beijing , China , 6 Nanjing Research Institute of Hydrology and Water Resources , Nanjing , China , 7 Soil and Water Science Department, Tropical Research and Education Center, IFAS, University of Florida , Homestead, Florida , United States of America
Understanding the carbon sources supporting aquatic consumers in large rivers is essential for the protection of ecological integrity and for wildlife management. The relative importance of terrestrial and algal carbon to the aquatic food webs is still under intensive debate. The Yangtze River is the largest river in China and the third longest river in the world. The completion of the Three Gorges Dam (TGD) in 2003 has significantly altered the hydrological regime of the middle Yangtze River, but its immediate impact on carbon sources supporting the river food web is unknown. In this study, potential production sources from riparian and the main river channel, and selected aquatic consumers (invertebrates and fish) at an upstream constricted-channel site (Luoqi), a midstream estuarine site (Huanghua) and a near dam limnetic site (Maoping) of the TGD were collected for stable isotope (d13C and d15N) and IsoSource analyses. Model estimates indicated that terrestrial plants were the dominant carbon sources supporting the consumer taxa at the three study sites. Algal production appeared to play a supplemental role in supporting consumer production. The contribution from C4 plants was more important than that of C3 plants at the upstream site while C3 plants were the more important carbon source to the consumers at the two impacted sites (Huanghua and Maoping), particularly at the midstream site. There was no trend of increase in the contribution of autochthonous production from the upstream to the downstream sites as the flow rate decreased dramatically along the main river channel due to the construction of TGD. Our findings, along with recent studies in rivers and lakes, are contradictory to studies that demonstrate the importance of algal carbon in the aquatic food web. Differences in system geomorphology, hydrology, habitat heterogeneity, and land use may account for these contradictory findings reported in various studies.
Funding: This research was supported by National Natural Science Foundation of China (NSFC grant No. 51179094 and 30700091) to JW and Chinese Academy
of Sciences through a Baren Project fund to GL and a Key Project of Knowledge Innovation Direction Program (Grant No. KSCX2-SW-109) to JH. Manuscript
preparation is supported by NSFC (grant No. 41376158) to BG. The funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
Understanding the relative importance of terrestrial and aquatic
sources of the carbon that supports food webs in large rivers is
essential for floodplain management and for wildlife conservation.
Three conceptual models have been proposed to examine the
contribution of various carbon sources to the lotic food webs. The
river continuum concept (RCC) proposes that the major source of
organic matter supporting the large river food webs originates
from terrestrial plants from the headwater and mid-streams, while
in-stream primary production is limited by turbidity and light
attenuation associated with depth [41,47]. The flood pulse concept
(FPC) emphasizes the importance of lateral river floodplain
exchanges and proposes that river food webs are more dependent
on production derived from the floodplain than on organic matter
transported from upstream . The riverine productivity model
(RPM) , however, highlights the importance of local in-stream
production (phytoplankton, benthic algae, and other aquatic
plants). Thorp et al  examined these three food web theories
within a floodplain and a constricted-channel reach of the Ohio
River. They argued that the RPM is the better model for
channelconstricted regions in the Ohio River, but could not be generally
applied until similar food web studies are conducted in different
types of rivers throughout temperate and tropical latitudes. The
RPM and the importance of in-stream autotrophs were further
verified in a study on the Upper Mississippi River using stable
isotope analysis . They argued that the nutritionally poor, and
sometime recalcitrant terrestrial organic matter cannot be the
major source of carbon to the river food web. Other studies also
identify algal carbon as the dominant energy source, fueling the
river and lacustrine food webs around the world [3,4,22,43,45].
By contrast, there is also strong evidence that
terrestriallyderived organic matter is an essential source of carbon to the
receiving waters and may dominate aquatic consumer production
in streams [38.48] and rivers [23,25,57]. About 70% to 90% of all
primary production eventually enters the detritus food web 
and exceeds autochthonous production . Wallace et al 
conducted a large-scale, three-year exclusion of terrestrial leaf
litter input to a forest stream. The exclusion of leaf litter led to the
decline of abundance, biomass or both of the invertebrate taxa
compared with those in a reference stream. This study
demonstrated the importance of riparian detritus as an essential carbon
source to the stream food web. Recently, whole lake 13C
enrichment revealed that terrestrial carbon played an important
role in zooplankton and fish production [7,33]. Zeug and
Winemiller  collected various production sources and aquatic
consumers from the Brazos River for stable isotope and IsoSource
mixing model analysis and identified terrestrial C3 plants as the
dominant carbon source. Doucett et al  and Cole et al 
found large differences in the natural abundance of hydrogen
stable isotopes between terrestrial plants and aquatic primary
producers and indicated that terrestrial carbon was a significant
contributor to zooplankton in small lakes and to benthic
invertebrates and fish in streams and rivers. Considering the
existence of contradictory findings on the relative contribution of
terrestrial and aquatic carbon sources to the river food web,
further work is required to gain greater insight into the energy flow
and nutrient cycling pathways in large rivers under vastly different
morphological, biological and ecological conditions.
The Yangtze River is the largest river in China and the third
longest river in the world, with a length of 6,300 km and a
drainage basin of 1,808,500 km2 (Fig. 1). Tremendous periodic
flood events occur in the summer with a dry period during the
winter and spring (Fig. 2). The average hydrological discharge is
30,166 m3 s21 with a maximum of 110,000 m3 s21 and a
minimum of 2,000 m3 s21 under natural condition. The Yangtze
River has a highly diverse freshwater fish community, with 361
fish species from 29 families and 131 genera, accounting for 36%
of all freshwater fish species in China [17,54]. The Three Gorges
Dam (TGD), which is the worlds largest dam, has been
constructed in the middle Yangtze River in south-central China.
As a result, the water depth upstream of the dam had increased by
66 meters since the Three Gorges Reservoir (TGR) began storing
water in late 2003. Since damming of rivers is the most dramatic
anthropogenic factor affecting freshwater environments
[13,14,15], construction of the TGD would also have severe
environmental consequences [1,34]. The formerly narrow channel
characterized by torrential water flow has been converted to an
extensive stagnant water body similar to the limnetic zone of a
large lake system, with corresponding increases in the water depth,
alternations of the flow rate, dissolved oxygen, light availability
and temperature in the TGR [31,40,50]. These changes likely
would result in an increase in phytoplankton production and a
decrease in large vascular plants. On the other hand, large
volumes of terrestrial organic matter are discharged into the TGR
from the upper Yangtze River and its distributaries during the wet
season and will remain in the TGR for a longer time than during
the pre-dam period, which also can change the composition of
organic matter and further convert the energy sources for the
aquatic biome in this region [34,42].
The combination of a large floodplain area, extreme hydrology
and rich fish resources makes the Yangtze River an ideal system
with which to examine the three food web models. We
hypothesized that terrestrial plants are the dominant source of
organic matter to the consumer production in the region of middle
Yangtze River where hydrology has not been greatly altered by
the TGD. This is due to the extreme flow events, high turbidity
and high water depth, which severely limit in-stream primary
production. We further hypothesized that the construction of the
TGD would increase the contribution of in-stream production to
consumer production, which would be especially apparent in the
immediate upstream region of the TGD. These hypotheses were
tested using stable isotope analyses of the dominant sources of
terrestrial and aquatic organic matter, and representative
consumer taxa collected at three study sites along the middle
Yangteze River, upstream of the TGD during two representative
months of the dry and wet periods. The IsoSource mixing model
 was used to determine the relative contribution of terrestrial
C3, and C4 plants, phytoplankton and benthic algal production.
Materials and Methods
Ethical approval was given by the ethical committee at the
Beijing Institute of Botany, Chinese Academy of Science. In the
current study, the use of tissue material from animals killed is a
part of routine commercial fishery production. The commercial
fishery complies with the local fishery law in Hubei Province. No
specific permits were required for the described field studies. The
locations studied were not privately-owned or protected in any
way and the field studies did not involve endangered or protected
The Yangtze River originates in the Geladandong Mountains
on the Tibetan Plateau, and follows a sinuous west to east route
before empting into the East China Sea at Shanghai (Fig. 1). The
main channel of the upper Yangtze River is 1,040 km long and the
TGD is located in the lower reaches of the upper Yangtze River.
The hydrological cycle in the Yangtze River region includes a dry
season from October to May and a wet season from June to early
October. This is illustrated by the monthly discharge at Luoqi in
2004 (Fig. 2). The drainage basin of the TGR covers 58,000 km2,
and encompasses 19 counties in Chongqing and Hubei Province,
with three major cities situated on the shorelines of the river
channel (Fig. 1). In addition to urban development, agriculture is
the other dominant human activity of the region.
The total surface area of the TGR is 1080 km2. Our three study
sites are located within the TGR from Chongqing to the TGD
(Fig. 1). Luoqi (29u 419100N, 106u 559150E) is about 500 Km
upstream of the TGD. It is a typical constricted-channel site with
predictable flooding during each summer and its hydrological
regime is not affected by the TGD during the study period.
Huanghua (30u 20980N, 108u 59300E) is about 300 Km upstream
of the TGD and is affected by the backwater from the dam.
Maoping (30u 509300N, 110u 589360E) is located immediate
upstream (,1 km) of the TGD and has displayed a hydrological
regime similar to the pelagic zone of a deep-water lake since the
completion of the TGD.
Sample collection and analysis
All samples were collected in two trips to the study sites in
September 2004 and May 2005 which represent the wet and dry
periods, respectively (Fig. 2). We collected organic matter from
eight to ten terrestrial plants (C3 and C4 plants) in the riparian
zone of each site. Leaves from two to three numerically dominant
and secondary tree species, seven grasses and agricultural crops
were collected. Plant samples were air dried, placed in unsealed
envelopes and transported to the laboratory for further processing.
The dominant C3 plants were Polygonum hydropiper, Conyza
canadensis and Alternanthera philoxeroides in Luoqi, Quercus
aliena var. acuteserrata, Pterocarya stenoptera, Nephrolepis
Figure 2. Monthly water discharge (solid line) and sand flux (dash line) at the Luoqi site on the upper Yangtze River in 2004.
auriculata and Ficus tikoua in in Huanghua and Pinus
massoniana, Quercus aliena var. acuteserrata, Conyza Canadensis
and Nephrolepis auriculata in Maoping. Cynodondactylon and
Echinochloa phyllopogon were dominant C4 plants in all sites with
Imperata cylindrica also abundant in Luoqi.
Epiphytic algae (EA) and filamentous algae (FA, mostly
Spirogyra) were collected from substrate surface and stored in
glass bottles. Samples were rinsed repeatedly in deionized water to
remove sediment particles and detritus. Particulate organic matter
(POM) was collected from several depths (surface, 5 m, 15 m,
30 m 50 m or deep depth) with a water bottle from two nearshore,
two mid-channel and one central station along five cross-transects.
Samples from the five stations of each transect were then
combined to form a composite sample. In the laboratory, water
was passed through coarse (100 mm) sieves and then filtered onto
precombusted Whatman GF/C (1.2 mm) glass fiber filters to
characterize the coarse POM (CPOM) and fine POM (FPOM;
1.2,100 mm), respectively. Microscopic examination showed that
CPOM was largely consisted of plant detritus while FPOM
samples were dominated by planktonic algae.
Four invertebrate taxa were collected. The invertebrates
included zooplankton, snails of several species of detritivores/algal
grazers, the stream crab (Sinopotamon yangtsekiense) and the river
shrimp (Macrobranchium nipponense) which are detritivores/
omnivores. Zooplankton were obtained from vertical hauls of a
100 mm plankton net from each of the five transects at each site.
All individuals captured were placed in distilled water at around
4uC for at least 24 h to allow the emptying of their gut contents
and then picked under a dissection microscope. All individuals
collected from each site of the same transect were combined into a
single sample. The snails, stream crabs and river shrimps were
hand-collected or captured with fish traps. Shells and gut contents
were removed and the individuals rinsed with distilled water. Five
individuals of crabs and 510 individuals of snails and shrimps
were composited into a sample with three replicates for each taxon
collected at each site during a sampling event.
A total of 27 species of fish were collected (Table S2, S3 and S4)
using various fishing tools including a seine net, fish trap, gillnet or
fishing pole. For each fish captured, the total length was measured
and three individuals of each species with similar length were
chosen for stable isotope analysis. The dorsal white muscle tissue
was removed and placed on ice in the field. Six fish species
available at all three study sites were selected for the model
estimates and comparative analysis. Among the fish species found
at all three study sites, the common carp (Cyprinus carpio) and the
crucian carp (Carassius auratus auratus) are omnivores. The
bronze gudgeon (Coreius guichenoti) is a benthic invertivore while
the yellow catfish (P. fulvidraco) is an insectivore/invertivore. The
Chinese minnow (Hemiculterella sauvagei) is a bentho-pelagic
feeder relying on the zoobenthos and zooplankton. The Chinese
perch (Siniperca chuatsi) is a demersal piscivore.
All samples were kept on ice in the field and 220uC in
laboratory prior to sample processing. Stable isotope analysis was
performed in the Stable Isotope Laboratory for Ecological and
Environmental Research at the Research Center for Plant Ecology
at the Beijing Institute of Botany, the Chinese Academy of
Sciences. Samples were dried at either 105uC for 24 h (plant
organic matter) or 60uC for 48 h (consumers) and ground into a
fine power using a mortar and pestle, or an agate mill for the
vascular plant samples. All laboratory equipment components
were cleaned with ethanol, distill water and dried between
The 13C/12C and 15N/14N ratios were determined using
Thermo Finnigan MAT DELTAplus XP isotope-ratio mass
spectrometers. Stable isotope data are expressed as the relative
difference ratios of the heavy to light isotopes relative to an
internationally accepted reference standard and calculated as:
dX ~ Rsam Rstd =Rstd x 1000
where X is the isotope of interest (13C or 15N) and; Rsam and Rstd
are the heavy to light isotope ratios of the sample and the standard,
respectively. The dX is expressed as the per mil (%) deviation of
that sample from working standards (glycine and cellulose for 13C;
urea and glycine for 15N) and the measurement precision was
approximately 0.1 and 0.3% for 13C/12C and 15N/14N,
The IsoSource software  was used to estimate the
contributions of terrestrial and aquatic carbon sources to the
consumer assemblage at each study site. Data from the two
collection events at each site were pooled prior to the IsoSource
analysis. This is because most consumers integrate stable isotope
signatures of the dietary resources over time  and there was no
significant difference in consumer d13C at each site between the
dry and wet period ( and see Results for more details). There
were several potential carbon sources available to the aquatic
consumers in the TGD area. These carbon sources included
terrestrial C3 and C4 plants, CPOM and FPOM, EA and FA. The
CPOM was largely comprised of terrestrial organic matter shown
by the intermediate d13C values (224.4%) between those of C3
and C4 plants at each site (Fig. 3 and Table S1) and was hence
excluded from IsoSource model estimates. We used the four
invertebrate taxa and six fish species available at all three study
sites for the model estimates.
Before isotope data were entered into the IsoSource model,
trophic enrichments of d13C and d15N values for each consumer
taxon were corrected based on the consumers trophic position
(TP) and the respective isotope enrichment factors. The TP of
each consumer taxon was estimated based on the difference in the
d15N values of consumers and the trophic baseline organisms, and
the isotope enrichment factor per trophic level (Post 2002):
TP~ d15Nconsumer d15Nbaseline =Dd15Nz2
where d15Nconsumer is the average d15N of the consumer whose TP
is estimated here; d15Nbaseline is the average d15N of the dietary
organism. We used the average d15N value of zooplankton (the
water column phytoplankton feeder) and the stream crab (the
benthic grazer) as trophic baseline organisms. The Dd15N is the
isotope fractionation level during assimilation of dietary nitrogen
by the animal. An average value of 3.4% from the literature 
was used to correct the isotope enrichment per trophic level.
To prepare isotope data for IsoSource model estimates, the
consumers TP above the primary producer (i.e., TP-1) was
multiplied by the isotope enrichment factor of 0.4% for d13C and
3.4% for d15N to obtain the taxon-specific isotope enrichment
levels which were then added to each production source. The
mean d13C and d15N values for each taxon and the five production
sources (corrected for isotope enrichment) were analyzed with the
IsoSource mixing model at an increment level of 1% and a
tolerance level of 0.1%.
Selected hydrological and environmental variables for the
predam (1999 to 2003) and post-dam (20042005) periods (this study
period) were collected from the literature and monitoring reports
to evaluate the hydro-ecological conditions at locations near the
Statistical analysis was performed using SigmaPlot software
(Version 12.5, Systat Software, Inc. San Jose, CA). Data set was
first examined for normality using the Shapiro-Wilk procedure.
One-way Analysis of Variance (ANOVA) was used following by
pairwise comparison using the Holm-Sidak method. When
normality test failed, non-parametric analysis (Kruskal-Wallis
ANOVA on Ranks) was used followed by Duncans multiple
comparison. Two-way ANOVA was used to examine the
interactive effects of study sites and stable isotopes. Statistical
difference was considered to be significant at p,0.05.
Environmental and isotopic variations
Selected eco-hydrological variables measured before (1999
2003) and after the TGD construction (20042005) were
compared to evaluate the impacts of damming to hydrological
regime and the water quality near the three study sites (Table 1).
Flow rates were lower at all sites following dam construction.
Interannual variation in rainfall may have contributed to the
observed pattern of flow reduction before and after dam
construction. Current studies indicate that the hydrological regime
at the Luoqi, which is 500 Km upper stream of the TGD, was not
impacted. However, the flow rates at the two downstream sites
decreased by about 90% compared to the pre-dam period. The
hydrological regime at Huanghua has changed from a typical
constricted channel to an estuarine habitat since the TGD started
storing water in 2003.The flow rate (the average from the wet and
dry seasons) at Maoping were 0.01 and 0.03 m s21 in 2004 and
2005, respectively. High nutrient concentrations have been found
at all study sites before and after TGD construction. Primary
productivity as Chl a concentration remained low at Luoqi and
was considerably elevated at the two downstream sites, especially
Conductivity, ms cm21
Data are taken from various reports [30,52,55,56].
at Maoping (Table 1), likely as the results of the decreased flow
rate and improved water clarity.
Stable isotope compositions of terrestrial organic matter, C3 and
C4 plants collected from different sites are shown in Fig. 3 and
Table S1. The average d13C values of the terrestrial plants
spanned a narrow range, from 229.8 to 228.6% for C3 plants
and 213.5 to 212.2% for C4 plants. The average d15N values of
these plants were more variable within study sites, ranging from
0.2 to 2.3% for C3 plants and from 24.3 to 2.1% for C4 plants.
Among the algal taxa, the average isotope compositions of FPOM
varied slightly from 225.5 to 223.8% for d13C and from 3.6 to
4.6% for d15N among sites and were more depleted at Huanghua
than other sites (Fig. 3). Epiphytic algae displayed the highest d13C
values (220.2 to 218.3%) and moderate d15N values (4.9 to
6.8%) among sites. By contrast, the FA displayed moderate d13C
values (223.7 to 221.7%) and the highest d15N values (9.4 to
11.1%). There was no consistent trend of isotope changes from the
upstream to the downstream site (Fig. 3).
Consumer d13C varied from 224.9 to 217.1% at Luoqi, 224.0
to 220.5% at Huanghua, and 223.6 to 217.1% at Maoping
(Fig. 3, and Table S2, S3 and S4). There were no significant
differences in consumer d13C between the dry and wet period
within each site (Two-tail paired T test, all p$0.05). Pooled data
from the dry and wet periods from each site also showed no
significant differences among sites (Kruskal-Wallis Analysis,
H = 1.912, df = 2, p = 0.38). Consumer d15N varied from 5.6 to
12.1% at Luoqi, 5.2 to 13.5% at Huanghua, and 5.1 to 14.9% at
Maoping (Table S2, S3 and S4). There was no significant
difference in consumer d15N at Luoqi between the dry and wet
period (Two-tail paired T test, p = 0.59). However, there were
significant differences in consumer d15N at Huanghua (Two-tail
paired T test, p = 0.001) and Maoping (Two-tail paired T test,
p = 0.001) between the dry and wet period. During the dry period,
the average consumer d15N increased from Luoqi (8.4%),
Huanghua (9.8%) to Maoping (10.1%) and displayed significant
differences among sites (ANOVA, F = 8.1, p = 0.003) although the
midstream site (Huanghu) and downstream site (Maoping) did not
differ significantly (Duncans method, p = 0.47). However, there
were no significant differences in consumer d15N among sites
during the wet period (ANOVA, F = 0.48, p = 0.62).
Carbon sources supporting river consumers
Consumer d13C at Luoqi had a range of 8.9% which is smaller
than the d13C range of the five production sources (Fig. 3). The
IsoSource model estimates (1st and 99th percentile) indicated that
terrestrial plants were the dominant carbon sources to consumer
production (Table 2). The average 1st and 99th percentile
contributions from C3 and C4 plants to the 10 consumer taxa
ranged from 3 to 43% and 20 to 33%, respectively. The C3 plants
supported between 1761% and 1258% of the production of
zooplankton and the Chinese minnow while contributions to the
other taxa were less significant. The contribution from C4 plants to
each consumer taxon was similar with a few exceptions. The 1st
and 99th percentile values were typically over 20 and less than
40%, respectively for the majority of consumer taxa. Zooplankton,
Chinese minnows and Chinese perch received less than 20% from
C4 plants. However, the stream crab received between 40 and
59% of its carbon from C4 plants. The 1st percentile values of algal
carbon for each consumer taxon were zero although the 99th
percentile values of FPOM were typically greater than 60%. The
FA displayed the lowest 1st and 99th percentile values (Table 2).
The consumer taxa at Huanghua had a considerably narrower
d13C range (3.2%) than at Luoqi (Fig. 3). Unlike the upstream site
where C3 plants showed sporadic support to consumer production,
the majority of the consumers at Huanghua were supported by
both C3 and C4 plants. This was especially consistent for the C3
plants as the 1st and 99th percentile values were all high with a
typical range between 40 and 50% (Table 3). On average, C3
plants contributed between 47 and 57% of the consumer
production at Huanghua while the average contribution from
C4 plants fell between 28 and 36%. The Chinese perch, which is a
piscivore, received the highest support (57 to 68%) from C3 plants
through trophic transfers. Similar to the upstream site, C4 plants
contributed the most to stream crab production with a range of 36
to 43%. Also similar to the upstream site, the model results
revealed that all algal sources had a 1st percentile value of zero.
One noticeable difference was the 99th percentile value of FPOM
(typically below 20%) at Huanghua that was considerably lower
than those at the upstream site.
The d13C values of the consumer assemblage collected in the
limnetic zone (Maoping) of the TGR had a range of 4.8% while
the production sources had a range of 16.6% (Fig. 3). Model
estimates showed that both C3 and C4 plants were the dominant
contributors to most of the consumer taxa (Table 4). The average
1st and 99th percentile values to the consumer taxa were 1742%
for C3 and 1732% for C4 plants which were considerably lower
than those at Huanghua. Zooplankton, the bronze gudgeon, the
yellow catfish and the Chinese perch all depended more upon C3
organic matter, directly or indirectly, than the other taxa (Table 4).
Again, the stream crab utilized more organic matter from C4 (34
to 51%) than C3 plants (0 to 15%). However, other taxa showed
similar percentile values for the contribution from C4 plants
(Table 4). Unlike at Luoqi and Huanghua where all algal sources
had a 1st percentile value of zero, the model estimates revealed
some usage of algal carbon at Maoping. In addition to terrestrial
plants, the crab also relied on planktonic carbon (1 to 38%) while
zooplankton had 1st and 99th percentile values of 2 to 31% for the
epiphytic algae. However, other consumer taxa showed no signs of
a significant use of algal carbon although the 99th percentile value
for FA and FPOM at Maoping increased significantly or remained
as high as those in other sites.
The dominance of terrestrial carbon sources
The IsoSource mixing model identified terrestrial organic
matter, both C3 and C4 plants, as the dominant carbon source
supporting consumer production at the three study sites upstream
of the TGD (Table 5). Our results support other studies which
demonstrated the essential role of terrestrial organic matter for
aquatic production in lakes [5,8], streams [12,38,48] and rivers
[12,23,25,57]. However, our results conflict with studies suggesting
that the terrestrial source is not important to the aquatic food web
[6,10,16,22,46]. We argue that in large rivers with high turbidity,
high flow and greater depth such as in the Yangtze River,
planktonic and benthic algal production is severely limited. The
production of aquatic macrophytes in the Yangtze River is also
limited as the result of a steep slope, a hard river bottom and little
littoral zone area for the development of macrophytes. These
Table 2. 1st and 99th percentiles of the contribution of the five production sources to the 10 consumers taxa in the upstream
(Luoqi) of the Three Gorges Dam.
Note: Percentiles were estimated using a five-source dual isotope mixing model in the IsoSource program (Philips and Gregg 2003). FPOM: Fine particulate organic
matter; EA: Epiphytic algae; FA: Filamentous algae.
Table 3. 1st and 99th percentiles of the contribution of the five production sources to the ten consumer taxa at the midstream
(Huanghua) site of the Three Gorges Dam.
Table 4. 1st and 99th percentiles of the contribution of the five production sources to the ten consumer taxa at the limnetic zone
(Maoping) site of the Three Gorges Dam.
Note: Same as in Table 2.
impacts result in an insufficient supply of autochthonous
production to meet the growth demand of the river consumers.
Consequently, terrestrial loading of organic matter becomes an
important or a dominant carbon source to aquatic consumer
production. The relative importance of terrestrial and algal carbon
to the aquatic food web is highly variable among rivers where the
geological location, geometry, hydrological pattern, system
productivity, floodplain size and land use characteristics differ greatly.
These factors may explain the contrasting findings in the literature
While the importance of terrestrial C3 plants to the river food
web has been demonstrated in a number of studies (although
conflicting findings exist), our model results also pointed to the
significant contribution of terrestrial C4 grasses to consumer
production at all three upstream sites of the TGD (Table 5). Our
results conflict with other studies that suggest that C4 plants
contributed little to the river food web [46,57]. However, a few
additional studies showed that the C4 contribution to the river
food web was more significant during the wet season, as in-stream
primary production is hampered by high flow and high turbidity
[24,29,49]. Model estimates indicate that C3 and C4 carbon were
assimilated simultaneously by primary consumers or other
Note: Same as in Table 2.
consumers at higher trophic positions through the food chains,
suggesting that both C3 and C4 plants play a comparatively large
role as a dietary source to river consumers. Another transfer
mechanism of C4 carbon to aquatic food webs is the consumption
of terrestrial invertebrates which feed on C4 plants by fish of
higher trophic positions. Several studies indicated that transfer of
invertebrates-derived nutrients from land can be a significant
contribution to aquatic ecosystem [1,27,39]. Given the mixed
results from different studies, it is likely that C4 grasses, the
availability of other resources, consumer composition and habitat
heterogeneity may influence the contribution of C4 plants to the
consumer production in different rivers.
The sum of the 99th percentile values from C3 and C4 plants at a
given site cannot account for 100% of the production sources to
consumer production (Table 5). This implies that autochthonous
production also contributed to the local food web. The low 99th
percentile value of the filamentous algae ruled out the possibility of
it having a supplementary role in consumer production. However,
the FPOM and EA may have served as contributors to consumer
production in the Yangtze River. This is supported by the high
99th percentile values of FPOM and EA, especially at Luoqi and
Maoping. Model estimates showed non-zero values for the 1st
Table 5. Averages of 1st and 99th percentile values of each production source contribution to the 10 consumer taxa in the three
study sites calculated from the IsoSource program.
Note: Same as in Table 2.
percentile associated with FPOM and EA in two invertebrates
from the downstream site. This may be an indication of increasing
algal production as the consequence of reduced flow near the
TGD. However, the contribution from algal carbon was low and
was limited to a few taxa, suggesting an insignificant role of
autochthonous production in supporting the aquatic food web in
the middle Yangtze River. This is at least the case during our study
period less than a year after TGD construction.
Tests of river conceptual models
The river continuum concept (RCC) hypothesizes that large
rivers receive the majority of the organic matter, supporting their
food webs, from terrestrial loading to headwater and mid-order
streams . The flood pulse concept (FPC) proposes that river
food webs are more dependent on production derived from the
floodplain than upon organic matter transported from tributaries
upstream . The FPC includes aquatic macrophytes as an
important input but algal carbon is less important. We do not have
direct evidence to evaluate the importance of organic carbon from
the head waters, but given the length of Yangtze River and the
presence of numerous tributaries along the main river channel, it is
unlikely that the carbon source described in the RCC model
provides the main energy flow to river consumers in the middle
The results of the current study support our hypothesis that
suggests the importance of terrestrial carbon for river consumers at
all three hydrologically different reaches of the middle Yangtze
River. We also found that the contribution from C4 carbon is as
important as that from C3 plants in the current study. Some
studies suggest that terrestrial carbon, due to its low nutrient value,
is supplemental at best to the primary consumers. We offer three
explanations to account for the consumer dependence of terrestrial
carbon in the Yangtze River and other large rivers. First, algal
production in the Yangtze River, as well as in some other large
rivers, is not sufficient to support the secondary production. This is
due to several well-recognized constraints (high turbidity, high
flow rate and a deep water column) which severely limit algal
production. For example, the pristine site (Luoqi) upstream of the
TGD had low Chl a concentrations before and after dam
construction despite high nutrient concentrations (Table 1). Wu et
al  reported the results of a stable isotope study on the sources
of carbon within the Yangtze River system from May 1997 and
May 2003. They found that the d13C values of the particulate
organic carbon (POC) pool were consistent with the isotope
signature of the terrestrial soil organic carbon in the Yangtze River
and that the planktonic algal carbon was a minor component of
the POC pool. Second, plant detritus from the floodplain, unlike
that from the headwater, can be of high nutritional value. For
example, some agriculture crops (legume species) had a C/N ratio
as low as that of the algal biomass (Table S1). On the other hand,
the nutritional value of terrestrial detritus can be improved by
microbial colonization upon entering the aquatic ecosystems
[2,32]. In addition, some plant parts, such as seeds and fruits,
which are apparently high in nutritional value, may be directly
consumed by tropical aquatic animal species . Finally,
terrestrial organic matter can enter the river food web through
the consumption of terrestrial invertebrates by fish, as discussed
above. In conclusion, environmental constraints of autochthonous
production and an ample supply of terrestrial organic matter, with
relatively high nutrient content, can lead to terrestrial carbon
sources dominating the aquatic production in many large rivers.
The riverine-productivity model  emphasizes the
importance of algal carbon to river consumers. Support of aquatic
consumers by algal carbon has been well documented in streams
and rivers. The existence of evidence for supporting both sides (the
RPM and FPC) in the literature, strongly suggests that given the
high variability of hydrology, geomorphology and taxonomy of
large rivers, a single riverine source model is not sufficient to
predict the carbon sources of the worlds rivers . In some cases,
both models are needed to account for the energy flow to the food
webs of large rivers. Habitat heterogeneity, fragmentation and
localized eutrophication may lead to the dominance of different
sources of organic matter, either by terrestrial contributions or by
algal production for consumer production within rivers [24,57].
This is especially true for large rivers whose drainage basins have
experienced intensive land-use development and hydro-dam
Damming is the single most important factor influencing river
connectivity and its ecological functions [14,15,42]. This study was
conducted at one and one and a half years after the TGD was
completed. The effects of damming to the upstream hydrology and
water quality were almost immediate (Table 1). Changes in fish
species composition after dam construction also occurred as the
numbers of species in Yangtze River decreased greatly . This
is also evidenced by the decrease in the number of fish species
collected at the three study sites (Table S2, S3 and S4). However,
significant increases in phytoplankton production at Maoping (Chl
a, Table 1) was not well reflected in the isotope composition of the
consumers. The importance of surface water algal blooms at
Huanghua and especially at Maoping to the consumers may have
been reduced as a consequence of a rapid decrease in light
penetration and the greater depth of the water column (80 m) in
the TGR. Under these circumstances, the amount of algal carbon
available to the consumer taxa appears to be much less significant
compared to the contribution of terrestrial carbon which is
expected to increase at the limnetic zone of TGR. In fact, model
estimates showed that limited amount of algal carbon to the
consumer production at Maoping was present in a few
invertebrate taxa. This is contrary to our second hypothesis which
suggested that the planktonic algal contribution would dominate
the food web in the river reaches affected by the TGD
construction. Since our study was conducted shortly after the
completion of TGD, further study is needed to examine if the
major carbon sources to the consumer food web have changed at a
longer time scale after TGD completion.
Constraints of stable isotope analysis and model
Our model estimates based on data from three upstream
reaches of the TGD for two hydrological events provide some
temporally and spatially integrated information on the importance
of terrestrial carbon to the river food web. At each site, the isotope
values of the terrestrial source were relatively consistent between
the dates while the isotope values of the autochthonous production
in rivers will likely change seasonally . Therefore, the isotope
signatures of the algal samples collected in May and September
may only provide a brief snapshot of the two hydrologically
contrasting periods. Future studies should consider sampling at an
increased frequency to better capture the temporal variations in
stable isotope values and thereby provide more accurate model
estimates. In addition, our estimates of planktonic algal
contributions might be affected by the appropriateness of FPOM as an
algal surrogate . The use of an algal concentration technique
such as colloidal silica centrifugation  can provide better
isotope signals for the algae. Recent research has revealed large
differences in hydrogen stable isotope values for terrestrial and
aquatic production, which are also reflected in the aquatic
consumers [8,12]. This technique could be used to further
elucidate the relative contribution of terrestrial and aquatic carbon
to the various consumers in large rivers.
Understanding energy flow and nutrient cycling pathways in the
food web of large rivers is essential in the planning for wildlife
conservation and environmental protection [14,18,34,51]. Our
study using stable isotope analysis and the IsoSource model
estimates revealed the important roles of terrestrial carbon to
consumer production in the middle Yangtze River. With rapid
urbanization in China and other developing countries, protection
of riparian integrity and stability is a prerequisite for better water
resource management. Our data may also serve as baseline
information on the initial effects of the construction of the TGD,
the world largest hydro-dam, which could help direct floodplain
management of large rivers around the world.
Average d13C, d15N and C:N ratios for the
production sources at the three study sites in the
ThreeGorges Reservoir area during the wet and dry periods
between 2004 and 2005.
Average, standard deviation (SD) of d13C, d15N
ratios for all consumer taxa at the upstream site (Luoqi)
in Three-Gorges Reservoir during the wet and dry
periods between 2004 and 2005.
Average, standard deviation (SD) of d13C, d15N
ratios for all consumer taxa at the midstream site
(Huanghua) in Three-Gorges Reservoir during the wet
and dry periods between 2004 and 2005.
Average, standard deviation (SD) of d13C, d15N
ratios for all consumer taxa at the upstream site
(Maoping) in Three-Gorges Reservoir during the wet
and dry periods between 2004 and 2005.
Conceived and designed the experiments: JW JH XH GL. Performed the
experiments: JW. Analyzed the data: JW BG FZ YL. Contributed
reagents/materials/analysis tools: JW. Contributed to the writing of the
manuscript: JW BG.
1. Allan JD , Wipfli MS , Caouette JP , Prussian A , Rodgers J ( 2003 ) Influence of streamside vegetation on inputs of terrestrial invertebrates to salmonid food webs . Can J Fish Aquat Sci 60 : 309 - 320 .
2. Bowen SH ( 1987 ) Composition and nutritional value of detritus . In: Conference Proceedings of the International Centre for Living Aquatic Resources Management. Volume 14. Philippines pp. 192 - 216 .
3. Brett MT , Kainz MJ , Taipale SJ , Seshan H ( 2009 ) Phytoplankton, not allochthonous carbon, sustains herbivorous zooplankton production . Proc Nat Aca Sci 106 : 21197 - 21201 .
4. Bunn SE , Davies PM , Winning M ( 2003 ) Sources of organic carbon supporting the food web of an arid zone floodplain river . Freshw Biol 48 : 619 - 635 .
5. Carpenter SR , Cole JJ , Pace ML , Van de Bogert M , Bade DL , et al. ( 2005 ) Ecosystem subsidies: terrestrial support of aquatic food webs from 13C addition to contrasting lakes . Ecology 86 : 2737 - 2750 .
6. Clapcott JE , Bunn SE ( 2003 ) Can C4 plants contribute to aquatic food webs of subtropical streams? Freshw Biol 48 : 1105 - 1116 .
7. Cole JJ , Carpenter SR , Pace ML , Van de Bogert M , Kitchell JL , et al. ( 2006 ) Differential support of lake food webs by three types of terrestrial organic carbon . Ecol Lett 9 : 558 - 568 .
8. Cole JJ , Carpenter SR , Kitchell JL , Pace ML , Solomon CT , et al. ( 2010 ) Strong evidence for terrestrial support of zooplankton in small lakes based on stable isotopes of carbon, nitrogen, and hydrogen . Proc Nat Aca Sci 108 : 1975 - 1980 .
9. Delong MD , Thorp JH , Greenwood KS , Miller MC ( 2001 ) Responses of consumers and food resources to a high magnitude, unpredicted flood in the upper Mississippi River basin . Regul River 17 : 217 - 234 .
10. Delong MD , Thorp JH ( 2006 ) Significance of instream autotrophs in trophic dynamics of the Upper Mississippi River . Oecologia 147 : 76 - 85 .
11. Doi H ( 2009 ) Spatial patterns of autochthonous and allochthonous resources in aquatic food webs . Popul Ecol 51 : 57 - 64 .
12. Doucett RR , Marks JC , Blinn DW , Caron M , Hungate BA ( 2007 ) Measuring terrestrial subsidies to aquatic food webs using stable isotopes of hydrogen . Ecology 88 : 1587 - 1592 .
13. Dudgeon D ( 2000 ) The ecology of tropical Asian rivers and streams in relation to biodiversity conservation . Annu Rev Ecol Syst 31 : 239 - 263 .
14. Dudgeon D ( 2010 ) Requiem for a river: extinctions, climate change and the last of the Yangtze . Aquat Conserv 20 : 127 - 131 .
15. Dynesius M , Nilsson C ( 1994 ) Fragmentation and flow regulation of river systems in northern third of world . Science 226 : 753 - 762 .
16. Finlay JC , Khandwala S , Power ME ( 2002 ) Spatial scales of carbon flow in a river food web . Ecology 83 : 1845 - 1859 .
17. Fu CZ , Wu JH , Chen JK , Wu QH , Lei GC ( 2003 ) Freshwater fish biodiversity in the Yangtze River basin of China: patterns, threats and conservation . Biodivers Conserv 12 : 1649 - 1685 .
18. Fu BJ , Wu BF , Lv YH , Xu ZH , Cao JH , et al. ( 2010 ) Three Gorges Project: Efforts and challenges for the environment . Prog Phys Geogr 34 : 741 - 754 .
19. Goulding M ( 1980 ) The Fishes and the Forest . Berkeley, CL, USA: University of California Press.
20. Gu B , Schelske CL , Waters MN ( 2011 ) Patterns and controls of seasonal variability of carbon stable isotopes of particulate organic matter in lakes . Oecologia 165 : 1083 - 1094 .
21. Gushing CE , Cummins KW , Minshall GW ( 2006 ) River and Stream Ecosystems of the World . Amsterdam: Elsevier.
22. Hamilton S , Lewis W , Sippel S ( 1992 ) Energy sources for aquatic animals in the Orinoco River floodplain: evidence from stable isotopes . Oecologia 89 : 324 - 330 .
23. Hoffman JC , Bronk DA , Olney JE ( 2008 ) Organic matter sources supporting lower food web production in the tidal freshwater portion of the York River estuary, Virginia . Estuar Coast 31 : 898 - 911 .
24. Hoeinghaus DJ , Winemiller KO , Agostinho AA ( 2007 ) Landscape-scale hydrologic characteristics differentiate patterns of carbon flow in large-river food webs . Ecosystems 10 : 1019 - 1033 .
25. Huryn AD , Riley RH , Young RG , Arbuckle CJ , Peacock K , et al. ( 2001 ) Temporal shift in contribution of terrestrial organic matter to consumer production in a grassland river . Freshw Biol 46 : 213 - 226 .
26. Junk WJ , Bayley PB , Sparks RE ( 1989 ) The flood pulse concept in riverfloodplain systems . Can Spec Publ Fish Aquat Sci 106 : 110 - 127 .
27. Kawaguchi Y , Taniguchi Y , Nakano S ( 2003 ) Terrestrial invertebrate inputs determine the local abundance of stream fishes in a forested stream . Ecology 84 : 701 - 708 .
28. Kendall C , Silva, Kelly VJ ( 2001 ) Carbon and nitrogen isotopic compositions of particulate organic matter in four large river systems across the United States . Hydrol Process 15 : 1301 - 1346 .
29. Leite RG , Araujo-Lima C , Victoria RL , Martinelli LA ( 2002 ) Stable isotope analysis of energy sources for larvae of eight fish species from the Amazon floodplain . Ecol Freshw Fish 11 : 56 - 63 .
30. Lin Y ( 2006 ) Studies on the eutrophication and spring bloom in Xiangxi Bay of Three-Gorge Reservoir . Ph.D thesis . Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, China.
31. Liu J , Cao W ( 1992 ) Fish resources of the Yangtze River basin and the tactics for their conservation . Resour Environ Yangtze Valley 1 : 17 - 23 .
32. Mann KH ( 1988 ) Production and use of detritus in various freshwater, estuarine, and coastal marine ecosystems . Limnol Oceanogr 33 : 910 - 930 .
33. Pace ML , Cole JJ , Carpenter SR , Kitchell JF , Hodgson JR , et al. ( 2004 ) Wholelake carbon-13 additions reveal terrestrial support of aquatic food webs . Nature 427 : 240 - 243 .
34. Park Y , Chang J , Lek S , Cao W , Brosse S ( 2003 ) Conservation strategies for endemic fish species threatened by the Three Gorges Dam . Conserv Biol 17 : 1748 - 1758 .
35. Phillips D , Gregg JW ( 2003 ) Source partitioning using stable isotopes: coping with too many sources . Oecologia 136 : 261 - 269 .
36. Pomeroy LR ( 1991 ) Relationship of primary and secondary production in lakes and marine ecosystems . In: Cole JJ, Lovett G , Findlay S , Editors. Comparative analyses of ecosystems: patterns , mechanisms, and theories. New York : SpringerVerlag, pp. 97 - 119 .
37. Post DM ( 2002 ) Using stable isotopes to estimate trophic position: Models, methods, and assumptions . Ecology 83 : 703 - 718 .
38. Reid DJ , Quinn GP , Lake PS , Reich P ( 2008 ) Terrestrial detritus supports the food webs in lowland intermittent streams of southeastern Australia: a stable isotope study . Freshw Biol 53 : 2036 - 2050 .
39. Robertson CR , Zeug SC , Winemiller KO ( 2008 ) Associations between hydrological connectivity and resource partitioning among sympatric gar species (Lepisosteidae) in a Texas river and associated oxbows . Ecol Freshw Fish 17 : 119 - 129 .
40. Salazar JG ( 2000 ) Damming the child of the ocean: the Three Gorges Project . J Environ Dev 9 : 160 - 1743 .
41. Sedell JR , Richey JE , Swanson FJ ( 1989 ) The river continuum concept: a basis for the expected ecosystem behavior of very large rivers . In: Proceedings of the International Large Rivers Symposium (LARS). Can Spec Publ Fish Aquat Sci . pp 49 - 55 .
42. Shen G , Xie Z ( 2004 ) Three Gorges project: chance and challenge. Science 304 : 681 .
43. Sobczak WV , Cloern JE , Jassby AD , Mu Ller-Solger AB ( 2002 ) Bioavailability of organic matter in a highly disturbed estuary: The role of detrital and algal resources . Proc Nat Aca Sci 99 : 8101 .
44. Thorp JH , Delong MD ( 1994 ) The riverine productivity model: an heuristic view of carbon sources and organic processing in large river ecosystems . Oikos 70 : 305 - 308 .
45. Thorp JH , Delong MD ( 2002 ) Dominance of autochthonous autotrophic carbon in food webs of heterotrophic rivers . Oikos 96 : 543 - 550 .
46. Thorp JH , Delong MD , Greenwood KS , Casper AF ( 1998 ) Isotopic analysis of three food web theories in constricted and floodplain regions of a large river . Oecologia 117 : 551 - 563 .
47. Vannote R , Minshall GW , Cummins KW , Sedell JR , Cushing CE ( 1980 ) The river continuum concept . Can J Fish Aquat Sci 37 : 130 - 137 .
48. Wallace JB , Eggert SL , Meyer JL , Webster JR ( 1997 ) Multiple trophic levels of a forest stream linked to terrestrial litter inputs . Science 277 : 102 - 104 .
49. Wantzen KM , de Arruda Machado F , Voss M , Boriss H , Junk WJ ( 2002 ) Seasonal isotopic shifts in fish of the Pantanal wetland , Brazil. Aquat Sci 64 : 239 - 251 .
50. Wu J , Huang J , Han X , Xie Z , Gao X ( 2003 ) Three-Gorges Dam- Experiment in Habitat Fragmentation? Science 300 : 1239 - 1239 .
51. Wu Y , Zhang J , Liu SM , Zhang ZF , Yao QZ , et al. ( 2007 ) Sources and distribution of carbon within the Yangtze River system . Estuar Coast Shelf Sci 71 : 13 - 25 .
52. Xu X ( 2011 ) Research on eutrophication effect of storing water in Three Gorges on Jialing River downstream and estuary section . MS Thesis . Chongqing University. Chongqing, China.
53. Xie P ( 2003 ) Three-Gorge Dam: risk to ancient fish . Science 302 : 1149
54. Yi Y , Wang Z , Yang Z ( 2010 ) Impact of the Gezhouba and Three Gorges Dams on habitat suitability of carps in the Yangtze River . J Hydrol 387 : 283 - 291 .
55. Zhang S ( 2005 ) Distribution of nutrient and plankton mass in the Three Gorges Reservoir water body . Southwest Agricultural University, Chongqing, China.
56. Zeng H ( 2006 ) Phytoplankton in Yangtze and Three Gorges Reservoir: dynamics and relationship with nutrients and hydrological conditions . Ph.D Thesis . Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, China.
57. Zeug SC , Winemiller KO ( 2008 ) Evidence supporting the importance of terrestrial carbon in a large-river food web . Ecology 89 : 1733 - 1743 .