Figures
Abstract
Benthic communities below the photic zone depend for food on allochthonous organic matter derived from seasonal phytoplankton blooms. In the Baltic Sea, the spring diatom bloom is considered the most important input of organic matter, whereas the contribution of the summer bloom dominated by diazotrophic cyanobacteria is less understood. The possible increase in cyanobacteria blooms as a consequence of eutrophication and climate change calls for evaluation of cyanobacteria effects on benthic community functioning and productivity. Here, we examine utilization of cyanobacterial nitrogen by deposit-feeding benthic macrofauna following a cyanobacteria bloom at three stations during two consecutive years and link these changes to isotopic niche and variations in body condition (assayed as C:N ratio) of the animals. Since nitrogen-fixing cyanobacteria have δ15N close to -2‰, we expected the δ15N in the deposit-feeders to decrease after the bloom if their assimilation of cyanobacteria-derived nitrogen was substantial. We also expected the settled cyanobacteria with their associated microheterotrophic community and relatively high nitrogen content to increase the isotopic niche area, trophic diversity and dietary divergence between individuals (estimated as the nearest neighbour distance) in the benthic fauna after the bloom. The three surface-feeding species (Monoporeia affinis, Macoma balthica and Marenzelleria arctia) showed significantly lower δ15N values after the bloom, while the sub-surface feeder Pontoporeia femorata did not. The effect of the bloom on isotopic niche varied greatly between stations; populations which increased niche area after the bloom had better body condition than populations with reduced niche, regardless of species. Thus, cyanobacterial nitrogen is efficiently integrated into the benthic food webs in the Baltic, with likely consequences for their functioning, secondary production, transfer efficiency, trophic interactions, and intra- and interspecific competition.
Citation: Karlson AML, Gorokhova E, Elmgren R (2014) Nitrogen Fixed By Cyanobacteria Is Utilized By Deposit-Feeders. PLoS ONE 9(8): e104460. https://doi.org/10.1371/journal.pone.0104460
Academic Editor : Mónica Medina, Pennsylvania State University, United States of America
Received: April 9, 2014; Accepted: July 1, 2014; Published: August 8, 2014
Copyright: © 2014 Karlson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files.
Funding: This research was funded by BEAM (Baltic Ecosystem Adaptive Management), the strategic marine research programme of Stockholm University and The Swedish Research Council for Environment, Agriculture and Spatial Planning (Contracts 217-2007-749 to EG and 215-2009-813 to RE). 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.
Introduction
Temporal changes in trophic interactions can be considerable, particularly in temperate aquatic environments, where primary production is strongly pulsed. In these systems, growth and population dynamics of deposit-feeders is tightly coupled to sedimentation of algal blooms [1]–[5]. Consequently, deposit-feeder growth is affected by the quantity, quality and species composition of settling bloom material [6],[7]. Short-term labeling experiments in the field [8] and laboratory [9],[10],[11], show that some species respond rapidly to the freshly deposited bloom material, whereas others rely on old organic matter and associated microflora [12]. Little is known, however, about the effects of a bloom event across different levels of biological organization, or the mechanisms of the observed responses.
Both phytoplankton and benthic communities in the Baltic Sea have recently undergone large-scale changes due to a combination of eutrophication, biological invasions, and climate change. The biomass of the spring bloom, which is the most important input of high quality organic matter to the sub-thermocline soft-bottom community [13],[4],[5], has decreased, while magnitude of the summer bloom of cyanobacteria has increased [14],[15]. Although deposit-feeders in experiments consume cyanobacteria [16],[10],[11], the diatoms largely constituting the spring bloom are superior in supporting growth of benthic animals, probably because the cyanobacteria contain toxins, have low content of essential fatty acids [7],[17] and lack sterols [18]. Cyanobacteria are, however, rich in nitrogen and phosphorous [19], amino acids [20] and vitamins [21], and could, therefore, be a complementary and, perhaps, crucial food source when the high quality spring bloom input to the sediments has been exhausted. In copepods, positive effects on recruitment were observed when cyanobacteria were offered as a supplementary food source [22]. Also, bivalves performed better if fed cyanobacteria than when food deprived [23]. Moreover, cyanobacteria partially processed by, for example, clams were found to become more palatable and nutritiously upgraded for amphipods [23], thus facilitating usage of cyanobacteria-derived organic matter by benthic communities. Whether settled cyanobacteria could be eaten when relatively fresh or only after some decomposition remains, however, unclear and may differ among species.
Stable isotopes are commonly used as tracers when studying trophic relationships. They provide time-integrated information on diet and habitat use [24],[25],[26]. Ratios of 15N to 14N (expressed as δ15N) exhibit stepwise enrichment with each trophic transfer, whereas ratios of carbon isotopes (δ13C) change little with trophic transfer and can be used to determine ultimate sources of the dietary carbon. The isotopic composition of primary producers may vary over a season and influence the isotopic composition of primary consumers and predators [27],[28]. In the Baltic Sea, the summer bloom dominated by filamentous, nitrogen-fixing cyanobacteria has a characteristically depleted δ15N signal of about −2‰ [27],[11] that is mirrored in zooplankton [27] and littoral consumers [28],[29], and can, therefore, be used as a tracer of dietary nitrogen. Phytoplankton δ13C usually becomes more enriched later in the season, due to discrimination against the heavy isotope [30], but diazotrophic cyanobacteria commonly are more depleted in δ13C than other phytoplankton [27]. A depleted δ13C-signal has therefore been proposed as a tracer of cyanobacterial input in benthos [31], although the signal is not strong and might be confounded by habitat-derived differences in δ13C and carbon processing by microbial communities [32].
Stable isotopes can also be used to quantify subtle changes in community- or population-level trophic structure [33],[34] by quantifying distances between relative positions of individuals in the δ13C- δ15N bi-plot space; for example, the distance between the most enriched and most depleted individual (or species) in δ13C is the niche width of the group (population or community) with regard to the carbon sources. Recently, Tecchio et al. [35] found that trophic niche width in benthos was positively correlated with microplankton biomass in the surface water. In this context, the settled cyanobacteria bloom material with its associated microheterotrophic community would provide a diverse nutrient source to benthic communities, which could be reflected in a wider isotopic niche after the bloom. Moreover, the additional resources and the diversification of the diet could decrease both intra- and interspecific food competition in these food-limited communities [36]. Thus, access to cyanobacteria-derived material during summer could improve growth conditions for benthic consumers.
This study uses stable carbon and nitrogen isotope data from macrobenthic communities sampled at several stations during the summer season to test four interrelated hypotheses addressing incorporation of cyanobacteria-derived nitrogen and carbon in the benthic food web and the importance of cyanobacterial blooms for feeding and nutritional condition of the deposit-feeders. The tested hypotheses were:
- In deposit-feeders, δ15N values will become more depleted after the cyanobacteria bloom, reflecting uptake of diazotrophic nitrogen. In addition, their δ13C will be also more depleted after the bloom, since cyanobacteria tend to have more depleted δ13C than other phytoplankton.
- After the bloom, the isotope niche area will increase in deposit-feeders because microheterotrophic communities supported by the bloom will become more abundant and diverse, which will increase and diversify the food available to consumers.
- Increased isotope niche-area will be positively related to the body condition of the animals, reflecting complementarity in the food sources, higher resource availability and lower intra- and interspecific competition.
- A decreased niche overlap among the species will be observed after the bloom, due to resource diversification and differential utilization of the bloom material by co-occurring competitors. This, together with the species-specific niche increase will widen the isotope niche of the total community.
Our results show that cyanobacterial nitrogen is utilized by deposit-feeders, and indicates that the settling bloom material induces trophic changes in the community.
Methods
Sampling And Test Species
We sampled sediment and deposit-feeding macrofauna at three coastal stations (stn) in the north-western Baltic proper (Figure 1); Håldämman (30 m bottom depth, 58°49′18 N, 17°34′58 E), Uttervik (20 m, 58°50′58 N, 17°32′77 E) and Mörkö (23 m, 58°54′13 N, 17°42′45 E). The sampling was scheduled around the cyanobacteria blooms in 2009 (May 26, June 2, August 13 and September 21) and 2010 (June 7, June 28 and- September 23). Stations Håldämman and Uttervik are located close (1 and 5 km, respectively) to the phytoplankton monitoring stn B1 (58°48′28 N, 17°37′60 E; Swedish National Marine Monitoring Program, data publicly available from www.smhi.se) and stn Mörkö right by the phytoplankton stn H3 (58°56′04 N, 17°43′81 E, Himmerfjärden Eutrophication study), 14 km from stn B1. No specific permission is required for sampling invertebrates in the Baltic Sea and the field studies did not involve any endangered or protected species. Figure 1 shows sampling schedule and cyanobacteria bloom development at both monitoring stations in 2009 and 2010.
Cyanobacteria bloom development in 2009 and 2010 at stn B1 (near sampling stations Håldämman and Uttervik), dotted line, and stn H3 (close to stn Mörkö), solid line. The X-axis is scaled for Julian days and benthos sampling dates are indicated by arrows. See text for the description of bloom composition. In 2010, stn Mörkö was sampled only in late June and September.
Macrofauna was sampled with a benthic sled, set to collect the top 1–2 cm of sediment, which was sieved through a 1 mm mesh to retain the macrofauna. The two most abundant species, the Baltic clam Macoma balthica and the non-indigenous spionid polychaete Marenzelleria cf. arctia [37], were found at all stations on all sampling occasions. The amphipod Monoporeia affinis was less abundant and absent from stn Mörkö in May and July 2009. The amphipod Pontoporeia femorata was never found at stn Uttervik and absent from stn Håldämman in May and July 2009. On each sampling event ∼10 individuals of each species of approximately 2 mg dry weight (shell-free for M. balthica) were collected for isotope analyses. Due to the unbalanced sampling design and the different timing of the bloom between years and stations, we pooled the two pre-bloom sampling occasions and those after the bloom (in 2009) in the statistical analyses (Table 1, but see Figure 2). This simplified comparisons among stations and between years and facilitated the statistical niche analysis, which requires relatively large sample size.
Temporal trends in δ15N, δ13C, C:N ratio and individual weight (top to bottom panels) for each species and station (solid line = stn Uttervik, broken line = stn Mörkö and dotted = stn Håldämman). The peak of the bloom is indicated by the grey bar. The pre-bloom samples were taken 7 and 1 weeks before the peak of the bloom and the post-bloom samples were taken 6 and 11 weeks after the bloom. Data are transformed to z-scores and pooled for 2009 and 2010. Error bars are 95% CI (n∼20 for each sampling occasion; see also Table 1).
Stable Isotope Analyses
Animals and bulk sediment (upper 1–2 cm, from the sled) from each sampling occasion were oven-dried (60°C), packed individually into tin capsules and analysed for elemental and stable isotope content (carbon and nitrogen) at the UC Davis Stable Isotope Facility, USA. The C and N isotope ratios are expressed in the ‰ notation, using the equation:where R is the ratio between the heavy and light isotopes (13C:12C or 15N:14N). The stable isotope ratio, denoted by δ, is defined as the deviation in ‰ from the international reference standard (Vienna PeeDee Belemnite for C, and atmospheric nitrogen gas for N). Higher δ indicates a higher proportion of the heavy isotope. Samples were run in continuous flow with a standard deviation of <0.2‰ among replicate standard samples both for C and N. The δ13C values in animals were corrected for lipid content according to Post et al. [38]. Rawdata on isotope and elemental composition in benthic fauna can be found in Table S1.
Data Analyses And Statistics
Biomass And Body Condition.
The body condition was assayed as C:N ratio (w/w), with higher values reflecting higher fat content [39]. All species were tested for possible correlation between C:N ratio and weight (Pearson product moment correlation) for each station and year, before and after the bloom. Only for M. affinis at stn Uttervik in 2009 and stn Mörkö in 2010, was it necessary to adjust C:N ratio for reduced individual weight: where A is the regression coefficient for the significant weight-C:N relationship; this adjustment is mathematically equivalent to using the residual C:N values of the weight-C:N regression. Adjusted values are used in the figures and all statistical tests, with the exception of the statistical models for isotope data, where raw data were used as explanatory variables for δ15N.
For each species, pre-bloom and post-bloom values for individual body weight and C:N ratio were compared using paired t-tests with station as grouping factor. Pearson correlations and t-tests were performed in STATISTICA 12 (StatSoft).
Generalized Linear Models.
Markedly higher δ15N values were observed at stn Mörkö, due to the nearby sewage treatment plant [40],[41]. Since differences among the stations were not the focus of our study, all data (δ15N, δ13C, C:N ratio, weight) were normalized to z-scores (zero mean, unit variance) for each station and year to remove the baseline differences in population means between samples collected in different years and at different stations, and to bring the distributions closer to normal.
The effect of the bloom on δ15N in the test populations was tested separately for each species in generalized linear models using the GLZ module in STATISTICA 12 (Statsoft) with a log-link error structure and AIC (Akaike information criterion) as the selection criteria. The full model included all categorical variables; bloom (2 levels, before and after), station (3 levels) and year (2 levels), and the three continuous variables: body weight, C:N ratio and δ13C values. Body weight and C:N ratio are indicative of ontogeny and body condition, respectively, and δ13C correlated strongly to δ15N for most species, as frequently observed [26]. The effect of the bloom on the δ13C was tested using the same categorical factors as for δ15N, but with the body weight as the single continuous variable. The response variables, δ15N and δ13C, were Box-Cox transformed prior to analyses. Goodness of fit and residuals were examined to confirm that the selected error distribution and link function were appropriate. Two outliers for M. affinis and M. arctia, respectively, were identified using Grubbs test and removed prior to the analyses. This removal did not affect the analyses outcome but improved normality of residuals.
Niche Metric Calculations And Statistics.
Isotope niche metrics [33],[34] were calculated for each species, station and year, before and after the bloom, using the SIBER (Stable Isotope Bayesian Ellipses in R) package for R v.2.10.135 [42], included in the package SIAR [43]. The metrics developed by Layman et al. [33] represent different aspects of trophic diversity. A large δ15N range (dN) indicates greater trophic diversity in terms of the trophic position and potential for omnivory, whereas a large δ13C range (dC) indicates multiple carbon resources (e.g., benthic, pelagic, terrestrial runoff) and potential for niche diversification. Mean nearest neighbour distance (MNND) describes how individuals are distributed relative to one another within the dietary niche space of the population or the community, with higher values indicating greater dietary divergence. The standard ellipse area, SEA or SEAc when corrected for small sample size [34] provides information about the core aspects of a group (population or community) niche and can also be used to calculate niche overlap between species.
These niche metrics were calculated both for the populations in question (for each species, station, year, before and after bloom) and for the entire community at the same spatio-temporal resolution. Bloom effects on the niche metrics were tested using paired t-tests, with stations as a grouping factor, or Wilcoxon matched pair tests when assumptions of normality were not met. The relationship between the overlap area between the species sharing the same niche and the community niche area was tested using Spearman rank correlation.
Linking Niche Metrics To Bloom Intensity, Body Condition And Inter-Specific Competition.
Bloom intensity was calculated as the area under the curve for the total cyanobacteria biomass from the mid May to the end of September using linear spline interpolation (XLfit, Microsoft Excel add-in). The relative change in SEAc following the bloom event was used as a dependent variable in the linear regression with bloom intensity and species as independent variables. We used SEAc because it is less affected by outliers and small sample size [44],[45] than the other metrics.
The mean body C:N ratio was calculated for all species, stations, and years, before and after the bloom. The relative change in the C:N ratio after the bloom event was also used as a dependent variable in the linear regression with bloom intensity and species as the independent variables. Finally, the relative change in C:N ratio after the bloom event was used as a dependent variable in the linear regression with relative change in SEAc (post-bloom to pre-bloom ratio) and species as independent variables. The significance of the categorical factor was tested using a homogeneity-of-slopes model. The amphipod Pontoporeia femorata was not included in these analyses since we found no evidence for uptake of cyanobacterial nitrogen by this species. To examine whether there is an exploitative competition between M. balthica and M. affinis (i.e. the species that overlapped most often), we tested if their SEAc values change in concert using Pearson product moment correlation. Residuals were examined to confirm normality prior to the analyses.
Results
Bloom Development And Species Composition Of Cyanobacteria
At stn B1 in 2009, the bloom started in late June, peaking on June 30, while in 2010, the bloom peaked only in mid-August. The bloom at stn H3 occurred in mid-July in both years. Bloom intensity was highest at stn B1 in 2009, with 30–35% lower values observed at stn H3 (both years) and stn B in 2010. In 2009, Aphanizomenon constituted 76 and 78% of diazotrophic biomass at stn H3 and stn B1, in 2010 the contribution was higher; 82 (stn H3) and 85% (stn B1). Nodularia spumigena was the second most abundant species at stn B1 (15 and 11% in 2009 and 2010, respectively), whereas Anabaena spp. contributed more (16 and 13%) at stn H3. At both stations in 2007–2009, seston δ15N mirrored nitrogen fixation, declining from a spring bloom maximum to a minimum (2–4‰) two weeks after the peak of the cyanobacteria bloom (coinciding with the third sampling event in 2009), and then increasing again [46].
Elemental Composition And Isotope Ratios Of Sediments And Test Species
Elemental content in the sediment varied from ∼3.5% C and 0.5% N at stn Mörkö to ∼5.5% C and 0.8% N at stn Uttervik, with intermediate values at stn Håldämman (Table 2). In 2009, sediment values of δ15N, C% and N% were lower after the bloom only at stn Uttervik, but in 2010 they were lower at all stations.
Macrofaunal body condition, indicated by the C:N ratio, changed over the season (Figure 2). In M. affinis, both individual weight and lower C:N ratio were lower after the bloom (Paired t-tests, biomass: t = 2.66, p = 0.057, C:N ratio t = 4.60, p = 0.01). In the other species, no change in the individual weight from the pre-bloom to the post-bloom were observed (M. balthica p = 0.82, M. arctia p = 0.53), but their C:N ratios declined (M. balthica, t = 3.41, p = 0.02, M. arctia, t = 2.11, p = 0.09). The small sample size for P. femorata precluded meaningful comparisons. Nitrogen content (%) increased after the bloom for the surface-feeding species (M. balthica and M. arctia: Z = 2.20, p = 0.028, n = 6, M. affinis Z = 2.02, p = 0.043, n = 5), with no change in C% (M. balthica Z = 1.15, p = 0.25, M. arctia Z = 0.94, p = 0.35, M. affinis Z = 1.75, p = 0.08). The carbon and nitrogen content for each species and period are shown in Figure 3.
Carbon and nitrogen (%) in the test species (left to right: Mac - M. balthica, Mz - M. arctia, Mon - M. affinis and Pon - P. femorata) before (white) and after (grey) the bloom. Values are median with 25 and 75% percentiles as well as 95% CI (n = 6 for Mac and Mz, n = 5 for Mon and n = 3 for Pon).
Bloom Effects On The Isotope Ratios Of The Test Species
For all species except P. femorata, the δ15N values became significantly depleted after the bloom (Figure 2, Table 3). The δ13C values were also significantly depleted after the bloom for M. balthica and the amphipods, but not for M. arctia (Figure 4, Table 3). In line with these observations, bloom was a significant explanatory variable in all winning models, with the exception of δ15N for P. femorata (Table 3).
Stable isotope bi-plots illustrating the isotopic niche of the four species, M. affinis (black), M. arctia (red), M. balthica (green), P. femorata (blue) at three study sites (top stn Håldämman, mid stn Uttervik and bottom stn Mörkö) before and after the bloom, for both years (2009, left panels, and 2010, right panel). The dotted lines enclose convex hull area for each species and solid line shows the standard ellipse area, SEAc, for the entire community. Note the presence of M. affinis post-bloom at stn Mörkö 2009 inflates the niche area, since it was not found pre-bloom. Excluding M. affinis at this station in 2009 results in a reduced niche area (as in Fig 5, top panel).
Population Level Niche Metrics And Bloom Intensity, Competition And Faunal C:n Ratio
Testing for a bloom effect across all niche metrics and species results in 12 tests, and when applying a Bonferroni corrected critical p-value, no test was significant. This was due to large variability between stations and years and the small sample size (n = 5 or 6). A significant negative correlation in SEAc change after the bloom was observed between M. affinis and M. balthica (Pearson product moment correlation r = −0.89, p = 0.04). The post-bloom to pre-bloom SEAc ratio explained 25% of the variation in the post- to pre-bloom C:N ratio (F1,15 = 6.42 p = 0.02, adj r2 = 0.25) - when pooling the three species that were found to respond to the cyanobacteria bloom (Figure 5). When M. affinis, which appeared to differ from the other two species, was removed from the model, the SEAc explained 54% of the C:N change (F1,10 = 14.12, p<0.01, adj r2 = 0.54). Bloom intensity explained 32% of the variation in C:N ratio change following the bloom (F1,13 = 9.84 p<0.01, adj r2 = 0.32), with no significant species effect (p = 0.68, Figure 5a). Bloom intensity explained little variation in SEAc change after the bloom (F1,15 = 3.65, p = 0.07, adj r2 = 0.14; Figure 5b).
Concurrent changes (post-bloom/pre-bloom) in C:N ratio and SEAc in relation to the bloom intensity (area under curve; top and mid panels) and to each other (bottom panel). Each data point represents one species, station and year.
Niche Overlaps And Community Niche Pre- And Post-Bloom
The isotope niche of M. arctia was distinct from the other species (Figure 4), but overlapped slightly (∼4% of the total niche area) with M. balthica on one occasion, after the 2010 bloom at stn Mörkö. For species that were found at a station before and after the bloom, the amount of overlap correlated negatively with the SEAc (Spearman R = −0.88, t = −5.28, p<0.01, n = 10). As such, post-bloom overlaps were recorded at stations that showed a decreasing or unchanged SEAc after the bloom, but not where the SEAc increased (Figure 6). At the stations with no cross-species overlap (stns Uttervik 2009 and Håldämman 2009) SEAc also increased considerably after the bloom (112 and 32% increase, respectively). Moreover, stns Uttervik and Håldämman showed increased dN range (56–19% and 6–9% increase) and MNND (25–34% and 37–100% increase) in 2009 and 2010 respectively, whereas stn Mörkö had nearly unchanged dN range, 29 and 39% reduction in SEAc and 9 and 29% lower MNND after the bloom in 2009 and 2010, respectively.
Isotopic niche overlap defined as an area common for the two populations in relation to the total isotopic space occupied by these populations (proportion; left y-axis) and total community niche (arbitrary units; right y-axis) calculated using pre-bloom and post-bloom datasets for each station and year (see also Fig. 4 for raw data). The grey bars show the overlap estimate for each pair of species whose niches were found to overlap, white bars are the total community niche, and pie charts on the top of the gray bars show overlapping proportions of the isotopic niche for each population in question. Datasets that have no overlap during the study period are not included.
Discussion
We found clear evidence that diazotrophic nitrogen derived from the summer cyanobacteria bloom is incorporated into the benthic food webs in the Baltic Sea, with significant differences among the test species. As hypothesized, depleted δ15N values were found in Monoporeia affinis, Macoma balthica and Marenzelleria cf. arctia after the bloom (Figure 2). These species are all surface-feeders, while Pontoporeia femorata, which did not show depleted values after the bloom, is primarily a sub-surface feeder [9]. In line with this, the observed lack of response in P. femorata likely reflects its lower use of bloom material newly settled on the sediment surface. The changes in carbon and nitrogen content in the test species provide supporting evidence for this explanation: all species, except P. femorata, increased their nitrogen content (%) without significant loss in C% (Figure 3). This increase in N% most likely reflected usage of cyanobacterial matter with relatively high nitrogen content and beneficial amino-acid composition [20]. All species also had more depleted carbon isotope values after the bloom, as hypothesized. The observed depletion in carbon values may, however, be also caused by uptake of other than of cyanobacterial C, e.g., aged sediment, which is more depleted in δ13C than seston [47], as a result of microbial activity [48]. Therefore, the observed decrease in δ13C of P. femorata, but unchanged δ15N, may reflect feeding on aged sediment rather than cyanobacteria.
Variations in isotopic fractionation can also contribute to the seasonal changes in isotopic signatures of consumers [49]. The depleted δ15N signal in benthic fauna after the bloom is, however, unlikely to be confounded by variation in fractionation. Despite the cyanobacterial input of nitrogen, the availability of nitrogen decreased during summer (Table 2). This should result in a lower growth rate (as is indeed the case for amphipods, [13],[36], and, hence, higher fractionation of 15N [49]. Microbial conditioning of the cyanobacterial bloom material can also increase δ15N values [50], complicating pre- and post-bloom comparisons of the isotopic signatures in bulk sediment samples. Hence, two main opposing processes may influence sediment and faunal δ15N values: (1) settling and incorporation of diazotrophically fixed nitrogen, which results in depleted δ15N values, and (2) trophic turnover in sediment microbial loops, which results in enriched δ15N values. Despite the latter process, the δ15N values in the key macrofauna benthic species became significantly depleted after the bloom.
Isotopic signatures do not, however, prove direct feeding on cyanobacterial filaments. Much of the diazotrophic nitrogen incorporated in the benthic fauna may have passed through other phytoplankton in the water column or through heterotrophic bacteria, both pelagic and benthic. A large proportion of fixed dinitrogen is known to leak from the cell as ammonium or dissolved organic nitrogen that can be assimilated by other phyto- and bacterioplankton [51]. Using isotope tracers, incorporation of cyanobacterial nitrogen by macrofauna has been estimated to contribute 5–10% of the animal nitrogen content [11]. Cyanobacteria have also been estimated to contribute significantly to zooplankton nitrogen content [27],[52]. On shallow sandy bottoms, Nordström et al. [28] found δ15N values depleted by several ‰ in late summer at all trophic levels studied; macroalgae, macrobenthos and fish. Similarly, Lesutiene et al. [29] found depleted δ15N after the cyanobacteria bloom throughout the shallow Curonian Lagoon food web, and estimated that diazotrophically fixed nitrogen may support up to 80% of secondary production during the bloom period. Interestingly, our most depleted values were also found at the shallowest station, indicating a stronger influence of cyanobacteria in these habitats. In shallow waters, the animals have generally higher growth rates (M. affinis: [36]), and higher incorporation rates would thus be expected. Taken together, these findings indicate that cyanobacterial blooms are likely to be important for secondary production of all benthic communities during summer, and particularly so in shallow habitats.
Our hypothesis that the niche area would increase as a result of diversified food supply after the bloom, which – at least to some degree – could alleviate nutrient limitation, found only partial support (Figures 4, 6). These responses appear to be complicated by the magnitude of bloom, sediment nutrient content and inter-specific competition. The community-level responses included reduced overlap between the species along with an increase in niche area (SEAc) following the bloom (Figure 6). This was further supported by increased MNND at stations Uttervik and Håldämman (both years) implying that, although no overlap was found throughout 2009, the dietary divergence between individuals increased after the bloom. However, the pattern was different at stn Mörkö, where we observed reduced SEAc and MNND and increased isotopic niche-area overlap between all species (Figure 6). Mörkö had the lowest sedimentary C and N content (Table 2). One can speculate that, due to food limitation, the settling bloom material might be more actively utilized by all animals, resulting in higher niche overlap throughout the season.
Interestingly, the isotopic niche of the non-indigenous species M. arctia was clearly different from those of potentially competing native species in all data sets, supporting previous experimental studies demonstrating resource partitioning in the invaded communities [53]. Moreover, M. balthica and M. affinis are known to compete for fresh phytodetritus [53] and seldom co-occur in high numbers [54]. The inverse relationship between the bloom-induced change in their niche sizes provides supporting evidence for interspecific food competition, highlighting the importance of detailed knowledge on trophic interactions when interpreting isotopic niche.
At the population level, the increase in isotopic niche following the bloom event was positively related to change in C:N ratio. Similarly, when the cyanobacterial bloom was large, we observed less reduction in the C:N ratio than after a bloom of lower intensity (Figure 5a). All species, including P. femorata, which did not utilize diazotrophic nitrogen, showed reduced C:N ratio at the end of the summer. This pattern was expected for the amphipods, especially M. affinis, which has its main growth period after the spring diatom bloom [1],[13],[39], with growth rate, body condition and lipid stores normally declining during summer [36],[39]. Amphipods need to build up lipid reserves before the winter and it is possible that they can benefit from cyanobacterial matter only after pre-conditioning (trophic upgrading [23]) for longer than the duration of this study or published experiments [16],[17]. This could, at least partially, explain the stronger relationship between SEAc and C:N ratio when M. affinis was excluded from the analysis. Populations that increased their isotopic niche throughout the season (i.e. increased generalist feeding) had a better body condition than populations with unchanged or decreasing niches, suggesting that supplemental feeding may convey fitness advantages, which is rarely documented (but see [55]).
In conclusion, the incorporation of the bloom material into the food webs, as evident from the depleted δ15N signals in fauna, is advantageous for these species during the critical summer season. The possible role of cyanobacteria in supporting not only primary, but also secondary production in the Baltic Sea deserves more attention.
Supporting Information
Table S1. Isotope and elemental composition (carbon and nitrogen) for each individual sampled.
https://doi.org/10.1371/journal.pone.0104460.s001
(XLSX)
Acknowledgments
Monitoring data on phytoplankton were kindly provided by Drs, S. Hajdu and U. Larsson (Himmerfjärden Eutrophication Study). D. Jones is acknowledged for help with fieldwork, language check and R-scripts. We thank Askö laboratory staff for help with field work.
Author Contributions
Conceived and designed the experiments: AMLK RE. Analyzed the data: AMLK EG. Contributed to the writing of the manuscript: AMLK EG RE. Performed the field research and sample processing: AMLK.
References
- 1. Elmgren R (1978) Structure and dynamics of Balthic benthos communities, with particular reference to the relationship between macro- and meiofauna. Kieler Meeresforsch Sonderh 4: 1–22.
- 2. Graf G (1992) Benthic-pelagic coupling: a benthic view. Oceanogr Mar Biol Annu Rev 30: 149–190.
- 3. Goedkoop W, Johnson RK (1996) Pelagic-benthic coupling: Profundal benthic community response to spring diatom deposition in mesotrophic Lake Erken. Limnol Oceanogr 41: 636–647.
- 4. Ólafsson E, Elmgren R (1997) Seasonal dynamics of sublittoral meiobenthos in relation to phytoplankton sedimentation in the Baltic Sea. Estuar Coast Shelf Sci 45: 149–164.
- 5. Lehtonen KK, Andersin AB (1998) Population dynamics, response to sedimentation and role in benthic metabolism of the amphipod Monoporeia affinis in an open-sea area of the northern Baltic Sea. Mar Ecol Prog Ser 168: 71–85.
- 6. Goedkoop W, Demandt M, Ahlgren G (2007) Interactions between food quantity and quality (long-chain polyunsaturated fatty acid concentrations) effects on growth and development of Chironomus riparius. Can J Fish Aquat Sci 64: 425–436.
- 7. Nascimento FJA, Karlson AML, Näslund J, Gorokhova E (2009) Settling cyanobacterial blooms do not improve growth conditions for soft-bottom meiofauna. J Exp Mar Biol Ecol 368: 138–146.
- 8. Levin LA, Blair NE, Martin CM, DeMaster DJ, Plaia G, et al. (1999) Macrofaunal processing of phytodetritus at two sites on the Carolina margin: in situ experiments using C-13-labeled diatoms. Mar Ecol Prog Ser 182: 37–54.
- 9. Byrén L, Ejdung G, Elmgren R (2006) The use of sedimentary organic matter by deposit-feeding amphipods, studied using three isotope tracers. Mar Ecol Prog Ser 313: 135–143.
- 10. Nascimento FJA, Karlson AML, Elmgren R (2008) Settling blooms of filamentous cyanobacteria as food for meiofauna assemblages. Limnol Oceanogr 53: 2636–2643.
- 11. Karlson AML, Nascimento FJA, Näslund J, Elmgren R (2010) Higher diversity of deposit-feeding macrofauna enhances phytodetritus processing. Ecology 91: 1414–1423.
- 12. Rudnick DT (1989) Time lags between the deposition and meiobenthic assimilation of phytodetritus. Mar Ecol Prog Ser 50: 231–240.
- 13.
Cederwall H (1979) Energy flow and fluctuations of deeper soft bottom communities in the Baltic Sea. Doctoral thesis. Askö laboratory and Department of Zoology, Stockholm University.
- 14. Raateoja M, Seppälä J, Kuosa H, Myrberg K (2005) Recent changes in trophic state of the Baltic Sea along SW coast of Finland. Ambio 34: 188–191.
- 15. Suikkanen S, Laamanen M, Huttunen M (2007) Long-term changes in summer phytoplankton communities of the open northern Baltic. Est Coast Shelf Sci 71: 580–592.
- 16. Karlson AML, Nascimento FJA, Elmgren R (2008) Incorporation and burial of carbon from settling cyanobacterial blooms by deposit-feeding macrofauna. Limnol Oceanogr 53: 2754–2758.
- 17. Karlson AML, Mozuraitis R (2011) Deposit-feeders accumulate the cyanobacterial toxin nodularin. Harmful Algae 12: 77–81.
- 18. Basen T, Rothhaupt K-O, Martin-Creuzburg D (2012) Absence of sterols constrains food quality of cyanobacteria for an invasive freshwater bivalve. Oecologia 170: 57–64.
- 19. Walve J, Larsson U (2007) Blooms of Baltic Sea Aphanizomenon sp (cyanobacteria) collapse after internal phosphorus depletion. Aquat Microb Ecol 49: 57–69.
- 20. Ahlgren G, Gustafsson IB, Boberg M (1992) Fatty acid content and chemical composition of freshwater microalgae. J Phycol 28: 37–50.
- 21. Sylvander P, Haubner N, Snoeijs P (2012) The thiamine content of phytoplankton cells is affected by abiotic stress and growth rate. Microb Ecol 65: 566–577.
- 22. Vehmaa A, Hogfors H, Gorokhova E, Brutemark A, Holmborn T, et al. (2013) Projected marine climate change: effects on copepod oxidative status and reproduction. Ecol Evol 3: 4548–4557.
- 23. Basen T, Gergs R, Rothhaupt K-O, Martin-Creuzburg D (2013) Phytoplankton food quality effects on gammarids: benthic–pelagic coupling mediated by an invasive freshwater clam. Can J Fish Aquat Sci 70: 198–207.
- 24. DeNiro MJ, Epstein S (1978) Influence of diet on the distribution of carbon isotopes in animals. Geochim Cosmochim Acta 42: 495–506.
- 25. Peterson BJ, Fry B (1987) Stable isotopes in ecosystem studies. Ann Rev Ecol Syst 18: 293–320.
- 26.
Fry B (2006) Stable Isotope Ecology. Springer Science + Business Media, LLC, New York.
- 27. Rolff C (2000) Seasonal variations in delta C-13 and delta N-15 of size fractioned plankton at a coastal station in the northern Baltic proper. Mar Ecol Prog Ser 203: 47–65.
- 28. Nordström M, Aarnio K, Bonsdorff E (2009) Temporal variability of a benthic food web: patterns and processes in a low-diversity system. Mar Ecol Prog Ser 378: 13–26.
- 29. Lesutiene J, Bukaveckas PA, Gasiunaite ZR, Pilkaityte R, Razinkovas-Baziukas A (2014) Tracing the isotopic signal of a cyanobacteria bloom through the food web of a Baltic Sea coastal lagoon. Est Coast Shelf Sci 138: 47–56.
- 30. Goering J, Alexander V, Haubenstock N (1990) Seasonal variability of stable carbon and nitrogen isotope ratios of organisms in a North Pacific bay. Est Coast Shelf Sci 30: 239–260.
- 31. Limén H, Ólafsson E (2002) Ostracod species-specific utilisation of sediment detritus and newly settled cyanobacteria, Aphanizomenon sp., in the Baltic Sea: evidence from stable carbon isotopes. Mar Biol 140: 733–738.
- 32. McGoldrick DJ, Barton DR, Power M, Scott RW, Butlera BJ (2008) Dynamics of bacteria–substrate stable isotope separation: dependence on substrate availability and implications for aquatic food web studies. Can J Fish Aquat Sci 65: 1983–1990.
- 33. Layman CA, Arrington DA, Montãna CG, Post DM (2007) Can stable isotope ratios provide for community-wide measures of trophic structure? Ecology 88: 42–48.
- 34. Jackson AL, Inger R, Parnell AC, Bearhop S (2011) Comparing isotopic niche widths among and within communities: SIBER – Stable Isotope Bayesian Ellipses in R. J Anim Ecol 80: 595–602.
- 35. Tecchio S, van Oevelen D, Soetaert K, Navarro J, Ramirez-Llodra E (2013) Trophic dynamics of deep-sea megabenthos are mediated by surface productivity. PloS one 8(5): e63796
- 36. Uitto A, Sarvala J (1991) Seasonal growth of the benthic amphipods Pontoporeia affinis and P. femorata in a Baltic archipelago in relation to environmental factors. Mar Biol 111: 237–246.
- 37. Blank M, Laine AO, Jürss K, Bastrop R (2008) Molecular species identification key based on PCR/RFLP for discrimination of three polychaete sibling species of the genus Marenzelleria, and their current distribution in the Baltic Sea. Helgol Mar Res 62: 129–141.
- 38. Post DM, Layman CA, Arrington DA, Takimoto G, Quattrochi J, et al. (2007) Getting to the fat of the matter: models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia 152: 179–189.
- 39. Hill C, Quigley A, Cavaletto JF, Gordon W (1992) Seasonal changes in lipid content and composition in the benthic amphipods Monoporeia affinis and Pontoporeia femorata. Limnol Oceanogr 37: 1280–1289.
- 40. Hansson S, Hobbie JE, Elmgren R, Larsson U, Fry B, et al. (1997) The stable nitrogen isotope ratio as a marker of food web interactions and fish migration - a comparison of three coastal Baltic Sea stations. Ecology 78: 2249–2257.
- 41. Savage C, Elmgren R (2004) Macroalgal (Fucus vesiculosus) delta 15-N values trace decrease in sewage influence. Ecol Appl 14: 517–526.
- 42.
R Development Core Team (2012) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. Vienna, Austria.
- 43.
Parnell A, Inger R, Bearhop S, Jackson AL (2008) SIAR: Stable isotope analysis in R.
- 44. Syväranta J, Lensu A, Marjomäki TJ, Oksanen S, Jones RI (2013) An empirical evaluation of the utility of convex hull and standard ellipse areas for assessing population niche widths from stable isotope data. PloS one 8(2): e56094.
- 45. Brind'Amour A, Dubois SF (2013) Isotopic diversity indices: how sensitive to food web structure? PloS one 8: e84198.
- 46.
Zakrisson A, Larsson U, Höglander H (2014) Do Baltic Sea diazotrophic cyanobacteria take up combined nitrogen in situ? J Plankton Res doi:10.1093/plankt/fbu053.
- 47.
Rolff C (1998) Stable isotope studies of contaminant and material transport in Baltic pelagic food-webs. Doctoral thesis. Stockholm University Sweden. ISBN 91-87272-62-8.
- 48. Hayes JM, Freeman KH, Popp BN, Hoham CH (1990) Compound-specific isotopic analyses: A novel tool for reconstruction of ancient biogeochemical processes. Org Geochem16: 1115–1128.
- 49. Wolf N, Carleton SA, Martínez del Rio C (2009) Ten years of experimental animal isotopic ecology. Funct Ecol 23: 17–26.
- 50. Goedkoop W, Åkerblom N, Demandt MH (2006) Trophic fractionation of carbon and nitrogen stable isotopes in Chironomus riparius reared on food of aquatic and terrestrial origin. Freshw Biol 51: 878–886.
- 51. Ploug H, Musat N, Adam B, Moraru CL, Lavik G, et al. (2010) Carbon and nitrogen fluxes associated with the cyanobacterium Aphanizomenon sp. in the Baltic Sea. The ISME Journal 4: 1215–1223.
- 52. Wannicke N, Korth F, Liskow I, Voss M (2013) Incorporation of diazotrophic fixed N2 by mesozooplankton — Case studies in the southern Baltic Sea. J Mar Syst 117–118: 1–13.
- 53. Karlson AML, Näslund J, Blomgren Ryden S, Elmgren R (2011) Polychaete invader enhances resource utilization in a species-poor system. Oecologia 166: 1055–1065.
- 54. Segerstråle S (1973) Results of bottom trawling in certain localities in the Tvärminne-Askö area (inner Baltic) with special reference to the so-called Macoma-Pontoporeia theory. Commentat Biol Soc Sci Fenn 67: 1–12.
- 55. Darimont CT, Paquet PC, Reimchen TE (2007) Stable isotopic niche predicts fitness of prey in a wolf–deer system. Biol J Linn Soc Lond 90: 125–137.