AEROBIC DECOMPOSITION OF SEDIMENT AND
DETRITUS AS A FUNCTION OF PARTICLE SURFACE
AREA AND ORGANIC CONTENT
Bawy T. Hargrad
Freshwater Biological Laboratory, The University of Copenhagen, Hillergd, Denmark
Oxygen uptake by microbial populations on mud, sand, and various types of detritus was
measured in short-term experiments in aerated water at 20C. Sample size had no effect
on oxygen consumption per unit weight, but stirring increased uptake. A 2% Formalin
solution completely stopped biological uptake of oxygen by microorganisms on sand and
detritus; lake muds showed various degrees of chemical uptake of oxygen.
Microorganisms on dead Phrugmites leaves consumed oxygen at an increasing rate during
the first few days of decomposition, followed by a decline to a rate comparable to uptake
by freshly collected detritus. Limnaea feces initially consumed oxygen three times more
rapidly than detritus, but after 5 days the rates were equal.
Detritus consumed up to three orders of magnitude more oxygen per dry weight than
sand; uptake rates were inversely related to particle diameter. The logarithm of oxygen
uptake was directly related to the logarithm of particle organic content. Particulate oxygen
uptake in this and previous studies fell between 0.1 and 10 mg 02 (g organic matter)-l hr-l,
a rate inversely related to particle diameter. Oxygen uptake per unit weight by particles
of ashed mud and sand exposed for 24 hr to decomposing detritus in a nutrient solution
and transferred to fresh solution was inversely related to particle size and similar to rates
measured with freshly collected samples. On an areal basis all particles consumed between
0.01 and 1.0 x lo-’ mg O2 cm-l hr-I.
The negative linear correlation on logarithmic axes of surface area, organic carbon and
nitrogen and bacterial plate counts with sediment particle size is similar to that observed
for measures of oxygen uptake. Bacteria cover only a few percent of particle surfaces. This
may result in the narrow range of measures of microbial community respiration on an
INTRODUCTION oxygen uptake by detritus, mud, and
Aquatic sediments and detritus consume sand sediments. Waksman and Hotchkiss
oxygen when stirred in aerated water, The ( 1938)) Teal ( 1962), and Rybak ( 1969)
oxygen uptake is an integrative measure suggested that oxygen uptake is related to
of all oxidative processes occurring in the the organic matter available for decompo-
sample, both chemical and biological. sition; Odum and de la Cruz (1967) and
Reducing substances are usually rapidly Fenchcl (1970) d cmonstrated an invcrsc
oxidized. Respiration of the organisms as- relation between detritus particle size and
sociated with particles is probably pri- oxygen consumption. Chemical oxidation
marily bacterial, although algae, protozoa, of reduced subsurface sediment may also
and fungi may also contribute. Whatever be related to the concentration of reduced
populations are involved, measurements of substances (Gardner and Let 1965; Har-
oxygen uptake reflect the metabolism of grave 1972).
communities of microorganisms involved Teal (1962) and Odum and de la Cruz
in the decomposition of natural substrates. ( 1967) found approximately similar rates
of oxygen consumption by Spartina detri-
No measure is provided, however, of an-
tus. Thalassia decays much faster than
oxic microbial decomposition. does Spartina (Wood et al. 1969)) yet oxy-
Many workers have measured rates of gen uptake by detritus from these plants,
freshly collected or artificially prepared, is
l Present address: Fisheries Research Board of
Canada, Marinc Ecology Laboratory, Bedford In- similar ( Fenchel 1970). Lake Esrom scdi-
stitute of Oceanography, Dartmouth, Nova Scotia, mcnts consume oxygen at rates similar to
LIMNOLOGY AND OCEANOGRAPHY 583 JULY 1972, v. 17( 4)
584 BARRY T. I-IARGRAVE
those observed with salt marsh mud (Har- was used immediately. Disks (l-cm diam)
grave 1972). were cut from larger pieces of decaying
I here compare rates of oxygen uptake vegetation with a cork borer and lo-20
by sediment and detritus particles of dif- mg (dry wt) were placed in bottles. In-
ferent sizes and organic content, These cubations lasted l-2 hr in the dark at 2OC,
results, combined with previous measurc- and bottles were rotated (4 rpm) on their
ments, show an inverse relation between long axis to prevent oxygen depletion
particle size and oxygen consumption. The around sedimented particles. All experi-
importance of surface area and organic ments were carried out during summer
content is examined by calculating oxygen and early fall when field temperatures
uptake of natural and recolonized parti- ranged between 10 and 22C.
cles on a surface area basis. A possible A buffered (pH 6.8) solution of 2% For-
mechanism causing similarities and differ- malin was used to fill some BOD bottles.
ences in rates of sediment oxygen con- ZoBell and Brown (1944) found that a
sump tion is suggested. 0.25% Formalin solution stopped bacterial
I gratefully acknowledge a NATO Sci- respiration, so oxygen consumed by parti-
ence Fellowship provided by The National clcs in this solution was assumed to be due
Research Council of Canada, Professors E. to chemical uptake. Additional blank bot-
Stecmann Nielsen and T. Fcnchel kindly tles corrected for small amounts of oxygen
read the manuscript and offered com- uptake by the Formalin.
ments. Mr. S. Bordon calculated the re- Dissolved oxygen was measured by a
gression lines. micro-Winkler technique (Fox and Wing-
field 1938) with samples taken in 5-ml
MATERIALS AND METI-IODS syringes; a solution of 1% NaN3 was added
Sediment and detritus from freshwater to the NaI solution to prevent interference
sources were collected from eutrophic from reduced substances. Syringe samples
Lake Esrom and Frederiksborg Castle of supernatant were taken after settling
Lake, from the humic-acid brown-water of particles. Mud particles, in particular,
Store Grib Lake, all located in northern appear to bind the iodide liberated dur-
Zealand (Denmark) (Berg 1938), and from ing the Winkler reaction and thus reduce
Grane Lang Lake, an oligotrophic lake in the apparent oxygen concentration. For-
Jutland (Whiteside 1970). Other samples malin soIutions were titrated slowly over
were taken from Hclsing@r, Kronborg, and 10 min for accurate end-point determina-
Julebaek beaches along the northern @re- tions. Measurements of dissolved oxygen
sund near Helsinger ( Fenchel 1969). were accurate to at least 0.1 mg/liter.
Either a Kajak corer (Brinkhurst et al. Changes as smalI as 1% of total oxygen
1969) or a glass cylinder pushed by hand present could be measured; changes were
into shallow sand sediment was used to usually much greater, but concentrations
take undisturbed cores of mud and sand. were never reduced more than 30%. Cor-
In the laboratory, samples of mud and fine rections were made for volume displace-
sand sediment wcrc taken from the surface
Mud and detritus were filtered onto pre-
of the cores by pipette, mixed, and drained ashed and preweighcd glass-fiber filters,
on filter paper. About 0.1 g wet wt was dried ( 8OC for 24 hr ), and weighed. Sand
removed from the filter paper with a grains and pebbles were dried in pre-
spatula and added to 28-ml Pyrex BOD weighed crucibles. Organic matter was
bottles filled by siphon with filtered aer- determined by loss on ignition at 550C for
ated water from the collecting area. Large 3 hr; carbonate dissociation is avoided at
pebbles, with a total weight of up to 5 g, this temperature if ashing times are not
were added to bottles with forceps. Detri- long (Rybak 1969). Particle size (mean
tus collected by hand from beach margins value of greatest dimension of randomly
DECOMPOSITION OF SEDIMENT AND DETRITUS 585
chosen particles) of sand grains was deter- diameter is Fenchel’s internal surface area
mined by microscopic examination; at least divided by 16.
100 were examined when their diameter Recolonization expcrimcnts were on
was less than 500 p. All pebbles greater mud, sand, and pebbles from Lake Esrom
than l-mm diameter in the bottles were sterilized by ashing at 550C for 3 hr. A
measured and the mean diameter dcter- small piece of decomposing Phragmites
mined, Mud sediments consisted entirely leaf in a solution of 1% each of glucose,
of silt and amorphous organic debris that, peptone, and K&P04 and mud-water ex-
with gentle agitation, passed through a tract (1 kg of wet profundal mud auto-
nylon screen with 100-p-mesh openings claved with 1 liter of Lake Esrom water)
but clogged screens of 20 p; an average was used for inoculum, with controls of
particle diameter of 50 p was assumed. sterilized particles in autoclavcd lake wa-
The detritus from decaying vegetation, tcr without detritus. After 24 hr of aera-
when not cut with a cork borer, was mea- tion, control and inoculated particles were
sured wet on graph paper or under a rinsed and placed in freshly filtered water,
binocular microscope. and oxygen uptake was measured.
Fecal pellets were collected from three
common benthic invertebrates in Lake Es- RESULTS
rom. Adult Ilyoclrilus hummoniensis ( Oli- Experimental variables
gochaeta) and fourth instar Chironomus
anthracinus (Diptera) larvae were held in Lake Esrom mud between 1 and 20 mg
freshly collected profundal lake mud, and ( dry wt ) per bottle gave no significantly
fecal pellets were collected from petri different rates of oxygen uptake per gram
dishes in which several animals had been dry weight. Eight samples of surface sedi-
held overnight. Limnaea palustris ( Mol- ment taken from four profundal mud cores
lusca ) , collected with decaying Phragmi- during June consumed 0.22 * 0.04 mg 02
tes communis, browses epiphytic material g-l hr-l regardless of sample size. Samples
and produces abundant feces. All fecal of sand grains (250-p diam), from a beach
pellets were kept in gently aerated lake surface in Lake Esrom, of dry weights
water when not used within a few hours, up to 1 g showed similar reproducibility
Rates of oxygen consumption per unit (220%). Two replicate samples were then
weight were compared on a surface area used for each determination and the mean
basis by dividing rates of oxygen uptake value recorded. Short incubations, l-2 hr,
by the number of square centimeters of ensured sufficient oxygen yet gave mea-
sediment. Area was calculated for detritus surable changes in concentration. The
particles larger than 1 mm before drying rapid oxygen uptake by some sediment
by doubling that traced on graph paper. and detritus particles necessitated agita-
Values for mud and detritus with a par- tion of samples. Profundal mud from Lake
ticle diameter less than 1 mm were taken Esrom consumed 0.32 mg 02 g-l hr.-l when
from Fenchel’s (1970) estimate of total bottles were slowly rotated and 0.08 when
surface area per gram dry weight of Tha- standing. Oxygen uptake of beach sand
lassia detritus. Sand grain and pebble sur- decreased by 15% when it was not stirred.
face area was estimated from a relation
between mean sand grain size and “in- Relative importance of chemical
ternal surface” arca derived by Fenchcl and biological oxidation
( 1969), based on 10 cm3 of sample and Oxygen uptake by sand and detritus
the assumption that sand grains were from beaches was always completely
spherical. If the specific gravity of sand stopped by 2% Formalin; oxygen uptake
(quartz) is assumed to be 1.6 (Kaye and by subsurface sediment from below 2 cm
Laby 1956), then the surface area per in cores from Store Grib Lake and Lake
gram dry weight for a given sand grain Esrom was unaffected by Formalin treat-
586 BARRY T. HARGRAVE
TABLE 1. Effect of Formalin treatment (+ = with, - = without) on oxygen uptake by surface (up-
per 5 mm) mud from four Danish lakes at various times of the year. All values are means of at least
two separate determinations
- + by chemical Biological Sediment mg 0,
oxidation respiration org matter (g org
(mg 0, g-1 hr-1) (%I (mg 0, g-1 hr-1) (%I matter)-1 hr-1
15 Jun 0.75 0.44 59 0.31 39.6 0.78
13 Nov 1.48 0.87 59 0.61 40.0 1.52
15 Dee 1.14 0.35 31 0.79 39.9 1.98
5 Sep 0.74 0.48 65 0.26 72.0 0.36
22 Sep 1.34 0.73 54 0.61 45.0 1.36
15 Mar 0.10 0.01 10 0.09 21.3 0.43
ment, indicating that oxidative processes of organic matter in various types of scdi-
were chemical, Surface muds (upper 5 mm ment was standardized by expressing oxy-
as taken by pipette) from various lakes, gen uptake on an ash-free weight basis.
however, showed varying degrees of sensi- In sand and detritus this measure reflects
tivity to poisoning (Table 1). Oxygen was the metabolic activity of communities of
present ( 3-5 mg/liter ) over all sampled microorganisms since these substrates
cores. Data from Lake Esrom show con- showed no chemical oxidation. Mud par-
siderable seasonal variation in total oxygen ticles did undergo chemical oxygen uptake
uptake and the amount susceptible to For- and this has been subtracted in the calcu-
malin poisoning. lations of biological respiration in Table
1. The range of values of oxygen uptake
Particle size and organic content by various particles was reduced by these
Oxygen uptake by sand grains, detritus calculations, and the effect of particle size
particles, and mud sediments was differ- was enhanced (Fig. 3). Oxygen uptake by
ent per unit weight and each varied on the all material fell within the range [O.l-10.0
basis of mean particle diameter (Fig. 1). mg 02 (g org matter)-l hr-l] observed for
With the logarithmic transformation of detritus ( Fig. 1) largely because values
both axes, Fig. 1 reveals a linear, inverse for oxygen uptake by sand grain commu-
relation between particle diameter and ox- nities, when expressed on an ash-free ba-
ygen uptake. Values for mud were omitted sis, were raised. Although differences are
from the regression calculations because of large, organic matter in various muds was
the inaccurate estimation of particle size. oxidized at much lower rates than detritus
Oxygen uptake for a given particle diam- particles of similar size.
eter was two to three orders of magnitude
higher for detritus than for sand. Sand Stage of decomposition
grains and pebbles contained less than 1% Fresh snail fecal pellets (about 0.5-mm
organic matter while mud and detritus diam and 2 mm long) contained 80-90%
particles contained 40-90% organic matter. organic matter. This always increased as
After logarithmic transformations, oxygen they aged, probably from organic matter
uptake was linearly related to organic con- removed from solution by microorganisms
tent (Fig. 2). colonizing their surfaces, although it could
The effect of differences in percentage also result from leaching of inorganic salts.
DECOhXTi’OSITION OF SEDIMENT AND DETRITUS 587
-4 -I 0 8c,““I 1 8 I II”11 1 18““‘I ’ “A?
0 1 2 3 4
loglo particle diameter
FIG. 1. Relation between mean particle diam-
eter and oxygen uptake per gram dry weight of FIG. 2. Relation between percent organic mat-
sand, mud, and detritus from various aquatic ter and oxygen uptake per gram dry weight of
habitats. different particles described in Fig. 1. Regression
Solid circle-detritus from Lake Esrom shore line, calculated omitting bacteria and mud sam-
(1, mean value for five species of deciduous tree ples, y = -2.15 + 0.90x (r = 0.91, sE of slope
leaves; 2, unidentified reed detritus; 3, Phrugmi- 0.075, significant at P = 0.05).
tes leaves); solid triangle-invertebrate feces (l-
day-old ) ( 1, Limnaea palustris; 2, Ilyodrilus
hammoniensis. 3, Chironomus anthrucinus); solid days the rate of oxygen uptake declined
square-sand and unknown seaweed detritus from
Julebaek beach; cross-sand, detritus, and mud
to 0.55 mg O2 g-1 hr-l, followed by a
from Store Grib Lake; open triangle-Frederiks- slow decrease to 0.25 after holding for 20
borg Castle lake mud; point-Grane Lang Lake days. The decline in oxygen uptake was
profundal mud; circled cross-IIelsinggr beach accompanied by successions of communi-
sand; circled point-Kronborg beach pebbles;
open square-sediment bacteria (Pseudomonas
ties of microorganisms associated with the
spp.) cultured from Lake Esrom surface profundal fecal material. Ciliates, abundant after the
mud; open circles-Fenchel (1970), Thalassia de- first day, declined after the fifth. Fungal
tritus; squared points-Odum and de la Cruz hyphac grew rapidly and changed the
( 1967), Spartina detritus. pellets from brown to white within 12 hr.
Regression lines-detritus: y = 1.45 - 0.52 x ( r
= 0.65, SE of slope 0.17); sand: y = -0.20 - 0.85 Phragmites leaves, collected dry from
x (r= 0.95, SE of slope 0.07). Slopes are signifi- dead plants in November, were placed in
cantly different from zero (P = 0.05). Data for aerated lake water in the laboratory, Disks
mud, bacteria, Thalassia detritus ( Fenchel 1970))
and Spartina detritus (Odum and de la Cruz 1967)
of leaf material tested immediately con-
not included in regression calculations. tained 82.5% organic matter and consumed
0.22 mg 02 g-l hr-l (Fig. 4). After 3 days
Oxygen uptake by microorganisms on fe- oxygen uptake had increased to 1.0 mg
cal pellets used within 6 hr of production 02 g-l hr-l and leaf material contained
was three times faster than when pellets 90% organic matter. During the next 2
were first held for 12 hr (Fig, 4). After 3 days there was little change in organic
588 BARRY T. HARGRAVE
c 0 fecal pellets
“,~~~~~‘, Phragmites detritus
0 10 20 30 40
log ,O particle diameter Time (days)
FIG. 4. Changes in oxygen uptake by Limnuea
FIG. 3. Relation between mean particle diam- feces and Phragmites detritus held at 20C in aer-
eter and oxygen uptake per gram organic matter ated Lake Esrom water. Samples removed for use
(calculated from Figs. 1 and 2 ). Symbols as in in short-term experiments at various times. a-
Fig. 1. Regression line, calculated omitting mud, Detritus prepared from plant leaves collected dry;
bacteria, and data (line joining points) from Odum b-detritus wet and partly decomposed when
and de la Cruz (1967), y = 2.07 -0.69x (r = 0.79, collected.
SE of slope 0.10, significant at P = 0.05).
content ( 90.5% ) , but oxygen consumption Odum and de la Cruz (1967) and
fell to the original level. Oxygen uptake Fenchel ( 1970) related oxygen uptake to
remained constant over the following 34 particle size of Spartina and ThaZussiu de-
days, while organic matter slowly in- tritus. Their results, plotted in Fig, 1, are
creased to 98.5%. of the same magnitude as mine for detri-
In a similar experiment with Phragmites tus from different sources.
I have used organic contents of sedi-
leaves collected as wet detritus from be-
ment and detrital material reported by
side Lake Esrom, oxygen uptake, initially
various workers as a basis for comparing
0.5 mg 02 g-l hr-l, decreased steadily, rates of oxygen uptake per unit dry weight
while organic matter increased from 94.3- (Fig. 5); sewage particles, artificial stream
98.0%. The final constant rate of oxygen periphyton, aquatic vascular plants, and
consumption by this material was similar aquatic bacteria are included for compari-
to those by freshly wetted leaves and son. When expressed on logarithmic axes,
Limnaea feces at the end of the holding these data show the same trend and span
period (Fig. 4). This value (0.25 mg 02 the same range as those from this study
g-1 hrl) falls within the range observed (Fig. 2). Odum and de la Cruz ( 1967)
for detrital material from other sources measured both organic content and parti-
(Fig. 1). cle size of decomposing Spartina detritus.
DECOMXPOSITION OF SEDIUENT AND DETRITUS 589
m mg O2 g-’ hr-’
0 percent organic matter
m mg O2 (g organic matter )-’ hr-’
m mg o2 x10-~ cm-* hr-’
particle diameter CJI)
mud sand pebbles
FIG. 6. Mean oxygen consumption and organic
log,o percent organic content of triplicate samples of ashed mud, sand,
matter and pebbles from Lake Esrom after 24 hr of
FIG. 5. Comparison of organic content and ox- exposure to Phragmites detritus in a nutrient-cn-
ygen uptake per gram dry weight for different riched solution. Control samples, not exposed to
types of sediment and detritus, aquatic vascular detritus and nutrients, contained no mensurable
plants, periphyton communities, and bacterial cells organic matter and consumed no oxygen.
from previous studies. Circles indicate average
values for measurements at about 20C. Lines rep-
resent range of published observations. Oxygen uptake, on an ash-free basis, fol-
l-Johnson ( 1936 ) , resting marine bacteria; lows the trend of my data and falls within
2-ZoBell and Stadler (1940), multiplying lake the same order of magnitude when com-
bacteria; 3-Liagino and Kusnetzov (cited in Zo-
Bell and Stadler 1940), lake bacteria; 4-Har-
pared with particle size ( Fig. 3).
grave (1969), estimated value for lake sediment
bacteria; 5-Gessner ( 1959), various aquatic vas- Particle surface area
cular plants; 6-McIntire ( 1966)) periphyton All particles considered in this study
communities in artificial streams; 7-Teal (1962),
Spartina detritus during 4-week decay period; 8- consumed between 0.01 and 1.0 x 1O-3 mg
Odum and de la Cruz ( 1967), Spartina detritus; 02 cm-2 hr-l with the exception of freshly
9-Fenchel ( 1970), Thalassia detritus. produced Limnaea feces (Table 2); oxygen
N-Water Pollution Research Rep. ( 1968),
consumption of fcccs fell to within the
suspended solids in treated sewage effluent; ll-
DiSalvo ( 1971), coral reef regenerative sediment; range observed for all other particles
12-I&rnerblad (1930), mean value for 14 Swed- within 3 days. Dctri tus particles con-
ish lakes ( assumed 2 g dry wt/flask); 13- sumed more oxygen on an arcal basis than
Miyadi (cited in ZoBell 1946), lake sediments;
14-Waksman and Hotchkiss (1938), marine sedi- pebbles, sand, or mud. There was no con-
ments from Woods Hole; 15-Kato ( 1956), ma-
rine bottom sediments from the northern Japan
Sea; 16-Teal and Kanwisher (1961), surface salt-
marsh mud; 17-Gardner and Let ( 1965), Lake (1969), black clots, Polish lakes of varying trophic
Mendota sediment (initial 3-day rate); 18-Rybak type.
590 BARRY T. IIARGRAVE
TADLE 2. Comparison of mean particle diameter recolonized pebbles, however, was an or-
(p), estimated surface area per gram dry weight
der of magnitude higher than that of
(A), and oxygen uptake per unit surface area (C)
for various types of particles. Data fur oxygen freshly collected material. There were few
uptake per gram (B) from Fig. 1 protozoa in the detritus used for inocula-
tion and thus oxygen consumption prob-
and source A B C
ably reflects only bacterial respiration.
No organic matter remained on particles
Pebbles after ashing, but after 24 hr of exposure
Kronborg beach 9,000 1.2 0.00012 0.10 to the nutrient solution and detritus par-
7,800 1.5 0.00030 0.20
ticles a loss on ignition was measured in
5,340 1.9 0.00655 0.2.9
4,000 2.7 O.OC@70 0.26 all samples (Fig. 6). The percentage or-
Esrom beach 1,914 5.6 0.0012 0.21 ganic matter was inversely related to par-
1,260 6.5 0.0011 0.17 ticle size but all particles contained an
Sand order of magnitude less organic matter
Esrom beach 550 21.9 0.0058 0.26 than naturally occurring samples of simi-
3G6 31.2 0.0028 0.09 lar type.
Helsing@r beach 350 32.0 0.0046 0.14 Oxygen uptake per gram organic matter,
185 62.,5 0.0064 0.10 which ranged from 9-28, was not related
Julebmk beach 248 50.0 0.0086 0.17
to particle diameter (Fig, 6). Oxygen
Lake mud consumption per unit surface area in-
Esrom 50 6,000 0.86 0.14 creased with increasing particle size; the
Store Crib 50 6,000 0.74 0.12 range of values was within that observed
Grane Lang 50 6,000 0.10 0.02
Frederiksborg Castle 50 6,000 1.34 0.22
with natural particles (Table 2).
Esrom shore debris - - 0.33 0.30
Wet elm leaves - - 0.18 0.42
The enumeration of microorganisms can-
Phragmites - - 0.24 0.42 not serve as a measure of their importance
E quisetum - - 0.10 0.26 in proccsscs of decomposition. There may
Limnaea feces, fresh 5001 650 3.50 5.38 be no clcarcut relation between numbers
3-day-old 500 650 0.55 0.85 and metabolic rate. The dynamic nature
Spartina 239 1,050 0.63 0.60
(Odum and de 150 1,800 1.62 0.90 of microbial populations also makes esti-
la Cruz 1967 ) 64 4,200 1.81 0.43 mates of population size of limited use;
Thalassia 1,000 420 0.40 0.95 grazing may stimulate or retard cell divi-
(Fcnchel 1970) 400 850 0.60 0.71 sion and substrate changes determined by
100 2,500 1.05 0.42
species succession may continuously affect
population size. These limitations prob-
ably account for the lack of correlation
sistent relation between oxygen uptake per between sediment organic content and
square centimeter and particle size. indices ( plate counts) of bacterial biomass
(Cooper et al. 1953; Volkmann and Oppen-
Recolonization experiments heimer 1962; Anthony and Hayes 1964).
Previously ashed pebbles, sand, and Also, the amount of total organic matter
mud, held in sterilized water for 24 hr, in sediment does not indicate its availa-
showed no measurable oxygen consump- bility to cithcr microorganisms ( Waksman
tion. After 24 hr of exposure to Phragrni- and Hotchkiss 1938) or invcrtcbratcs (Har-
tes detritus in an enriched medium, rinsed grave 1970a).
samples of mud and sand consumed oxy- Measurement of oxygen consumption
gen at rates inversely related to particle appears to be a more useful basis for com-
size and in the range of similarly sized paring microbial activity in different types
particles with natural communities of mi- of sediment and detrital material. Hayes
croorganisms (Fig. 6). Oxygen uptake by and MacAulay (1959) suggest that oxygen
DECOMPOSITION OF SEDIMENT AND DETRITUS 591
uptake by mixed mud may represent an minerals (Gardner and Lee 1965) and
integrative measure of lake productivity. slow microbial oxidation of organic matc-
Although they measured oxygen consumed rial such as lignin and humic complexes.
by restratified mud-water systems, there Whatever the cause of the time-dependent
was a linear relation between oxygen up- decrease in oxygen uptake, short-term ex-
take and “lake quality index,” a value pcrimcnts give maximum rates of oxygen
derived from an estimate of fish yield. consumption, which might reflect rates oc-
curring in nature if samples were taken
Experimental variables from well-aerated areas.
The interpretation of oxygen uptake The depth from which a sample is taken
necessitates standardization of those exper- can be important. The variable proportion
imcntal variables which might affect mea- of chemical oxidation in lake muds (Lon-
sured rates. For example, Waksman and ncrblad 1930 and Table 1) differs from
Hotchkiss (1938) found shaking to have marinc sand sediment where sterilization
no effect, but total oxygen uptake de- completely stopped oxygen uptake (Waks-
creased with the amount of sediment per man and Hotchkiss 1938). The absence
bottle; I observed opposite effects of shak- of chemical oxidation of beach sand and
ing and sample size. They left sediment detritus probably results from the well-
standing for up to 30 days, stirring it by oxygenated conditions on wave-washed
hand several times daily. In long-term beaches where few reducing substances
experiments, oxygen dcplction around set- exist ( Fenchel 1969). Usually, however,
tled particles would lower the rate of oxy- sediments contain an oxidized surface
gen uptake. The effect would be greater layer overlying anoxic reduced sediment
with larger samples ( Kato 1956). Oxygen ( Hayes 1964; Fcnchcl and Ricdl 1970).
uptake by mud sediments, because of the Oxygen uptake by subsurface sediment
small particle size and often high propor- may bc cntircly chemical ( Moore 1931))
tion of chemical oxidation ( Table I), may but thin layers of surface sediment show
bc more affected by agitation and size of varying proportions of chemical and bio-
sample than sand sediments. The small logical oxygen uptake (IIargrave 1972 and
range of weights used hcrc (1-28 mg dry Table 1) . Biological oxidation accounted
wt/28 ml) may not have been sufficient for an average of 30% of oxygen consumed
to show such an effect. by salt marsh mud (Teal and Kanwisher
1961), with no clear effect of depth,
Sample variability Samples of sediment and detritus held
Changes in rates of oxygen consumption in acrated water rapidly become dcoxy-
by sediment and detritus with time can genatcd and anaerobic bacteria appear.
be expected as easily oxidized substances This is not surprising since oxygen uptake
disappear. Previous studies with various by aquatic bacteria at 20C may range
types of scdimcnt (Anderson 1939; Kato from NO-500 ~1 02 mg-l hr-l ( Fig. 5). A
1956; Teal and Kanwisher 1961; Gardner 1-g sample of detritus at 20C consumes
and Lee 1965; Hamilton and Greenfield about 1 mg 02 hr-l; thus, all of the oxygen
1967; Rybak 1969; Fenchel 1969; DiSalvo in 100 ml of water could bc consumed
1971) all show a time-dependent decrease within 1 hr. The rapid rate of oxygen con-
in oxygen consumption although the time sumption by bacteria in sediment may de-
scale is different. The immcdiatc effect oxygenate bottom deposits ( ZoBell 1946).
may represent oxygen uptake by reduced Dissolved oxygen may only penetrate the
minerals, such as ferrous iron and ferrous top few millimeters of mud sediments not
carbonate, and microbial decomposition of exposed to water turbulence ( Hargrave
easily oxidized organic substances. Long- 1972). Oxidative decomposition in such
term slow rates of oxygen consumption habitats must bc limited by oxygen supply.
may bc due to acid-insoluble iron sulfide The supply of easily oxidized organic
592 BARRY T, HARGRAVE
matter must also affect rates of sediment composition. The time-course of detritus
oxygen uptake. The higher rates of oxy- decomposition as measured by oxygen up-
gen uptake by surface sediment in Lake take seems similar for different substrates.
Esrom in November and Dcccmber may It is not clear why different particles of
reflect the decomposition of organic mat- a similar size should have even an approx-
ter newly deposited after phytoplankton imately similar rate of oxygen uptake
blooms ( Table 1). (Fig. 3). Decomposition processes in dc-
The stage of decomposition of detritus tritus and mud, where the substrate is
is important. Organic content of fcccs may metabolized, may differ from those on
increase or decrease with time (Newell sand and pebbles where mineral surfaces
1965; Hargrave 1970b ) , indicating bactc- serve largely as a source of attachment.
rial colonization ( Fenchel 1970). Feces If, however, only organic matter adsorbed
were rapidly colonized (Fig. 4) and within to surfaces is available to heterotrophic
4 days, oxygen uptake decreased from 3.5- bacteria ( ZoBell 1946)) oxygen consump-
0.4 mg 02 g-l hr-l. Phragmites detritus tion by microorganisms associated with
did not reach as high rates of oxygen up- sand and pcbblcs also reflects substrate
take as feces but also showed a decrease decomposition.
with time. Presumably initially high rates Oxygen uptake by mud particles tends
result from an increase in microbial respi- to be lower than might be expected on
ration as labile substrates, such as mucous the basis of its particle size relative to that
secretions are utilized. Grazing organisms of detritus, and at least some of the oxy-
such as protozoa, which appeared in both gcn is used for chemical oxidation. The
feces and detritus during initial maximum somewhat arbitrary estimate of mud par-
rates of oxygen uptake, would also con- ticle diameter makes comparison on the
tribute to increased oxygen consumption. basis of particle size difficult. Organic
Teal (1962) observed a peak in oxygen matter in lake sediment ranged from 20-
uptake in freshly produced Spartina dctri- 70%; detritus always contained more than
tus during the first week of decomposi- 70%. The substrates in mud may bc less
tion, similar to that noted with Phragmites easily oxidized. Grane Lang Lake mud
(Fig. 4). Explanations for differences in consumed the least oxygen ( 0.10 mg 02
the maximum rate obtained by Spartina g-l hr-l ) and also had the lowest organic
detritus ( 14 mg O2 g-l hr-l ) , Limnaea fe- content (21.3%) of the lake sediments ex-
ccs (3.5), and Phragmites detritus ( 1.0) amined. Low annual primary production
during the initial period of decomposition in this lake means that its surface scdi-
as well as the varied final rates of oxygen ment is older than that of the more pro-
uptake (0.25 mg O2 g-l hr-l for feces and ductive Lake Esrom or Frcderiksborg
Phragmites, 2.0 for Spartina) may reflect Castle Lake (Whiteside 1970). Rybak
the species of microorganisms involved (1969) reported that lake sediments that
and the residual substrates. consumed less oxygen are poorer in or-
The difficulty of following decomposi- ganic matter; data summarized from other
tion by only observing changes in organic work (see Fig. 5) shows a similar trend.
matter is illustrated by the increase in These observations support Hayes and
organic content of Phragmites during de- MacAulay’s ( 1959) suggestion for sedi-
composition, in contrast to Spartina (Teal mcnt oxygen uptake as an index of lake
1962 ) . Changes in organic content in- productivity.
cludc changes in both organic substrate Particles of various types have diffcr-
and communities of colonizing microorga- ent rates of oxygen uptake per unit sur-
nisms, but changes in rates of oxygen con- face area, but the range of values is not
sumption arc more easily interpreted as large (Table 2). Estimates of surface area
they reflect the metabolic activity of all per gram for various particles span four
organisms associated with aerobic de- orders of magnitude, although oxygen up-
DECOMPOSITION OF SEDIMENT AND DETRITUS 593
take per square centimeter only has a ties of bacteria (Rubcntshik et al. 1936).
range of two. Surface area was only mea- ZoBell (1943) noted a logarithmic increase
sured for large pieces of detritus, and the extending over 18 days of bacteria at-
values assumed for other particles from tached to glass surfaces; 24 hr is probably
previous studies could introduce errors; insufficient for populations of colonizing
however, the cstimatcs are probably accu- microorganisms to reach a stable level of
rate enough for present calculations and density and metabolism.
suggest some constancy in total commu-
nity metabolism per unit surface for very Bacterial density on particle surfaces
different types of particles. Fenchel (1970) obscrvcd that both oxy-
gen uptake by Thalassia detritus and the
Recolonization experiments number of bacteria, small zooflagellatcs,
Oxygen consumption per unit weight and diatoms wcrc proportional to surface
and the adsorption of organic matter were arca. The number of organisms per sur-
both inversely related to particle size in face area, about 3 x lo6 bacterial cells
recolonization experiments, but the dif- cm-2 by direct counts, was always similar
ferences were not as great as would be on intact dead leaves, various sized parti-
expected on the basis of surface area. Ox- cles, and artificial detritus after 4-6 days.
ygen uptake per unit weight and surface Tsernoglou and Anthony (1971) calculated
area was within the range found with the density of bacteria by direct micros-
freshly collected samples. All recolonized copy on freshwater sediment particles and
particles had an approximaetly similar rate summarized previous estimates of bactc-
of oxygen uptake on the basis of organic rial density on sand, pebbles, and soil
content, higher than that found with the surfaces; depending on the source of the
smallest detritus particles used by Odum sediment, there was one bacterial cell for
and de la C* luz ( 1967). Sincc respiration every 70-300 p2. If an average bacterial
of rinsed particles was measured after ccl1 is assumed to cover 1 p2, then <I%
transfer to fresh nonenriched solutions, or- of the particle surface arca was colonized.
ganic matter on the surfaces must have The bacterial density observed on detritus
undergone relatively rapid oxidation simi- by Fcnchel (1970) would cover about 3%
lar for each particle type; the organic mat- of the surface area.
ter and microbial populations adsorbed to Batoosingh and Anthony ( 1971) also
the various particles in these experiments calculated bacterial density on pebbles di-
may also have been similar. rcctly by fluorescent microscopy. Single
Recolonized particles should all have cells accounted for 70% of the bacterial
similar rates of respiration per unit sur- populations, When colonies were present
fact if oxygen uptake by communities of many seemed to be chance aggregations
microorganisms on particle surfaces is de- rather than colonies, as the individuals
pendent only on surface area, However, were generally quite separate from each
uptake increased with increasing particle other. An analysis of the dispersion of
size ( Fig. 6). There was a disproportion- the direct counts showed that the bacteria
atcly high organic content on pcbblcs; were slightly over-dispersed on the pcbblc
mud contained less organic matter than surfaces. There was an average of 1 cell
expcctcd on the basis of its surface arca. /300 /.L!
These differences may result from altcra- These calculations show, however ap-
tion of the surface characteristics of the proximately, that bacterial cells do not
particles in ashing, Badcr (1962) did not completely cover these sediment and de-
find simple adsorption phenomena bc- trital particles. Other habitats, such as the
twecn dissolved organic compounds and surface film of stagnant water, may have
mineral particles. Various sediments differ higher ccl1 densities ( Fenchel, personal
in adsorbing properties for different spc- communication), but natural sediment
594 BARRY T. HARGRAVE
tcrial numbers (Batoosingh 1964). Detritus,
more flocculant and with a higher organic
content than sand or pebbles, should of-
fer a more suitable substrate. Still, from
Fenchel’s ( 1970) observations, bacterial
2 cover in detritus is far from continuous.
Burkholder (1963) has shown that growth-
regulating substances are released by scdi-
mcnt bacteria. ZoBell (1943) described a
faintly staining film which surrounds bac-
terial cells attached to glass slides, two
or three times the dimension of the cells
themselves, whose size increases with age.
The nature of this material is unknown.
\\ Whatever the cause of spacing between
percent ‘. \
organic ‘* bacterial cells, only a fraction of the par-
-2- \ ‘. \
nitrogen ‘. ‘. ticlc surface arca is covered and the eco-
‘. logically available space is considerably
-3 - mg O2 g-’ hr-‘x less than the total arca. Such a relation-
ship bctwecn density and surface area
\ would impart some degree of similarity
-4 -I I I , 1\ to measures of community respiration per
0 1 2 3 4
unit area (Table 2). An upper limit to
loglo particle diameter
cell density may be determined by avail-
able space, with metabolism per cell
FIG. 7. Comparison on logarithmic axes of related to the prcscncc of oxidizable sub-
mean particle diameter and internal surface area
strates, the two factors interacting to pro-
of sand (from Fenchel 1969), percent organic
carbon and nitrogen in beach sediments (from duce a relatively constant areal rate of
Longbottom 1970), bacterial plate counts in dif- community metabolism. This would ex-
ferent sediments (from ZoBell 1946), and mea- plain why although detritus consumes more
sures 0I particle oxygen consumption taken from oxygen per unit surface than do sand,
Figs. 1 and 3. Solid lines are calculated regres-
sion lines presented in the respective references.
pcbblc, or mud particles, all rates vary
The regression for bacterial numbers on particle within narrow limits.
diameter is y = 4.00 - 0.94x, T = 0.95. Dotted Longbottom ( 1970) reported that in
line indicates a slope of -1.0. sediments from gently sloping, moderately
well-drained beaches, both organic nitro-
particles of various types do not appear gcn and carbon content varied inversely
to bc heavily colonized. Cell counts per with median particle diameter after loga-
unit weight give a false impression of high rithmic transformations. The slope coeffi-
numbers; the spatial distribution of micro- cicnts, about -1.0, and the intercept values
organisms is given pcrspectivc when den- of the regression lines showed little sca-
sity is related to surface area. sonal change and were similar for sedi-
Many factors are involved in the spacing ments from diffcrcnt arcas. Fcnchel (1970)
of bacterial cells. Grazing and physical has also observed an inverse linear rela-
abrasion can remove cells. Algal and bac- tion (h = -1.0) bctwecn median particle
terial cells are usually found in depressions diameter and surface area of sand grains
and crevices of sand grains and pebbles on a logarithmic basis. These curves are
(Meadows and Anderson 1966), suggest- redrawn in Fig. 7. Apparently organic
ing physical removal from raised surfaces. matter adsorbed or attached as micro-
Particles removed from these influences, organisms to these sediment particles is
however, do not show an increase in bac- proportional to surface arca and the rcla-
DECOMPOSITION OF SEDIMENT AND DETRITUS 595
tionship is approximately similar for sam- mining rates of oxidation of sediment and
ples from different arcas. detrital organic matter.
The relationship between sand grain The narrow range in rates of oxygen
size and oxygen uptake per unit weight uptake per square centimeter by different
also approaches a slope value of -1.0 (Fig. particles of a similar size suggests that
7)) perhaps reflecting the close correlation communities of microorganisms colonizing
between sand grain diameter and surface various surfaces may have similar rates of
area; mud and detritus particles are not metabolism. In aerobic habitats, coloniza-
spherical, so they need show no similar tion and succession may occur to the
relation. point where all of the oxygen supply is
Detritus oxygen consumption per unit fully used, Measurement of oxygen up-
weight is also inversely related to particle take by communities of microorganisms
size but with a slope less than -1.0 (Fig. associated with sediment and detritus par-
7). The relationship bctwccn particle di- ticlcs offers a simple method for investi-
ameter and external surface area per gram gating these processes.
dry weight of Thdassia detritus too has a
slope of less than -1.0 on logarithmic axes REFERENCES
( Fcnchcl 1970). ANDERSON, D. Q. 1939. Distribution of organic
Bacterial counts per unit dry weight of matter in marine sediments and its availabil-
sediment incrcasc with dccrcasing particle ity to further decomposition., J. Mar. Res.
size (ZoBell 1946), and the relationship ANTHONY, E. I-I., AND F. R. HAYES. 1964. Lake
(b = -0.94) resembles that described for water and sediment. 7. Limnol. Oceanogr.
surface area in Fig. 7. Those data were 9: 35-41.
derived from plate counts and thus only BADER, R. G. 1962. Some experimental studies
with organic compounds and minerals. Oc-
an index of bacterial numbers is provided, cas. Publ. 1, p. 42-57. Grad Sch. Oceanogr.,
although a relatively constant proportion Univ. RI., Kingston.
of the total bacterial population may have BATOOSINGII, E. 1964. The bacteriology of ma-
been measured in diffcrcnt sediment types. rine pebbles. MS. thesis, Dalhousie Univ.
For regression calculations, sand, silt, clay, 67 p.
-, AND E. H. ANTHONY, 1971., Direct and
and colloidal particles have been assumed indirect observations of bacteria on marine
to have diameters of 500, 50, 5, and 1 p; pcbblcs. Can. J. Microbial. 17: 655-664.
these are reasonable, but arbitrary, esti- BERG, K. 1938. Studies on the bottom animals
mates of particle size within the range of Esrom Lake. Kgl. Danske Vidensk. Selsk.
Skr. Nat. Math. Afd. 9 8: 1-225.
obscrvcd in the different types of scdi-
BRINKHURST, R. O., K. E. CHUA, AND E. BATOO-
mcnt cxamincd by ZoBell. SINGEI. 1969. Modifications in sampling
procedures as applied to studies on the bac-
CONCLUSIONS teria and tubificid oligochaetes inhabiting
aquatic sediments. J. Fish. Res. Bd. Can,
There is apparently a basically simple 26 : 2581-2593.
relation between surface area, organic BUJXKHOLDER, P. R. 1963. Some nutritional re-
content, and the size and metabolism of lationships among microbes of sea sediments
communities of microorganisms associated an d waters, p. 133-150. In C. H. Oppen-
heimer fed.], Symposium on marine microbi-
with aquatic sediment and detritus parti- ology. Thomas.
clcs ( Fig. 7). Undoubtedly, tcmpcrature, COOPER, B. A.,E. G.D. MURRAY, AND II. KLEERE-
oxygen supply, and the nature of the or- KOPPER. 1953. The bottom sediments oE
Lake Lauzon, 2. Rev. Can. Biol. 12: 457-
ganic substrate can modify any such rela- 494.
tionship; variability introduced by these DISALVO, L. H. 1971. Regenerative functions
factors, howcvcr, should not detract from and microbial ecology of coral reefs. 2. Can.
the observation that particle size, and thus J. Microbial. 17: 1091-1100.
FENCIIEL, T. 1969. The ecology of marine mi-
the surface arca available for oxygen ex- crobcnthos, 4. Ophelia 6: 1-182.
change, is an important factor in deter- -. 1970. Studies on the decomposition of
596 BARRY T. HARGRAVE
organic detritus derived from the turtle grass water sand grains. Nature 212 : 1959-1060.
Thalassia testudinum. Limnol. Oceanogr. MOORE, II. B. 1931. The muds of the Clyde
15: 14-20. area. 3. J. Mar. Biol. Ass. U.K. 17: 325-
- AND R. J. RIEDL. 1970. The sulfide 358.
syitcm : a new biotic community underneath NEWELL, R. 1965. The role of detritus in the
the oxidized layer of marine sand bottoms. nutrition of two marine deposit feeders, the
Mar. Biol. 7: 255-268. prosobranch Hydrobia ulvae and the bivalve
Fox, II. M., AND C. A. WINGPIELD. 1938. A Macoma balthica. Proc. Zool. Sot. London
portable apparatus for the determination of 144: 25-45.
oxygen dissolved in a small volume of water. ODUM, E. P., AND A. A. DE LA CRUZ. 1967.
J. Exp. Biol. 15: 437-445. Particulate organic detritus in a Georgia salt
GARDNER, W., AND G. F. LEE, 1965. Oxygena- marsh-estuarine ecosystem, p. 383-388. In
tion of lake sediments, Int. J. Air Water G. II. Lauff [ea.], Estuaries. Publ. Amer.
Pollut. 9: 553-564. Ass. Advan. Sci. 83.
GESSNER, F. 1959. Hydrobotanik, 2. VEB RUBENTSHIK, L., M. B. ROISIN, AND F. M. BIEL-
Deut. Wiss. 701 p. JANSKY. 1936. Adsorption of bacteria in
HAMILTON, R. D., AND L. J. GREENFIELD. 1967. salt lakes. J. Bacterial. 32: 11-31.
Manometric assay of the metabolic activity RYBAK, J. I. 1969. Bottom sediments of the
of marine sediment micro-biota. Z. Allg. Mi- lakes of various trophic type. Ekol. Pol.
krobiol. 7: 19-27. Ser. A 17 : 611-662.
HARGRAVE, B. T. 1969. Epibenthic algal pro- TEAL, J. M. 1962. Energy flow in the salt
duction and community respiration in the marsh ecosystem of Georgia. Ecology 43 :
sediments of Marion Lake. J. Fish. Res. Bd. 614-624.
Can. 26: 2003-2026. AND J. KANWISHE~. 1961. Gas exchange
-. 1970a. The utilization of benthic mi- in ‘a Georgia salt marsh. Limnol. Oceanogr.
croflora by Hyalella axteca (Amphipoda). J. 6 : 388-399.
Anim. Ecol. 39: 427437. TSERNOGLOU, D., AND E. H. ANTHONY. 1971.
-. 197Ob. The effect of a deposit-feed- Particle size, water-stable aggregates, and
ing amphipod on the metabolism of benthic bacterial populations in lake sediments. Can.
microflora. Limnol. Oceanogr. 15 : 21-30. J. Microbial. 17: 217-2'27.
-. 1972. Oxidation-reduction potentials, VOLKMANN, C. M., AND C. II. OPPENHEIMER.
oxygen concentration and oxygen uptake of 1962. The microbial decomposition of or-
profundal sediments in I,ake Esrom. Oikos ganic carbon in surface sediments of marine
23 : 167-177. bays of the central Texas Gulf Coast. Publ.
HAYES, F. R. 1964. The mud-water interface. Inst. Mar. Sci. (Texas) 8: 80-96.
Oceanogr. Mar. Biol. Annu. Rev. 2: 121-145. WAKSMAN, S. A., AND M. HOTCI-IKISS. 1938. On
AND M. A. MACAULAY. 1959. Lake the oxiclation of organic matter in marine
water and sediment 5. Limnol. Oceanogr. sediments by bacteria. J. Mar. Res. 1: lOl-
4 : 291-298. 118.
JOIINSON, F. II. 1936. The oxygen uptake of WATER POLLUTION RESEARCII REPORT. 1968.
marine bacteria. J. Bacterial. 31: 547-556. H. M. Stationery, London. 53 p.
KATO, K. 1956. Chemical investigation on ma- WHITESIDE, M. C. 1970. Danish chydorid Cla-
rine humus in bottom sediments. Mem. Fat. docera : modern ecology and core studies.
Fish. Hokkaido Univ. 4: 91-209. Ecol. Monogr. 40: 79-118.
KAYE, G. W. C., AND T. II. LABY. 1956. Tables WOOD, E. J. F., W. E. ODUM, AND J. C. ZIEMAN.
of physical and chemical constants and some 1969. Influence of sea grasses on the pro-
mathematical functions. Longmans. ductivity of coastal lagoons, p. 495-562. In
LONGBOTTOM, M. R. 1970. The distribution of A. Ayala-Castafiares and F. B. Phleger [eds.],
Arenicola marina (L. ) with particular refer- Coastal lagoons, a symposium. Univ. Mexico.
ence to the effect of particle size and organic ZOBELL, C. E. 1943. The effect of solid sur-
matter of the sediments. J. Exp. Mar. Biol. faces upon bacterial activity. J. Bacterial.
Ecol. 5: 138-157. 46: 39-56.
L~NNERBLAD, G. 1930. Ober die Sauerstoffab- -. 1946. Marine microbiology. Chronica
sorption des Bodensubstrates in cinigen Sec- Botanica. 240 p.
typen. Bot. Notis. 1930: 53-60. -, AND B. F. BROWN. 1944. Studies on the
M~NTIRE, C. D. 1966. Some factors affecting chemical preservation of water samples. J.
respiration of periphyton communities in lotic Mar. Res. 5: 178-184.
environments. Ecology 47: 918-930. AND J. STADLER. 1940. The effect of
MEADOWS, P. S., AND J. G. ANDERSON. 1966. Mi- oxigen tension on the oxygen uptake of lake
cro-organisms attached to marine and fresh- bacteria. J. Bacterial. 39: 307-332.