ABSTRACT
A. Clark Gaither. BELOWGROUND BIOMASS IN SIX SALT MARSH PLANT
COMMUNITIES. (Under the direction of Vincent J. Beilis.) Department of
Biology, May, 1985.
Belowground standing crop, net annual production, distribution of
macro-organic matter with depth, turnover time, and belowground/
aboveground biomass ratios were estimated for six salt marsh plant
communities bordering estuarine tributaries of South Creek in Beaufort
County, North Carolina. Estimates were derived from macro-organic
matter (MOM) values derived from analysis of core samples obtained at
monthly intervals throughout 1982.
Mean belowground standing crop values (kg*m~2.yr-1) were:
herbaceous annuals 5.53, Cladiurn jamaicense 7.44, Spartina cynosuroides
8.34, Typha domingensis 2.65, Juncus roemerianus 7.80, and Distich1 i s
spicata 2.61. MOM within these plant communities was not evenly
distributed throughout the soil profile. The bulk of this material
resided within the 0 to 30.0 cm depth increment which contained most of
the living roots and rhizomes.
Three indices of net annual production (NAP) were also evaluated.
Estimates of NAP based on maximum minus minimum values from periodic
regression analysis (PRA) were judged to be most reliable. NAP values
(kg.m-2) based on PRA were: herbaceous annuals 1.34-, Cl adi urn jamai cense
2.33, Spartina cynosuroides 2.16, Typha domingensis 3.55, Juncus
roemerianus 1.52, and Distichlis spicata 3.69. Generally, these values
fell within the range of belowground production values for comparable
marshes along the Atlantic and Gulf Coasts. Belowground NAP within
these communities ranged from two to four times greater than aboveground
production.
BELOWGROUND BIOMASS
IN SIX SALT MARSH PLANT COMMUNITIES
A Masters Thesis
Presented to
The Graduate Faculty of the Department of Biology
East Carolina University
In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Biology
by
A. Clark Gaither
May, 1985
BELOWGROUND BIOMASS
IN SIX SALT MARSH PLANT CONWUNITIES
by
A. Clark Gaither
APPROVED BY:
SUPERVISOR OF THESIS
Vincent J/. Beilis, Ph.D.
THESIS COMMITTEE
Robert R. Christian, Ph.D.
Graham J. Davis, Ph.D.
Lee i. Otte, Ph.D.
CHAIRMAN OF THE DEPARTMENT OF BIOLOGY
/ Joseph G. Boyette, Ph.D.
DEDICATION
For their continued love and support, I would like to dedicate this
work to my Mom and Dad, Mena Bel 1 and Dent Turner Gaither.
ACKNOWLEDGMENTS
This work was made possible by grants from North Carolina Phosphate
Corporation, Aurora, North Carolina. Many thanks go to employees Rusty
Walker, Page Ayres, David Bradshaw, and the hobbit Duke Whedbee.
In addition I would like to thank the many other individuals
who provided valuable and indispensible assistance: field hand and king
of the MOM washers, Tom James; MOM washers and field hands first class.
Chip Green and Charles Saunders; graphics champion, Susan Finch; word
processor masters, Sandy Tomlinson, Steve Harlan, and Carl Hayes;
computer genius John Cramer, statistics expert Jim Yucha; and C. T.
Hackney for his help with the periodic regression analyses.
For their advice and critical reviews of this work, I would like
to extend gratitude to thesis committee members Robert Christian, Graham
Davis, and Lee Otte.
Special thanks go to senior advisor Vincent Beilis whose rugged
determination, intestinal fortitude, bulldog tenacity, and sometimes
infinite patience saw this graduate student through the successful
completion of a Masters program in Biology.
CONTENTS
Page
LIST OF FIGURES i
LIST OF TABLES ii
INTRODUCTION 1
DESCRIPTION OF STUDY AREA 5
MATERIALS AND METHODS 14
RESULTS AND DISCUSSION 21
Belowground Standing Crop 21
Net Annual Production 23
Computations 23
Comparison With Other Studies 26
Belowground Biomass Pathways 35
General Seasonal Pattern 37
Herbaceous Annuals Community 38
Cladium jamaicense Community 39
Spartina cynosuroides Community 39
Typha domingensis Community 40
Juncus roemerianus Community 41
Pi Stic hi is spicata Community 41
Distribution of Macro-organic Matter with Depth 42
Turnover Time 45
Belowground/Aboveground Biomass Ratio 47
SUMMARY AND RECOMMENDATIONS 50
LITERATURE CITED 52
LIST OF FIGURES
Figure 1. Location map of South Creek study area.
Figure 2. Vegetative cover map of Jacks Creek tributary marsh shov/ing
sampling sites for herbaceous annuals (A), Spartina
cynosuroides (S), Cladium jamaicense (C), Typha domingensis
(T), J uncus roemen anus ( J), and the marsh water level
recorder (R).
Figure 3. Vegetative cover map of Jacobs Creek tributary marsh showing
the sampling site for Distichlis spicata (D).
Figure 4. Coring apparatus for sampling marsh type sediments, modified
from Gallagher (1974).
Figure 5. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
herbaceous annuals community.
Figure 6. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
Cladiurn jamaicense community.
Figure 7. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
Spartina cynosuroides community.
Figure 8. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
Typha domingensis community.
Figure 9. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
Juncus roemerianus community.
Figure 10. Living aboveground biomass, actual macro-organic matter
biomass, and predicted macro-organic matter biomass for the
Distichlis spicata community.
Figure 11. Pathways open to belowground biomass, modified from Good et
al. (1982).
1
LIST OF TABLES
Table 1. Vegetative cover of Jacks Creek marsh.
Table 2. Vegetative cover of Jacobs Creek marsh.
Table 3. Estimates of belowground standing crop for six salt marsh
plant communities. Estimates derived from this study are
based upon yearly means in kg.m-2 to a depth of 30.0
centimeters.
Table 4. Comparison of belowground productivity among six salt marsh
plant communities using four methods of estimating net annual
production.
Table 5. Periodic regression models for macro-organic matter in six
salt marsh plant communities.
Table 6. Belowground productivity estimates for salt marsh plant
communities along the western Atlantic and Gulf Coasts
modified from Good et al. (1982).
Table 7. Distribution of macro-organic matter (MOM) with depth in six
salt marsh plant communities.
Table 8. Belowground biomass turnover times (years) for six salt marsh
plant communities.
Table 9. Be 1owground/aboveground biomass ratio for six salt marsh
plant communities.
INTRODUCTION
Over the past twenty years, salt marsh ecosystems have received
considerable attention from biologists. Salt marsh productive capacity
became of interest to estuarine ecologists following the publication of
a paper by John Teal ( 1962) describing energy flow in a salt marsh at
Sapelo Island, Georgia. Teal concluded that 45% of the marsh's
production was removed by tidal flux to estuarine waters, thus
contributing to the detrital based food web. This view was expanded by
Eugene P. Odum (1963) who six years later introduced the term
"outwelling" to describe the transport of nutrients from the salt marsh
to estuarine based consumers. Since that time, numerous research
efforts on salt marsh ecosystems have produced a moderate body of data
on aboveground production. These data have been summarized in
literature reviews by Keefe (1972), Turner (1976), and most recently by
Nixon (1980) in which he questioned earlier assertions that all salt
marshes are exporters of organic matter to coastal waters. Nixon's
assessment of available data indicated that salt marshes must be
considered individually, some as exporters and others as sinks of
nutrients and/or organic carbon. How then can salt marshes be
simultaneously considered as both exporters and sinks of organic carbon?
This question is currently being debated among estuarine ecologists.
After the conceptual model of "outwelling" was introduced,
ecologists directed a greater amount of research effort toward measuring
aboveground primary production in salt marsh ecosystems. This usually
entailed the use of clipped plot techniques originally developed for use
in terrestrial plant communities. Results from these studies have shown
southern (Gulf Coast) marshes to be one order of magnitude more
1
productive than those in the extreme northern latitudes. Net annual
aboveground production values of 1-2 kg.m-2 were obtained from southern
marshes while annual production in northern Canada was estimated at 0.3
kg-m“2. a gradient of production values has been shown to exist between
these two extreme locations, supporting the contention that greater
productivity is correlated with longer growing seasons (Turner, 1976).
Most early investigators of salt marsh primary productivity have,
however, failed to consider the relative contribution of belowground
components (stem bases, roots, and rhizomes) to overall productivity.
Good et al. (1982) concluded that this failure has occurred simply
because aboveground components are more accessible. If one is to gain
an understanding of the salt marsh and marsh/estuarine interactions,
should not the total vegetative component be considered when assessing
primary productivity? Kucera et al. (1967), in a study of a tall grass
prairie, stressed the importance of assessing both above and belowground
components in order to fully appraise plant productivity. In a study of
a New Jersey tidal marsh. Smith et al. (1979) noted the need for data on
biomass, turnover time, and nutrient value of the total vegetative
component as a necessary prerequisite for understanding the ecological
dynamics of salt marshes and their degree of interaction with estuarine
food webs.
It is only within the past ten years that ecologists have turned
their attention to salt marsh production belowground. John L. Gallagher
was the first to advance our ability to measure the belowground
vegetative component in a salt marsh, a component difficult to measure
and evaluate even in terrestrial plant communities (Dahlman and Kucera,
1965). Marsh soils often remain, or are intermittently, inundated and
2
can be quite soft. As a result, traditional terrestrial sampling
techniques are inappropriate. However, Gallagher (1974) devised a
coring apparatus and technique for sampling the belowground biomass
(BGB) in salt marsh soils which aliowed the collection of suitable
samples.
After Gallagher's 1974 paper was published, other ecologists
(Stroud, 1976; de la Cruz and Hackney, 1977; Whigham et al. 1978; Smith
et al., 1979; and others) began to report the significance of this
vegetative component with respect to overall primary productivity and
total standing crop. Good et al. (1982) have recently reviewed the
available literature on BGB in salt marsh ecosystems. They concluded
that the belowground marsh component is at least equal to, and often
exceeds, the aboveground component in terms of production and standing
crop. For instance, de la Cruz and Hackney (1977) reported belowground
standing crop 4-5 times greater than average aboveground standing crop
in a Mississippi Juncus roemerianus Scheele marsh, while Gallagher
(1974) in a study of'short form Spartina al terni flora Loisel at Sapelo
Island, Georgia, reported a belowground/aboveground biomass ratio of
48.9:1. Good et al. (1982) further reported that the potential
significance of the typically large belowground component to food chains
or webs, nutrient cycl ing, and structural maintenance of the marsh is
"enormous." For example, de la Cruz and Hackney (1977) sampled
estuarine seston from the aforementioned roemerianus marsh in
Mississippi which contained root and rhizome material indicating that
marsh rhizospheric components were entering the estuary, thereby
contributing to the detrital food web. They suggested that this organic
input might be significant in marsh-estuarine trophic interactions and
3
concluded that physical, chemical, and biological parameters should be
considered along with the distribution of belowground components with
depth when studying total primary production of marsh vegetation. Also,
Gardner (1975) reported that through seepage there was a probability of
rhizospheric productivity leaving the marsh in the form of dissolved
organic matter. That belowground components may serve as a potential
energy and nutrient source for estuarine-dependent consumers was
recognized by Hackney and de la Cruz (1980), prompting a decomposition
study of roots and rhizospheric materials from two species of tidal
marsh plants in a Mississippi Gulf Coast marsh. Their findings revealed
a large BGB of decaying plant material which maintained a stable, highly
nutritious energy supply available for consumption.
Before any role may be defined for salt marshes as either sinks or
sources of organic carbon, the belowground component must be more
adequately assessed. Currently, the amount of data available on BGB
within these important ecosystems is small and fragmentary. Considerably
more information concerning aboveground and belowground biomass dynamics
from a variety of salt marshes must be obtained in order to define the
relative contribution of belowground biomass production to total salt
marsh productivity.
The primary purpose of this investigation was to expand the
available data base on belowground biomass (BGB) by simultaneously
measuring this component in each of six different salt marsh plant
communities with accompanying computations of turnover times and
belowground/aboveground biomass ratios. Secondary objectives were to
determine the distribution of BGB with depth among vegetation types and
to assess whether observed seasonal patterns could be explained in terms
4
of the life history of the plants involved. Coincidentally, several
alternative methods for estimating belowground net annual production
were tested and compared.
DESCRIPTION OF STUDY AREA
For this investigation two marshes were selected for study. These
marshes border Jacks and Jacobs Creeks and are located in Beaufort
County near Aurora, North Carolina (Figure 1). Both creeks are small,
lateral, brackish tributaries of South Creek. Each watershed exhibits a
vegetative cover gradient beginning with upland forest at the
headwaters, passing through swamp forest, then abruptly changing to
brackish marsh. Both creeks open to the larger estuarine system of
South Creek and then into the Pamlico River.
The historical development of the study streams has been
established from our stratigraphic studies and carbon-14 age
determinations of subfossil organic matter recovered from marsh peat
(Beilis and Gaither, manuscript in preparation). Both Jacks and Jacobs
Creeks originated from drainage patterns established by erosion of the
Yorktown formation between two and six million years ago. This was
followed by several success!ve transgress!ve/regressive cycles and
depositional episodes. The present basin, sculptured during the
Wisconsin glaciation some thirty to fifty thousand years ago, assumed
its present level within the last five to fifteen thousand years (early
Holocene) (Eames, 1983 and Copeland et al., 1983).
Wetland vegetation encroached the basins of Jacks and Jacobs Creeks
some 4,000 years before present with brackish marsh in residence for
5
Figure 1. Location map of South Creek study area (Lat. 35°20', Long. 76°46').
less than 3,000 years. Vertical accumulation of marsh peat within the
study marshes has averaged approximately 0.8 meters per thousand -years.
Organic sediment accretion coupled with slow lateral expansion of marsh,
faster upstream expansion, and downstream erosion should persist as long
as sea level continues to rise.
The open water channel of Jacks Creek (Figure 2) is approximately
850 m long and 90 m wide at its mouth but only 7-15 m wide along the
greater portion of its length. Jacobs Creek (Figure 3) has an open
water channel 975 m long and a breadth which gradually widens from 4.5 m
at the headwaters to 106 m at the mouth. Channel depths varied from
1.3-1.5 m at their mouths to 0.6-0.9 m in the upper reaches. Each marsh
is situated in a basin having relatively steep lateral slopes (rise to
run ratios of 1:4-1:6 for Jacks Creek) of mineral soil which may reach
3.0 m in elevation above mean sea level.
Aerial photos supported by ground truth observations of Jacks Creek
have revealed a vegetative cover consisting of a mosaic pattern of
discrete vegetative communities. In contrast, the vegetative cover of
Jacobs Creek presented a much less conspicuous mosaic pattern with a
tendency for the plant cover to form complex, continuously intergrading
mixtures of vegetation. Also, the vegetative cover of Jacobs Creek
marsh exhibited a greater number of species than were encountered at
Jacks Creek marsh.
Eight vegetative communities, based on dominance of aerial
coverage, were identified at Jacks Creek (Figure 2). These were Juncus
roemerianus, Spartina cynosuroides (L.) Roth, Cladiurn jamaicense Crantz,
Typha domingensis Persoon, Hibiscus moscheutos Linnaeus, mixed marsh
(mixtures of the aforementioned types), mixed herbaceous annuals
7
Cladium iamaicense Mixed Marsh
Spaitina cvnosuroides Hibiscus moscheutos
Typha domingensis Woody Trees and Shrubs
Juncus roemerianus Mixed Annuals
10 0 10 20 30 40 50 100 METERS
Figure 2. Vegetative cover map of Jacks Creek tributary marsh showing sampling sites
for herbaceous annuals (A), Spartina cynosuroides (S), Cladium jamaicense (C),
Typha domingensis (T), Juncus roemerianus (J), a~nd the marsh water level
recorder (R).
iiiiiiiiiiiil!
--•J Qo—•• •: 'j” Cladium jamaicense M 'I'l'I'X'i'I'l '¡l"'l Distichiis spicata
*
Spartina cynosuroides •* *• *•* ** * Panicum virgatum
Typha domingensis Scirpus spp.
Juncus roemerianus Hibiscus moscheutos
10 o 10 20 30 40 50 100 METERS
Figure 3. Vegetative cover map of Jacobs Creek tributary marsh showing the
sampling site for Distichlis spicata (D).
(usual ly Aster subu1atus Michaux, Amaranthus cannabinus (L.) J. D.
Sauer, Eleocharis obtusa (Willdenow) Schultes, Osmunda regal i s
(Wi 1 1 denow) Gray, and P1uchea purpurascens (Swartz) de Candol le), and
woody trees and/or shrubs (usually Pinus taeda Linnaeus and Baccharis
halimifolia Linnaeus). Relative aerial coverage by each vegetation type
is given in Table 1. Within the stream itself, open water covers an
area of 2.49 hectares with 2.88 hectares of bordering marsh. The marsh
border was variable in width and ranged from 1.5 - 4.6 m at the
headwaters to 30 m near the midpoint of the creek.
The vegetative cover of Jacobs Creek (Figure 3), although similar
in composition to Jacks Creek, was a more heterogeneous mixture of
vegetation. Only relatively small areas of marsh were dominated by any
one vegetation type while large areas of mixed vegetation were apparent.
Seven aerially important cover types were identified on Jacobs Creek in
addition to those previously listed for Jacks Creek. These were
Spartina patens (Aitón) Muhlenberg, S. al terniflora, Distichlis spicata
(L.) Greene, Scir p u s spp., Panicum virgatum, A. cannabinus, and
Phragmites communis Trini us. Figure 3 shows aerial coverage for the
dominant species only. Minor plant communities in terms of aerial
coverage were not omitted but were incorporated in to the closest
dominant vegetation type. Relative aerial coverage by each vegetation
type is given in Table 2. The open water surface area of Jacobs Creek
consisted of 4.91 hectares- with an adjacent 5.85 hectares of bordering
marsh. Marsh border varied in width from 6-30 m along the length of the
stream.
10
Table 1. Vegetative cover of Jacks Creek marsh.
Marsh community type^ Area Relative coverage
(ha) (percent of total)
1. Juncus roemerianus (stand) 1.10 38.13
Juncus roemerianus (fringe) 0.08 2.92
2. Spartina cynosuroides (stand) 0.70 24.21
3. Cladium jamaicense (stand) 0.27 9.53
Cladium jamaicense (fringe) <0.01 <0.05
4. Typha domingensis (stand) 0.26 8.97
5. Mixed (no dominant) 0.15 5.08
6. Hibiscus moscheutos (clumps) 0.13 4.36
7. Woody trees and shrubs 0.12 4.31
8. Herbaceous annuals2 0.07 2.44
SUBTOTAL 2.88 100.00
Open water 2.49 46.31
Marsh 2.88 53.69
TOTAL 5.37 100.00
^named after dominant species comprising 51 percent or more of living
biomass of sample
^principal 1 y Amaranthus cannabinus and Aster subiatus
11
Table 2. Vegetative cover of Jacobs Creek marsh.
Marsh community typel Area Relative coverage
(ha) (percent of total)
1. Juncus roemerianus 1.51 25.79
2. Distichlis spicata 1.01 17.29
3. Cladium jamaicense 0.75 12.77
4. Scirpus spp. 0.57 9.76
5. Spartina cynosuroides 0.46 7.89
6. Woody trees and shrubs 0.45 7.60
7. Typha domingensis 0.36 6.14
8. Panicum virgatum 0.32 5.49
9. Hibiscus moscheutos 0.30 5.15
10. Spartina patens 0.06 1.08
11. Phragmites communis 0.04 0.68
12. Spartina alterni flora 0.01 0.20
13. Amaranthus cannabinus 0.01 0.16
SUBTOTAL 5.85 100.00
Open1 water 4.91 54.39
Marsh 5.85 45.61
TOTAL 10.76 100.00
^named after dominant species comprising 51 percent or more of living
biomass of sample
12
Surface water movement within these two stream systems was
controlled by several factors including lunar tides, wind tides, and
freshwater runoff. Wind tides (primarily northeasterly) exerted the
greatest influence on water level with each stream. Due to the
essentially flat topography of these marsh systems, complete flooding
occurs when wind-driven tides push a sufficient amount of water from the
Pamlico estuary into the study streams such that storage capacity is
exceeded. During the period of study these marshes were inundated well
over half of the time with single flood events lasting from several
weeks to more than a month. The extreme range in recorded water level
was from approximately 1.5 m, one meter above the marsh surface to one-
half meter below. Water level was also recorded in a well located at
Jacks Creek at a point 40.0 m north of the stream channel within a stand
of J. roemerianus (Figure 2). Thi s instrument recorded water level
within the marsh soil during non-flood conditions. Between November,
1981 and April, 1983, this recorder documented a water level range from
0.94 m above the marsh surface to 0.11 m below its surface. Both Jacks
and Jacobs Creeks are meso-ol igohaline with salinities ranging from 0
°/oo to 15 o/oo.
The term "tributary marsh" has been proposed (Beilis and
Gaither, unpublished manuscript) to describe these and similar brackish
marsh systems exhibiting the following characteristics:
(1) narrow configuration of the marsh basin,
(2) presence of a single open water channel draining adjacent
uplands as well as the marsh itself,
(3) relatively high ratio of marsh edge per unit area,
(4) mosaic pattern in the distribution of plant communities, and
13
(5)a tendency for these mosaic patterns to form as bands parallel
to the long axis of the marsh.
These characteristics serve to distinguish the tributary marsh from the
more extensive "estuarine fringe marshes".
MATERIALS AND METHODS
Procedures for collecting and processing belowground biomass (BGB)
samples have been reported previously by Gallagher (1974). These
procedures were used in this investigation with slight modification of
the coring apparatus. The coring device (Figure 4) was constructed from
a 0.5 m section of 7.4 cm diameter aluminum irrigation pipe. The bottom
end of the core tube was sharpened and serrated to facilitate coring. A
handle was attached to the opposite end with adjustable thumb screws.
Notches were filed into one side of the coring tube at one decimeter
intervals to facilitate control over the depth of insertion into the
marsh sediments. A three inch diameter (7.6 cm) plumber's test plug was
used to cap the corer during extraction from the marsh. A vent pipe was
constructed from a 1.0 m section of half-inch electrical conduit. The
bottom end of this pipe was tightly plugged with a rubber stopper above
which a series of small holes (1/16 in. dia.) were drilled.
Coring was accomplished by inserting the core tube into the marsh
sediments while alternately twisting the handle clockwise, then counter-
clockwise, and exerting a downward force. The core tube was inserted
into the marsh in one decimeter increments. After each decimeter
increment of penetration, a meter stick was placed inside the top end of
the corer until it came to rest atop the sediments inside. From the
14
plumber's
test plug
wooden
vent core
pipe extractor
rubber
stopper
Figure 4. Coring apparatus for sampling marsh type sediments, modified
from Gallagher (1974).
15
length of the core tube (0.5 m), the depth of insertion, and by
observing the meter stick reading at the tube's rim, the amount of core
compression was computed for each successive increment of penetration.
The core was inserted to a final depth of four decimeters in every
sampling attempt.
Gallagher (1974) noted that the barrel of the core tube acts as a
resonator such that any change in sound indicates a change in soil
density. Cutting thorough underground stems and rhizomes could not only
be heard but felt through the hands as well. These observations helped
determine the proper amount of downward force applied to the top end of
the coring tube, thereby serving to lessen the effects of core
compaction.
After coring was completed but before the core tube was extracted
from the sediments, a pi umber's test plug was inserted into the open top
end and tightened to form an air tight seal. The vent pipe was then
introduced parallel to, and along the entire length of, the core tube.
The small vent holes at the sediment end of the vent pipe allowed air to
enter the void created beneath the core tube during extraction, thereby
equalizing the pressure. In this manner, the core tube was more easily
removed with minimal loss of core material.
Upon recovery of the core, a wooden piston type core extractor was
used to remove the intact core from the corer. Cores were pushed out
into a trough constructed from a 1.0 m long split section of 10.8 cm
(outside dia.) polyvinylchloride pipe. Final core length and core hole
depth were recorded in order to document core compaction. All cores
were trimmed to a final length of 30.0 cm and sectioned into three 10.0
cm increments. Good et al. (1982) has reviewed the literature
16
concerning measurements of marsh belowground biomass and concluded that
the bulk of all living vascular plant material resides 30.0 cm or less
in depth within the sediments, and in a study of a New Jersey ^
alternif1 ora tidal marsh. Smith et al. (1979) found half as much
belowground material at the 20-30 cm increment of depth as compared to
the 0-10 cm increment. Based on these findings, living vascular plant
biomass below 30.0 cm of depth was considered to be insignificant and
was not sampled in this study.
Breakage of a core occasionally occurred within the 20-30 cm
increment. Recovered cores less than 30.0 cm, but longer than 20.0 cm,
in length were retained and trimmed to a final length of 20.0 cm
resulting in only two 10.0 cm increments after sectioning. It was
concluded that these cores were satisfactory for analysis because
breakage usually occurred at a well defined interface between
contiguous, intertwined, living roots and rhizomes and decomposing plant
tissues having less structural integrity. Thus, nearly all of the living
material present was recovered within these "short cores". Cores less
than 20.0 cm in length before sectioning were discarded and replaced by
other cores of suitable length.
After sectioning, each core increment was placed into an
individually labeled plastic bag (Dow Ziplock sandwich bags, 16.5 cm x
14.9 cm), sealed, and returned to the laboratory. Cores were kept at 4
C until further processing.
No attempt was made to separate living from dead tissues.
Separation techniques based on differential staining (Knievel, 1973) and
differences in color and turgidity (Valiela et al., 1976) were found
inappropriate for some plant species (de la Cruz and Hackney, 1977).
17
Because such methods are also tedious, dependent upon judgements by
technicians, and time consuming (especially when large numbers of
samples are involved), many investigators have reported BGB as a
combination of living and dead tissues (Gallagher, 1974; de la Cruz and
Hackney, 1977; Good and Frasco, 1979; and Smith et al., 1979).
Processing entailed washing each core increment free of very fine
mineral and organic matter. Core increments were hand washed with tap
water over a Wildco brass screen sieve having a 520 urn pore size.
Washing was considered complete when water passing through the material
atop the sieve ran clear. All material retained by the sieve has been
defined as macro-organic matter or MOM (Gallagher and Plumley, 1979) and
represents the BGB within the salt marsh soils sampled. This material
consisted of both living and dead roots, stem bases, and rhizomes.
MOM from each core increment was oven dried for not less than 48
hours at 85 C, weighed to the nearest milligram (Mettler P120 N single
pan balance), and recorded as mg.cm“3 (a 10.0 cm core increment obtained
from a core tube having an inside diameter of 7.4 cm contains 430 cm3).
After MOM from each core increment had been dried and weighed the
material was stored in individually labeled plastic bags should future
access be required. Incremental biomass values in mg.cm"3 were summed
for each individual core and were extrapolated to a per meter square
basis to a depth of 30.0 cm using a conversion factor of 232.56, which
is the number of cores potentially obtainable from 1.0 m2 using a corer
with an inside diameter of 7.4 cm. A similar conversion was used by de
la Cruz and Hackney (1977).
Core samples were collected monthly from six different vegetative
cover types from January through December, 1982. Five sample sites were
18
located on Jacks Creek (Figure 2) and included the vegetation types
herbaceous annuals, jamaÍcense, S. cynosuroides, T. domingensis, and
J. roemerianus. An additional sample site containing D. spicata located
on nearby Jacobs Creek (Figure 3) comprised the sixth vegetative
community selected for study.
Kucera et al. (1967) reported that seasonal development of standing
crop [above or belowground], cessation of growth [senescence], and
transfer to the litter stage are complex occurrences and that within a
multispecies community the phenological development, and thus the time
of maximum biomass, may not coincide for all components. For this
reason large, uniform, nearly monospecific stands of vegetation
representative of the dominant marsh vegetative types were selected for
study.
Cores for the assessment of BGB were recovered from plots
simultaneously sampled by another investigator (Beilis, personal
communication) assessing aboveground biomass. This procedure
facilitated estimates of total (above and belowground) biomass and
annual production based on concurrent sampling. In practice, all cores
for BGB determinations were drawn from within the last of three
contiguous 0.5m2 aboveground clipped plots within each vegetative
community. Gallagher and Plumley (1979) selectively sampled marsh plant
communities by taking cores either over stems or between stems resulting
in two measures of BGB for the same stand of vegetation. Because a
measure of mean BGB was desired for each community type independent of
their respective, relative stem densities, no such selectivity was
applied in this study.
For the first six months of 1982, five cores were removed each
19
jnonth from each vegetative community. In July, 1982, the number of
cores recovered monthly was increased to ten in order to reduce the
effects of statistical variation between consecutive cores. This
sampling regime continued for the remainder of 1982. Altogether, a
total of 90 cores were removed from each plant community giving an
annual total of 540 cores.
Aboveground samples were individually placed in plastic trash bags
p
(Hefty two-ply) at harvest, returned to the laboratory, and stored at
4 C until processed. Live and dead vegetation within each bag was
separated by species and weighted to the nearest 10.0 gm using a Pesóla
1 kg hanging scale. After recording total wet weight of a sample, each
portion was subsampled (approximately 100 g). Subsamples were placed in
#8 brown paper bags, weighed to the nearest gram (Mettler P 120 N single
pan balance), oven dried to constant weight at 85 C, and reweighed to
the nearest gram. Total dry weight of the original sample was computed
and reported as g-m-2.
Statistical analyses were performed using packaged program
procedures accessible through the Statistical Package for the Social
Sciences (SPSS) including summary statistics, and one-way analyses of
variance (Nie et al., 1975). Periodic regression analysis (Hackney and
Hackney, 1977 and Hackney and Hackney, 1978) was performed on BOB data
using release 79.6 of SAS (SAS is a registered trademark of the SAS
Institute, Cary, North Carolina) computer programs. The one-term
periodic regression model used was:
=
ag + sin (CXi ) + cos (CXl )
where
y-\ = dependent variable (BGB)
20
= constant parameter
cti,Bi= coefficients of the harmonic function of X.
C = Zir/n
= ith independent variable (numbered day of the year)
RESULTS AND DISCUSSION
Belowground Standing Crop
The surface texture of salt marsh sediments appeared to be closely
correlated with plant stem density and the configuration of the roots
and rhizomes. Mean macro-organic matter standing crop varied from a low
of 5.53 kg-m-2 for the herbaceous annuals community to approximately
twice that value (12.63 kg.m-2) for the D. spicata community (Table 3).
Highest macro-organic matter standing crop values were obtained for the
Distichlis and Juncus communities which grew from sha 11 ow rhizomes
bearing short internodes. The closely spaced, upright stems and
associated adventitious roots produced a fibrous, firm surface.
Conversely, Cladiurn and Typha communities were often associated with a
discontinuous!y textured surface characterized by scattered clumps of
stems and associated rhizomes. The Spartina community was intermediate
in both standing crop (8.34 kg.m“2) and texture between that of
Cl adi um/Typha and Juncus/Di sti ch 1 i s. Herbaceous annual s produced the
lowest belowground standing crop values as would be expected from a
plant community which is non-rhi zomatous and therefore stores no food
reserves.
Few published estimates are available for comparisons of
belowground standing crop data (Table 3). Of the values that are
21
Table 3. Estimates of belowground standing crop for six salt marsh
plant communities. Estimates derived from this study are
based upon yearly means in kg.m-2 to a depth of 30.0
centimeters.
Belowground
Plant community Standing crop1' Location Source
Annual Mean
(kg.m-2)
Herbaceous annuals 5.53 North Carolina This study
Cladium jamaicense 7.44 North Carolina This study
Spartina cynosuroides 8.34 North Carolina This study
10.901 Mississippi de la Cruz, 1974
Typha domingensis 2.651 Texas McNaughton, 1966
9.19 North Carolina This study
Juncus roemerianus 7.80l Mi ssissippi de la Cruz, 1974
8.532 Georgia Gallagher, 1974
11.38 North Carolina This study
Distichlis spicata 2.612 Georgia Gallagher, 1974
12.63 North Carolina This study
Ho a depth of 20.0 cm
^published values that were adjusted to reflect a depth profile of
30.0 cm
22
available, some represent only a single sampling effort (McNaughton,
1966 and Gallagher, 1974) and/or sampling-was completed to a depth of
only 20.0 cm (McNaughton, 1966 and de la Cruz, 1974). These
inconsistencies make strict comparisons impossible. Currently there is
simply insufficient information to support broad conclusions concerning
possible floristic or geographic general patterns of belowground biomass
distribution.
Although belowground standing crop biomass estimates obtained from
this study (5.53-12.63 kg*m“2) were generally of the same order of
magnitude as those reported elsewhere (2.61-10.90 kg-m“2), estimates for
the same plant species at different locations were quite divergent. For
instance, McNaughton (1966) reported a standing crop biomass for
domingensis of 2.65 kg.m-2, an estimate nearly four times greater (9.19
kg.m-2) was obtained from this study.
Net Annual Production
Computations
Several methods were used to estimate net annual production (NAP)
for each vegetation type. All NAP estimates (Table 4) were derived from
variations of the maximum/minimum technique in which NAP is defined as
the difference between the maximum and the minimum macro-organic matter
measurements obtained during one complete growth cycle.
The maximum/minimum technique using monthly means has been the
most frequently used method for estimating NAP. The main disadvantage
of this type of data treatment is that it relies on only two extreme
values taken from an entire collection cycle. A NAP value derived from
23
Table 4. Comparison of belowground productivity among six salt marsh plant communities using four methods
of estimating net annual production.
Belowground production
.(kg-m*•2.yr-l to a depth of 30. 0 cm)
Marsh community type MOM^ standing crop Max. - min. of Running
annual mean monthly means PRA^ means Duncan's range
(kg*m“^)
Herbaceous annuals 5.53 2.72 1.34 1.01 0.68
Cladium jamaicense 7.44 6.33 2.33 2.38 5.12
Spartina cynosuroides 8.34 4.88 2.16 1.86 0.75
Typha domingensis 9.19 3.39 3.55 2.81 1.26
Juncus roemerianus 11.38 3.28 1.52 2.46 2.41
Distichlis spicata 12.63 3.89 3.69 2.75 0.51
^macro-organic matter
Aperiodic regression analysis
this method could not be expected to adequately account for variation
inherent in sampling.- Periodic regression analysis (PRA) has the
advantage of using all values in a data set, not just extreme or
averaged values. A periodic curve fitted to all data points will
produce maximum and minimum predicted estimates derived from all data
points. One disadvantage of the PRA technique is that more harmonics
may be required to gain a statistically significant fit than can be
explained biologically in terms of seasonal development. If this
occurs, other methods of data analysis may be more appropriate (Hackney
and Hackney, 1978).
The use of running means effectively reduced the influence of
extreme values, thereby, smoothing out cyclical patterns. In this
method of analysis individual monthly means were averaged for successive
three month periods. In each case the resultant mean was reported for
the central month. NAP was then estimated from the difference between
annual maximum and minimum running mean values. This type of analysis
generally yielded lower NAP values (Table 4) than periodic regression
analysis or the maximum/minimum of monthly means. The most conservative
estimates of NAP were obtained from Duncan's multiple range test. This
method often grouped chronologically non-sequential months into two or
more statistically related subgroups, thus interrupting the continuum of
the natural growth cycle and making biological interpretation difficult.
Duncan's range test was also employed to estimate NAP. This test
arranged monthly means into two or more subgroups which were
significantly different from one another statistically ( = 0.05). NAP
was then estimated as the difference between the means of the highest
and lowest subgroups. NAP estimates generated using Duncan's range test
25
were among the lowest obtained (for four of the six vegetation types)
and were judged to represent gross underestimates. While it is true
that any method which fails to account for decomposition will
underestimate NAP, this method seemed especially conservative. It was
concluded that predicted NAP values generated by PRA provided the most
accurate assessment of BGB. At the very least this method allowed for a
more realistic assessment of sample variation even where statistical
significance was lowered in terms of actual data to curve fit.
Given the high degree of variation among monthly macro-organic
means (Figures 5-10), any estimate of NAP based on the simple
maximum/minimum technique using the monthly means was considered to be
statistically inappropriate. Less value was placed on NAP values
derived from the running means and Duncan's multiple range test because
these values seemed too conservative for use. However, their use is not
precluded where variability is low and where cyclical patterns are well
defined.
Table 5 presents the periodic models for each vegetation type.
All prediction curves were carried through two harmonica as these were
the only models which could be readily interpreted biologically in terms
of a seasonal cycle of development.
Comparison With Other Studies
NAP values obtained in this study were compared to published values
available in the literature (Table 6). The NAP values listed here for
the present study were not generated using the preferred PRA technique,
but rather from the maximum/minimum of monthly means - values which
26
PREDICTED MOM VALUES
— MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC ABOVEGROUND
MATTER
(dry BIOMASS
weight (dry
kg* weight
kg*
MONTH
Figure 5. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organic matter biomass for the herbaceous annuals community. Vertical bars represent
±1 standard error.
—' PREDICTED MOM VALUES
— MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC ABOVEGROUND
MATTER
(dry BIOMASS
weight (dry
kg- weight
kg-
1982 3.
MONTH
Figure 6. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organic matter biomass for the Cladium jamaicense community. Vertical bars represent
±1 standard error.
— PREDICTED MOM VALUES
MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC ABOVEGROUND
MATTER
(dry BIOMASS
weight (dry
kg* weight
kg*
1982
MONTH
Figure 7. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organiz matter biomass for the Spartina cynosuroides community. Vertical bars represent
±1 standard error.
—' PREDICTED MOM VALUES
MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC 11.010.0 ABOVEGROUND
MATTER 9.0
(dry BIOMASS8 0
weight 1.0 (dry7.0 0.5
kg* weight0
kg*
Figure 8. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organic matter biomass the the Typha domingensis community. Vertical bars represent
±1 standard error.
PREDICTED MOM VALUES
— MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC ABOVEGROUND
MATTER
(dry BIOMASS
weight (dry
kg- weight
kg-
1982 3
MONTH ^
Figure 9. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organic matter biomass for the Juncus roemerianus community. Vertical bars represent
±1 standard error.
PREDICTED MOM VALUES
— MONTHLY MEAN MOM VALUES
LIVING ABOVEGROUND BIOMASS
LIVE
MACRO-ORGANIC ABOVEGROUND
MATTER
(dry BIOMASS
weight (dry
kg* weight
kg*
1982 3
MONTH ^
Figure 10. Living aboveground biomass, actual macro-organic matter biomass, and predicted macro-
organic matter biomass for the Distichlis spicata community. Vertical bars represent
±1 standard error.
Table 5. Periodic regression models for macro-organic matter in six salt marsh plant communities.
Marsh community type Probabi1ity Periodic model Annual mean
of > F (y=g*m~^)
Herbaceous annuals 0.0437 0.11 y=5494.3-295.9sin(CX. )+93. lcos(CX^-)- 5,532
49.0sin(2CX^- )-467.4cos(2CX^- )
Cladium jamaicense 0.0671 0.10 y=7367.4-359.6sin(CX-)+612.1cos(CX-)+ 7,436
694.3si n ( 2CX^- ) -45.5cos ( 2CX^- )
Spartina cynosuroides 0.1708 0.07 y=8379.0-57.7sin(CX-)+222.2cos(CX,-)+ 8,335
743.3sin(2CX^-)-57^.5cos{2CX^-)
Typha domingensis 0.0009 0.20 y=9089.1-443.5sin(CX^ )+1012.7cos(CX^-)- 9,194
474.4sin(2CX^-)-797.8cos(2CX^-)
Juncus roemerianus 0.1895 0.07 y=11266.7-522.6sin(CX,-)+121.4cos(CXi-)- 11,381
129.0sin(2CX^-)+344.0cos(2CX^-)
Distichlis spicata 0.0001 0.29 y=12466.2-939.7sin(CX,- )-117.5cos(CX^ )- 12,629
751.2sin ( 2CX^- )+991.8cos ( 2CX^- )
C=2V365
Xj=the number of the specific day of the year (1-365) for which a predicted value is desired
Table 6. Belowground productivity estimates for salt marsh plant communities along the western Atlantic
and Gulf Coasts, modified from Good et al. (1982).
Net annual
Marsh community type product!vity Location Source
(kg‘m“^)
Herbaceous annuals 2.72 North Carolina This study
Cladium jamaicense 6.33 North Carolina This study
Spartina cynosuroides 2.20 Mississippi de la Cruz and Hackney, 1977
3.56 Georgia Gallagher and Plumley, 1979
4.88 North Carolina This study
Typha domingensis 3.39 North Carolina This study
Juncus roemerianus 1.36 Mississippi de la Cruz and Hackney, 1977
3.28 North Carolina This study
3.36 Georgia Gallagher and Plumley, 1979
4.4-7.6 Alabama Stout, 1978
Distichlis spicata 1.07 Georgia Gallagher and Plumley, 1979
2.78 New Jersey Good and Frasco, 1979
3.40 Del aware Gallagher and Plumley, 1979
3.89 North Carolina This study
tended to be zero to three times larger than PRA estimates. These
values, although considered less reliable, are presented because all
other published figures are based on the maximum/minimum technique and
comparisons should be made free from the bias imposed by a different
technique. Estimates of NAP for the S. cynosuroides (4.88 kg-m-2) and
D. spicata (3.89 kg.m-2) communities were greater than values reported
in any previous study of these two species (Table 6). NAP for the
Juncus community (3.28 kg*m“2) was most comparable to an estimate from a
Juncus community in Georgia. Both of these values were intermediate
compared to extreme NAP estimates from Mississippi and Alabama. No data
are available in the literature for comparison of NAP values obtained
for T^ domingens is and jama Ícense (3.39 and 6.33 kg.m-2,
respectively).
As with the belowground standing crop, limited data precludes
comparisons, generalizations, and predictions concerning NAP estimates
versus geographic location and associated climatic conditions. Good et
al. (1982) has reviewed the literature on the NAP of these and other
salt marsh species and also concluded that the avaliable data is too
fragmentary and variable to enable the delineation of trends. However,
they did conclude that belowground production does not seem to follow
the trend of gradual increase from north to south apparent in
aboveground biomass production. Data from this study when compared with
other studies (Table 6) also supports this conclusion.
Belowground Biomass Pathways
Several biological processes could cause belowground biomass to
35
HËB8IV0RY
WINTERING WATERFOWL
MUSKRATS
OTHER CONSUMERS INCLUDING INVERTEBRATES
DECOMPOSITION
— AEROBIC PATHWAYS^
ANAEROBIC PATHWAYS
SEDIMENTS ESTUARI NE
FOOD CHAINS/WEBS
SHORT-TERM SINK
^ » REMOVAL OP MATERIALS
BY LEACHING, STORM ACTION
DEEP SEDIMENTS
LONG TERM SINK
Figure 11. Pathways open to belowground biomass, modified from
Good et al. (1982).
36
increase or decrease (Figure 11). Excess photosynthate might be stored
belowground as reserves. Senescence of aboveground plant parts might
also contribute to storage. Both processes serve as input modes to
belowground biomass. The leaching of roots and rhizomes, decomposition
of belowground biomass, new growth of plant parts aboveground,
environmental stress on above or belowground vegetation, inflorescence,
and seed production could constitute modes of output, all decreasing
belowground biomass. Additional reported output modes such as herbivory
by geese (Smith and Odum, 1981), muskrats (Lynch et al., 1974) or the
burrowing activity of invertebrates like the fiddler crab (Katz, 1980)
were not evident within the marshes studied.
General Seasonal Pattern
A single pattern relating actual and predicted BGB values to
living aboveground biomass (AGB) was apparent for most of the vegetation
types. S. cynosuroides (Figure 9) was a good example of this general
pattern. Exceptions to this pattern will be discussed separately.
In late spring living AGB increased as BGB declined. Translocation
of nutrients stored in belowground rhizomes upward to support newly
initiated shoots could account for the observed spring decrease in BGB.
As ABG approached a summer maximum, BGB began to increase, presumably
because excess photosynthate was being stored belowground. Maximum BGB
typically occurred during summer or late summer. A decline in BGB then
ensued coincident with inflorescence production. Translocation of
nutrients to reproductive structures from belowground could explain this
decrease in BGB. With reproduction completed, AGB rapidly declined with
37
the coming of winter. As aboveground components senesced, BGB began to
increase. With the onset of winter, BGB declined and continued to do so
as storage products were lost to leaching, decomposition, or both.
Predicted BGB values based on the PRA model sometimes failed to
correspond closely to actual data points based on monthly mean BGB
values. In several instances slight increases in BGB were apparent from
the PRA model during periods of the year when little or no living
aboveground vegetation was present, as in S. cynosuroides (January -
March). It was concluded that these apparent increases in BGB should be
considered as distortions resulting from extreme variation among actual
monthly mean values.
Herbaceous Annuals Community. The herbaceous annuals community,
located adjacent to the swamp forest at the headwaters of Jacks Creek
(Figure 5), exhibited comparatively high stem density but lower BGB than
any of the other vegetation types. Lower BGB values were expected
because this was a non-rhizomatous plant community. Initial sampling
showed an unusually high BGB content in January (Figure 7) followed by a
sharp decline. While decomposition of the previous seasons root mass
could account for this decline, the increase in BGB from February to
April cannot be accounted for because of the complete absence of living
AGB. In fact, sporatic increases and decreases in BGB were seen
following the appearance of living AGB in April. This inconsistent
pattern was attributed primarily to the more than occasional presence of
varying amounts of subfossil wood in the lower core segments. If the
steady increase in BGB from July to September is real it appears to
correspond to the growth peak of living ABG. Living ABG began to
decrease rapidly in September until none remained in November.
38
Coincident with this decrease was the steady decline in BGB beginning in
September, presumably due to decomposition.
PRA of the data is represented in Figure 7. Given a probability of
a greater F (P>F) of 0.0437 the model (Table 5) should be a good
predictor of BGB in this vegetative community, subfossil wood not
withstanding.
Cl ad i urn jamaicense Community. This plant community typically
consisted of unevenly spaced clumps or culms and exhibited relatively
low stem density. This vegetation type is one of two studied here in
which portions of the leaves remain green throughout the winter months.
Belowground stem bases and rhizomes were typically large and tough.
Rhizomes were found to be evenly spaced between culms and were often
continuous with adjacent culms.
This plant community followed the general seasonal pattern
previously described with respect to BGB and living ABG (Figure 6), the
only notable exception being January's comparatively high mean value.
Variation was great but not great enough to account for this initially
high BGB value. Reinspection of the stored MOM obtained from each of
the five cores revealed that a greater number of cores were taken over
stem bases rather than between stems. Two of these stem bases were
especially large in size. It was felt that the weight contribution of
these plant parts raised total BGB to a higher value than would have
been obtained otherwise.
PRA of the data offered a reasonably good fit (P>F=0.0671). It
would seem that the model (Table 5) might be sensitive to the effects of
January's initially high BGB value and that a better fit would be
obtained had sampling been more homogeneous.
39
Spartina cynosuroides Community. The relatively large sample
variance that occurred in this vegetative community (Figure 7), was
attributed to its habit of belowground growth. Although plants were
widely but uniformly spaced aboveground, belowground rhizomes were not.
Rhizomes, typically deeply buried, were horizontally elongate and ran
from culm to culm as previously noted by Eleuterius (1981). Successful
recovery of viable rhizomatous material was mainly dependent upon
sampling in close proximity to a stem base or directly over one.
Overall this vegetative community followed the general seasonal
pattern with respect to BGB and living AGB. The major deviations in
actual BGB values from predicted values (Figure 7) occurred where
variation was greatest as in the months of March and August. Due to
this variation, PRA yielded a P>F of 0.1708 (Table 5) and, therefore,
the data does not fit the model well. Additional harmonics did produce
a better fit but the model did not then fit biologically logical
seasonal patterns in BGB in relation to living ABG.
Typha domingensis Community. The greatest amount of variation
occurred in this vegetative community (Figure 8) for the same reasons
discussed for S. cynosuroides in regard to its habit of belowground
growth. Stem bases were comparatively tougher, larger, more fiborous,
and heavier. Disproportionate sampling over or between stem base could
cause a great deal of variation among replicate cores. Aboveground,
culms were unevenly spaced while stem density was low.
It is interesting to note that even with extreme variation, PRA
yielded an exceptionally good fit to a two harmonic model with a P>F of
0.009 (Table 5). The first cosine term of the model explained most of
the data in terms of fit. It would appear that a large amount of
40
variation among replicate samples worked for the model where the fit may
not have been as good at other points throughout the year. One sample
value could shift the phase angle of the curvilinear plot sufficiently
to gain a good fit. In this particular instance, the model should be
used with reservation for prediction purposes.
BGB followed the same general patterns in relation to living AGB as
as previously described with no notable exceptions.
Juncus roemerianus Community. The J. roemerianus community
exhibited a high stem density aboveground and turgid, tightly interwoven
roots and rhizomes belowground. This plant community did deviate from
the general pattern at several points. Typically, this community type
experienced two peaks in living ABG during the summer months, and a
large portion of this component remained alive thoughout the winter
months (Figure 9). The prediction curve for BGB did not always
correspond closely to the actual data (P>F = 0.1895). Divergence from
the prediction model was greatest when variation among replicate core
samples was greatest, as in the months of April through June. Given the
compactness of the cores from this community coupled with an inherently
high stem density, one might not have expected such great variability
among replicate cores. One possible explanation for the observed
variation during spring could be that relatively small variations in
aboveground stem density may result in large changes in BGB. This
cannot be confirmed as we did not record stem density at the time of
sampling.
Pi Stichi is spicata Community. D. spicata was characterized by
dense growth and high stem density aboveground and small tightly
intertwined roots and rhizomes belowground. Such growth characteristics.
41
indicative of this vegetative community, resulted in the largest annual
mean standing crop (12.63 kg.m-2.yr-l) of all the plant communities
sampled here.
Unlike the other vegetation types, the prediction curve for this
plant community corresponded closely to actual monthly means throughout
the entire year (Figure 10). The prediction model exhibited a P>F of
0.0001.
D. spicata also differed from the other plant communities in that
as living ABG increased BOB did also. It wasn't until July that BOB
started to decline. By that time the aboveground vegetation had become
well established. The dec 1ine in BOB continued unti1 September when
living AGB reached a maximum. The sharp increase in living ABG between
August and September could have been caused by an upward translocation
of stored reserves to support inflorescence and seed production
consequently decreasing BGB. Belowground reserves depleted and
reproduction complete, living ABG declined rapidly from September to
November. It was hypothesized that as aboveground plant parts senesced,
nutrients were translocated into the rhizomes, thereby restocking
depleted reserves. Decomposition could account for the slight decline
in BGB between November and December.
Distribution of Macro-Organic Matter with Depth
Macro-organic matter (MOM) was not evenly distributed throughout
the soil profile within the various vegetative community types (Table
7). MOM decreased with depth within the first twenty centimeters in the
herbaceous annuals community, but exhibited an increase below the root
42
Table 7. Distribution of macro-organic matter (MOM)l with depth in six
salt marsh plant communities.
Annual mean MOM values with depth (kg.m-2)
Marsh community type 0-10 cm 10-20 cm 20-30 cm
Herbaceous annuals 2.51(0.07) 1.26(0.06) 1.76(0.12)
Cladium jamaicense 2.52(0.11) 2.36(0.09) 2.62(0.08)
Spartina cynosuroides 3.58(0.13) 2.87(0.11) 1.67(0.06)
Typha domingensis 3.09(0.12) 3.07(0.13) 3.16(0.10)
Juncus roemerianus 3.82(0.10) 3.83(0.09) 3.72(0.09)
Distichlis spicata 5.54(0.14) 4.30(0.08) 2.71(0.06)
^mean (standard error)
43
zone in the 20.0 to 30.0 cm increment. This discontinuous pattern of
MOM distribution appeared to have resulted from the fact that these
samples were collected near the headwater swamp forest at Jacks Creek.
Marsh vegetation has only recently encroached this area and little peat
had formed within the silty swamp forest soils. The sudden apparent
increase in MOM within the 20.0 to 30.0 cm increment was caused by an
abundance of subfossil wood buried at this level which was not a factor
at other sample sites.
For J. roemerianus, T. domingensis, and C. jamaicense, MOM values
appeared to remain relatively uniform with depth through the 30.0 cm
profile. This would be a Type I profile as described by Gallagher and
Plumley (1979). Conversely, S. cynosuroides and D. spicata MOM values
declined continuously with increasing depth to 30.0 cm, Gallagher and
Plumley's (1979) Type 2 profile. Differences in MOM distribution
belowground could reflect varying rates of decomposition at different
depths in the peat as was demonstrated by Hackney and de la Cruz (1980),
to differing habits of growth belowground, or to changing patterns of
vegetative cover. A combination of these three factors would be the
most likely explanation.
Except in the herbaceous annuals community most of the seasonal
variation in MOM occurred within the 0-20.0 cm increment, and it was
within this zone of the soil profile that most of the living roots and
rhizomes were found. Living roots and rhizomes held cores intact.
Because few rhizomes penetrate further than 20.0 cm, breakage of the
core sometimes occurred within the 20.0 to 30.0 cm increment as it was
extracted from the marsh.
44
Turnover Time
Belowground biomass turnover times were derived by dividing the
maximum monthly MOM value by the annual biomass increment of growth (NAP
based on the maximum/minimum of the monthly means) giving a turnover
time in years. While turnover rates may give insight concerning
relative rates of decomposition, they do not offer details of the
process itself (Good et al., 1982). Turnover times for each vegetation
type studied are given in Table 8. NAP estimates based on the
maximum/minimum of the monthly mean MOM values rather than those based
on periodic regression analysis were used to calculate turnover times,
because the former method was used by all previous investigators and,
thus, should yield turnover times more comparable to those already
published.
Generally, plant communities in this study which exhibited
relatively low stem density also exhibited a lower MOM turnover time.
Such communities included C. jamaicense (1.85 years), S. cynosuroides
(2.34 years), and T. domingensis (3.23 years). A1 tenati vel y, the _D^
spi cata (3.59 years) and roemeri anus (3.74 years) communities
exhibited comparatively high stem density and revealed longer turnover
times. The herbaceous annuals (2.60 years) exhibited the third longest
turnover time, however, this elevated value was considered an artifact
caused by an abundance of subfossil wood in the 20-30 cm increment.
Some data on turnover time were available in the literature for
purposes of comparison. Hackney and de la Cruz (1980) obtained turnover
times of 2.33 years for cynosuroides and 3.67 yrs. for ^ roemerianus
from salt marshes in Georgia. These values are in close agreement with
45
Table 8. Belowground biomass turnover times (years) for six salt marsh
plant communities.
Marsh Max. monthly Annual increment Turnover time
community type MOM value of growth (years)
(kg*m-2 ) (kg.m-2.yr-l)
Herbaceous annuals 7.07 2.72 2.60
Cladium jamaicense 11.68 6.33 1.85
Spartina cynosuroides 11.42 4.88 2.34
Typha domingensis 10.95 3.39 3.23
Juncus roemerianus 12.28 3.28 3.74
Distichlis spicata 13.98 3.89 3.59
46
those derived from this study (Table 8). For D. spicata, studies in
Georgia and Delaware (Gallagher and Plumley, 1979) revealed a turnover
time of 3.33 years at each location. Again, these values are quite
comparable to the turnover time derived from this study for spicata
(3.59 years). One might expect a longer turnover time from the Delaware
study where generally cooler ambient temperatures might slow the rate of
decomposition. In a study of New Jersey D. spicata marsh, Good and
Fransco (1979) did, in fact, obtain a longer turnover time of 4.50
years. Currently, such differences cannot be readily explained, but it
would seem that physical and chemical properties of individual salt
marshes might significantly influence decomposition rate, not just
temperature.
Belowground/Aboveground Biomass Ratio
Concurrent sampling of belowground biomass (BGB) and aboveground
biomass (AGB) has al lowed the computation of BGB/AGB ratios for each
plant community studied (Table 9). While BGB/AGB ratios may not be
strictly comparable to root/shoot ratios available in the literature
(Good et al., 1982) they do confirm the findings of previous
investigators that greater amounts of fixed carbon reside within the
soils of salt marshes than in the aboveground vegetative component. For
the salt marsh plants studied here, observed ratios ranged from 4.9 for
C. jamaicense to 10.3 for T^ domingensis. The herbaceous annuals
exhibited a BGB/AGB ratio of 13.8:1, a high value probably due to the
presence of subfossil wood within the 20.0 to 30.0 cm increment.
Studies of D. spicata in Georgia (Gallagher, 1974) and New Jersey
47
Table 9. Belowground/aboveground biomass ratio for six salt marsh
plant communities.
Marsh Belowground biomass Aboveground biomass B:A
community type annual mean annual mean ratio
(kg-m-2) (kg.m-2)
Herbaceous annuals 5.53 0.04 13.8
Cladium jamaicense 7.44 1.51 4.9
Spartina cynosuroides 8.34 1.62 5.1
Typha domingensis 9.19 0.89 10.3
Juncus roemerianus 11.38 1.60 7.1
Distichlis spicata 12.63 1.97 6.4
48
(Good and Frasco, 1979) revealed root/shoot ratios of 7.2 and 4.5,
respectively. An intermediate BGB/AGB ratio of 6.4 was obtained from
this study of Pistichlis. The BGB/AGB ratio of 7.1 obtained from the ^
roemerianus community was most comparable to a root/shoot ratio of 8.2
obtained from a Georgia Juncus marsh (Gallagher, 1974). Much lower
ratios were obtained from Juncus marshes in Mississippi (de la Cruz and
Hackney, 1977) and Alabama (Stout, 1978), 0.80 and 3.26, respectively.
Investigations by Good et al. (1982) have shown that large
root/shoot ratios within salt marsh plant communities result from a
disproportionately larger amount of belowground biomass as compared to
aboveground biomass. Data from this study supports this observation.
Why is the belowground component so disproportionately large? It
has been suggested that a portion of this component might be necessary
for the structural integrity and maintenance of the marsh (Good et al.,
1982). Roots and rhizomes containing stored food reserves needed for
the initiation of new growth, maintenance of aboveground tissues or
during periods of extreme environmental stress form another significant
fraction of the belowground component. Yet, superficially, it would
appear that much more photosynthate is stored than is necessary for
winter survival and subsequent rejuvenation. Unfortunately, there are
not yet enough data on the belowground component of the salt marsh to
adequately assess its complete function and interactions with the
estuarine environment.
49
SUMMARY AND RECOMMEDATIONS
Estimates of belowground standing crop and net annual production from
two low salinity marshes of Jacks and Jacobs Creeks fell within the range
of values reported for similar marshes along the Atlantic and Gulf Coasts.
Net annual production estimates derived from periodic regression analysis
were judged to be more realistic than those derived from the simple
maximum/minimum technique using monthly means because all data points were
used to generate the prediction models.
Macro-organic matter was not evenly distributed throughout the soil
profile within the various plant communities. The bulk of this material
resided within the 0-20.0 centimeter depth increment which contained most
of the living roots and rhizomes.
Relatively lower turnover times were obtained from plant
communities which exhibited relatively low stem density while higher
turnover times were obtained from plant communities having comparatively
high stem density. Turnover times derived from this study were
comparable to other published values.
While BGB/AGB ratios are not strictly comparable to root/shoot
ratios, they did confirm reports by previous investigators that greater
amounts of fixed carbon are produced within the soi 1s of salt marshes
than in the aboveground component.
Based on these and other studies it would appear that much more
photosynthate is stored belowground than seems necessary for initiation
of new growth the following season. Currently, the functional
50
significance for this apparent excess storage is unclear. Additional
studies needed to clarify this phenomenon include: "in situ"
decomposition of roots and rhizomes, measurement of belowground biomass
and its distribution with depth, the "in situ" growth response to
selected stress factors (salinity, hydroperiod, photoperiod, etc.), and
quantification of the contribution of living tissues to total
belowground biomass for many more salt marshes in different geographic
locations. In order to account for the variation inherent in sampling
BGB in narrowly confined 0.5 m2 plots, it is recommended that conçurent,
random sampling be performed to evaluate variation. Also, confirmation
of predicted MOM values would be indicated in order to fully evaluate
periodic regression analysis modeling.
Studies such as these would help to elucidate the role of the
belowground component in the salt marsh and any function the total
vegetative component may have in marsh-estuarine interactions.
51
LITERATURE CITED
Beilis, V. J. Personal Communication. Department of Biology, East
Carolina University, Greenville, North Carolina.
Beilis, V. J. and A. C. Gaither. Unpublished manuscript. Department
of Biology, East Carolina University, Greenville, North Carolina.
Copeland,B. J.,R. G.Hodson,S. R. Riggs and J. E. Easley, Jr. (1983).
The ecology of Albemarle Sound, North Carolina: an estuarine
profile. U.S. Fish and Wildlife Service, Division of Biological
Services, Washington, D.C. FWS/OBS-83/01. 68 pp.
Cruz, A. A. de la (1974). Primary productivity of coastal marshes in
Mississippi. Gulf Res. Rep. 4:351-356.
Cruz, A. A. de La and C. T. Hackney ( 1977). Energy value, elemental
composition, and productivity of belowground biomass of a Juncus
tidal marsh. Ecology 58:1165-1170.
Dahlman, R. C. and C. L. Kucera (1965). Root productivity and turnover
in native prairie. Ecology 46:84-89.
Eames, G. B. (1983). The late Quaternary seismic stratigraphy,
1 ithostratigraphy, and geologic history of a shelf-barrier-
estuarine system. Dare County, North Carolina. Masters Thesis.
East Carolina University, Greenville, North Carolina.
Eleuterius, L. E. (1981). Tidal Marsh Plants. Mississippi - Alabama
Sea Grant Consortium, Ocean Springs, Miss, publication number
MASGP-77-039. 131 pp.
Gallagher, J. L. (1974). Sampling macro-organic matter profiles in
salt marsh plant root zones. Soil Sci. Soc. Amer. Proc. 38:154-
155.
Gallagher, J. L. and F. Gerald Plumley ( 1979). Underground biomass
profiles and productivity in Atlantic coastal marshes. Amer. J.
Bot. 66(2) :156-161.
Gardner, L. R. (1975). Runoff from an intertidal marsh during tidal
exposure-recession curves and chemical characteristics. Limnol.
Oceanogr. 20:81-89.
Good, R. E. and B. R. Frasco (1979). Estuarine evaluation study; a four
year report on production and decomposition dynamics of salt marsh
communities; Manahawkin marshes. Ocean County, New Jersey. Report
to N.J.D.E.P. Div. Fish, Game and Shelfisheries. 105 pp.
52
Good, R. E., N. F. Good, and B. R. Frasco (1982). A review of primary
production and decomposition dynamics of the belowground marsh
component, pp. 139-157. In: Estuarine Comparisons, (Ed.) Victor
S. Kennedy. Academic Press, New York. 709 pp.
Hackney, C. T. and A. A. de la Cruz (1980). In situ decomposition
of roots and rhizomes of two tidal marsh plants. Ecology 61(2):
226-231.
Hackney, C. T. and 0. P. Hackney (1978). Animproved, conceptually
simple technique for estimating the producti vityof marsh vascular
flora. Gulf Research Reports 6(2):125-129.
Hackney,0. P.and C. T.Hackney (1977). Periodic regression analysis
of ecological data. Journal Miss. Academy Sci. XXI 1:25-33.
Katz, L. C. (1980). Effects of burrowing of the fiddler crab, Uca
pugnax (Smith). Estuar. Coastal Mar. Sci. 11:233-237.
Keefe, C. W. ( 1972). Marsh production: a summary of the literature.
Contrib. Mar. Sci. 16:163-181.
Knievel, D. P. (1973). Procedure for estimating ratio of live to dead
root dry matter in root core samples. Crop Sci. 13:124-126.
Kucera, C. L.,R. C. Dahlman, and M. R. Koelling (1967). Total net
producti vityand turnover on an energy basisfor tallgrass prairie.
Ecology 48(4) :536-541.
Lynch, J. J., T. O'Neil and D. W. Lay (1974). Management significance
of damage by geese and muskrats to Gulf Coast marshes. J. Wildl.
Man. 11:50-76.
McNaughton, S. J. (1966). Ecotype function in the Typha community type.
Ecol. Monogr. 36:297-325.
Nie, N. H., C. H. Hull, J. Jenkins, K. Stei nbrenner, and D. H. Bent
(1975). Statistical Package for the Social Sciences. McGraw-Hill,
New York. 675 pp.
Nixon,S. W. (1980). Between coastal marshes and coastal waters - a
review of twenty years of speculation and research on the role of
salt marshes in estuarine productivity and water chemistry, pp.
437-525. In: P. Hamilton and K.B. MacDonald (Eds.) Estuarine and
Wetland Processes, Plenum Press, New York.
Odum, E. P. (1968). A research challenge: evaluating the
productivity of coastal and estuarine water. In: Proceedings of
the Second Sea Grant Conference. Univ. of Rhode Island, October,
1968. pp. 63-64.
53
Smith,K. K'.,R. E.Good,and N. F.Good( 1979). Production dynamics
for above and belowground components of a New Jersey Spartina
al terni flora tidal marsh. Estuarine and Coastal Mar. Sci. 9:109-
wr.
Smith,!. J., Ill and W. E. Odum (1981). The effects of grazing by Snow
Geese on coastal salt marshes. Ecology 62(1):98-106.
Stout, J. P. (1978). An analysis of annual growth and productivity of
Juncus roemerianus Scheele and Spartina alterniflora Loisel in
coastal A1 abama. Doctoral di ssertati olïT The Uni ver si ty of
A1abama.
Stroud, L. M. (1976). Net primary production of belowground material and
carbohydrate patterns of two height forms of Spartina alterniflora
in two North Carolina marshes. Doctoral dissertation. . North
Carolina State University, Raleigh, North Carolina.
Teal, J. M. (1962). Energy flow in the salt marsh ecosystem of Georgia.
Ecology 43:614-624.
Turner, R. E. (1976). Geographic variation in salt marsh macrophyte
production: a review. Contrib. Mar. Sci. 20:47-68.
Valiela, I., J. M. Teal, and N. Y. Persson ( 1976). Production and
dynamics of experimentally enriched salt marsh vegetation:
Belowground biomass. Limnol. Oceanogr. 21(2):245-252.
Whigham, D. F., J. McCormick, R. E. Good, and R. L. Simpson (1978).
Biomass and primary production in freshwater tidal wetlands of the
Middle Atlantic Coast. In: R.E. Good, D.F. Whigham, and R.L.
Simpson (Eds.), Freshwater Wetlands: Ecological Processes and
Management Potential. Academic Press, New York, pp. 3-20.
54