f. herbert bormann gene e. likens pattern and process in a forested...

F. Herbert Bormann

Gene E. Likens

Pattern and Process in aForested EcosystemDisturbance, Development and the Steady StateBased on the Hubbard Brook Ecosystem Study

Springer-Verlag

Reorga&zation- a- - -zar-cutting, duringte accumulation of1 century when the-.. __at the end of the

vhich total biomassI fluctuates about a

:quent chapters, to3ment that couples

Hubbard Brook i slel and the smal lstudy are presented

CHAPTER 2

Energetics, Biomass,Hydrology, andBiogeochemistry of theAggrading Ecosystem

A logical place to begin our discussion of ecosystem development mightbe with the Reorganization Phase that immediately follows clear-cutting.However. for a number of reasons that seem to outweigh the risk of atemporary discomfiture to the reader’s sense of time. we shall begin ourdiscussion with the &radation Phase.

The Aggradation Phase is the one we have most intensively studied. 11is characterized by a storage of biomass and nutrients and by maximum-.- ._biotic regul&ion over ez.rgy, nutrient, and h&logic flux. Its outputrelationships are the most predictable of all of the proposed phases. It isour perception that the biogeochemistry and ecology of all the otherphases is best understood in terms of departures from the highlypredictable parameters of the Aggradation Phase.

Aggradation begins about 15 years after clear-cutting, when the totalbiomass curve of the ecosystem (Figure I-lO)sJ&s_from netUoss_to netaccumulation. This pattern continues until about Year JZ& when totalbm&hes a peak for the ecosystem. The period of aggradation is-_ --not one of uniform development or biomass.accumulation. The rate ofbiomass accumulation is at first relatively rapid as accumulation occurs inthe living biomass, dead wood. and forest floor compartments (Figure2-1). Later the rate slows as accumulation is limited primarily to livingbiomass, which at that time is also accumulating at a lesser rate. Markedchanges in forest structure, species composition, and dominance alsooccur during this 155yr period (these changes are discussed in Chapter 3).Compared to the Reorganization Phase (from 0 to 15 years after clear-cutting), when there is a net loss in total biomass and fairly.rapid anddrastic changes in other ecosystem parameters, the Aggradation Phase is

42 Pattern and Prccess in a Forested Ecosystem

Living Biomass

Years Atter Clear-Cutting

Figure 2-1. The accumulat ion of biomass during the Aggradation Phase ofecosystem development.

one of relat ive stabil i ty. That is . i ts energetic. hydrologic. and bio-geochemic relat ionships remain fairly constant and predictable. Manyof the stability relationships examined in this chapter are for the mostp a r t t h e l e a s t s u b t l e o n e s . More subtle aspects of the long-termrelationships are discussed in the chapter that deals \vith reorganization(Chapter 3).

Forests in the Aggradation Phase currently are found over most ofnorthern New England and are the base from which we have obtained ourempirical and experimental data. Since this phase represents a cont inuumof change. some of the conclusions about ecosystem functions such asweathering. nitrogen fixation. and control of stream-water chemistrybased on young aggrading stands (i.e.. 50-60 years old) may needmodification before they arc fully applicable to older stands within theAggradation Phase. On the other hand._conclusions on hydrology anderosion would seem to apply with little modification to the entire phase.

-Throughout this section we will draw on information presented in an earliertext. ~iopeochemis~: of N Forested Ecosystem (Likens et al.. 1977).

The Aggrading Ecosystem 43

SOLAR ENERGY FLOW

The stable conditions that characterize the aggrading ecosystem result inlarge part from its capacity to.,u$lize or control the flow of radiant ene_rgy : .impinging on it. This is done by reflection, heat loss, eyapot.ramiration.-_--

md-j56oxiynthetic use. The ecosystem has its greatest effect on energy )flow during the growing season (June through September), when 15% of ‘- -$ejn_cQming.radiation is reflected. 41% is lost as heat, 42%_is used in the- - - -evaporation of water (transpiration plus evaporation), and about 2% is /3ia-T; photosynth&is (Gosz et al., 1978).

--. - - - -

--.Sola_r- energy fixed in ~photosynthesis not only provides most of theenergy necessary to drive thebioiogical functions of the ecosystem but isalso stored within the ecosystem in the form of the carbon compoundsthat make up ecosystem-structure (Table 2-l). The 55yr-old forest atHubbard Brook contains aboutzl@-Kcal/ha in living biomass, whichaccumulates at a net rate of 1.2 5.lQ’Kcal/yr. About 1.4 X lO’Kcal/ha isstored in dead biomass, which accumulates at a net rate of about1.2 x 10’ Kcal/yr. Living and dead biomass provide the organic structure

Table 2-l. Energy Relationships of a 55Yr-Old Aggrading ForestEcosystem of the Hubbard Brook ExperimentalForest’

Kcal x 10Y/ha

Energy stored in ecosystem biomassLiving biomassDead woodDead biomass (organic matter in soil)Total

Incident solar energyAnnualGrowing season (June-September)

Annual conversion of solar energyPhotosynthesis (GPP)TranspirationTotal

0.710.151.302.16

10.104.43

0.101.751.85

Efficiency (%)b

Photosynthesis aloneAnnualGrowing season

Photosynthesis and transpirationAnnualGrowing season

1.02.3

18.341.8

‘Modified from Gosz et al., 1978.bRatio of conversion/incident energy.

Air andPrecipitation

ContamingNatural andPollutinqSubstances

L

oi r-Processes That

Modifv Climate /‘Y-’ ,&’ a,‘”

./J’\ I

-

processes That Chemicallycleanse Through-flowing

Root arAdsorption of Chemicals Promote lntdtratmrl 01

Uptake of Chemicals on Sod Particles Wdter into Soil1by HOO’Z -

Figure 2-2. Some of the direct and indirect uses of solar energy in the aggrading forest (Bormann, 1976).

rNa,er-holdingCapacity of Sol1

,

Cleaner AIIstream. LessIJust & Pollutantr

tMore PlearIngSUIWV~ Climate

Cleaner water10 stream* andGroundwater

* hlore Plraslngstreams and,,Cd’h~~!~Wats, S”ppller

Ab.NMlLYltof Flood Peaks

46 Pattern and Process in a Forested Ecosystem

sag.runoff psriod. (Figure - 33-‘). This temporal pattern suggests that

E~rhrcnium acts as a short-term.-nk .o.r_U:vernal dam.” with nutrientsincorporated in accumulating biomass during the spring flushing perrod

and released by shoot decomposition during the early summer . Such a

mechanism would reduce nutrient losses in spring streamflow. make themavailable to carlv summer growth. and thus preserve the nutrient capitalof the ecosystem (kluller and Bormann, 1976).

! ; During one year. about 1.0 X IO’” Kcal ofsolar energy_ are received by,I

11 each hectare at Hubbard Brook (Table 2-i). About 1:/o is-converted-h);’ _ photosynthesis into gross primary production. Approximately 55% of the;.:. GPP is<used to sustain green plant respiration. the remaining 15% (NPP)

supports all heterotrophic activities and storage in living and deadv L,

,__ ;L biomass.7.; :.: The photosynthetic conversion efficiency of l%/vr suggests a relatively--.----

Snowmelt

E0.-m

Ic 2 0

E

i IOc)

0

March April May June

Figure 2-3. (C) Stream-water nitrate concentrations and (3) biomass Increase inErythronium americanurn in 1972 and 1973. Biomass increase was calculated asthe rate of biomass change between any two consecutive harvest dates. Periodo f snowmelt and closure of the overstory c a n o p y a r e s h o w n (Muller a n dBormann. 1976).


modest-capacity~ of the ecosystem to capture and use solar energy.However, the utilization of solar energy in photosynthesis supports lifeand growth of the green plants and makes it possible for the ecosystemto wze-.&large amount of solar energy in the evaporation of waterby the process of transpiration. As a result, water that would leavethe ecosystem as liquid water in streamflow is converted to a gas andleaves as water vapor. The 1.75 x 10” KcaVha-yr of solar energy used intranspiration (the calculations of energy used in transpiration will be ”

.:

given in Chapter 3) greatly enhances the capacity of the ecosystem toresist external destabilizing forces to which it is continually subjected.such as wind. water. and gravity.

The use of energy in transpiration represents about 17% of all incidentenergy received by the ecosystem during 1 year. When energy used intranspiration is added to that of photosynthesis. annual efficiency is 18%.However. if we calculate efficiency on the basis of incident energy in Junethrough September. the months when almost all production and tran-spiration occurs. efficiency is about 37%.

Solar energy channeled into photosynthesis and transpiration may bethought of as sustaining active and passive ecosystem processes which ,;/” ,(greatly affect the overall biogeochemistry of the ecosystem. Transpiration, r ..:. .> ,carbon assimilation, nitrogen-fixation-and wat.er and nutr&n~uptak$ may---.. ---.-~~ 1’ C.--dbe thought of as active processes that require a continuous expenditure of >,‘. . .biologically fixed or biologically mediated solar energy. Whereas passive ‘V~~l,~..., .processes may be thought of as nonbiological (physical or chemical) ’ ,~ ;. _processes that require little immediate expenditure of biologically fixed orbiologically mediated solar energy, such as _r_efle_c_tion of sunlight. impac- : ~1” ’tionofaerasols. filtJation. and exchange p.rocesses in-the soil. These pro- ; -.cesses remove chemicals or particles from the streams of air and water 1 ‘~~*. !:moving through the ecosystem (Figure 2-2). The relationship between 5 +

active and passive processes is complex. For example, soil exchange ,:-. _-A =.- :processes may be viewed as passive, yet biologically derived exchange* :sites on humus play an important role in the process. Impaction, the _ , ,.+L~‘:process of removal of aerosols and dust from the airstream, presents r! ,i ‘,-another interesting case. Although this is a physical process, the biological r \-(.i^.* iAstructure and composition of the ecosystem exercises considerable control ‘+

3”’ ’over the intercepting surface and thus affects the rate of impaction. The l+r,;

architecture of the major impaction sites, i.e., the vegetation, changesI

;both OR a seasonal basis with deciduousness and on a longer-term basis

; ~-

with ecosystem development. : ij c ,, .‘ ‘-!I

Through regulation of the flow of solar energy and the use of solarenergy to support both passive and active processes, the aggrad&-~eco?yssm_ gains ~substantial -control- over internal microclimate as well

l+c .!J

as biogeochemistry. Regulation of internal humidity and temperature ,.;:; :I.::;

regimes in summertime is achieved by the reflective capacity of the upper I/.

.I,canopy and by transpiration, which evaporates approximately 3 x 106 ’ ,:

liters of water/ha-yr at Hubbard Brook. The highly buffered microclimate$8

-“l;‘:‘.,-~I-.., ,. 1, r; .- 7.’

.’

48 Pattern and Process in a Forested Ecosystemd

J “fi ’ />., 1, within the system plays a major role in regulating the decomposition

*:,5:‘? regime within the soil, and in this way is probably closely related to the

0” i’ highly predictable biogeochemistry of the aggrading ecosystem (Likens etal., 1977).

., ,C*.Ii-

;’ ^‘ 1BIOMASS: DEVELOPMENT OF REGULATIONAND INERTIA

The amount and distribution of &ving~ a&dead b&mass.are among themost important structural features of northern hardwood ecosystems and- - .~~are closely and functionally linked to-th&tabi!ity.

Analysis of the second-growth forest at Hubbard Brook (Table 2-3)provides data roughly equivalent to a 55yr-old stand m our hypotheticalsequence (Figure l-10). Total biomass is about 42,000 g/m’. Of this. about16,000 g/m’ is living plant biomass distributed about 82% aboveground

and 18% belowground. Eighty percent of the fine roots (<3-mm) areconcentrated in the upper 30 cm of the soil (Safford. 1974). Although theforest has a leaf area index of 5.8, only 315 g (2.4%) of the abovegroundbiomass is in leaves (Whittaker et al.. 1974).De-bulk@%) of.thefo.talb&masssis_ dead.- However, inspection of Figure

2-1 suggests that hvmg

biomass predominates from about 80 years after clear-cutting to the endof the Aggradation Phase. It should be noted, however, that accumulationof living biomass is primarily the accumulation of wood m the interiorof living trees.

Concentrations of both living and dead biomass vary markedly

Table 2-3. Living and Dead Biomass in a 55-Yr-Old Aggrading Forest Ecosystemat Hubbard Brook”

Biomass (g/m’)Vertical Distribution

Category of Biomass (4 Living Dead

Above soil surfaceLiving plant biomassDead wood

YC,“.. . . . ..- ------

Living rootsDead woodForest floorMineral soil

o.oc-20.00.0040.0

13,2816440

0.00-W 2.8Xb0.00-0.5 2.9000.00-0.09 4,8000.09-0.45 17.300

16,107 25.440Total

aFrom Dominski. 7971: Whittaker et a/., 7974; Gosz et al.. 1976: Me/i/lo, 1977.

bLJoes not include herbs and shrubs, but these are -=z 7%.


throughout the vertical dimension of the ecosystem. The concentration ofbiomass per unit volume in the upper 45-cm of the soil is about 90 timesgreater than average concentrations aboveground. This distributionreflects the functional requirements of distributing leaves through a largevolume of air to facilitate energy interception and gas exchange andof concentrating roots in a relatively small volume of soil containingavailable nutrients and water.

Dead Biomass and the Forest Floor

About 60% of the organic matter below the soil surface (Table 2-3) isincorporated in the mineral soil, but concentrations are quite variablespatially owing to pitand-mound topography and the generally rockynature of the profile. Most (90%) of the organic matter in the upper 45 cmof the soil is dead and is largely the product of microbiological alterationsof plant material. The mineral soil is generally covered with a well-developed forest floor which averages about 8.6 cm (Dominski, 1971;Melillo. 1977) in depth and contains about 4800 g/m’ of organic matter(Figure l-7). gganic matter in the forest floor and in the B-horizon of themineral soil are not the same. The organic matter of the forest floor isconsidered more labile and nutrie.nt:riih, whiIe most of the organic matterin the deeper horizons is thought to be more refractory and less subject-torapid change (see Chapter 1).

Dead biomass in the soiI has very strong effects on both hydrologic and,LL&<

nutrient-cycling characteristics of the ecosystem.&anicmatter reduces2 ‘3 z--’

2 da ’&density, increases water-holding and cation-exchange capacity, andserves as a reserve store of plant nutrients (Hoyie, 1973). Because of the

p ,,;, ,

low bulk density of the forest floor (0.2 kg/liter; Dominski, 1971), these .”soils, in common with most other forest soils (Lull and Reinhart. 1972), ” “-“”

,

have ~mous infiltration capacities (ca. 76-cm/hr; Trimble et al., 1958) ’ 7’ ’as well as excellent percolation rates. Because of the relative lack of clay

in our northern hardwood soils, humus also serves .as the principal -- ‘,,‘. _

&on-exchange site within the soil profile, and the highest concentrations:_.,_ -i

of exchangeable cations are found within the forest floor (Hoyle, 1973; : “,- *- ::Pilgrim and Harter. 1977). ‘?

___The_growth of the forest floor during the Aggradation Phase. from 3L- - -26 t/ha ‘at 15jyears after-cle~ar-cutting to; 57 t/haiatJO years after cutting- - - - - r,d’-T..

(Covington, 1976). indicates that both available water-storage and ‘~ ’St&n-exchange capacity grow as weli. We estimate that between Years ”

15 and 60 water retention capacity (mostly available water storage; i ’Trimble and Lull, 1956) .d,r F

grows by 0.7-cm and that the cation-exchangecapacity of the forest floor increases by about 100%. Since water and

fib>,

;;..7.

nutrients are vital to production. it would appear that the productivecapacity of the aggrading ecosystem increases during the first third of the

I Aggradation Phase coincident with the growth of tde forest floor..z _7-e.q,&

P‘


DETRITAL-GRAZING CYCLES

The detrital cycle is overwhelmingly dominant in terms of energy and- - -nutrient flux within the aggrading northern hardwood ecosystem. In anaverage year we estimate that about 26% of the net primary productivityin a 55yr-old aggrading forest is stored within the ecosystem in the formof a net gain in living biomass. mostly wood and bark (Figure 2-J).

Seventy-four percent of NPP plus a small addition of dissolved organic- matter in precipitation represents that proportion of the current crop of

-? .,energy available to support the sum of all heterotrophic activities. Less-; ,,. than 1% of this energy is consumed by grazing animals in an average

- year. The principal herbivores. in approximate order of importance. are: chipmunks, mice. foliage-eating insects, birds. deer. and hares (Gosz et

al., 1975; Hanson, 1977). The remaining energy is added to the dead- - -

SolarEnergy

Organic MatterIn Precipitation

I

.I Z‘ t;,! _-\; ’ _

‘. -‘, ilr:

.,:..+ 4 /

.o; s “,P“.‘- -/i

5720 +(Heat)

3380 -(Heat)

Ecosystem

Net Primaryf Productivity

4880

1Gross PrimarylOi 34ko

1Respiration

1 2 1 0 ’Stoied In Living

Biomass

Remainder For All

r3510 - Other Uses

Grazing Cycle < I%Detrital Cycle >99%

Heterotrophic Stored In Dead

Respiration , Biomass- 120

Organic Matter In*IO

Stream-water

Figure2-4. The flow of energy in carbon compounds through a 55-yr-old

aggrading forest at Hubbard Brook; numbers give energy values in kilocaloriesper square meter (after Gosz et al., 1978).

The &grading Ecosystem 51

wood, forest floor, and mineral soil compartments in various forms ofJtbe_r+ leacha&. and exudates. There it is subject to utilization in thedetrital cycle (Figure 2-4). In an average year. an amount of energyequivalent to about 95% of that added to the detrital cycle (but notnecessarily the same energy) is released as heat due to heterotrophicrespiration. while about 5% is added to net dead biomass accumulation.-~Enargy consumed in all heterotrophic respiration. grazing and detrital,

is coupled to the biogeochemical process of mineralization or the releaseof ions previously locked in organic compounds. In effect, these ions areadded to the available nutrient compartment in our nutrient flux andcycling~model (Figure 1-14).

We have not identified the principal saprobic plants (microflora) orinvertebrate detritivores at Hubbard Brook; however. the major groups j;’

~

have been reported for acid mor soils elsewhere in the Northeast (Eaton ,. ,.and Chandler. 1942; Lutz and Chandler. 1946),F_ungi are thought ,to-be T<he principal agents of .decomposition since the relative importance of , L.bacteria declines markedly in acid soil. Major invertebrate detritivoresare probably similar to those reported for mor soils in New York Statewhere, in terms of numbers, the following groups predominated:Arachnida (mites, false scorpions, and spiders), Collembola (springtails),Coleoptera and Diptera (beetles and flies), Hymenoptera (ants), Sym-phyla, and Annelida (earthworms). Earthworms, millipedes, and isopodswere absent in the New York State sample, but all have been found insmall numbers at Hubbard Brook (R. Holmes, personal communication).

The complex relationship and relative importance of the microflora anddetritivores in decomposition remains to be determined at HubbardBrook. However. it would appear that as soon as soft tissues such asleaves or bud scales fall to the forest floor they are subject to acoordinated attack (Lutz and Chandler, 1946; Jackson and Raw, 1966;Schaller, 1968; Brady, 1974). Fungi and bacteria initiate the action but are-.soon joined by springtails, bark lice, and various larvae which eat or tear__~~~ho&s-in the tissue (fenestration). pp_ening _it..~to more rapid microbialattack. larger larvae and mites bring about further perforation andskeletonization. Large amounts of feces, or frass, are produced, whichmay be consumed again by other fauna. The activities of the soil faunaand microflora are thus closely linked. Chewing, ingestion, and digestionby fauna not only result in decomposition of the organic matter butsimultaneously create surface and moisture conditions more favorable tomicrobial action both within the fauna1 gut and in the resu!tant frass. Itseems likely that the &&itivores-obtain their principal energyguppliesirom the ~qa><y$ecomposable substances within the litter such as sugars,starches, and simple and crude proteins. Exoenzymes of fungi and

--_- _.bacteria not only attack these easily decomposable substances but arelargely responsible for the decomposition of the more resistant com-pounds, such as hemicellulose, cellulose, and lignins, which compose thebulk of the leafy and wood litter.

52 Pattern and Process in a Forested Ecosystem The Aggrading Ecosystem 53

The rate of detrital decomposition by the microflora and the detriti-vores is regulated by many factors including the dimensions and physicaland chemical structure of the substrate. In general. boles of trees,particularly large ones. decay slowly, while leaves decay fairly rapidly.

Litterbag studies at Hubbard Brook (Gosz et al., 1973) indicate thatvellow. birch leaves decompose more rapidly than sugar maple&v+

GGch in turn decay more rapidly than beech; we estimate that 5, 9, and11 years. respectively. are required for 95% of the original dry weight ofthe above species to disappear after a leaf falls to the forest floor.

Soil invertebrates also play a role in the physical transport of

decomposing organic matter. However. owing to the relatively smallpopulations of earthworms and other larger invertebrates at Hubbard

Brook, there is relatively little. mixing activity within the forest floar-orwith the underlving mineral soil. The relatively happy circumstance of afairly clear dis&ction between the forest floor and the mineral soil madeit possible for us to carry out the study of the decline and growth of theforest floor discussed in Chapter 1.

The microflora and detritivores are themselves food sources for

parasites and predators. including carnivores such as mites, centipedes,

salamanders, and shrews, as well as microfloral feeders. To some degree,mites and springtails regulate fungal populations by feeding on hyphae(Mitchell and Par mson, 1976), and protozoa may exercise some controlk’over bacterial populations (Lutz and Chandler, 1946).

Lack of knowledge of the details of saprobic decomposition representsone of the major gaps in our knowledge at Hubbard Brook, yet this isof key importance in understanding the ecosystem’s biogeochemistry,and particularly changes through time. For example, on the basis of

input-output budgets (Figure 1-14) we know that w&r! an aggr&ng

, ecosystem is clear-cut the output of nitrate may increase severalfald.ThisIs due primarily to a marked increase in the populations and-ac&&-of-autotrophic nitrifying bacteria (Smith et al., 1968) and other heterotrophicn&fying organisms (J. Duggin, personal communication) within the soil.

I What we do not know is how the rest of the saprobic plant-detritivoresystem responds. Do the new conditions result in marked expansion orcontraction of various detritivore and microfloral populations? A com-parative study of changes in saprobic populations in the forest floor of

undisturbed and recently clear-cut forests would seem to be an informa-tive way of attacking this problem.

The division of energy flow between the detrital and grazing food websis not constant from year to year. and in some years there is a dramaticrise in herbivory, principally due to rapid increases in populations ofleaf-eating insects. High rates of defoliation occurred at Hubbard Brookin the years 1969-71 when a species of defoliating caterpillar, thesaddled prominent (Heterocampa gurtivita Walker), was in outbreakphase. During the peak year of defoliation, about 44% of the total leaftissue was consumed, while in local areas patches of forest were totally

stripped of leaves (Holmes and Sturges. 1975; Gosz et al.. 1978). Leavesof yellow birch were preferred by the larvae of the defoliating insects (R.Holmes, personal communication). Although these insects consume laroe

,

quantities of leaf tissue, the amount of energy they assimilate is relativelyy:‘,

small, about 14% of the total contained in the ingested tissue. TheL:‘.

remainder (86%) is added to the detrital cycle in the form of frass.The potential of these organisms to&:ct system biogeochemistry uoes

far beyond their energy utilization, because their grazing activities szke_.

directly at the ecosystem’s primary source of energy fixation and ‘.hydrologic and energy regulation. which is- leaf tissue. Not only is there astrong potential to reduce gross primary productivity temporarily and

‘s ‘/ ~

hence diminish the amount of energy available to support all organisms y ‘+Gwithin the ecosystem (Figure 2-3). but by reducing leaf area it seems likelythat the ecosystem_ w_ould undergo proportionate reductions in transpira-

r +2’~ V -:

Fh_and nutrient uptake by green plants. According to Hibbert’s review(1967) of various forest-thinning experiments. streamflow is increased .5 ” ’approximately in direct proportion to the reduction in leaf area. Inundisturbed hardwood forests of North Carolina. it has been observedtha t defo l ia t ion by the fa l l cankerworm (Alosophilu pometuriu) isassociated with concentrations of nitrate in stream water that are aboutfive times higher than those measured in nondefoliated watersheds(Swank and Douglass, 1975). Increased-nitrate loss is thought to result from Aar$uence of defoliation on the forest floor, for example. from higher soil b _temperatures or lower nutrient uptake by plants. At Hubbard Brook, no .__significant increase in nitrate concentration in stream water was observed c_during 1970, the year of peak insect defoliation.

The following effects emphasize the large destabilizing potential thatsevere outbreaks of leaf-eating insects may have on an ecosystem: (1) xconversion of living leaf tissue to CO,, insect biomass, and frass; (2)increased output of stream water, the chief vehicle for removingnutrients; (3) decreased uptake of nutrients by green plants, perhapsleading to increased nutrient concentrations in soil and stream water. (4)

’more radiant energy flow to the forest floor; (5) increased transfe; ofliving biomass to dead organic matter (Figure l-5); and (6) possiblealterations in microbiologic activities in the soil leading to increasednitrification (implications to be discussed later). Production may betemporarily lowered, and the system may become more open in terms ofstream-water and nutrient loss. Not only does defoliation affect majoraspects of the system’s biogeochemistry, but the detxital cycle also mustbe affected. During heavy outbreaks of defoliators, leaves are consumedin the canopy, and we wonder what effect this has on the detritivorepopulations. For example, certain populations of soil invertebratescarrying out fenestration and skeletonization rely heavily on newly fallenleaf litter (Schaller, 1968).

It is exceedingly difficult to evaluate the long-term role of defoliatinginsects in regulating the biogeochemistry of natural ecosystems. On the


. .one hand, insects may constitute a malor destabihzmg force causing

numerous rate changes and an increased loss of nutrients from the

ecosystem. On the other hand. because of differential grazing and otherinternal ecosystem responses. they may act to channel the flow of radiant

energy. nutrients. and water to better-adapted individuals and species and

thus function something like a cybernetic regulator of primary produc-tion (Mattson and Addy, 1975). At Hubbard Brook these defoliatmg or-ganisms apparently operate within bounds that allow the aggrading

--_- ~..ecosystem to continue its accumulation of living biomass at the f.airly

predictable pace reported by Whittaker et al. (1974).Holmes and Sturges (1975) and Gosz et al. (1978) report a behavior

pattern of predatory animal species at HubbardBrook that may illustrateanother animal-based biogeochemical regulatory mechanism. They pointout that many of the major predators are almost entirely opportunistic intheir diet preferences and that these animals cannot be identified witheither the grazing or detrital cycles. In most years, however. it seemslikely that major predators such as birds, shrews.

salamanders, and

rodents and invertebrates such as centipedes. beetles. and spiders gamtheir major energy inputs by feeding on a variety of detritivores and theirinvertebrate predators. They also consume defoliating insects, and duringvears when they are common many of these predators have the capacityto substantially shift their diet to the newly available food source.

The intensive study of birds by Holmes and Sturges (1975) providesdata for one group at Hubbard Brook. In general, birds consume only a

small part, about 0.2%, of the annual net primary productivity occurring

aboveground and probably have no major direct effect on ecosystem

structure or overall energy flow rates. During average years,when most

energy is flowing to the detrital cycle, the summer diet of the bird

community consists largely of adult Dipterans whose larvae are detriti-vores in the forest floor and in streams. adult Coleopterans whose larvae

prey on soil-dwelling organisms, and adult Hymenopterans which are

themselves secondary and tertiary consumers. In most years,t h.e-n>e

bird community is closely linked by energy. flow with the forest flog&HoGvex-Holmes and Sturges (1975) have shown that during the yearswhen Heterocampu was in outbreak phase, many birds concentrated onthis new and abundant source of food. During the years of outbreak, 1969to 1971, and for one year afterward. the total number of individuals in

breeding bird populations increased progressively, presumably because ofa more abundant and easily available food supply. Studies of animalbehavior during outbreaks of gypsy moths suggest that other predatorypopulations such as mice (Peromyscus) and beetles (Culasomu) behave in afashion similar to the bird population at Hubbard Brook (Campbell,1975).

These behavior patterns suggest that these vertebrate and invertebratepopulations of opportunistic predators, although playing a minor role in

energy flow patterns, may perform important biogeochemical regulatory


functions in both the detrital and grazing cycles. They must play somerole in regulating the process of mineralization through effects ondetritivore populations and the processes of transpiration, nutrientcycling, and primary production through effects on populations of insectdefoliators. Holmes-(personal communication) believes that populationsofdefoliating insects at Hubbard Brook are kept at low numbers by n, .:&vo factors: (1) defense systems of the plants themselves, including the

ilow food quality of plant tissues and (2) predators. When plant defenses - ,

are lowered because of stress or other factors. the insects consume moreplant t issue. reproduce and grow faster to a point where they areuncontrollable by the predators. and thus enter an outbreak phase.Campbell (1975, personal communication) suggests that once an outbreak

,-“:begins other factors such as parsitoids. disease, and starvation are muchmore important in bringing about its termination. It is most interesting l--’

that these dhe..predators are firmly attached to the~detrital cycle,the ’most predictable aspect of energy flow within the ecosystem, and in a _I- - ~~~~.sense are facultative with regard to the less predictable grazing cycle.-

BIOTIC REGULATION OF BIOGEOCHEMICAL FLUX

Regulation of the Hydrologic Cycle

Personnel of the U.S. Forest Service have monitored and maintainedaccurate records of p:ccipitation and streamflowvat Hubbard Brook since1926 (Likens et al., 1977; Figure 2-5). On the average the area receives130-cm/yr of precipitation, of which 70% is rain -and 300/ is snow In_ _-0 .~ ---.aggrading forests, about_62%. runs off as liquid water in stream channels,while the remainder is lost as vapor through evapotranspiration. >A

In any particular year,,. _. - -,

monthly inputs of precipitation may showrandom extremes, but for the longer term the monthly pattern is quiteregular (Figure 2-5). In contrast to seasonal uniformity of precipitation, -

j,’ /i,:.”rno~thlistreamflow is widely variable and in fact defines four seasons ?(Figure 2-5). Streamflow during the “gyo-~~ng._S_e_aason,” approximately ,: : FJune through September, isvery low. Most of the annual streamflow, D _ ’54%: occu=during the spring snowmelt period, March through May with I,’ =30% in April alone. A second minor peak in streamflow occurs ‘frommid-October through mid-December, afterleafffall but before~ heavydevelopment of the snowpack; whereas streamflow is diminished dt&g ,‘-‘.” rthe winter, January through mid-March, as cold winter conditions prevailand precipitation accumulates as a snowpack. - ’ ”

The living and dead biomass of the Aggradation Phase exerciseconsiderable and fairly predictable control over the hydrologic cycle forthe ecosystem. Control is expressed largely by effects on the storage-ofwater within the ecosystem and on the pathways by which water moves~----~through and out of the ecosystem.

I


25r

I m Precipitation

The effect of transpiration is so strong that when 18 years of data onannual evapotranspiration (which includes evaporation throughout theyear as well as transpiration during the growing season) and annualstreamflow are graphed against annual precipitation (Figure 2-6), evapoT)transpiration remains relatively constant despite major fluctuations in ’precipitation, while streamflow is highly correlated with amount of ?precipitation.-This indicates that the aggrading ecosystem has a remarka-bly consta$ transpirational demand over a wide range of environments! LvAriabIes.

When precipitation crosses the boundary of an ecosystem, it has bothpotential energy and potential solvent power which, as it passes throughand out of the ecosystem, could be uti l ized in the erosion andtransportation of particulate matter and in the removal of dissolvedsubstances. The vegetation of the aggrading ecosystem acts to minimizethese degrading activities by the utilization of solar energy in transpira-tion. Thus streamflow is reduced, and a variety of other mechanisms allow

0 Streamflow

Ecological Seasons, Months

Figure 2-5. Monthly average distribution of precipitation and streamflow in cen-timeters of water. Ecological seasons are the growing season: autumn: winter,accumulating snow; and spring snowmelt and runoff period (after Likens et al.,1977).

The living biomass of the aggrading ecosystem affects the streamflow- -characteristics during all seasons (Figure 2-5). Durmg the. autumn,increased streamflow is correlated with the cessation oftransp!rat?on.the hard winter, and in spring snowmelt periods. timing of streamflow isaffected by the _sh_ade-of standing leafless trees, but this is of minorimportance. “Growing season”not only timing but amount of streamflow. In this. the biological process of

Hydrologic data from June through September. when the forest is iny full leaf, show that. while precipitation ranges from 32 to %)-cm, stream-

flow averages only 5-cm. A S is well known (Kittredge. 1948; Colman,

1953; Hibbert. 1967). durjng the growing season evapotranspiration

(mostly transpiration) hasfirst call on wat& stored in soil, thus evacuat-ing storage space. Only after storage space is refilled by precipitation

does streamflow occur.During the Aggradation Phase, no destruction of the forest canopy is

visualized and thus it seems probable that the hydrologic conditions justdescribed are generally applicable to the entire period. although minorvariations due to variable proportions of evergreen trees might occur

lCwank and Douglass. 1974).

12-12-

ll-ll-

lO-lO-

9-9-

a-a-

7-7-

6 -6 -Evapotronsp~rotionEvapotr~~nSP~r~‘~o~

5-5- .00

0 0000

L -L - OO

II ,, II1010 1111 1212 II1313 1L1Lt

1515 1616 1717 1818 19Annual Precipitation (e x lo6 /ho)

19Annual Precipitation (e x lo6 /ha)

Figure 2-6. Relationship between precipitation, streamflow, and evapotranspira-tion for the Hubbard Brook Experimental Forest during 1956-74 (Likens et al.,1977).


the ecosystem tqabsorb the kinetic energy of flowing water while yielding.a minimum of eroded material.

Physiologists often portray transpiration as a necessary evil. or at thevery .most as useful in cooling leaf tissue on hot days. F_cqm the broaderperspective of the ecosystem, in humid systems transpiration is a maj_or

regulatory process exercising impqrtantcont ro l over a var ie ty -of

ecosystem responses. Solar energy, lmpmging on the forest canopy: isutilized to increase evaporation over what it would be if no vegetation

were present. This transfer of liquid water to vapor has these actual or

potential ecosystem effects:1. Transpiration markedly reduces stream-water

output during the

: growing season. We estimated an average annual reduction, of

jl stream-water output of -35% at Hubbard Brook due to transpiration

(Pierce et al.. 1970).,’ 2. This reduction acts as a_nutrient~conservation mechanism because

reduced streamflow means less nu t r i en t s ca r r i ed out of the~ I ;’

.--’ ‘,_

3.

:‘,.i

-..,.-/-: _:

;;i ,

,. 1

3.

ecosystem.Transpiration not only reduces streamflow, but also dimin&@

summertime-peak discharge rates during storms (Remhartet al.,

1963; Pierce et al., 1970; Hornbeck, 1973b; Bormann et al., 1974).

This regulation has important effects on the process of erosion.

since erosion is closely related to storm peaks. Obviously, during,exceedingly heavy and/or prolonged periods of rain, transplratlonwould have relatively little effect on flood peaks.Transpiration “powers” a large part of the circulation of nutrients

within the ecosystem. ‘Some nutrients are drawn to root surfaces inthe mass flow of water immediately after rainfall and some arelifted to the canopy by transpirational “pull.” and in contrast, tostreamflow. which carries nutrients out of the system, transplratlonmay be viewed as a “distillation” process in which nutrients are left

behind in the leaves where they are eventually recirculated by leafdrop, resorption. or leaching.

5. Finally, the complicated positive-feedback relationship betweentranspiration and availability of soil water plays an important role in

1 regulating primary production of the ecosystem through well-knowneffects if available water on photosynthesis (Bormann, 1953:

. Kramer and Kozlowski. 1960) and, indirectly. in regulating decom-

position within the soil . Transpiration also may be of someimportance as a soil-aerating mechanism by removing soil water__~_ ~_._after rainfall and, in effect, opening air passages into the soil. This

effect is well known for wet sites (Dansereau. 1957).Biomass not only exerts control over the hydrologic cycle by transpira-

tion but in several additional ways. -_Lilring&leaves a n d forest-

intercept and disperse the energy of falling drops..of water. The heav)l

-concentrations of dead biomass in the A,,-horizon and in the mineral sol1--nn+,,, lllornent both retention and detention water storage in the soil


and create excellent conditions for infiltration and percolation of water.This latter factor determines the route of liquidwater pa&g through theecosystem, because it permits almost_unlimited infiltration. Overland flowis negligible in forested ecosystems at Hubbard Brook (Pierce, 1967); as aconsequence, water moves to the stream by lateral flow within the soil.This relationship exists even during the winter, because organic layerscoupled with snow accumulation insulate the soil, and normally there ismailfreezingd.uring the winter (Hart et al., 1962). This routing of waterhas important implications for both ecosystem biogeochemistry anderosion. Finally, interception and evaporation of both rain and snow bydormant and nondormant northern hardwood forests at Hubbard Brookreduce the amount of precipitation reaching the ground by about 12%(Leonard. 1961).

Control of Particulate Matter Export

In terms of overall ecosystem stability,al-_e_rosiqn is perhaps the mostimportant potential destabilizing force for terrestrial ecosystems. ChronicI --- --~~-erosion of exposed mineral soil tends to remove the finer inorganicfractions and organic matter. These fractions are rich sources for availablenutrients and also contribute to the moisture-holding capacity of the soil-hence erosion has the potential to seriously diminish ecosystem producltion (Brady, 1974). In its most severe form. erosion can remove a significantproportion of both organic and inorganic exchange surfaces within theecosystem. Not only will this have immediate drastic effects on theproductivity of the ecosystem, but it may have substantially long-termeffects. The rebuilding of exchange surfaces to pre-erosion levels by theproduction of clay minerals through weathering and accumulation of soilorganic matter and the reacquisition of lost nutrients may take severalcenturies. In essence, the developed ecosystem may be forced back to amore primitive level of development with lower production and lesscontrol over its immediate environment. Loss of contro1 may for a timeact as a positive feedback and delay the rate at which the ecosystemdevelops after serious erosion. It is interesting to note that this most basic- ;:*:and import_ant-perturbation ~to ecosystem stability, erosion, long known to ”gg!gri$turalists and foresters, hardly enters the current ecological debate ’

’on stability-of ecosystems. In fact, the word erosion is not even found inthe subject index in a number of recently published ecology texts.

Aggrading northern hardwood ecosystems in the Hubbard Brook areahave low rates of erosion due to geological circumstances (Hunt 1967;Lull and Reinhart, 1972), but the blQgical_fraction of the.ecosyste’m also

J&Y_ an irnppr>Gtl-egulatory role. Rates of-e;i$& of particulate matterec!ond-growth forests are among the very lowest recorded for humid -ySosystems. Losses average 3.3 F 1.3 t/km’-yr, despite the -fact that- theseforests occur on rather st-gep~ slop&, -12. to 13” and are subject to anaverage annual precipitation of 130-cm (Figure 2-5). This output is small


compared to other forests and represents an extremely low rate when

compared to potential outputs of thousands of metric tons per year fromseriously disturbed sites (Table 2-4).

At Hubbard Brook we have found, as have others elsewhere, that-partj.c.ulate mat te r losses are. dire.~t_ly-~~~~-.nonllnearrelated to the

discharge rate o f the-- s t r eam (F igure 2 -7 ) . The shape o f theconcentration-flow rate curve is determined by the erodibility of the

ecosystem and the fact that the capacity of running water to do work

increases roughly as the square of its velocity (Bormann et al., 1969).

Erodibility is a complex function of the ecosystem, and disturbance of theecosystem can shift the curve (Figure 2-7) to the left. making low flow

rates more effective in the erosion of particulate matter.

Table 2-4. Particulate Matter Output (Sediment Yield) from Forested Ecosys-tems, Small River Watersheds with Mixed Cultural Conditions, andSmall Man-Manipulated Ecosystems in the Northeastern UnitedStates”

Watershed SedimentSize Yield

Location (km’) (t/km’-yr)

Forested ecosystemsMature forest ecosystem, Hubbard Brook

(8-yr average)Protected second-growth forest. mountains.

West VirginiaWooded watershed, Somerset, Kentucky

Selected small riversTributary basins

Susquehanna River, 88% forest coverPotomac River, 88% forest cover

Streams draining New England upland in New JerseyScantic River, ConnecticutGeorge Creek. Maryland (rural and wooded)Tributary basins

Susquehanna River, 25% forest coverPotomac River, 15% forest cover

Small man-manipulated ecosystemsDeforested ecosystem, Hubbard Brook

(4-yr Average)(Maximum annual)

Careless clear-cut, mountains, West VirginiaCiear-cut, followed by farm and pasture,

mountains, North Carolina (maximum yield)Construction site, Baltimore. Maryland

‘Modified from firmann et al.. 1974.

0.13

0.3919-.-

3

?

3

253188

??

0.16 190.16 380.30 302

0.090.006

3.3

<3.25.3

1212

3.5-352672

85160

3,69049,000

An analysis of particulate matter output and flow rates for theaggrading forest at Hubbard Brook shows the importance of theoccasional storm peak in the removal of particulate matter (Bormann etal.. 1974). During a XI-month period. 77% of the total stream water wasdischarged at rates between 0 and 30 liters/set. but only 14% of theparticulate matter was removed. At discharge rates of >85 liters/set.3.7% of the water was discharged. while 52% of the total particulate f t- ”matter was lost. Sixteen percent of the particulate matter was eroded byone flow ranging from 310 to 340 Iiters/sec. This flow carried only 0.2% of

Jthe total streamflow and comprised less than 1 hour’s time during the ’entire 50-month period (Figure 2-8). Clearly. the annual variability in 2 : ’particulate matter output from the aggrading forest is due primarily to theoccurrence of large storms.

T h e bi-ological component of the aggrading ecosystem regulates -4 2 ’ ’particulate--matter- erosion by its effect on both discharge rates and !’ -’erodj-bi!ity. As already noted. transpiration during the growing season_..~tends to reduce the size of storm peaks. This is of special note since T-*“!’ -Ierodibility shows seasonal variability and is greatest during the growingseason (Bormann et al., 1969, 1973).

Flow, ft’/sec (28.3 Iiterskec)Figure 2-7. Composite curve for second-growth aggrading forests at HubbardBrook (Bormann et al., 1969) showing the relationship between particulate matterconcentration in stream water and flow rate.

62 Pattern and Process in a Forested EcosystemThe Aggrading Ecosystem 63

Figure 2-8. (B) Photograph taken 24 October 1959, when the weir on W4 wasunder construction and during the highest flow recorded at Hubbard Brook during its24-yr history. The peak discharge measured on Watershed 3, 1.2 times larger thanW4, was 68.3 fY/sec (1934 liters/set). According toa log-Pearson plot, the storm of24 October 1959 has a recurrence interval of about once every 50 years. One stormof this magnitude would probably remove an amount of particulate matter equivalentto many years of removal at flow rates not marked by unusually high storm flow. Thescale in the foreground of both photos is about the same. (Photograph B courtesy ofR. S. Pierce, U.S. Forest Service.)

64 Pattern and Process in a Forested Ecosystem The Aggrading Ecosystem 6.5

Erodibility, a measure of the ease with which earth materials give way__--(Leopold et al., 196-l), is a function of both geologic substrate and

biomass. Whether or not the kinetic energy of &&ng water within theecosystem is strongly coupled to the process of erosion and transportationof particulate matter or is dissipated in other ways is often determined bythe biotic condition of the ecosystem. That the biota and its organic debrisexert considerable control over this phenomenon has been shown experi-mentally and will be discussed in more detail in Chapter 3.

1 /Biotic control of erodibility in the aggrading forest is achiev_ed-by:

(1) qrganic-matter enhancement of infiltration and percolation,_ which

converts surface flow to subsurface flow (surface flow does the work ofparticulate matter erosion); (2) the binding together of soil and organic

debris in drainage channels and along streambanks by roots; (3)

protection (interception) of mineral soil from direct action of falling

raindrops by canopy and forest floor layers; (4) orgamc-debns dams m&;am channels which tend to regulate streamflow and retain particulatematter within the ecosystems: and (5) Jardwood leaves.

which form a

protective shield against erosion of exposed stream banks (Bormann et

al., 1969; Fisher, 1970).

Regulation of Streamflow Chemistry

An important characteristic of the aggrading hardwood forest ecosystemis its ability to regulate the chemistry of drainage water. The chemistry ofwater entering the ecosystem as rain or snow is altered in highly

predictable ways as it passes through the system and Issues forth as

stream water.The measurement of precipitation chemistry at Hubbard Brook rep-

resents the longest comprehensive record of precipitation chemistry in

the United States (Likens et al., 1977). Pre_cipitation is-acid. and sulf_a_te

land hydrogen ions dominate (Table 2-5). The chemical composition of

precipitation is highly variable on a weekly basis, but. in general.

relatively lower concentrations occur with snowfall (Likens et al.. 1977).Several long-term trends. apparently related

to _aiLpollution, have

emerged f rom these s tud ies . _ _Nitrate concenfrations. havupprotiat_el_y

doubled-since 1955. The input (concentration times amount of precfpita-tion) of nitrate and hydrogen ions has increased by

2.3 and 1.4 times.

respectively. during the last decade (Likens et al.. 1977).What happens to the chemistry of precipitation as it passes through the

ecosystem? Because of the dense forest canopy, the heavy concentrationof organic matter in the forest floor, and the high infiltration capacity of

the soil, almost all precipitation entering Hubbard Brook watersheds

comes into intimate contact with the living and dead biomass and the

mineral soil of the ecosystem. During the growing season, first contact ofthe incoming water (precipitation) is made with the forest canopy. Thechemistry of the resulting throughfall and stemflow is markedly changed.

i

Table 2-5. Volume-Weighted Concentrations of Dissolved Substances in Precipi-tation and Stream Water from ‘Aggrading Forest Ecosystems atHubbard Brook with Values as Means for the Period 1963-74~

PrecipitationSubstance

Streamflow(mg/liter) (mg/liter)

StreamflowiPrecipitationRntin- .-_.-

,-. 7+La-Mg’+K+Na’A13+NZ4’H’sod2-NO3-Cl-PoJ3-HC03-Dissolved silicaDissolved organic

carbon

0.16 1.650.04 0.380.07 0.230.12 0.87

ll

0.220.0732.91.470.470.008

-0.006b

2.4

0.240.040.0126.32.010.550.0020.924.5

1.0 0.42

10.39.53.37.2-

0.180.162.21.41.2

- 0.25 1s

153-

‘From Likens et al., 1977.bkt determined, Pace quantities.

in part owing to cation-exchange reactions between precipitation and- -foliage. Some ions wiate-&d potassium are greatly enriched in the

&oughfall and stemflow, whereas others like hydrogen ion are depleted(Eaton et al.? 1973; Wood and Bormann, 1974; Likens et al., 1977).

Lack of soil frost during the winter (Hart et al., 1962) and the highinfiltration and percolation capacity of the soil mass guarantees thatduring all seasons almost all water moving through the ecosystem to thestream channel comes into intimate contact with the exchange surfaceswithin the soil. The chemistry of water that follows this route isremarkably altered; scentrations of some ions rise more than lo-foldwhereas those of others drop more than 6-fold (Table 2-5). Increases in--~ .concentration result in part from evapotranspiration losses which reducethe amount of liquid water leaving the ecosystem. Theoretically, thiscould increase concentrations by 1.6 times, but decreased concentrationsand the size of some increases (Table 2-5) show that factors in addition toevapotranspiration are operational. In fact, various complex biogeochemi-cal mechanisms are involved in determining the chemical composition ofwater as it passes through the ecosystem. Some of these mechanisms willbe discussed below.

In spite of the numerous and marked differences in the chemistry ofwater entering the ecosystem as precipitation and that leaving as stream-


flow, stream-water concentrations are highly regulated by the aggradingforested ecosystem. Concentrations of the various chemicals in stream waterare either nearly constant or highly predictable (Johnson et al., 1969; Likenset al 1977). This is remarkable since streamflow varies by three or fourorders of magnitude during the year. The maintenance of this relativelystable stream-water chemistry is related to the intimate contact of the waterwith exchange surfaces as it travels toward the stream channel and to anabundance of exchangeable ions in relation to ions lost in transient waters(Likens et al., 1977). Factors controlling ultimate stream-water concentra-tions, however, are very complex and involve regulation of and interactionwith numerous aspects of both the hydrology and biogeochemistry of theecosystem. There is little doubt, however, that activities or conditions

associated with living and dead biomass play a major role in regulating thisstable and predictable behavior of the aggrading ecosystem.

Predictability of Stream-Water Chemistry

Johnson et al. (1969) developed a dilution-concentration..model forHubbard Brook which explained transient changes in the chemistry ofstream water by means ofmixing processes. (i.e., mixing of precipitationwith soil water). The rate of the mixing process is proportional to thedischarge of stream water from the watershed. The model showed three

/i ..<Ipatterns of behavior for chemicals moving through the ecosystem:

1. those that were diluted with high discharge-Na- and dissolved,-silica;

I ,,, i. 2. those that were more concentrated at high discharge-H+, A13’, and:;,.

,p ;.’ :, ’

/. r, !

: ,‘,I L.’ I

4”

‘, I,’

: ii*-”

:7.:I

” N O , - ; a n d3. those whose concentrations changed relatively little in relation to

.- s t r e a m discharge-Mg?‘, Ca”. K’, S04”-, and Cl - . Dur ing the

growing season. concentrations of NOj- and K- in stream waterwere markedly reduced.

Before proceeding further, attention should be directed to the c&se

relationship that exists between annual gross output of dissolvedsubstances (kilograms per hectare) in stream water and the annual output_pf stream water in area centimeters (Figure 2-9). This relationship

represents one of the most predictable aspects of the aggradingecosystem. At first glance it might appear to be at odds with thedilution-concentration behavior of the various elements at different

discharge rates described above. However, changes in nutrient concentra-

tions in stream water, although significant for some ions. are relatively

small (generally less than three times minimum values). while changes instream discharge range over four orders of magnitude. This, plus the factthat the bulk of the streamflow (77%) occurs at flow rates between 0 and30 liters/set (Bormann et al.. 1974). provides the strong predictable

relationship between annual gross output of dissolved substances and

total annual streamflow.

.

I

,


0 10 20 30 ‘0 M 65 7080 90 100 110 120 130 1‘0 7543

Annual Streamflow I cm)

Figure 2-9. Relationship between total annual streamflow and annual grossoutput of Calcium, sodium, magnesium, and potassium (in kilograms per hectare)during 1963-74 for forested Watersheds 1-6 at the Hubbard Brook ExperimentalForest (Likens et al., 1977).

Nutrient Input-Output Budgets

Eleven years of data from six small watershed-ecosystems at HubbardBrook indicate that nutrient losses-from the aggrading.forested ecosystemare relatively small (Table 2-6). Also, it may be observed that for someelements, meteorologic inputs are -very important in offsetting geologicaloutputs. In fact. the input-output budgets suggest that absolute amountsof hydrogen ions. phosphorus, and nitrogen are increasing within theaggrading ecosystem at Hubbard Brook: i.e., amounts in bulk .precipita-tion exceed amounts exported in stream-water. Although the averagedata (Table 2-6) indicate that chloride is increasing within the ecosystemin the long term, data for annual bulk precipitation are highly variable. -and thus inputs are not statistically greater than outputs (Likens et al..1977).

,I

, , l,‘, .!

Inp.@outputt budgets take no account- of gaseous exchanges forelements like sulfur and nitrogen. It can be seen from Table 2-6 that bulkprecipitation does not account for all losses of sulfur in stream water.Since sulfur is only a minor constituent of bedrock and till at HubbardBrook, it thus seems probable that net losses cannot be made up byweathering within the ecosystem. In fact this is the case. Weatheringreleases only about 0.8 kg of S/ha-yr (Likens et al., 1977). So to providethe sulfur necessary for annual biomass (living and dead) accretion (Table

Table2-6. Average Annual Input and Output (LOSSeS) of Nutrients for a 55-Yr-Old Aggrading ForestedEcosystem at Hubbard Brook with Values in Kilograms per Hectare per Year During the Period1963-74”

Geologic Output-streamnowh30,

Meteorologic Input aOrganic Inorganic -u

Isulk Net Gas anti Dissolvctl I’;lrticul;ile I’i~rtictlliltC ar-l

Element Precipitation Aerosol sl~l~st;llKxs” Matter” Matter” Net Los lnVI-.

1.4 3.42

Aluminum c c 2.0 <!a

c c 17.6 0.1 6 . 1 23.x!

Silicon ”2.2 ,I 13.7 0.06 0.17 11.7

2Calcium ;i;Magnesium 0.6 d 3.1 0.0s 0.13 2.7 2

Potassium 0.9 d I .o 0.0 I 0.5 I I .s %

cl 7.2 CO.01 0.25 5.0m

Sodium 1.6 8Iron c

c c0.0 I 0.63 0.64 u”

-

Hydrogen ion ’ .;s

c - -

(

-0.86

Phosphorus ‘,; ‘, .:,’ ,I 0.06 0.04 ‘I 0.10 0.01 ‘Co.01 0.0 I -0.02 2 3?

Nitrogen6.5 14.2’ 3.9 0.1 1 c -16.7 ‘,‘,>

0.02 0.01 -1.2/

Sulfur 12.7 6.1’ 17.6.Chloride 6.2 d 4.6 ;

~~...--_------

‘After Likens et al., 1977.bDlsso/ved substances and organic particulate matter constitute net output of elements that occur or have occurred in

Ionic form within the ecosystem; inorganic particulate matter is largely composed of unweathered primary or secondary .. ‘4 tminerals.

‘Nearly zero. ‘_i k

‘\ I dRelatively small.*PaMy due to biological activity within the ecosystem, i.e., nitrogen fixation, gaseous absorption, and impaction on plant

‘X,\I

,. .I

- ._I__ ---._ - -____c

70 Pattern and Process in a Forested Ecosystem IA

fixation in this 55yr-old aggrading ecosystem (Bormann et al., 1977). The&mate is minimum because denitrification. which would remove N fromthe ecosystem in a gaseous form. would not be detected by our budgetmeasurements. The estimate is avera,oe because the input-output budgets

are averaged over 10 years (Likens et al.. 1977) and because the biomassaccretion value is an average of two pentad periods. 1956-60 and

1961-65. In a more complete budget. nitrogen losses due to denitrificationwould have to be added to losses in stream water. and hence estimatednitrogen fixation could be larger than that predicted. We also have noestimate of the input of nitrogenous =oases or aerosols. If these are present

in significant amounts. the input due to nitrogen fixation would bereduced accordingly.

Of the two types of nitrogen fixation. symbiotic and free living, thelatter- is more likelv since among the higher plants common in the>igra-dizi-forest (Bormann et al.: 1970; Siccama et al.. 1970) none is

known to have symbiotic nitrogen fixation. Free-living nitrogen fixation inmixed hardwood forests has been reported by Todd (1971). and this topic hasbeen studied at Hubbard Brook by J. Roskoski (1977). Results based on theacetylene-reduction technique (Hardy et al., 1968) indicate that decayingwood, lying on the surface of the ground, is an active site of free-living

nitrogen fixation by as yet unidentified microbes. Roskoski’s preliminaryestimates indicate that small amounts of N (- 1 kg of N/yr) are added to theecosystem by fixation in woody litter within the ecosystem.

All other elements (Table 2-6) show net losses from the ecemrn.Since the ecosystem is aggrading, or growing, and these elements have noprominent gaseous phase, we can conclude that these net losses from the

intrasystem nutrient cycle must be made up by the weathering of primaryand secondary minerals in bedrock and till. In our budgetary scheme fornutrient flux and cycling, bedrock and till undergoing weathering areconsidered to be within the ecosystem’s boundaries.

NUTRIENT RESERVOIRS WITHIN THEAGGRADING ECOSYSTEM

Nutrients stored in reservoirs or sinks within the ecosystem provide

inertia and ameliorate change in functional characteristics -ofthe

s--ecosystem.enhance the rate of ecosystem development. It is apparent that nutnentlosses from the aggrading ecosystem (gross-output of &ssol_vedand organic particulate matter; Table 2-6), plus-immobilization of


if the ecosystem is to continue to grow and develop. As we have already

sinks = sources). provides a useful analytical tool for probing variousaspects of the biogeochemistry of the aggrading northern hardwood

Our nutrient flux and cycling model (Figure l-13) indicates that there Fare three potential nutrient sinks-within the aggrading ecosystem: (lb -;T.

:

biomass. (2) in available nutrients, and (3) in secondary minerals forming ., _within the ecosystem. For any element. the sum of the net accretion ofnutrients in these compartments would equal the amount immobilized ’ “’ ,within the ecosystem during some period of time (Table 2-7). _ ,x

The aggrading ecosystem by our definition is accumulating biomass in ’ : ”three subcompartments-l iving- b&mass, dead wood, and forest floor _ lT;,i(Figure 2-l); and we have estimates of nutrient accretion in eachsubcompartment (Table 2-7).

we haye no complete measure--of exchangeable or available nutri_en_tsfor the @&soi! profiled-at Hubbard Brook. However, Hoyle (1973) “2’ ”indicates that, for soils similar to Hubbard Brook, a large proportion ofthe aaIable~_nutrients are located in-the forest-floor. Since our estimates ,, ~I’of elements in the forest floor includes atoms incorporated in the ‘.’ -. 1’molecular structure of biomass and exchangeable nutrients (Covington. ’ :1976), a large part of any net change in available nutrients is already -;,,,;i-incorporated in the nutrient data associated with the total biomass ‘. ...-,,’accretion values in Table 2-7.

-The -netaccumulation of..e!ements in secondary minerals -in the soil iS-----~-- Ivery difficult to quantify directly on a short-term basis. The deeper soils atHubbard Brook are filled with rocks an&b&ders of various sizes, and

;- ,-I 3.,--j 1

the soil depth is highly variable (Figure l-7). Such heterogeneity makes 9 ,<.!~J 1 Iquantitative sampling extremely diflicult. However, a comparison of

;/

elements removed from the ecosystem [net output of dissolved substancesi .i

.and particulate organic matter (Table 2-6) plus that immobilized inaccumulating biomass (Table 2-7)] with their abundance in the weatheringsubstrate (Table 18 of Likens et al., 1977) provides some indication aboutsecondary mineral formation. Output of inorganic particulate matter isnot included because of its unweathered nature (Table 2-6) Evaluation ofL-~

I

the-removal abundance ratio shows that aluminum has the lowest ratio-at iHubbard Brook, indicating that in proportion to its abundance in rockand till it is the least removed. T h e orde_r.is Al<Si<Fe<K I:<Mg<Na<Ca. ~_.

_---df ;r,, c L F,,P,/ *I./, a & L :/c-d G-c- i;

The low ratios for aluminum, iron, and silicon, coupled with the fact j;:that aluminum, iron, and silicon oxides are major secG.qdn minerals inpodzolic soils at Hubbard Brook (Johnson et al., 1968), are suggestivethat secondary mineral formation is occurring and that mineral soilprofiles are still developing.


SOURCES OF NUTRIENTS FOR THEAGGRADING E C O S Y S T E M

Undisturbed watershed-ecosystems have two major sources of nutrients:(1) m+orologic input of dissolved, particulate, and gaseous chemicals

froxoutside the ecosystem’s boundaries and (2) release by weatheringof nutrients from primary and secondary minerals stored within the

B’Oecosystem’s boundaries (Figure 1-13). ~Biological- activity within tiw

P,“L i : system may affect the rate of both meteorologic input and weathering..!‘i For example. the architecture of the forest canopy may determine the

i<,. - i- rate of removal of airborne aerosols moving through the ecosystem. whilethe release of nutrients by weathering is influenced by a variety of

_‘,C I intrasystem biotic and abiotic factors.

--, ,-I.‘. - _ <, -:_ -2,) ,-. /?,’I

Weathering‘..,i g- .I ,T 1 When the various aspects of flux, cycling, and biomass ac+mula{j~~for__- _- ___--

_ .I ,‘I the ecosystem are known with some precision. a budget of thesecomponents may be used to estimate unknown parameters such as-net

_jweathering release or cationic denudation (Likens et al.. 1977). This is

i directly analogous to what we did with the two atmophilic elements N and_ ‘; .,-iv: S t~&&a~e net meteorologic inputs of gases and aerosols. In that case it

.,I,_ was possible to use the budget method because the concentrations ofthese two elements in the weathering substrate were negligible or very

r L,; :‘I ’small and the other components were known. Using the budget method

o*7’ i. (Figure l-14), we can estimate _ne?_we&heringrelease_f~reach~elementby

4 the relationship (streamwater output f biomass immobilization - meteo-1 ,{’ i’. rologic i<$t =-weathering). Weathering for Ca calculated by the budget

7 I. method is a reasonable estimate of the total amounts actually released by”,;*” ’ weathering, since calcium-rich minerals are thought-to be ~the least. st_a&

and calcium is not thought to be significantly.involved.in the formationof_secondary minerals at Hubbard Brook (Johnson et al., 1968; Likens et al.,1977). Assuming that calcium is completely extracted from primary min-erals in the weathering process, we calculate that about 1500 kg/ha of rockor till are weathered each year (Likens et al., 1977).

Net weathering release calculated by the budget method does notdistinguish between the processes of differential rates of release and

formation of secondary minerals. As we have already discussed.aluminum, silicon. and iron give strong indications of being involved insecondary mineral formation. Hence, the actual weathering rates forthese elements are probably higher than those estimated by difference.lucid-soils-phosphorus is ~thought to form a rariety_of_insQ!ub1-=p_our&through -reactions. with. soluble and hydrous oxides of .X&-ALand Mg (Brady, 1974); hence, the actual weathering rate of phosphorus islikely to be higher than estimates of net release of phosphorus based onthe budget method. PotaSsium and Mg may be involved to a lesser extent

2. I, , I I- 1, 4 ;- _,


in secondary mineral formation and, along with Na, may have adifferential rate of release from the weathering substrate at HubbardBrook (Johnson et al.. 1968; Likens et al.. 1977).

Relqtiveknportance of Weathering andMeteorologic Input

The relative importance of the various sources for biologically importantnutrients for the aggrading ecosystem are given in Table 2-8. Weatheringis the major source of several elements-Fe. Ca, K. Mg,Na, Al,and 6Meteorologic input is the most important source of-.-h- and S and asecondary source of Na, Mg, K, and Ca.

yrs .-ys,- , 3 *’ r (>,’ J ,,.J i5 _ _..,_ ,t2 !. L _‘,

CIRCULATION AND RETENTION OF NUTRIENTS

A fundamentally important characteristic of the aggrading forestedecosystem is its ability to store agd-recvcle nutrients. Thisfrugalh@+<ry cf. vital resources assures the constant growth and develop-m”“n-- of. the -ecosystem during .-this per&l, Losses are minimal, anda-yailable-andstored .re_sources accumulate. This may be illustrated mostclearly by an examination of the nitrogen cycle.

The Nitrogen Cycle

The biogeochemistry of nitrogen requires special consideration, since thiselement plays a key role in the metabolism of the ecosystem. The eco-

---~ .~- -syste-rn-accumulates nitrogen throughout the Aggradation Phase > 1.5through.170 .years after clear-cutting (Figure 2-10). During. the first thirdof the period, the rate of net accumulation is high, ca. 20 kg/ha-yr butaccumulation tapers off to about 5.0 kg/ha-yr during the last 30 years.‘?;hechange in rate is due to the coincidence of growth in the living biomassand forest floor during Years 15 to 75 and the fact that accumulation-in

e--living biomass alone sustajns.yhe increase in the standing crop after Year7.5 and rate of living biomass accumulation begins to level off after Year140. It is interesting to note that P_ead~-wood has iittle. direct influence.eonthe ecosystem’s standing crop of nitrogen.

As we have already discussed, there are two possible sources of i/initrogen to sustain accumulation--fhe igput of.nitrogen.in.bulk pr.eci-p_i_ta-km and the_fixatio.n-n_of_gas_eouS_ nitrogen .by_m_icroorganisms within the r *1.!-*,

iecosystem. We can speculate about the relative importance of thesesources throughout the period by using detailed biogeochemical datafrom small watersheds at Hubbard Brook. The input.of nitrogen in bulk

-~~precipitation h&s-averaged 6.5 kg/ha-yr. Th-ides the~~p~~-limit ofnitrogen that might be accreted from precipitation; however.

somenitrogen is inevitably lost in drainage water, and if we subtract the Il-yr


T&g 2-8. Relative Importance of Various %Ur0%S O~&~~e~!s~oLtiLHubbard Brook Experimental Forest’

Nutrient Source (%)

Nutrient

IronCalciumPotassiumMagnesiumSodiumNitrogenSulfur

Precipitation Net Gas or Weathering

Input Aerosol Input Release

b 6-- loo

9 - 91

11 - 89

15 - 85

2.2 - 78b-31 69’

65 31d 4

‘Data from Likens et al., 1977.bNeariy zero.‘Nitrogen fixation.d/mpacfion and gaseous absorption.

3 0 0 0 I I I I I

(Excluding Mineral

Live Biomass

D e a d W o o d

0 ’ !20 40 60 80 100 120 140 160

Age Of Ecosystem After Clear-Cutting

Figure Z-10. Estimated accumulation of .nitrogen during the Aggradation Phasebased on living and dead biomass projections multiplied by appropriate nitrogenconcentrations. The estimate of total nitrogen does not include nitrogen in themineral soil.


average stream-water loss, 4.0 kg/ha-yr, the net precipitation contribution--.- ~

to nitrogen accumulation is about 2.5 kg/ha-yr (Table 2-6). If we subtractthis net contribution from the yearly net accumulation of nitrogen (Figure2-lo), we obtain an estimate of the net accumulation due to nitrogenfixation. This number varies from about 28 kg/ha-yr in Years 20 to 40 toabout 2 kg/ha-yr during Years 140 to 170.

However, there is a problem with this method of calculation since itcannot take into account denitrification, Thus, it is conceivable thatdenitrification fluctuates greatly during the aggrading period, whereasnitrogen fixation does not; this would cause the balance within the systemto be affected more by changes in denitriiication than by changes in nitro- Jgen fixation. Because dead wood is an important site of nitrogen fixation, ,Lit is intere+ng~ to note that the standing crop of dead wood (Figure 2-l) -1””remains about the same from Year 50 to 170. DifEculties involved in .-., , ‘!long-term evaluations of nutrient behavior will be discussed more fully inChapters 5 and 6.

! I’

Several important features of the nitrogen cycle emerge from the ,: .overall pattern as seen in a Syr-old ecosystem at Hubbard Brook (Figure ’2-11). These are: (1) ?O%-of the nitrogen added to the ecosystem is added&-nitrogen Jixation, 30% is added in-precipitation, and little, if any, is ..- -_added by weathering; (2) of the estimated 20.7 kg/ha entering the system. i+~-8~0% is held or accreted within the ecosystem; (3) of the estimated83.6 kg/ha_of nitrogec compounds added to the inorganic nitrogen pool,-only 5%-leaks out of the ecosystem; and (4) of the 119 kg of nitrogen

?

.estimated to be used in growth processes of the plants. 33% is withdrawnt,- r

from storage locations within plant tissues in the spring and utilized in Ii. “’growth and a like amount is withdrawn from the leaves and stored inmore permanent tissues shortly before leaf senescence in the fall (Ryan, .1978).

The latter point is of particular significance since it suggests that az----good portion of the early spring growth is sustained by nitrogen (andphosphorus) withdrawn fro~GZ?ient pool-within living plant tissue.This internal reserve gives the plants some independence from soilsources during the critical early-growth period (Ryan, 19% The internalresorption pool would be almost completely destroyed in clear,-cutting! /;> r.,v:, i ; ,, ,:‘) .: ; ‘.~i; .h ’

Niaogen-incorporated in organic compounds within the aggradinge$-Gsystem~is ultimately decomposed in a number of steps to ammonium:Ammonium ions may be strongly held on cation-exchange sites. fixed onorganic matter or clay, or taken up directly by plants. Ammonjum, alongwithsome-other nitrogenous compounds, may be used as a substrate forthe proce$s~pf nitrification,the..oxidation of nitrogen by a variety ofchemoautotrophic and heterotrophic microorganisms with the productionof nitrate and hydrogen ions (Table 2-9). Nitrate is very soluble and maybe easily leached from the ecosystem if-Z:; not taken up by plants.


I 614.2

I Aggrading Ecosystem

Figure 2-l 1. Major nitrogen transfers into,out of, and within a 55-yr-old aggrad-

ing northern hardwood ecosystem. Numbers are kilograms of nitrogen per

hectare per year (Bormann et al., 1977).

Ammo&m, on the other hand. as a cation,is more tightly retained in the

.-- -soil profile.

1 --f-- -*xac*;nn rpoarding nitrogen flow in the aggrading ecosystem is

u~gLr” cy ..---___ -~ nonium is nitrified to nitrate. Th.e forest floor of

le podzol soils, particularly under conifers, may show little tendency tonitrify (Rome11 and Heiberg, 1931; Corke, 1958), and it may be presumed

that _ammonium is the major nitrogen ion utilized by .the..\rege&&~.

.,

B‘ ’

, ,\ ,. \. ./ i/ ! ;. .,’

Table 2-g. Chemoautotrophic Nitrification”

Step 1

NFL,+ + l’iZ Oz+NOz- + 2H’ + Hz0 66 Kcal

Step 2

LvaNOz + I’/2 02+NaN03

=After Alexander (1961).

18 Kcal


Many other podzol and podzolic soils, particularly under northernhardwood forests, have fores_t_floors that have a strong tendency-to nitrifywhen a sample is removed and kept moist in the laborator_y_(Romell andHeiberg. 1931). In these latter soils. however. there is little informationon the amount of nitrification that occurs within the intact forestecosystem.

Soils under aggrading northern hardwoods at Hubbard Brook fit into \the latter category since samples brought into the laboratory rapidly nitrify L /. ”(J. Duggin. personal communication): Smith et al. (1968) have isolated , ._chemoautotrophic-nitrifying bacteria from these soils. Application of a~~~~ ~- ~..chemical inhibitor of autotrophic nitrification reduced the rate of nitrifica- /tion in these laboratory samples by about 50%. This indicates that -.I,’nitrification is not simply the result of chemoautotrophic nitrification but ‘.i

. _probably has a heterotrophic component as well (J. Duggin. personal ’communication).

,e’

Levels of exchangea&ammonium and nitrate measured throughout : “\ “,the-year at Hubbard.Brook- in the soil and in soil placed in plastic bags - 6 2’and replaced in the soil (the Solling Method) suggested that no more than10 to 20% of the ammonium is nitrified to nitrate during the year 7 , [(Melillo, 1977). Dominski (1971) estimated that about 16% of the <SC, =- -ammonium was nitrified and that about lZkg/ha of N03-N were producedeach year. Thus, despite the strong tendency of samples from the forest

r L/ :.

floor at Hubbard Brook to undergo nitrification in the laboratory, it is ic”/,-.,‘,apparent that under field conditions only a small proportion of thenitrogen cycling within the forest is converted to nitrate. , I!i

From this, we conclude that, although a strong pGtentia1 for nitrification ‘1 c__-- -. ~. _----._ (. ., .~exists, that potential is not realized ins the intact- aggrading ecosystem. ,Sbme factor such as competition for ammonium ions between green ”plants and nitrifying microorganisms. or- some set of factors, partiallyinhibits- the prodess. Another possibility is allelopathic inhibition -of *“”nittifyitii-organisms by the dominant vegetation of the ecosystem (Rice, r1974). Studies by Melillo (1977) indicate that root exudates may cause [ *‘: I,--~-_~~-some allelopathic inhibition of nitrification in the soil at Hubbard Brook. r;,i.~i

The partia1 ecosystemhas interesting _energetic -implications. Nitrification is an oxidative -’ 7 u(

/.‘J),

state,--the ‘*/“‘-’~- -.~~utjlization of a nitrate ion by a green plant requires a considerably greaterexpenditure of the plant’s energy (to reduce the nitrogen) than utilizationof an ammonium ion, which is already in a reduced state. This sugge&s

)L h!//:,- - - - -

that inhibiting the nitrification process in.-n.et~ I’-“’ p,~---_ ~---.--primary productivity. We can make a rough calculation of this & ‘- ‘I’* ,~,In r’ ‘/

**-c cG’.L-’ .*

10s ‘:i,!~ i-i “.--


biochemically usable form. This amounts to about 1% of the net primaryproduction of the 55yr-old aggrading ecosystem (Table 2-l).

Nitrogen relationships within the aggrading _ecosygem are reflected in ahighly reproducible pattern in the stream water draining these systems.

Nitrate concentrations, which are always low, i.e.. <6mg/liter, fall totheir lowest levels in late May when vegetation growth begins (Figure 2-3)and remain low until October when the vegetation becomes dormant.Thereafter, concentrations gradually rise to a maximum in late winter andearly spring. This pattern is related to the heavy utilization of nitrogen bythe biological fraction of the ecosystem during its active season (Johnsonet al.. 1969). The stream-water pattern provides a useful monitor of thenitrogen cycle within the ecosystem, and, as we shall see,_rhispat&rn is _~

wholly disrupted by clear-cutting.Thus, we conclude that within the aggrading ecosystem~most -nitrogen

compounds are decomposed to ammonium,which is directly taken up bygreen plants. The process of nitrification, although it occurs, does notreach its full potential because of the action of unknown factors or inhib-iting substances within the aggrading ecosystem. Stream water draining theaggrading ecosystem reflects these conditions and shows relatively smalllosses of nitrate and ammonium from the ecosystem. On this basis, wemay consider the cycling of nitrogen within the aggrading system to betight. Specifically. only about 5% of the nitrogen added to the inorganicpool (Figure 2-11) is lost in drainage water.

Summary

I. Our studies of the aggrading forested ecosystem show that the-&?-fwater_,nutfienls, and energy_ is highly regulated by biotic and abioticcomponents of the ecosystem..

2. Stability of the, aggradmg ecosystem !s characterized by:-y--ma. modest but rather constant excess of primary production over

‘t. .̂d e c o m p o s i t i o n ,w h i c h p r o d u c e s a p r e d i c t a b l e r a t e - & a d -

,I , J >- ;,-,/,;, tion of living and dead biomass;b. cl&e regulation of both -the-chemistry of drainage waters and-/e-y; and

c. an ability to exert considerable control over intrasystem aspectsof the hydrologic cycle, such as losses of water by evapotranspira-tion.

-These processes, integrated within&nitsstiy.xv,- top.ography,-biota,. and. leveLof_ecosysteti~elopment&~ermine the-

-&e &nutrient . reservoirs land produce-nutrientcycl_e_s tvplfied b y-r&Gmum outputs-ofdissolved-substancesand particulawatterandby maximum resistance tq erosion.____. ~~._.

:.;..:


3. The ability of the biotic fraction of the ecosystem to utilize externalenergy sources to convert liquid water to vapor through transpirationand interception results in the conservation of nutrients by its effect onthe volume of flow and on discharge rates.

a. .The reduction.~of streamflow has the direct &ect of holdingnutrients.within the ecosystem because, in the aggrading ecosys-tem, concentrations of dissolved substances in drainage water arehighly predictable and are only modestly affected by flow rate.Hence, the output of dissolved substances is closely related tothe amount of streamflow exiting from the ecosystem. Factorsregulating the concentrations of nutrients in drainage water areonly partially understood, but the following may be important:

quantity_;incl-quality-~ of_inputs, rates of- wea ther ing , micro-meteorological conditions at the level of the forest floor whichregulate soil chemistry, and allelopathic control of soil micro-organisms by living vegetation.

b. ,Particulate matter losses are a nonlinear direct function ofdischarge- rates. In the aggrading ecosystem. transpiralion andinterception can damp +harger_ates by lowering.!& quantity ofwater stored within the system. Thus, a heavy rain when thestorage volume of the system is relatively empty will producerelatively little flow and modest discharge rates so long as storagecapacity is unsatisfied. Transpiration -tends .to-keep.-hydroiogic-storage-at a mi.nimum duringthe growing season-and streamflowduring summer months is typically low even though precipitationis evenly distributed throughout the year.

c. P2K&culate matterlosses are a function not only of discharge ratebut -also- of. erodibility or the capacity of the ecosystem to resisterosion. Although geologic and other abiotic factors are irn-portant in determining the erodibiiity of an ecosystem, biotaand its inorganic debris exerts considerable control over thisphenomenon.

4. The effectiveness of all of these mechanisms in maintaining the--.-- --__..in tegr i ty of the aggrad ing ecosys tem aga ins t ex te rna l fo rces?

geological-degradation is seen in the fact that the 55yr-old forestedecosystems at Hubbard Brook have average gross dissolved-substanceand particulate-matter losses of 13 and 2.5 t/km’-yr, respectively.These losses constitute only a small fraction of the elements stored inthe various compartments of the ecosystem. This condition existsdespite the location of the Hubbard Brook ecosystem-watersheds onrelatively steep slopes subject to an average rainfall of 130cm/yr.

5. As an example of nutrient cycling in the aggrading ecosystem, thenitrogen cycle is discussed. Salient features are:

a. The system is accumulating nitrogen both from precipitationinput and nitrogen fixation.

b. The rate of accumulation changes with time and is not simplyrelated to biomass accumulation.

c. Nitrogen cycling is tightznly about 5 /OO o f t h e mkeralized- ..~.. _ ._ -~ ~-~-nitrogen is lost as geologtc*Qut.

d. The soils under the northern hardwood ecosystem have a strongpotential to nitrify, but under the aggrading condition relativelylit t le nitrification occurs , and mos t n i t rogen i s cyc led asammonium.

f. herbert bormann gene e. likens pattern and process in a forested...

Documents