Stuart G. Fisher

Energy Flow in Bear Brook, New Hampshire: An Integrative Approach to Stream Ecosystem Metabolism

An annual energy budget is presented for Bear Brook, a small undisturbed second-order stream in northeastern United States. The ecosystem approach, in which all input and output fluxes of potential energy as organic matter are considered, is used to describe the dynamics of energy flow in a 1,700-m segment of the stream. The annual input of energy to the system is 6,039 Kcal/m2. Over 99% of this is allochthonous, from the surrounding forested watershed or from upstream areas. Autochthonous primary production by mosses accounts for less than 1% of the total energy available to the ecosystem. Algae and vascular hydrophytes are absent from the stream. Meteorologic inputs (litter and throughfall) from the adjacent forest account for 44% of annual energy input. Most of this is in particulate form. The remaining 56% of input enters by geologic vectors (inflowing surface and subsurface waters). Eighty-three per cent of the geologic input and 47% of the total input of energy occur as dissolved organic matter. Approximately 4,730 Kcal/m2 of organic detritus, nearly equally divided between leaves and branches, is stored within the system. The size of this detritus reservoir is stable from year to year. The turnover time of the branch compartment is about 4.2 years; of the leaf compartment, about 1 year. Although much of the annual input of energy is in a dissolved state, dissolved organic matter does not tend to accumulate in the system and displays a very rapid rate of turnover. Sixty-six per cent of annual energy input is exported to downstream areas in stream water. The remaining 34% is lost as heat through consumer activity. Bear Brook is a strongly heterotrophic steady-state system in which import and export of organic matter play a significant role. A conceptual scheme is presented by which import, export, photosynthesis, and respiration may be used to describe the functional dynamics and developmental processes of ecosystems.

TEMPORAL SUCCESSION IN A DESERT STREAM ECOSYSTEM FOLLOWING FLASH FLOODING

Recovery of a desert stream after an intense flash flooding event is described as a model of temporal succession in lotic ecosystems. A late summer flood in Sycamore Creek, Arizona, virtually eliminated algae and reduced invertebrate standing crop by 98%. Physical and morphometric conditions typical of the preflood period were restored in 2 d and the biota recovered in 2—3 wk. Algal communities responded rapidly and achieved a standing crop of nearly 100 g/m2 in 2 wk. Community composition was dominated by diatoms early in succession and by filamentous greens and blue—greens later. Macroinvertebrates also recolonized denuded substrates rapidly, largely by immigration of aerial adults and subsequent oviposition. Growth and development were rapid and several generations of the dominant mayfly and dipteran taxa were completed during the 1st mo of recovery. Invertebrate dry biomass reached 7.3 g/m2 in 1 mo. Gross primary production (Pg) measured as O2 increased in a similar asymptotic fashion and reached 6.6 g°m—2°d—1 in 30 d. Pg exceeded community respiration (R) after day 5 and Pg/R averaged 1.46 for the remainder of the 2—mo sequence. This ecosystem is thus autotrophic and exports organic matter downstream and by drying, laterally. Uptake of nitrate and phosphorus were proportional to net primary production and exhibited a marked downstream decline in concentration during both light and dark periods. Temporal trajectories of various community and ecosystem attributes are compared with those suggested by Odum (1969) to be diagnostic of successional status. Agreement was poor in attributes which are especially modified in open, frequently disturbed ecosystems such as streams.

Ecosystem Expansion and Contraction in Streams Desert streams vary in both space and time and fluctuate dramatically in size

St re<l ms are hydrologically diverse and dynamic ecosystems. Flow may vary between extremes, from high-discharge floods to periods when surface water is absent. Although much is known about the role of floods in shaping ecological processes, far less is known about the biological and chemical changes that occu r du ring periods of water loss in stream ecosystems (Boulton and Suter 1986, Stanley and Fisher 1992). Nowhere is this lack of knowledge more apparent than in desert streams; these lotic ecosystems exist in a setting defined by water limitation, and periods of declining or absent flow are common. However, water loss is by no means unique to desert streams, because intermittent streams are found in many different environments. Moreover, escalating demands on a finite water supply arc increasing the likelihood of drying in streams and rivers worldwide. Irrigation, impoundment, diversion, and groundwater abstraction reduce streamflow in mesic and xeric regions alike. In arid and semiarid areas, large rivers that are devoid of water are common, and in more mesic locales, profligate water use decreases the total amount of surface water present

Exchange between interstitial and surface water: implications for stream metabolism and nutrient cycling

Metabolism of a Sonoran Desert stream was investigated by both enclosure and whole system oxygen techniques. We used recirculating chambers to estimate surface sediment metabolism and measured deep sediment respiration in isolated sediment cores. Metabolism of the stream ecosystem was determined for a 30-m reach as dark and light oxygen change with and without black plastic sheeting that darkened the stream and prevented diffusion. Average ecosystem respiration for two dates in August (440 mg O2 m-2 h-1) exceeded respiration of either the surface sediment community (155 Mg O2 m-2 h-1) or the hyporheic community (170 mg O2 m-2 h-1) alone. Deep sediments show substantial oxygen and nitrate uptake when isolated. In the stream, this low nitrate interstitial water is exchanged with surface water. Metabolism of the isolated surface community suggests a highly productive and autotrophic system, yet gross production is balanced or exceeded by community respiration when ecosystem boundaries include the hyporheic zone. Thus, despite high rates of gross primary production (600–1200 mg O2 m-2 h-1), desert streams may be heterotrophic (PG < R) during summer.

Nitrogen limitation in a Sonoran Desert stream

Four nutrient enrichment bioassay experiments were conducted in Sycamore Creek, Arizona, during summer and autumn 1983. In two experiments, nitrogen and phosphorus were added alone and in combination while in the other experiments nitrogen was added singly. In experiments involving enrichment of both nutrient-diffusing substrates (clay flowerpots) and streamwater overlying tile/gravel artificial substrates, nitrogen enrichment significantly enhanced rates of chlorophyll a accrual, primary production, and nitrogen uptake. Addition of phosphorus either singly or in combination with nitrogen did not result in significant responses of these parameters; thus ambient concentrations of phosphorus were above limiting levels, even when excess nitrogen was supplied. Nitrogen additions stimulated periphyton growth when background nitrate-N concentrations were ≤0.055 mg/L. We propose that nitrogen limitation is common in the desert Southwest since concentrations lower than this and atomic nitrogen to phosphorus ratios <16 occur in most (82% and 87%, respectively) previously surveyed southwestern streams (n=92). Temporal patterns of chlorophyll a accrual suggest that availability of nitrogen limited the rate of algal increase, but not the ultimate periphyton standing crop. If true, this hypothesis predicts that algal recolonization rates should vary depending on nitrogen supply. In desert streams, flood disturbances reduce algal standing crops to near zero, but postflood recovery periods may be quite long. Nitrogen limitation in desert streams thus may exert strong influence on rates and patterns of algal recolonization following floods.

Material Spiraling in Stream Corridors: A Telescoping Ecosystem Model

ABSTRACT Stream ecosystems consist of several subsystems that are spatially distributed concentrically, analogous to the elements of a simple telescope. Subsystems include the central surface stream, vertically and laterally arrayed saturated sediments (hyporheic and parafluvial zones), and the most distal element, the riparian zone. These zones are hydrologically connected; thus water and its dissolved and suspended load move through all of these subsystems as it flows downstream. In any given subsystem, chemical transformations result in a change in the quantity of materials in transport. Processing length is the length of subsystem required to “process” an amount of substrate equal to advective input. Long processing lengths reflect low rates of material cycling. Processing length provides the length dimension of each cylindrical element of the telescope and is specific to subsystem (for example, the surface stream), substrate (for instance, nitrate), and process (denitrification, for example). Disturbance causes processing length to increase. Processing length decreases during succession following disturbance. The whole stream-corridor ecosystem consists of several nested cylindrical elements that extend and retract, much as would a telescope, in response to disturbance regime. This telescoping ecosystem model (TEM) can improve understanding of material retention in running water systems; that is, their “nutrient filtration” capacity. We hypothesize that disturbance by flooding alters this capacity in proportion to both intensity of disturbance and to the relative effect of disturbance on each subsystem. We would expect more distal subsystems (for example, the riparian zone) to show the highest resistance to floods. In contrast, we predict that postflood recovery of functions such as material processing (that is, resilience) will be highest in central elements and decrease laterally. Resistance and resilience of subsystems are thus both inversely correlated and spatially separated. We further hypothesize that cross-linkages between adjacent subsystems will enhance resilience of the system as a whole. Whole-ecosystem retention, transformation, and transport are thus viewed as a function of subsystem extent, lateral and vertical linkage, and disturbance regime.

Secondary Production, Emergence, and Export of Aquatic Insects of a Sonoran Desert Stream

Aquatic insect secondary production, emergence, and export of adults to the adjacent terrestrial ecosystem were assessed in Sycamore Creek, Arizona, by means of benthic sampling, emer- gence traps, and catch-nets that passively sampled adults falling into the stream. Annual secondary production was 120.9 ? 18.0 g-m-2 yr-t and emergence was 23.1 g.m-2 yr-t (in dry mass units). The ratio of annual emergence to annual production (E/P) varied among taxa and ranged from 2 to 29%. Chironomids comprised 48.2% of production and 59.7% of emergence and mayflies accounted for 45.9 and 19.2%, respectively. Approximately 3% of emergent insect biomass returned to the stream; thus 22.4 g m-2 yr-I was transferred to the adjacent terrestrial ecosystem. The transfer of a significant portion of aquatic insect biomass to the terrestrial habitat reduced insects available to stream insec- tivores while providing prey for insectivores in neighboring terrestrial ecosystems.

Vertical Hydrologic Exchange and Ecological Stability of a Desert Stream Ecosystem

The influence of hydrologic linkage between hyporheic and surface subsys- tems was investigated in sand-bottomed reaches of a desert stream. Direction of hydrologic exchange was measured as vertical hydraulic gradient (VHG) using mini-piezometers. Maps of VHG indicated upwelling (discharge from the interstitial regions into surface water) at the bases of riffles and heads of runs; downwelling (infiltration of surface water into the hyporheic zone) occurred at the bases of runs. Dissolved NO3-N in surface water was higher over or immediately downstream from upwelling zones. Loss of continued supply from the hyporheic zone and intense assimilatory demand by surface autotrophs generated longitudinal declines in NO-N and lower nutrient concentrations in downwelling zones. Algal standing crop (as chlorophyll a) was significantly higher in upwelling zones than in areas without positive VHG. Postflood trajectories of chlorophyll a indicated that algae at upwelling zones recovered from disturbance significantly faster than those at downwelling zones. Recovery rate was related to supply of NO-N from enriched interstitial water in the hyporheic zone. Hydrologic linkage integrates surface and hyporheic subsystems and increases ecosystem stability by enhancing resilience of primary producers following flash flood disturbance.

Denitrification in a nitrogen-limited stream ecosystem

Denitrification was measured in hyporheic, parafluvial, and bank sediments of Sycamore Creek, Arizona, a nitrogen-limited Sonoran Desert stream. We used three variations of the acetylene block technique to estimate denitrification rates, and compared these estimates to rates of nitrate production through nitrification. Subsurface sediments of Sycamore Creek are typically well-oxygenated, relatively low in nitrate, and low in organic carbon, and therefore are seemingly unlikely sites of denitrification. However, we found that denitrification potential (C & N amended, anaerobic incubations) was substantial, and even by our conservative estimates (unamended, oxic incubations and field chamber nitrous oxide accumulation), denitrification consumed 5–40% of nitrate produced by nitrification. We expected that denitrification would increase along hyporheic and parafluvial flowpaths as dissolved oxygen declined and nitrate increased. To the contrary, we found that denitrification was generally highest at the upstream ends of subsurface flowpaths where surface water had just entered the subsurface zone. This suggests that denitrifiers may be dependent on the import of surface-derived organic matter, resulting in highest denitrification rate at locations of surface-subsurface hydrologic exchange. Laboratory experiments showed that denitrification in Sycamore Creek sediments was primarily nitrogen limited and secondarily carbon limited, and was temperature dependent. Overall, the quantity of nitrate removed from the Sycamore Creek ecosystem via denitrification is significant given the nitrogen-limited status of this stream.

Physical and chemical characteristics of the hyporheic zone of a Sonoran Desert stream

The hyporheic zone of three reaches of Sycamore Creek, Arizona consisted of an average 63 cm depth of predominantly sand or fine gravel (0.5-5 mm). Sediments were highly porous (19-23% interstitial space) and interstitial water volume was 3-4 times that of surface water. Spatial distribution of temperature, sediment organic matter, interstitial nutrients, and subsurface oxygen indicate that physical-chemical conditions vary greatly within the hyporheic zone. Much of the observed variability may be due to repeated disturbance by flash floods. Organic matter content of sediment was low (0.08% by weight), variable, and generally declined with depth in shallow portions of the hyporheic zone. Hyporheic water temperature was higher than surface temperature in regions beneath the wetted perimeter in summer. Nutrient concentrations of interstitial water were enriched compared to surface water; ammonium-N, SRP, and nitrate-N were 269%, 174%, and 327% of surface concentration, respectively. Sub-surface velocity was low (0.62 mm/s), but vertical exchanges were pronounced. Interstitial oxygen was high in regions of infiltration (downwelling), and was generally reduced in discharge regions (upwelling), but subsurface patterns were otherwise complex. Vertical linkages between surface and hyporheic zones provide a mechanism for mutual influences. Chief among these are replenishment of interstitial oxygen by downwelling (and enhancement of aerobic respiration), and nutrient enrichment of surface water at upwelling sites.