Research Summaries

Food webs in aquatic ecosystems

Stable-isotope techniques have recently been used to provide evidence of spatial structure within aquatic food webs. Across various habitats and ecosystems, the factors determining the major resources of aquatic food webs are primarily phytoplanktonic productivity, benthic algal productivity, and amount of subsidization from terrestrial habitats. Autochthonous and allochthonous resource availability in food webs shifts with gradients in water depth, nutrient concentrations, degree of canopy cover, and distance from terrestrial habitats (Numbers in Publications: 13, 17, 28). Size of lake and river ecosystem (i.e., lake volume and river width) also affects the relative contribution of the resources to the food webs, as this factor determines the ecosystem productivity and linkage to terrestrial habitats (35, 36). Human activities on river and lake ecosystems have subsequently modified the structure of food webs (28, 45).


Lake and pond ecosystems: Food chain length (FCL) has been studied extensively, and numerous hypotheses to predict FCL; productivity, ecosystem size, and productive-space have been proposed. We estimated FCL in 15 ponds using stable isotope techniques to test the most common hypotheses for predicting FCL. We found that productive-space (edible carbon + pond volume) was the best model to predict FCL. Therefore, our results suggest that both resource availability and ecosystem size predict FCL in pond ecosystems and play significant roles in maintaining longer FCLs (35). Ecosystem history may set important evolutionary constraints on community composition and food web structure. FCL has long been recognized as a fundamental ecosystem attribute. We found that food chains in the world’s ancient lakes were significantly shorter than in recently formed lakes and reservoirs, despite the fact that ancient lakes harbored much higher species richness, including many endemic species. Our results highlight a counter-intuitive and poorly-understood role of evolutionary history in shaping key food web properties such as FCL (65).  

    The deposit feeders inevitably assimilate food from different origins in a selective manner from sediment, including benthic and planktonic microalgae. We conducted the microcosm experiments with chironomid larvae and algae to determine selective assimilation by deposit feeder from mixed diets. The results clearly showed that the deposit feeding chironomid larvae selectively assimilated phytoplankton and benthic diatoms as fresh deposits from bulk sediments, and assimilated phytoplankton more readily than benthic diatoms (13, 17).


Stream food webs: Dam-reservoir system can subsidize the downstream food webs and alter community structure of macroinvertebrates. We estimated the subsidization using stable-isotope mixing model by carbon and nitrogen isotopes. Along with the distance from the dam, the subsidy effects on the food webs decreased and changes in macroinvertebrate community, especially filter-feeding groups (28).

    Reach-scale effects of canopy cover on trophic pathways to a net spinning filter-feeder and a grazer were investigated using carbon and nitrogen isotopes in a mountain stream. The results indicated that the main food sources of the grazer and the filter-feeder were derived from the in-situ benthic algal source at open reaches and from the terrestrial source at canopy covered reaches (20). In a headwater stream, we estimated the contribution of chemoautotrophic production to freshwater snails in a headwater stream using carbon, nitrogen, and sulfur isotopes (16, 18).


Estuarine and marine food webs: We carried out carbon and nitrogen stable isotope analyses on estuarine macrozoobenthos in order to examine relationships between their feeding habits (feeding mode and food selectivity) and the spatial shifts in food sources from upstream to downstream in an estuary. These results show that species-specific feeding habits should be considered when evaluating the roles of macrozoobenthos in regulating estuarine material flows (9). We performed comparisons of food sources among different gastropod species on sub-tropical and temperate tidal flats (26° and 38°N, respectively), and used isotope mixing models using carbon and nitrogen stable isotopes. A mixing model for stable isotopes, IsoSource, revealed that main food sources for three snails were macroalgae (50-56%) and seagrass (39-45%) at temperate site. At the sub-tropical site, snails fed mainly on macroalgae. It would thus appear that macroalgae and seagrass play an important role in the food webs not only in their own habitats but also on the adjacent tidal flats (38). 

    In coastal areas, aquaculture has accelerated in recent decades, and attached algae and invertebrates proliferate on the farming cages. We hypothesized that the attached algae on the farm structures is important to planktonic food webs. Our results show that attached microalgae from sea farms are important food sources for planktonic food webs in areas with fish farms and that the two food webs, attached and pelagic, are coupled through zooplankton grazing (23). We estimate two food source compositions for the cultured pearl oyster using stable isotope and stomach contents analysis. The results indicated that the oyster feed on a mixture of phytoplankton and attached microalgae on farm structures, and that the attached microalgae on pearl cages can serve as an important food source for them as well as phytoplankton (22). Oysters are long-term integrators of isotope ratios from their diet and carbon isotope signature of oysters can serve as a more accurate bioindicator of isotopic baselines than nitrogen for marine ecological study (30).


Prey-predator interactions

Our experiment results showed that a significantly greater number of the insect grazers were attracted to thin and thick algal mats than to the untreated ceramic plate. Thus, the insect grazers can recognize and respond to the abundance of microalgae through microalgal cues, which induce the movement of the insect grazers to habitats with high microalgal biomass (12). We attempted to detect differences in the patterns of top-down effects of herbivores among multiple spatial scales in a stream ecosystem. The relationship patterns between the insect grazer and periphyton were detected more clearly at larger scales than at smaller scales. Since specific interactions between Glossosoma larvae and periphyton may occur at a microhabitat scale, we considered multiple spatial scaling to understand the top-down effect of herbivores (27)

    Shift in the feeding behavior based on the intensity of microalgal cues would be beneficial, as it would allow the grazer to

feed on more periphyton (microalgal mat on stone). Using a caddisfly grazer, we conducted a laboratory channel experiment with upstream experimental plates having levels of periphyton abundance, and we recorded the movement behavior of the larvae. As periphyton abundance increased, the orientation of the crawling path significantly increased. That is, larvae crawled straight to the abundant periphyton patch. The behavior change was likely due to the detection of some microalgal cue (42).

Meta-analyses to synthesize ecological and biological themes

Metabolic scaling of organisms: Metabolic theory proposes that individual growth is governed through the mass- and temperature-dependence of metabolism, and ecological stoichiometry posits that growth is maximized at consumer-specific optima of resource elemental composition. gross growth efficiencies (GGEs) for elements are defined by the ratio of metabolism-dependent processes. We used data from 95 published studies to evaluate these metabolic-dependencies of GGEs from unicells to vertebrates. We showed that GGEs commonly decline as power functions of asymptotic body mass and exponential functions of temperature. (46).

    Long-term behavior of radionuclides of organisms is an important issue for estimating possible associated risks to human beings and ecosystems. We performed meta-analysis to collect published data of the long-term 137Cs decay process of fish species to estimate how biological (metabolic rate) and ecological (trophic position, habitat, and diet type) influence on the long-term decay process of 137Cs concentration in fish. We found that 1) the trophic position could predict maximum day of 137Cs concentration in fish, 2) metabolic rate of the fish species and water temperature in the environments could predict biological half lives and decay rate of fish species (60).

Plant and animal phenology and climate change

 In Japan, phonology of >120 animal and plant species were recorded across Japan by Japanese Metrological Agency (JMA). Using JMA dataset, the flowering date of apricots has been advanced, with a notable shift in regimes between 1953-1989 and 1990-2005. Winter flowering of Japanese apricot has been influenced by recent climate changes, especially by drastic climate-related shifts in the timing of key processes (19). Budburst dates advanced during 1953-2005 only in those localities with an increase of temperatures during spring (25). The appearance of the first adult dragonfly has significantly shifted to later in the spring in the past five decades (32).

    Flowering of cherry tree tended to occur earlier over the last three decades, whereas the appearance of the butterfly was delayed. The effects of climate and the timing of the sensitive period differ between both trophic levels. The plants were strongly affected by temperature 30–40 days prior to flowering, whereas the butterfly was less affected by temperature, and the effects mainly occurred during the 15 days prior to its appearance. The phenologies of the plants and butterfly are changing in opposite directions because they use different climatic cues with different temporal trends (31).

    We hypothesized that the phenological response of plants varied with latitude. To test the hypothesis, we estimated the phenological response to long-term climate change using autumn events of phenology were delayed by recent climate change. Our results showed that the single regression slopes of the phenological responses at lower latitude were larger than those at higher latitudes. We found the negative relationships between the leaf phenological responsiveness and the latitude. The findings would be important to predict the phenological timing with global climate changes (29).

    We estimated the regional variation across Japan in flowering and leaf budburst dates of plants based on a dataset of phenological timings from 1953 to 2005. The observed plants’ genetic diversity varied according to human cultivation. The within-species variations of phenological response to temperature as well as regional variations were less in the plant populations with lower genetic diversity. Under increased temperatures, low variation in phenological responses may allow drastic changes in the phenology of plant populations with synchronized phenological timings (44).