'\ 
nitrogen fixation (C2H2 reduction)0 
AN ABSTRACT OF THE THESIS OF 
Roger D. Blew for the Master of Science Degree 
in Biology presented on May 11, 1984 
Title: Rhizosphere Nitrogen Fixation (C2H2 Reduction) Associated 
with the Major S 
Abstract Approved: 
Three aspects in 
the tall grass prairie were investigated in this study. First, a sur­
 vey was done to determine how many plant species have nitrogen fixing 
ability associated with their roots. Over 40 species from the POACEAE,
 -
 FABACEAE and ASTERACEAE were tested. More than 30 of those species
 -
 gave a positive response. Second, the effect of moisture stress on 
rhizosphere nitrogen fixation was studied. Xylem Pressure Potential 
(XPP) and rhizosphere nitrogen fixation were measured through the summer 
for Andropogon gerardi, ~. scoparius, Sorghastru~ nutans, and Panicum 
virgatum. As XPP decreased, the rate of nitrogen fixation also decreased. 
This may be a response to decreased photosynthesis or an increased de­
 activition of nitrogenase from exposure to oxygen. Third, the response 
of rhizosphere nitrogen fixation to spring burning was measured. Adja­
 cent burned and unburned sites were chosen for this study. The rate of 
nitrogen fixation in the burned prairie was consistently higher than in 
the unburned prairie for all four of the major tall grass species. 
RHIZOSPHERE NITROGEN FIXATION (C2H2 REDUCTION) ASSOCIATED
 
WITH THE MAJOR SPECIES OF THE TALL GRASS PRAIRIE
 
A Thesis
 
Submitted to
 
the Division of Biological Sciences
 
Emporia State University
 
In Partial Fulfillment
 
of the Requirements for the Degree
 
Master of Science
 
by 
Roger D. Blew 
May. 1984 
~J 1:J-y>n'>..6
 j 
iv 
ACKNOWLEDGEMENT
 
I am indebted to a number of people for their assistance and sup­
 port during this research. Among them are Dr. Wes Jackson and Laura 
Jackson of The Land Institute. Dr. L. C. Hulbert, Director of the Konza 
Prairie Research Natural Area, Dr. Dwight Spencer, Director of the Ross 
Natural History Reservation, and Dr. G. R. Marzolf for arranging 
financial assistance. Special thanks go to Dr. James Mayo, whose 
guidance, understanding and friendship made this study possible. 
v 
TABLE OF CONTENTS 
LIST OF TABLES. 
PAGE
 
vi
 
LIST OF FIGURES 
INTRODUCTION.
 
vii
 
MATERIALS AND METHODS 8
 
Study areas ... 8
 
Acetylene reduction technique. 9
 
Root collection. . . 10
 
Acetylene injection. 11
 
Incubation .... 12
 
Gas chromatography 13
 
Drying samples. 18
 
Weighing samples 18
 
Xylem pressure potential 19
 
RESULTS AND DISCUSSION. 23
 
Survey . 23
 
Response of nitrogen fixation to spring burning. 23
 
Ecotypes of Andropogon gerardi 28
 
Moisture stress .. 31
 
Effect of moisture stress on nitrogen fixation 39
 
Effect of 02 concentration on nitrogen fixation. 48
 
SUMMARY . . . . . 54
 
LITERATURE CITED. 57
 
vi 
LIST OF TABLES 
TABLE PAGE 
1. Type of material tested and number of samples 
positive (+) and negative (-) for acetylene 
reducti on by speci es of POACEAE. . . . . . . . 24 
2. Type of material tested and number of samples 
positive (+) and negative (-) for acetylene 
reduct ion by speci es of FABACEAE . . . . . . . 25 
3. Type of material tested and number of samples 
positive (+) and negative (-) for acetylene 
reduction by species of ASTERACEAE ..... 26 
4. Number of samples positive (+) and negative 
(-) for acetylene reduction by accessions of 
Tripsacum dactyloides and genetic crosses. 27 
5. Rhizosphere acetylene reduction by four 
prairie grasses in annually burned (B) and 
unburned (UB) prairie . 29 
6. Rhizosphere acetylene reduction and XPP of 
Andropogon gerardi ecotypes growing 
sympatrically in annually burned prairie. 34 
vii 
LIST OF FI GURES 
FI GURE PAGE 
1. Tallgrass prairie in spring showing differences 
between burned and unburned sites . 6 
2. Gas chromatograph, innertubes for acetylene 
and ethylene and flasks used for calibration 15 
3. XPP of leaves of Andropogon gerardi, A. 
scoparius, Panicum virgatum, and Sorghastrum 
nutans stored in plastic containers .. 22 
4. Potential ecotypes of Andropogon gerardi 33 
5. Dawn and mid-day XPP for Andropogon gerardi 
from all burned and unburned prairie sites 36 
6. Dawn and mid-day XPP for Andropogon gerardi 
from adjacent burned and unburned sites. 38 
7. Dawn and mid-day XPP for Andropogon scoparius 
from adjacent burned and unburned sites. 41 
8. Dawn and mid-day XPP for Sorghastrum nutans 
from adjacent burned and unburned sites. 43 
9. Dawn and mid-day XPP for Panicum virgatum 
from adjacent burned and unburned sites. 45 
10. Nitrogenase activity and XPP for Andropogon 
gerardi from burned prairie througout the summer 47 
11. Nitrogenase activity vs. XPP 
gerardi from burned prairie. 
for Andropogon 
50 
12. Time course of nitrogenase activity for 
ecotypes of Andropogon gerardi incubated in 
air and in four percent 02 52 
INTRODUCTION
 
The tall grass prairie has long been noted for its great produc­
 tivity of forage grasses. This indicates that large amounts of nitro­
 gen must be available to the plants (Risser et al. 1981). Even though 
the atmosphere is comprised largely of nitrogen gas, it is unavailable 
for use directly by plants. In order for plants to utilize nitrogen, 
it must be in a combined form as either NH; or NO;. Although a good 
nitrogen budget for the tall grass prairie does not exist, some of the 
inputs of available nitrogen are known. There are two major pathways 
for converting unavailable N2 gas into nitrogen compounds that can be 
used by plants. One of these pathways is the conversion of N2 to NO; 
by lighting. This NO; is transferred from the atmosphere to the soil 
by precipitation or dryfall. The second pathway is through biological 
nitrogen fixation. The process of biological nitrogen fixation can be 
divided into three distinct types; symbiotic nitrogen fixation, non­
 symbiotic nitrogen fixation, and nitrogen fixation by free living bac­
 teria asso~iated with the rhizosphere of higher plants. 
Symbiotic nitrogen fixation is that fixation occurring in bac­
 teroid nodules within the roots of certain species of plants. The 
occurrence of bacteroid nodules is common among species of the legume 
family (FABACEAE), and has also been found in several other families. 
The amounts of combined nitrogen entering the ecosystem through biolog­
 ical nitrogen fixation has not been fully quantified. With bacteroid 
nitrogen fixation, the difficulty in quantification is due to the 
scarcity of nodules (Risser et al. 1981) and the rooting depth of some 
of the prairie legumes. Lead plant (Amorpha canescens), for example, 
may root to a depth of five meters (Weaver, 1954). 
2 
Nonsymbiotic nitrogen fixation is associated with the Cyanobacteria 
(blue-green algae) and some free living bacteria. With the development 
of the 15N and acetylene reduction assays~ several free living micro­
 organisms once reported as having nitrogen fixing ability were found 
to be unable to fix nitrogen. Among them are some of the yeasts and 
fungi (Mulder~ 1975). Only prokaryotic microorganisms have been proven 
to have the ability to fix nitrogen. Nitrogen fixation by cyanobacte­
 ria may be important to the prairie ecosystem~ but the amounts of nitro­
 gen fixed are generally small. In the Canadian prairie~ Alexander 
(1979) determined the input of nitrogen from cyanobacteria to be 1-2 
Kg ha-lyr- l . The cyanobacteria Nostoc occurs in the tall grass prairie 
as an algal crust on the soil surface. 
The major emphasis of study in nonsymbiotic nitrogen fixation has 
involved free living bacteria. The first free living~ nitrogen fixing 
bacteria to be isolated was Clostridium pasteurianum by Winogradsky in 
1893. The free living~ nitrogen fixing bacteria belong to only a few 
families and genera. They include: (a) the aerobic azotobacters; (b) 
the facultative anaerobic klebsiellas; (c) the facultative anaerobic 
bacilli of the Bacillus polymyxa and ~. macerans group; (d) most of the 
saccbarolytic clostrida; (e) the anaerobic sulfate-reducing bacteria of 
the genera Desulphovibria and Desulphomaculum; (f) the photosynthetic 
bacteria (Mulder~ 1975). 
In recent years there has been considerable interest in the asso­
 ciation of some of these free living~ nitrogen fixing bacteria with the 
rhizosphere of higher plants~ particularly the grasses. This association 
is due to a zone of enrichment around plant roots. Plant roots exude a 
variety of compounds into the surrounding soil. Among the compounds 
3 
found in root exudates are sugars, amino acids, peptides, enzymes, vita­
 mins, organic acids, nucleotides, and many other miscellaneous compounds 
(Rovira, 1969). According to Mulder (1975), if the amount of combined 
nitrogen in the soil is low and the amount of combined carbon is high, 
the presence of free living, nitrogen fixing bacteria will be favored 
over those microorganisms not capable of nitrogen fixation. 
The best known association of this type is that of the tropical 
grass Paspalum notatum with the bacteria Azotobacter paspali 
(D8bereiner~ 1970). Similar associations have been found in temperate 
grasses. Tjepkema and Burris (1976) reported finding nitrogen fixation 
associated with the roots of Andropogon gerardi, ~' scoparius, Spartina 
pectinata, Stipa spartea, Poa pratensis, Sprobolus heterolepsis, and 
Panicum virgatum in Wisconsin. Weaver et al. (1980) found associative 
nitrogen fixation in Cyndon dactylon, Paspalum dilatum, f. notatum, f. 
urvillei~ f. plicatalum, Axonopus affnis, Andropogon gerardi~ and~. 
scoparius in Texas. Associative nitrogen fixation has also been re­
 ported for Scripus atrovirens in Massachusetts (Kana and Tjepkema, 
1978)~ Oryzopsis hymenoides in Colorado (Wullstein, 1980) and many other 
native grasses and forbs. Some agricultural species have also been re­
 ported to fix nitrogen. Among them are certain varieties of wheat 
(Neal and Larson, 1976), corn (Raju, Evans, and Seidler, 1972), and 
rice (Trolldenier, 1977). 
This research had three major objectives. The first objective was 
to survey a broad range of prairie grasses and forbs for nitrogen fixing 
ability. There have been no reports of rhizosphere nitrogen by prairie 
plants in Kansas and more specifically in the Flint Hills region of 
Kansas. The aim of this survey was to confirm that species reported to 
4 
have ni trogen fi xi ng abil ity in other geograph i c regi ons also have that 
ability in the Flint Hills. This survey also contains species that have 
not been previously tested for nitrogen fixing ability. 
The second objective was to determine the effect of spring burning 
on rhizosphere nitrogen fixation associated with the major tall grass 
species. The tall grass prairie is not a true climatic climax community, 
but is a fire maintained sub-cl"imax community. In the absence of fire, 
the grassland is quickly invaded by cool-season grasses, woody shrubs 
and trees. Historically, the prairie was burned periodically by fires 
started by lightning or by the Plains Indians. Fire was used by these 
early men to attract large herbivores, bison and elk, in order to hunt 
them. Spring burning is now used to control woody vegetation, to in­
 crease forage production and quality, and to improve distribution of 
use by cattle. These range practices result in increased cattle gains 
(Owensby and Smith, 1973). Figure 1 illustrates the difference in 
appearance between burned and unburned prairie in the spring. 
Hulbert (Pers. Comm. 1982) found that applying nitrogen fertilizer 
increased flower stalk and leaf productivity in the major tall grass 
species. This indicates that increases in soil nitrogen following burn­
 ing could be responsible in part for the increase in forage production. 
This increase in nitrogen may be due to an increase in rhizosphere 
nitrogen fixation. It should be noted, however, that fertilizing 
native tall grass prairie is not an accepted practice. 
The third objective of this study was to measure the effect of 
moisture stress on rhizosphere nitrogen fixation. Moisture stress is 
a common environmental factor affecting the prairie ecosystem. Adequate 
moisture is not only essential for plant growth, but has also been 
Fi gu re 1.	 Tallgrass prairie in sprin9 showing difference 
between burned (foreground) and unburned (back­
 bround) sites. 
9 
7 
shown to have an influence on rhizosphere nitrogen fixation (Vlassek, 
Paul, and Harris, 1973). Moisture stress in that study as well as in 
other similar studies was measured as percent soil moisture content. 
Measuring soil moisture content does not accurately reflect moisture 
stress. Since water moves from an area of high energy to an area of 
low energy, not from high concentration to low concentration, a better 
measure of moisture stress would be to measure the free energy of 
water. The free energy of water is referred to as water potential (~). 
Percent soil moisture is also a poor measure of moisture stress because 
soils exhibit hysteresis. This is especially true of the clay soils 
common in the Flint Hills. When comparing percent soil moisture to ~, 
the soil does not wet and dry on the same curve. Instead, it exhibits 
a series of curves. Thus, one measure of soil moisture content could 
represent more than one ~, resulting in large errors. 
In this study, moisture stress was measured as the ~ p in the xylem 
of the plants. In the xylem: 
~ = 111 + ~
 . P 1T 
where ~ p is the pressure potential and ~ 1T is the osmotic potential. 
Since ~1T is negligible in the xylem: 
~ = ~ p 
~ p can be determined by measuring the Xylem Pressure Potential (XPP). 
,
 
MATERIALS AND METHODS 
Study Areas 
This research was done at three study areas. The majority of the 
research was done at the Konza Prairie Research Natural Area near 
Manhattan, KS. The Konza Prairie covers approximately 3500 ha, most of 
which is native, unplowed tall grass prairie. The area also includes 
small areas of farmland and riparian woods. Management of the area is 
designed to approximate the tall grass prairie as it was before settle­
 ment by European man. 
The Konza Prairie is divided into four replicates of a number of 
burning treatments. Those treatments are: burning after one, two, 
four, and ten years; burning after a wet year; burning after a dry year; 
and unburned. The treatment areas are divided such that each area is a 
small, but complete watershed. These burning treatments offered an 
excellent opportunity for sampling different burn treatments that have 
similar soils and are of the same range type. 
The F. B. and Rena G. Ross Natural History Reservation (RNHR) near 
Emporia, KS, was also used as a study area. The Ross Reservation 
offered not only burn treatments, but also allowed for sampling of a 
grazed and burned area and an adjacent area that was unburned and un­
 grazed. The Ross Reservation was also used to sample plants for the 
survey portion of the research. 
The largest part of the survey was conducted at The Land Institute, 
Salina, KS. The Land Institute is owned and operated by Dr. Wes Jackson. 
The research at The Land Institute is aimed at developing a perennial 
grain crop. A better understanding of nitrogen fixation by native 
grasses and forbs would be highly advantageous in developing a 
9 
self-sustaining perennial grain crop. For this reason, Dr. Jackson 
allowed sampling in The Land Institute herbary. The Land Institute's 
herbary is a large collection of prairie forbs and grasses offering an 
excellent opportunity to sample a variety of prairie plants for the sur­
 vey portion of this research. This sampling was done with the help of 
Ms. Laura Jackson. 
Acetylene Reduction Technique 
Nitrogen fixation was measured using the acetylene reduction tech­
 nique. This assay is based on the fact that the nitrogenase enzyme sys­
 tem not only reduces NZ to NH~, but will also reduce acetylene (CZHZ) to 
ethylene (CZH4). This was first shown by sch811horn and Burris (1966) and 
by Dilworth (1966). They were studying the mechanism of nitrogenase and 
were testing a number of gases as nitrogenase inhibitors. When testing 
acetylene as an inhibitor, they found that it did inhibit nitrogenase, but 
after a time, the acetylene disappeared. It was then discovered by using 
14C labelled acetylene that the acetylene was being reduced to ethylene. 
Ethylene is not reduced further and does not inhibit nitrogen fixation or 
acetylene reduction. The nitrogenase enzyme system has such a high affin­
 ity for acetylene that it is not necessary to exclude NZ' The presence 
of ethylene is easily detected using gas chromatography. 
The ace~llene reduction assay has allowed nitorgen fixation to be 
studied not only in the laboratory but also in the field. Samples can 
be taken in remote areas, injected with acetylene, and returned to the 
laboratory for analysis with a gas chromatograph. In this study a port­
 able gas chromatograph was constructed and taken to the study area. 
Field studies of nitrogen fixation have been done in remote areas such 
as the Arctic (Stutz and Bliss, 1970) and in the South American tropics 
(D8bereiner, 1970). 
10 
Root Co 11 ect ion 
Previous researchers using the acetylene reduction assay collected 
samples as intact cores of soil (Weaver et al. 1980; Kana and Tjepkema, 
1978). This method does not ensure sufficient live root material to 
detect nitrogenase activity or measure it accurately. This method also 
makes it difficult to be sure of the identity of the species of roots 
within the core. This is especially true in the grasslands because of 
the tendency for roots of neighboring plants to become intertwined. It 
is also possible for ethylene to be metabolized by soil organisms before 
it diffuses out of the soil core (Witty, 1979). 
The method of root collection chosen for this research involved 
digging up the root system of a plant and removing only those roots 
obviously alive. Only live roots and adhering soil were taken in the 
sample in order to insure the presence of sufficient root exudate in 
the enrichment zone to support nitrogen fixing bacteria. By using only 
living roots, the results could be reported on a per gram of root basis. 
If dead roots had been included in the sample, it would have lowered the 
reported rate of nitrogenase activity. 
As the roots were removed from the plants, they were placed in 
sample vials and immediately closed with serum stoppers to prevent 
dessication of the sample material. In some cases, a few drops of water 
were added to the vial to maintain the humidity in the vial. 
On each sampling date, an attempt was made to collect at least six 
samples of a species from each of two burn treatments. One sample from 
each set of samples was designated as a control and was not injected 
with acetylene. This control was designed to measure any ethylene that 
might normally be produced by the roots or soil microorganisms. 
11 
Acetylene Injection 
Within two hours after roots were collected, acetylene was injected 
into the vials through the serum stopper. Transfer of acetylene from 
the bottle of compressed acetylene was accomplished by using a small 
innertube as a low pressure tank. The valve stem core of the innertube 
was removed and a serum stopper was fitted over the valve stem. Epoxy 
was used to seal the stopper to the valve stem to prevent leaking of 
the acetylene. To remove any gases originally in the innertube, a 
hypodermic needle was fastened to the end of a vacuum line. The needle 
was then inserted through the serum stopper on the innertube. The vacuum 
then pulled out any gas that was in the innertube. By fitting a hypo­
 dermic needle onto the outlet of the regulator on the tank of compressed 
acetylene and inserting that needle through the serum stopper of the 
innertube, an easy transfer of acetylene could be accomplished. The 
innertube was then filled to a very low pressure. 
Before injecting acetylene into the sample vials, a five ml sample 
of air was removed from the vial with a syringe. Then an equivalent 
volume of acetylene was injected into the vial. This was done to main­
 tain the same pressure in the sample vial. 
Early in the research, it was discovered that the acetylene in the 
innertube was contaminated with a small amount of ethylene. The source 
of contamination was not known but was believed to be the rubber of the 
innertube. To correct for the contamination by ethylene, an acetylene 
"blank" was prepared for each set of samples. This was done by removing 
five ml of air from an empty vial and then injecting five ml of acetylene 
into the empty vial. This provided a control that could be subtracted 
from the results for each sample. This is a common problem since 
12 
ethylene is nearly ubiquitous (Mayo, Pers. Comm.). 
Incubation 
As soon as possible after acetylene injection, the samples were 
incubated in darkness. Dark incubation was chosen since it most closely 
represents light conditions in the soil. It is not known whether light 
affects any of the processes involved with nitrogen fixation. 
Incubation directly in the soil at the site of collection was con­
 sidered. This incubation would have been the closest possible simulation 
of the natural conditions. Incubation in the laboratory was chosen, how­
 ever, because of the logistics involved with ~ situ incubation. Incu­
 bation directly in the soil would have involved returning to each col­
 lection site after the incubation time had elapsed. This would mean 
marking each sample site and then relocating them following incubation. 
Because of the distances between study areas and between sample sites, 
in situ incubation was rejected. 
An incubation time of two days was chosen. Two days allowed suf­
 ficient time for a measurable concentration of ethylene to be produced 
without using up all of the available carbohydrate. The available 
carbohydrate was considered to be the limiting factor in the process of 
nitrogen fixation in the sample vial. 
The samples were incubated in air at a temperature of 23-26 C. 
Previous research (van Berkum and Sloger, 1982) has indicated that 
greater rates of acetylene reduction can be achieved by incubating in 
an atmosphere of reduced oxygen. This is because nitrogenase is readily 
inactivated by 02' Since the samples in this study were incubated in 
air the results are a conservative estimate of nitrogen fixation in the 
soi 1. 
13 
The literature is unclear as to a reasonable estimate of 02 concen­
 tration at the root surface. Mulder (1975) incubated samples at an 02 
concentration of four percent. In the laboratory, one set of samples 
was incubated at an 02 concentration of approximately four percent. 
This was done to give an indication of what the results would have been 
if all the samples were incubated at reduced 02 concentration. 
To do this, the sample vials containing root material were flushed 
with N2 gas and then stoppered. Oxygen was then added to the vial to 
approximate a four percent concentration. A set of samples incubated in 
air was run at the same time. One ml samples of gas were removed from 
the vials at intervals throughout the incubation period and measured 
for ethylene production. 
Gas Chromatography 
Production of ethylene in the sample vials was measured using gas 
chromatography. The gas chromatograph used in this study was a portable 
model constructed from plans designed by Mallard et al. (1977). This 
/ 
gas chromatograph (Figure 2) was designed specifically for the purpose 
of measuring nitrogen fixation by the acetylene reduction technique. 
It uses a 12 volt power supply that makes it quite portable and very 
useful in a field study situation. 
The three major components of this gas chromatograph are the dif­
 fusion column, the sensor, and a microampmeter. The column used in this 
study was 3.0 mm 00 by 44 cm stainless steel, packed with 22 cm of 
Poropak Rand 22 cm of Poropak N (Alltech, Deerfield, IL). A tee 
fitting at one end allows for an inlet for the carrier gas, N2' and 
also an injection port. The other end of the column is attached to a 
Swagelok fitting that has been soldered to a brass detector block. The 
Figure 2. Gas chromatograph. innertubes for acetylene and 
ethylene and flasks used for calibration. 
SL 
16 
brass block serves as a heat sink and as a housing for the sensor. The 
sensor is one commonly used in many smoke detectors. This sensor has a 
sintered semi-conductor molded around a small filament heater. When the 
filament is heated in the presence of combustible gases, the resistance 
of the semi-conductor is lowered. The change in the resistance is then 
indicated by the microampmeter. 
The gas chromatograph was housed in a metal instrument case. For 
added protection and to increase portability, a wooden case lined with 
styrofoam was constructed and the gas chromatograph was then placed in­
 side. Openings in both cases were provided for access to controls, 
microampmeter, carrier gas inlet, injection port, power terminals, and 
recorder output. It was soon discovered that the cases were too well 
sealed. This prevented the test gases from escaping the unit rapidly. 
This increased the recovery time of the sensor between injections because 
the trapped gases remained in contact with the sensor for a longer time. 
This problem was overcome by placing a small fan in the gas chromato­
 graph such that it drew fresh air from outside the case and directed it 
on the sensor and the brass detector block. This not only kept the 
test gases clear of the sensor but also pressurized the air inside the 
unit and forced the test gases out more rapidly. The fan did not affect 
the sensitivity of the detector, but it did speed up the recovery of the 
sensor between injections. 
This gas chromatograph is equipped with two sensitivity ranges. 
Range #1 measures concentrations between 10-10 moles ml- l to about 4.0 
X 10- 9 moles ml- l . Range #2 measures concentrations from 0.5 X 10-9 to 
about 3.0 X 10-8 moles ml- l . The higher sensitivity of range #1 was re­
 quired in this study because of the very low rate of nitrogen fixation 
17 
by nonsymbiotic bacteria. When testing for nitrogen fixation by legume 
nodules, range #2 was used because of the greater rate of nitrogen 
fixation by the symbiotic bacteria in the nodules. 
The operation of this gas chromatograph was quite simple. First, 
the carrier gas was turned on and allowed to run for several minutes be­
 fore the power was turned on. After the power was turned on, the unit 
was allowed to warm up for several minutes. When the gas chromatograph 
had warmed up and the microampmeter needle had stopped drifting, a sen­
 sitivity range could be selected and the meter zeroed. The gas chromato­
 graph was then ready for use. 
A one ml sample was taken from the sample vial using a one ml hypo­
 dermic. The sample was then injected into the gas chromatograph through 
the injection port. The time of injection was then noted. Gas chromato­
 graphy is based on the rate at which a specific gas will move through 
the column while interacting with the packing. Since different gases 
have different reaction rates, different gases from the same injection 
arrive at the sensor at different times. Ethylene moves through the 
column after about 60 seconds. Acetylene takes about 90 seconds to 
reach the sensor. By noting the time from injection to the peak response 
on the microampmeter, the gas can be identified as either ethylene or 
acetylene. The concentration of the gas was determined by the maximum 
needle deflection (i.e. peak height) of the microampmeter. 
Gas chromatography calibration was achieved using a series of 
ethylene dilutions. Five 250 ml volumetric flasks were made up to final 
volume of 200 ml by using glass beads. This was done by filling the 
flask with 200 ml of water and then adding glass beads until the water 
level just reached the serum stopper. 
18 
To make the series of dilutions, one flask was made to a concen­
 tration of 10- 7 moles ml- l by adding 0.496 ml of ethylene to 199.5 ml 
of air. The rest of the dilutions were made from this standard by the 
following dilutions. 
2.0 ml of 10-7 moles ml- l in 198.0 ml = 1.0 X 10-9 moles ml- l 
1.0 ml of 10- 7 moles ml- l in 199.0 ml = 0.5 X 10-9 moles ml- l 
0.5 ml of 10- 7 moles ml- l in 199.5 ml = 0.25 X 10- 9 moles ml- l 
0.2 ml of 10- 7 moles ml- l in 199.8 ml = 0.10 X 10- 9 moles ml- l
 
By injecting these dilutions in series and recording the meter deflec­

 tion (peak height), a calibrating curve could be determined. The instru­

 ment was calibrated once during each day of use.
 
Drying Samples
 
After each sample had been tested for ethylene with the gas 
chromatograph, the sample material was removed from the sample vial and 
placed in a small paper bag. Each bag was then labelled with date of 
collection, date of testing, species, type of material (root, rhizomes, 
nodules, or soil), study area, location in study area, and sample num­
 ber. The samples were then dried to a constant weight at 70 C and then 
weighed. 
Weighing Samples 
When the samples had sufficiently dried, 10-12 sample bags were re­
 moved from the oven and placed in a dessicator with dry-rite. This 
allowed the samples to cool to room temperature before weighing without 
allowing them to take up moisture from the air. After the samples had 
cooled, one sample was taken from the dessicator at a time. The sample 
was quickly separated into root material and soil. The root material 
was placed in a covered petri dish and weighed. 
19 
Xylem Pressure Potential 
Moisture stress was measured as Xylem Pressure Potential (XPP). 
Water in the xylem of plants is under tension. When a stem or leaf is 
cut; the water in the xylem is drawn back from the cut surface. This 
is due to the tension on the water in the xylem. By exerting an exter­
 nal pressure on the severed plant part (eg. a leaf), water in the 
xylem can be forced back to the cut surface. The negative of the pres­
 sure required to force the water back to the surface is the Xylem Pres­
 sure Potential. Since the water is under tension, the XPP is reported 
as a negative number. 
The Scholander Pressure Bomb (P.M.S. Instruments, Corvallis, 
Oregon) works on this principle. A cut leaf is placed in the pressure 
chamber with the cut end exposed through the lid. The chamber is then 
pressurized with N2 gas while observing the cut end of the leaf with a 
magnifying lens. When water is seen at the cut surface, the N2 gas is 
shut off and the pressure in the chamber is read from the pressure 
gauge. The negative of this pressure is the XPP. 
In this study seven leaves from seven different plants of each 
species being studied were collected at each sampling site. The leaves 
were then placed in pint plastic containers with wet paper towels. This 
kept the samples from losing any significant amounts of water before 
the samples were read (Figure 3). Each container held only leaves from 
one species from one site. The containers were labelled according to 
species and location. The samples were then taken to the laboratory to 
measure the XPP. 
Leaf samples were taken at dawn and mid-day along two transects. 
Each transect was run once every two weeks. Leaf samples were also 
DZ
 
Figure 3.	 XPP of leaves of Andropogon erardi (Ange), ~. 
Scoparius (Ansc) Panicum virgatum Pavi), and 
Sorghastrum nutans (Sonu) stored in plastic con­
 tainers. Time course shows little change in XPP 
for leaves stored for up to five hours. Vertical 
bars are 95 %confidence intervals. 
Xylem Pressure Potent iol (-M Po) a ~~ 
In b In 
oc-r~~~b:::!:~~E::!:z-:::-r--l 
::) 
lJ) 
n 
-f 
N ~ 
rn 
:r 
0 
C -, 
lJ) w 
~ < ~6'" to 
ro 
zz
 
RESULTS AND DISCUSSION
 
Survey 
Throughout the summer of 1982, a survey to determine the extent of 
nitrogen fixing ability associated with the rhizosphere of prairie 
plants was undertaken. This survey was designed to test a large number 
of prairie species, accessions and genetic crosses for nitrogen fixing 
ability. The survey centered on three families of prairie plants: the 
POACEAE, FABACEAE and ASTERACEAE. Tables 1, 2, and 3 list the plants 
tested in each family and the number of samples having a positive re­
 sponse (i.e. acetylene reduction) and the number having no response. 
The tables also indicate the type of material tested. Table 4 lists 
accessions of Tripsacum dactyloides and genetic crosses tested. The 
accession numbers were assigned by The Land Institute for plants col­
 lected from different areas. The crosses are also from The Land 
Institute. 
A negative response in this type of survey does not indicate that 
the plant has no rhizosphere nitrogen fixing association. A negative 
response only indicates that at the time of sampling, nitrogenase ac­
 tivity was at or near zero. The plant may not have a rhizosphere 
association for nitrogen fixation or environmental factors may have 
limited nitrogenase activity. Nitrogenase activity can be restricted 
by moisture stress, low amounts of combined carbon, high amounts of 
combined nitrogen~ temperature extremes, and other factors. 
Response of Nitrogen Fixation to Burning 
The nitrogenase activity of Andropogon gerardi, ~. scoparius~ 
Panicum virgatum, and Sorghastrum nutans in burned and unburned areas 
24 
Table 1. Type of material 
and negative (-) 
POACEAE. 
tested and number of samples positive (+) 
for acetylene reduction by species of 
Type of
 
Speci es Materi a1 #(+) #( - )
 
Agropyron elongatum 
A. smith i i 
Andropogon gerardi 
A. hall i i 
A. ischaemum 
A. scoparius 
Bouteloua gracilis 
Elymus canadensis 
Koleria cristata 
Panicum virgatum 
Phalaris arundinaceae 
Sorghastrum nutans 
Spartina pectinata 
Tridens flavus 
Tripsacum dactyloides 
Zea diploperennis 
root 
root 
root 
root 
root 
root 
soil 
root 
root 
root 
root 
root 
root 
root 
root 
root 
root 
1 
2 
59 
o 
9 
o 
o 
1 
9 
13 
o 
7 
o 
9 
o 
2 
o 
o 
2 
o 
2 
o 
2 
2 
Material from The Land Institute, Konza Prairie, and RNHR 
25 
Table 2. Type of material 
and negative (-) 
FABACEAE. 
tested and number of samples positive (+) 
for acetylene reduction by species of 
Type of
 
Species Material #(+) #(-)
 
Amorpha canescens 
Astragalus sp. 
Baptisa australis 
var. minor 
Cassia marilandica 
Desmanthus illinoense 
Desmodium illinoense 
Lathyrus sylvestris 
Lespedza bicolor 
L. capitata 
L. stuevei 
Melilotus officianalis 
Petalostemon purpureum 
P. candidum 
Psoralea sp. 
Schrankia nuttallii 
root a 
nodules a 
nodules 1 a 
root a 
root cortex a 1 
nodules a 
roots a 2 
nodules a 
nodules a 
nodules a 
nodules 1 a 
nodules a 
nodules 1 a 
nodules a 
nodules 1 a 
nodules a 
nodules a 
nodules a 
Material from The Land Institute, Konza Prairie, and RNHR 
26 
Table 3. Type of material 
and negative (-) 
ASTERACEAE. 
tested and number of samples positive (+) 
for acetylene reduction by species of 
Type of 
Species Material #(+) #(-) 
Aster ericoides root 1 a 
Englemannia £innatifida root a 
Helianthus grosseratus root a 
H. maximi 11 iani root a 
Helio2sis helianthoides root a 
Liatrus pycnostachya root a 
Ratibida f>.innata root 1 a 
Silphium actinatum root a 
~. ~eciosum root a 
Solidago canadensis root 2 
S. 9l9.antea root 2 a 
S. petiolaris root 
Material from The Land Institute and Konza Prairie 
27 
Table 4. Number of samples positive (+) and negative (-) for acetylene 
reduction by accessions of Tripsacum dactyloides and genetic 
crosses. 
Species or cross Accession # #(+) #(-) 
Tripsacum dactyloides 09 o 1 
18 1 o 
27 o 1 
37 o 
39 1 o 
84 o 1 
86 1 o 
94 o 
100 o 
320 o 1 
(lea mays X Trda) X o 
lea mexicana 
(lea mays X Trda) X o 1 
lea mays 
Trda - Tripsacum dactyloides 
Material from The Land Institute 
28 
is shown in Table 5. Nitrogenase activity associated with the rhizo­
 sphere of Andropogon gerardi and A. scoparius from burned areas was 
significantly greater than those plants from unburned areas. For 
Panicum virgatum and Sorghastrum nutans, nitrogenase activity in the 
burned areas was also greater, but these differences were not statisti­
 cally significant, perhaps due to insufficient sampl-ing. They do indi­
 cate a trend of increased nitrogenase activity following burning. 
The effect of spring burning of rhizosphere nitrogen fixation has 
not been previously reported in the literature. As shown by this study, 
nitrogen fixation is increased by burning. This increase may be respon­
 sible in part for the increased production of the tall grass prairie 
following burning. The cause of this increased nitrogenase activity is 
not fully understood. However, there are some possible explanations. 
During root collection, it was noted that plants in burned areas 
began new root growth earlier in the season than did plants from the 
unburned areas. This is probably due to earlier warming of the soil 
and less water stress in the spring in the burned areas (Knapp, 1984). 
The new root growth indicates increased photosynthesis resulting in 
more carbohydrate exuded into the rhizosphere which will increase the 
number of free living, nitrogen fixing bacteria in the rhizosphere. 
Increased nitrogenase activity may be due to this increase in number 
of nitrogen fixing bacteria. 
Ecotypes of Andropogon gerardi 
On July 8, 1982, I discovered what appeared to be two ecotypes of 
Andropogon gerardi growing in a burned unit of the Konza Prairie. The 
two different big bluestems were growing sympatrically on an upland 
site. One was dark green in color and had wide leaves. This ecotype 
29 
Table 5. Rhizosphere acetylene reduction by four prairie grasses in 
annually burned (B) and unburned (UB) prairie. (n moles 
C2H4/g root) 
SPECIES 
Ange* Ansc* Pavi* Sonu* 
Date B UB B UB B UB B UB 
5-26 
6-08 
6-14 
6- 15 
6-16 
6- 17 
6-25 
6-30 
7-08 
7- 12 
7-14 
7-19 
14.79 
20.63 
7.80 
17.25 
14.46 
35.56 
8.80 
7.50 
14.62 
48.39 
7.37 
3.36 
9.09 
13.44 
13.37 
13.28 
9.50 
4.57 
2.87 
32.35 
11.54 
34.70 
43.95 
22.73 
0.00 
12.73 
9.37 
20.75 
-
 0.00 
7.47 
9.6 
5.0 
14.60 
1. 95 
1. 59 
5.50 
12.67 
13. 17 
1. 96 
8.32 
14.52 
8.28 
24.57 
27.91 
-
 0.00 
0.00 
3.07 
2.24 
13.35 
12.52 
10. 14 
47.23 
7.25 
-
 0.00 
0.00 
2.37 
7.20 
8.73 
23.72 
16.64 
30.83 
3.82 
0.00 
5.41 
9.87 
6.74 
30
 
Table 5. (Conti nued) 
SPECIES 
Date 
Ange* 
- ­
 B UB 
Ansc* 
B UB 
Pavi* 
B UB 
Sonu* 
B UB 
7-21 1.09 
7-24 11.74 23.38 
21.20 15.66 
68.40 
9-29 1. 57 6.13 
Mean 16.9 + 
-
 5.4a­
 8.48 + 
3.9 
16.7+ 
-
 11.3 
1. 3a 18. 1 2.4 16.8 5.5 
*Ange 
Pavi 
= 
= 
Andropogon gerardi; Ansc 
Panicum virgatum; Sonu = 
= Andropogon scoparius; 
Sorghastrum nutans 
aSignificant difference at the 95 % level 
31 
was designated Ange Og. The second was lighter green and had narrow 
leaves. This ecotype was designated Ange Lg. Figure 4 shows both eco­
 types growing together. 
XPP and acetylene reduction of these potential ecotypes of A. 
gerardi were sampled separately. These results are shown in Table 6. 
These results show that Ange Og had a higher rate of nitrogenase activ­
 ity through most of the sampling period. The XPP of Ange Og was lower 
than the XPP for Ange Lg. This indicates that Ange Og was experiencing 
greater moisture stress. This probably is due to the stomates of Ange 
Og staying open longer thus allowing for increased C02 assimilation. 
The increase in photosynthesis could allow for more photosynthate to be 
translocated to the roots and then exuded. This might increase the 
energy available to the rhizosphere for nitrogen fixation. 
Ange Lg had a higher XPP. This suggests that it is conserving 
water by closing its stomates sooner. This could reduce C02 assimi­
 lation and in turn reduce the amount of energy available for nitrogen 
fixation. This suggests that Ange Lg is adapted to survive better in 
a dry year. In a wet year like 1982, Ange Og would do better. 
Moisture Stress 
Figure 5 shows dawn and mid-day XPP for Andropogon gerardi from 
all burned and unburned sites. Throughout the season, the plants were 
not under mucb moisture stress in both burned and unburned prairies. 
Plants from the unburned prairie did have lower XPP than did plants 
from the burned prairie in the early part of the season (Knapp, 1984). 
Figure 6 shows dawn and mid-day XPP for ~. gerardi from adjacent burned 
and unburned sites. The burned area had been burned annually and the 
unburned area had not been burned in recent years. Early in the season, 
·50 a5uV Aq papunoJJns (Ja1UaJ) 
·~pJ~Ja5 u060doJpuV JO SadA10Ja L~~1ua10d 
££
 
--
 Table 6. Rhizosphere acetylene reduction and XPP of Andropogon gerardi ecotypes growing sympatrically in 
annually burned prairie. 
n moles C2H4/9 root XPP (-MPa) 
Date Dark green Light green Dark green Light green 
7-08 11.73+2.8 1.54 + 1.95 Mean max 
Mean min 
XPP 
XPP 
= 
= 
.17 MPa 
1.34 MPa 
7- 12 17.21 + 1.9 
-
 4.03 + 10.9 1.31+0.6 .84 + .109 XPP s i gnifi cant 
7- 14 15.9 + 2.2 3.29 + 5.5 No XPP taken 
7- 21 
-
 1.09 3.22 1.12 + .16 1.20 + . 16 Only single samples of acetylene 
reduction 
7-23 0.56 + .09 .22 + .02 Dawn Significant1.82 +" .06 1.48 +" .08 Midday 
7-29 .18 + .10 
1--:-60 
.13 + .07 
1--:-65 
Dawn 
Midday -­ Single Samples 
-
 8-06 .125 + .07 
1.6 +" .24 
.20 + .09 
.68 +" .31 
Dawn 
r~i dday 
w 
oj:::> 
Figure 5.	 Dawn and mid-day XPP for Andropogon gerardi from 
all burned and unburned prairie sites. Vertical 
bars are 95 %confidence intervals. 
2,0 tr 0 Dawn, Burned 
0- ­ -­ -0 Dawn, Unburned 
• • Mi d-day, Burned 
... ---.. Mid-day, Unburned 
~ 1,S 
~ 
I
 -
 0
 .­
 -
 \ 
\ 
c I 
~ 1,0 
0 
/ \ 
\ 
Q.. 
Q) 
L­
 :J 
\ 
\ 
III 
III~q.S 
Q.. 
\ I 
i 9 
E 
Q) /~\~ 
>-. 
X r;-, 
1 10 20 30 10 20 30 1 
JUNE JULY AUG 
\, 
w 
m 
Figure 6.	 Dawn and mid-day XPP for Andropogon gerardi from 
adjacent burned and unburned sites. Vertical 
bars are 95 %confidence intervals. 
2,Or	 D C1 Dawn. Burned
 
0- - --0 Dawn, Unburned
 
w Mi d-day, Burned
• 
... -- ... Mid-day, Unburned 
-
 0...	 ~--+2::: a 1.+ ~ I I I - -- - " 
a " "­
 .....
 - /	 "­
 c -f---t ---	 " " , ,1,0 I Q) 
-+-'	 I ,0 
0... I 
" IQ) 
l- I '1::J
 
VI
 
VI 0,5Q)
 
I ­
 0... 
E 
Q)
 -
 >. 
X 
, , , , , , , , , I , I , I I , I I	 , I • I , 
1 10	 20 30 10 20 30 1 
I JUNE I	 JULY I AUG 
, , - ­ I , , , ' , , , 
w 
(Xl 
39 
plants from the unburned area were under greater moisture stress than 
those from the burned area. On June 8, new root growth was noted on 
plants from the unburned area. After that time, plants from both areas 
experienced similar moisture stress until the middle of July. During 
July there was periods of drought after which plants from the burned 
site experienced greater moisture stress than did plants from the un­
 burned area. 
Andropogon scoparius exhibited a similar pattern of XPP during the 
season on the same burned and unburned sites (Figure 7). Plants from 
the unburned prairie had a lower XPP early in the season than did A. 
gerardi during the same period. 
Sorghastrum nutans and Panicum virgatum also had a similar pattern 
of XPP on the same burned and unburned sites (Figures 8 and 9). How­
 ever, plants from the unburned area remained drier than those from the 
burned prairie until later in the summer. 
All four species had lower XPP in the unburned area until later in 
the summer when XPP tended to be lower in the burned area. This is 
probably due to slower root growth in the unburned area because of lower 
soil temperature. As root growth of unburned plants increased~ the XPP 
increased. 
Effect of Moisture Stress on Nitrogen Fixation 
Nitrogenase activity (CZH Z reduction) and XPP for Andropogon 
gerardi were followed through the summer. Figure 10 indicates that 
nitrogenase activity does respond to moisture stress. At the end of 
May, XPP is high indicating little moisture stress. Nitrogenase activ­
 ity at that time was also high. As XPP began to decrease after June 8, 
nitrogenase activity also decreased. This pattern continues through 
Figure 7.	 Dawn and mid-day XPP for Andropogon scoparius 
from adjacent burned and unburned sites. 
Vertical bars are 95 %confidence intervals. 
2,Or 0 c Dawn, BurnedAnsc 
0- - --0 Dawn, Unburned
 
a • Mid-day, Burned
•0.... ~ I ...---. Mi d-day, UnburnedL 
I 
-
 --1 1,5 
« 1\I-
 z 
w -...,------- ­
 I ­
 -0 ­
 0.... 1,0 
\ 
+-; " ........
w 
IX 
:J \ 
If) 
w
 \ 
r 
/ 
A", 
'" "-
 '" ~¥ 
/// 
........
 
\If) 
w \ 
IX 
0.... 0,5 \ .......
 
L / 
'" 
--1 
>- ~ '" X ,l ,I 
1 10 20 30 10 20 30 10 
JUNE I JULY I AUG 
+::> 
Figure 8.	 Dawn and mid-day XPP for Sorghastrum nutans from 
adjacent burned and unburned sites. Vertical 
bars are 95 %confidence intervals. 
w
  
o
  
Figure 9.	 Dawn and mid-day XPP for Panicum virgatum from 
adjacent burned and unburned sites. Vertical 
bars are 95 %confidence intervals. 
2.0r 
~ ~ L 
I
 =: 1.5 
<: 
....... 
z 
w 
....... 
~ 1,0 
w 
a: 
Pavi 
0 0 Dawn, Burned 
0-- --0 Dawn, Unbu rned 
• • 
Mid-day, Burned 
.­ -­ ..... Mid-day, Unburned 
'"
 f \ 
t-­
 ,f'
 / 
~ 
If) 
If) 
w 
~0.5 
L 
w 
--.J 
>­
 X -~---~¢==--I -­ I I I 
1 10 20 30 10 20 30 
JUNE JULY 
/ 
/ 
/ 
/ 
I 
rI ~, 
+- ----t-­
 V 
.s:::­
 U"l 
Figure 10.	 Nitrogenase activity (ethylene production) and 
XPP for Andropogon gerardi from burned prairie 
throughout the summer. 
20 
15 
-0 
e 
0' 
-10
 -£
 
~ 
III 
Q) 
o 
E 5 
c 
21 
dawn XPP\ 
20 
15 
"'in 
L­
 a 
.0 
I 
10 ­
 a.. 
D... 
x 
5 
~ 
'-J 
48 
the summer. On July 14, the dawn XPP was quite low. This was followed 
by a decrease in nitrogenase activity to near zero. 
Figure 11 more directly illustrates the effect of decreasing XPP 
on nitrogenase activity. These are single data points of ethylene pro­
 duction and XPP for ~. gerardi. Even though these are single data 
points, they show the tendency for nitrogenase activity to decrease as 
XPP decreases. 
Effect of 02 Concentration on Nitrogen Fixation 
Figure 12 shows the effect of 02 concentration on nitrogenase 
activity for root samples of the two potential ecotypes of ~. gerardi. 
The samples incubated in air had a linear increase in ethylene concen­
 tration through the incubation. Ethylene concentration increased 
exponentially for the first 24 hours in the samples incubated in four 
percent 02. After 24 hours, ethylene concentration leveled off. Samples 
incubated in reduced 02 had a final ethylene concentration as much as 
eight times greater than those samples incubated in air. This indi­
 cates that nitrogenase activity is enhanced be reducing the 02 tension. 
These results raise the question of which type of incubation 
atmosphere more closely represents what is happening in the field. 
Soil air and atmospheric air have nearly the same 02 concentration 
(Brady 1974). The problem is knowing what the 02 concentration is at 
the root surface. van Berkum and Sloger (1982) used concentrations 
ranging from 0.0 to 20.0 percent. Nowhere in the literature is it 
indicated that the 02 concentration used for incubation simulated the 
02 concentration at the root surface of plants growing in soil. 
Kristensen and Lemon (1964) developed a model to determine 02 
concentration at the root surface. This model requires that certain 
Figure 11. Nitrogenase activity (ethylene production) vs. 
XPP for Andropogon gerardi from burned prairie. 
• • 
(SJoq -) loqualod aJnssaJd walAX 
Ol O~ 
• 
:J 
3 0
 ­
 ro 
lJ) 
n 
N 
l'-I
 (Q
 ­
 .,
 O~ 0 
0
 -
 • 
Ol
 • 
os
 
Figure 12.	 Time course of nitrogenase activity (ethylene 
production) for two ecotypes of Andropogon 
gerardi incubated in air and in four percent 02. 
900
 
nmoles C2H4 
ANGE Og 4% 021500 
~-----------------~~-. 
ANGE Lg 
ANGE Lg 4% 
./,
 / 
/
 •I 
I 
I 
I, 
I 
I 
I 
I 
t 
__ - --e 
021200
 
groat
 AIR
 
600
 
300
 
.............. _._----------- ANGEDg AIR
 
8 16 24 32 40 48 56 64 72
 
HOURS 
Ul 
N 
53 
root and soil characteristics must be known. The three root character­
 istics are its oxygen use rate, the root radius and the critical level 
of 02 concentration for root growth. The three soil characteristics 
are the apparent liquid diffusion path length (the distance from the 
liquid-gas interface to the root surface), the diffusion coefficient 
for the solid-liquid matrix surrounding the root and the 02 concentra­
 tion at the liquid-gas interface. Determining 02 concentration at the 
root surface is beyond the scope of this research, but the model does 
indicate that the 02 concentration at the root surface will be much 
less than the 02 concentration of air. This indicates that the re­
 ported values of nitrogen fixation in this study are conservative 
estimates only. These values are adequate for comparison within the 
study. 
SUMMARY 
There are several conclusions that can be drawn from this re­
 search. 
Rhizosphere nitrogen fixation is associated with many species of 
prairie plants. Of the 43 species tested, more than 30 had nitrogenase 
activity associated with their roots. 
In measuring XPP, leaves of prairie grasses can be kept in 
humidity chambers for several hours before measuring without signifi­
 cantly changing the XPP. This allows for a large number of samples to 
be collected in a short period of time and then taken to the laboratory 
for XPP measurement. 
During the 1982 measurement period, XPP of prairie grasses was not 
particularly low. The plants were under only moderate stress. Early 
season water potentials were lower in the unburned areas than in adja­
 cent burned areas. This may be due to lower soil temperature in the 
unburned areas at that point in the season. This results in reduced 
early root growth. Later in the season, water potentials were lower 
in the burned prairie. 
Nitrogen activity decreases as XPP decreases. As the season pro­
 gressed and moisture stress increased, the rate of acetylene reduction 
decreased. This may be due to lower amounts of carbohydrate in the 
root exudate. As the plant experiences lower XPP its stomates remain 
open for only a limited time. This results in a decreased photosynthe­
 sis rate and less photosynthate being exuded by the roots. 
This phenomenon may also be an affect of 02 concentration. The 
nitrogen fixing bacteria in the rhizosphere occur on or near the root 
surface. Normally the root is surrounded by a liquid matrix. This 
55 
matrix controls the 02 concentration at the root surface. As this ma­
 trix dries~ the 02 concentration increases due to contact with the soil 
air. The nitrogenase is then inhibited by the oxygen. 
Potential ecotypes of Andropogon gerardi were found at the Konza 
Prairie. They have differing rates of rhizosphere nitrogen fixation 
and differing Xylem Pressure Potential. These differences appear to be 
the result of adaptations to moisture conditions. 
Rhizosphere nitrogen fixation was consistently higher in the 
burned prairie. This~ in part~ helps to explain the greater productiv­
 ity associated with spring burning of the tall grass prairie. 
0311J 3Cln1'u'Cl 31 Il 
LITERATURE CITED
 
Alexander, M. 1977. Introduction to soil microbiology. ,John Wiley 
and Sons, New Ybrk. 
Brady~ N.C. 1974. The nature and property of soils. 8th ed. Mac­
 Millan Publishing Co., Inc., New York. 
Dilworth, M.J. 1966. Acetylene reduction by nitrogen-fixing prepara­
 tions of Clostridium pasteurianum. Biochem. Biophys. Acta. 
127:285-295. 
Dobereiner~ J. 1970. Further research on Axotobacter paspali and its 
variety specific occurrence in the rhizosphere of Paspalum notatum. 
Flagge, Zentble. Bakt., Parasitke. Abt. 11. 124:224-230. 
Kana, M.T. and J.D. Tjepkema. 1978. Nitrogen fixation associated with 
Scirpus atrovirens and other nonnodulated plants in Massachusetts. 
Can J. Bot. 56:2636-2640. 
Knapp, A.K. 1984. Post-burn differences in solar radiation, leaf 
temperature and water stress influencing production in a lowland 
tallgrass prairie. Amer. J. Bot. 7(2):220-227. 
Kristensen, K.J. and E.R. Lemon. 1964. Soil areation and plant root 
relations. III. Physical aspects of oxygen diffusion in the 
liquid phase of the soil. Agron. J. 56:295-301. 
Mallard, T.M., C.S. Mallard~ H.S. Holfeld, T.A. LaRue. 1977. Inexpen­
 sive portable gas chromatograph for the acetylene reduction assay 
for nitrogenase. National Research Council of Canada. Prairie 
Regional Laboratory. Saskatoon, Saskatchewan. 
Mulder~ E.G. 1975. Physiology and ecology of free-living, nitrogen­
 fixing bacteria. In: Nitrogen fixation by free-living micro­
 organisms. W.D.F. Stewart editor. Cambridge University Press. 
Cambridge. 
Neal, J.L. and R. Larson. 1976. Acetylene reduction by bacteria iso­
 lated from the rhizosphere of wheat. Soil Biol. Biochem. 8;151­
 155. 
Owensby, C.E. and E.F. Smith. 1973. Burning true prairie. Proc. 
3rd Midwest Prairie Conf. 
Raju~ P.N., H.J. Evans, and R.J. Seidler. 1972. An asymbiotic nitro­
 gen-fixing bacterium from the root environment of corn. Proc. 
Natl. Acad. Sci. USA. 69:3474-3478. 
Risser, P.G., E.C. Birney, H.D. Blocker, S.W. May, W.J. Parton~ and 
J.A. Weins. 1981. The true prairie ecosystem. Huchinson Ross 
Publishing Co., Stroudsburg, Penn. 
58 
Rovira, A.D. 1969. Plant root exudates. Bot Rev. 35:35-57. 
Scholl horn, R. and R.H. Burris. 1966. Study of intermediates in nitro­
 gen fixation. Fedn. Proc. 25:710. 
Stutz, R.C. and L.C. Bliss. 1973. Acetylene reduction assay for nitro­
 gen fixation under field conditions in remote areas. Pl. Soil. 
38:209-213. 
Tjepkema, J.D. and R.H. Burris. 1976. Nitrogenase activity associated 
with some Wisconsin prairie grasses. Pl. Soil 45:78-94. 
Trolldenier, G. 1977. Influence of some environmental factors on N2­
 fixation in the rhizosphere of rice. Pl. Soil. 47:203-217. 
van Berkum, P. and C. Sloger. 1982. Physiology of root associated 
nitrogenase activity in Oryza sativa. Plant Physiol. 66:1161-1164. 
Vlassek, K., E.A. Paul and R.E. Harris. 1973. Assessment of biolog­
 ical nitrogen fixation in grassland and associated sites. Pl. 
Soil. 38:637-649. 
Weaver, J.E. 
Lincoln, 
1954. North American Prairie. 
Nebraska. 
Johnson Publishing Co., 
Weaver, R.W., S.F. Wright, M.W. Varanka, O.E. Smith, E.C. Holt. 1980. 
Dinitrogen fixation (acetylene reduction) by established forage 
grasses in Texas. Agron. J. 72(6):965-968. 
Witty, J.F. 1979. Overestimate of N2-fixation in the rhizosphere by 
the acetylene reduction method. In The soil root interface. J.L. 
Harley and R.S. Russell editors. ~cademic Press. New York. 
Wullstein, L.H. 1980. Nitrogen fixation (acetylene reduction) asso­
 ciated with rhizosheaths of Indian Ricegrass used in stabilization 
of the Slick Rock, Colorado tailings pile. J. Range Mgt. 33(3): 
204-206.