Ecological Research

Food webs and river drainages are both hierarchical networks and complex adaptive systems. How does living within the second affect the first? Longitudinal gradients in productivity, disturbance regimes and habitat structure down rivers have long interested ecologists, but their effects on food web structure and dynamics are just beginning to be explored. Even less is known about how network structure per se influences river and riparian food webs and their members. We offer some preliminary observations and hypotheses about these interactions, emphasizing observations on upstream–downstream changes in food web structure and controls, and introducing some ideas and predictions about the unexplored question of food web responses to some of the network properties of river drainages.

The leaves of woody plants at Harvard Forest in Central Massachusetts, USA, changed color during senescence; 70% (62/89) of the woody species examined anatomically contained anthocyanins during senescence. Anthocyanins were not present in summer green leaves, and appeared primarily in the vacuoles of palisade parenchyma cells. Yellow coloration was a result of the unmasking of xanthophyll pigments in senescing chloroplasts. In nine red‐senescing species, anthocyanins were not detectable in mature leaves, and were synthesized de novo in senescence, with less than 20 µg cm⁻² of chlorophyll remaining. Xanthophyll concentrations declined in relation to chlorophyll to the same extent in both yellow‐ and red‐leaved taxa. Declines in the maximum photosystem II quantum yield of leaves collected prior to dawn were only slightly less in the red‐senescing species, indicating no long‐term protective activity. Red‐leaved species had significantly greater mass/area and lower chlorophyll a/b ratios during senescence. Nitrogen tissue concentrations in mature and senescent leaves negatively correlated to anthocyanin concentrations in senescent leaves, weak evidence for more efficient nitrogen resorption in anthocyanic species. Shading retarded both chlorophyll loss and anthocyanin production in Cornus alternifolia, Acer rubrum, Acer saccharum, Quercus rubra and Viburnum alnifolium. It promoted chlorophyll loss in yellow‐senescing Fagus grandifolia. A reduced red : far‐red ratio did not affect this process. Anthocyanins did not increase leaf temperatures in Q. rubra and Vaccinium corymbosum on cold and sunny days. The timing of leaf‐fall was remarkably constant from year to year, and the order of senescence of individual species was consistent.

We investigated the patterns of soil nitrogen (N) and forest floor light availability, forest structure and tree species distribution along a topographic gradient on a 200‐m mountain slope in a cool‐temperate deciduous broad‐leaved forest in Japan. Rates of soil N mineralization and nitrification decreased from lower to upper slope positions, revealing that N availability decreased up the slope. Maximum tree height and above‐ground biomass were greater on the lower than the upper parts of the slope. Canopy openness, an index of light availability at the forest floor, increased up the slope. Tree species could be placed into three groups according to their distribution patterns on the slope. ‘Ridge’ and ‘valley’ species were distributed on the upper and lower parts of the slope, respectively, whereas ‘uniform’ species were distributed over the entire slope. Topography‐mediated resource gradients of soil N and light may be important determinants of species distribution patterns and forest regeneration, and the results of this study imply that the determinants of the regeneration process differ between the lower and upper parts of a slope. The former may be relatively light limited and the latter may be soil N limited. Valley species may have a greater ability to compete for light, whereas ridge species have a greater ability to compete for soil N. The broad distribution of uniform species probably reflects an ability to effectively compete for both light and soil N.

Regeneration traits of six co‐occurring Salix species were studied on a floodplain of the Sorachi River, central Hokkaido, Japan, and their colonization success and coexistence in a local habitat were discussed. Mixed Salix communities contained six Salix species; dominant:S. sachalinensis; four subordinates:S. rorida, S. pet‐susu, S. miyabeana and S. subfragilis; rare species:S. jessoensis. Their phenology, falling velocity and longevity of seeds, and the effects of microtopography and soil texture on seedling establishment were studied. The six Salix species had overlapped seed dispersal periods that coincided with the decrease of water level after a predictable spring flood. This coincidence was crucial for the colonization success because the seedlings were established on wet soils left by the decreasing water level. They showed two types of regeneration trait, specialization and generalization. S. rorida and S. subfragilis showed contrasting regeneration traits; early vs. late seed dispersal, large vs. small seeds, seedling distribution on coarse vs. fine soils, respectively. These two species rarely co‐occurred. On the other hand, the dominant S. sachalinensis had an intermediate seed size and dispersal timing, and a wide range of seedling distribution from coarse to fine soils. These results revealed that flooding seasonality influenced the colonization success together with the regeneration traits of Salix species, and that coexistence of the Salix species was facilitated primarily by regeneration niche separation related to flooding seasonality and soil heterogeneity.

Soil respiration was measured throughout the year (June 1992 to May 1993) in a mature, deciduous, broad‐leaved forest and an adjacent, clear‐felled stand which was made in November 1991, in Hiroshima Prefecture, west Japan. The same soil temperature and soil moisture content as those in the forest stand were maintained in two frame boxes covered with sheets of white netting in the clear‐felled stand to observe soil respiration. A herbicide was applied to the cut end of all stumps in one of the two frame boxes in order to kill the root system. There was no significant difference in the aboveground biomass and soil environmental conditions between the forest and the frame boxes in the clear‐felled stands. The difference in soil respiration rate between the forest and the frame box, in which the root system was killed by the herbicide, was considered to be due largely to the contribution of root respiration. Taking into consideration CO₂ evolution due to the decomposition of roots killed and the change in A₀ layer respiration rate after clear‐felling, the proportion of root respiration to the total soil respiration before clear‐felling was estimated to be 51% annually, which coincides closely with those values estimated previously in mature forests by other methods. The difference in the soil respiration rate between the two frame boxes (one with killed roots and the other with undisturbed roots) suggested that the annual root respiration rate just after clear‐felling dropped to about two‐thirds (70%) of that before clear‐felling.

Ontogenetic diet shifts are a widespread phenomenon among vertebrates, although their relationships with life history traits are poorly known. We analyzed the relative importance of body size, age and maturity stage as determinants of the diet of a marine top predator, the copper shark, Carcharhinus brachyurus, by examining stomach contents using a multiple‐hypothesis modeling approach. Copper sharks shifted their diet as size and age increased and as they became sexually mature, incorporated larger prey as they grew, and had a discrete shift in diet with body size, with only individuals larger than ≈200 cm total length able to prey on chondrichthyans. Body size was the most important trait explaining the consumption of chondrichthyans, while age determined the consumption of pelagic teleosts. Pelagic teleosts were consumed mostly by medium‐aged sharks, a result, probably, of a risk‐reducing feeding strategy at young ages coupled with either a senescence‐related decline in performance or a change in sensory capabilities as sharks age. Copper sharks of all sizes were able to cut prey in pieces, implying that gape limitation (i.e., the impossibility of eating prey larger than a predator's mouth) did not play a role in producing the diet shift. Our results suggest that, contrary to the current practice of setting minimum but not maximum size limits in catches, any plan to conserve or restore the ecological function of sharks, through their predatory control of large prey, should aim to maintain the largest individuals.

We studied the autumn food habits of the Asiatic black bear (Ursus thibetanus) from 1993 to 1997, based on 202 fecal samples in the Chichibu Mountains, central Japan. Nuts occupied the highest proportions of autumn foods (59.9–85.8% important values). Although the proportion of nuts of Quercus crispula, Fagus crenata and Fagus japonica varied greatly between the years, acorns of Q. crispula were most prevalent in four of the five years. We also determined the relative nut production of these three species by counting the number of nuts or cupules on the ground. Black bears consumed the nuts according to their relative availabilities. Nuts of Q. crispula appeared to be the most important food because: (i) these nuts were eaten in the highest proportion in four of the five study years; and (ii) even in poor years the bears consumed acorns of Q. crispula, whereas nuts of Fagus spp. were not consumed. We discuss the significance of alternative foods for black bears in relation to food tree diversity in the forest and the necessity for long‐term studies examining the food habits of Asiatic black bears.

A one‐male group (BE‐Group) of proboscis monkeys was studied along the Menanggul River, a tributary of the Kinabatangan River, Sabah, Malaysia, from May 2005 to 2006. It has generally been assumed that proboscis monkeys only set up their sleeping sites along the riverbank; however, when more than 1 m of water covered the forest floor for more than 700 m inland from the riverbank during the seasonal flood, the BE‐Group slept inside the forest. It seems that the sleeping‐site selection of the BE‐Group was not influenced by food availability during the flooded months because the food availability by the vegetational survey did not vary much between flooded and non‐flooded months. In addition, feeding behaviors of the focal monkey in the BE‐Group also did not vary much between flooded and non‐flooded days. On the other hand, the water level statistically influenced the sleeping‐site selection. The proboscis monkeys remained in inland forest during the flooded days because of the reduced predation threat, as terrestrial predators such as clouded leopards are prevented from foraging by deep water covering the forest floor. On non‐flooded days when the BE‐Group slept at the riverbank, they frequently slept close to other one‐male groups on the riverside trees. Contrastingly, when the group slept inside the forest on flooded days when the water level was high, they slept away from other groups. This difference in the need for one‐male groups to sleep close to each other might be attributed to the decreased predation threat during high water level in the flooded days.

Twelve salt marshes in south Louisiana (USA) were classified as either ‘impaired’ or ‘healthy’ before a summer sample collection of above‐ and below‐ground biomass and determination of sediment accretion rates. The above‐ground biomass of plant tissues was the same at both impaired and healthy salt marshes and was not a good predictor of marsh health. However, below‐ground root biomass in the upper 30 cm was much lower in the impaired marshes compared to the healthy marshes. Compromises to root production apparently occur before there is an obvious consequence to the above‐ground biomass, which may quickly collapse before remedial action can be taken. The subsequent change in vertical position of the marsh surface may be equivalent to many years of accretion, and be irreversible within decades without considerable effort. These results are consistent with the hypothesis that it is the plant's below‐ground accumulation of organic matter, not inorganic matter that governs the maintenance of salt marsh ecosystem in the vertical plane. Reversing the precursor conditions leading to marsh stress before the collapse of the above‐ground biomass occurs is therefore a prudent management objective and could be easier than restoration.