Giant cane [Arundinaria gigantea (Walter) Muhl.] is a native bamboo species that was once widely distributed within bottomland forests and as extensive monotypic stands (canebrakes) along waterways of the southeastern United States. Land conversion to agriculture greatly decreased the distribution of canebrakes. Limited to less than two percent of its historic range, canebrakes are now considered an endangered ecosystem. A 0.24 hectare low maintenance experimental nursery of giant cane was established at Southern Illinois University to examine the effect of planting stock type, soil amendments, and four collection sources on cane survival and growth (number of culms, height and diameter of the tallest culm, spread between furthest two culms) after two growing seasons. All treatments, plus interactions of source by soil, and source by stock, were significantly affected by survival. Collection source significantly affected all growth measurements. Stock type affected the number of culms, height, and spread. Soil amendments did not significantly influence any growth measurement but affected soil chemical properties. The interactions between source and stock affected the number of culms, height, and spread. The interaction between stock and soil significantly affected spread, but no other interactions significantly affected any growth measurements. Of all treatment combinations, the Cypress Creek West source, when grown in containers, tended to have the greatest survival and overall growth after two years, regardless of soil amendments. This study highlights factors that are important in establishing a giant cane restoration nursery. This nursery will also generate growing stock for future canebrake restoration and rehabilitation projects.

The effectiveness of giant cane [Arundinaria gigantea(Walt.) Muhl.] as riparian buffer vegetation has been demonstrated through research and has gained interest from state and federal agencies to support restoration efforts. Unfortunately, little is known about the physical and chemical properties of the soils below canebrakes and how soil characteristics influence aboveground and belowground biomass production. To determine what physical attributes of the plant influence its success as riparian buffer vegetation and also to determine the interactions with underlying soils, fourteen canebrakes were sampled throughout Southern Illinois. Objective one was to develop an allometric equation to quantify belowground biomass based on aboveground parameters of canebrakes. Previous research found that successful propagation was dependent on rhizome length, the number of internodes and the number of rhizome buds present, but no data exists regarding the yield of rhizomes for a given area. By harvesting all aboveground biomass (culms and leaves) and belowground biomass (roots and rhizomes) to a depth of 25 cm from a 1-m2 plot at each site, morphometric characteristics were quantified and biomass allocation throughout the plant was determined. A significant linear relationship between total aboveground biomass (live and dead) and belowground biomass in giant cane was evident (R=0.865, p<0.001). Although this is a strong relationship, it may be impractical for a manager to harvest, process, and weigh all of the aboveground biomass to speculate the biomass below ground. Therefore, metrics were explored for predicting the length of rhizome, number of rhizome internodes and number of rhizome buds an area will yield using multiple regression and models were developed that estimate these parameters. Using the equation that predicts the number of rhizome buds for a given area, the yield of propagules can then be estimated. Although this equation does not account for all variation of belowground characteristics, it will provide a general guideline for land managers restoring giant cane. The second objective was to estimate biomass allocation of giant cane roots/rhizomes beneath canebrakes by depth (i.e., at 25-cm increments to a depth of 150 cm). Results showed that 67% of giant cane's belowground biomass was within the top 25 cm of the soil profile and accounted for 65% of all belowground biomass encountered at that depth. Giant cane rhizomes were documented to a depth of 51-75 cm deep while cane roots existed in the deepest cores at a depth of 126-150 cm with an average density of 0.08 kg m-3. Giant cane belowground biomass declined with increasing depth, but was still the dominant species at 26-50 cm, comprising 61% of all biomass encountered at that depth. These results support the utility of giant cane as an effective riparian buffer species by increasing the soil porosity and promoting infiltration while contributing a significant source of carbon to the soil profile. Chemical and physical soil properties were measured to determine if they related to canebrake characteristics. Significant correlations were found between various soil properties and canebrake characteristics, implying there is an interaction between giant cane and the underlying soil. Results from this research will improve our understanding of the dynamics of giant cane and supplement existing information to help guide restoration efforts.

Giant cane (Arundinaria gigantea) is a native bamboo that forms large monodominant stands called canebrakes in bottomlands in the southeastern US. Canebrakes are valuable habitat for wildlife and function as riparian buffers to protect soils and water quality. Currently, only 2% of canebrake ecosystems remain. Thus there is interest in establishing new canebrakes as well as maintaining and expanding existing canebrakes. For field restoration, using cane rhizomes to produce transplants is possible but it is unknown when propagules should be collected and grown. For rehabilitation of existing canebrakes, preliminary studies suggest that fertilization and disturbance such as fire can be beneficial but additional broader-ranging studies are warranted. This research reports on three giant cane studies involving producing transplants from rhizomes and involving managing existing canebrakes with disturbance in southern Illinois. The main objectives of study 1 were to ascertain a) if collection season influences the proportion of rhizome propagules that grow at least one culm (culm production success) and the resultant culm growth when transplanted in a greenhouse b) if specific rhizome characteristics influence an individual rhizome’s ability to produce a culm c) if aboveground biomass could be a predictor of the amount of belowground propagules and d) if aboveground biomass, total rhizome length, number of rhizome nodes, or number of rhizome buds could be a predictor of how many culms could be produced when transplanting. Monthly, for a year, sample plots were randomly selected in the SIUC giant cane nursery. In sample plots aboveground culm measurements were collected included live culm density (#/0.25m2), dead culm density (#/0.25m2), height of the tallest culm (cm), diameter of the tallest culm (mm), and total aboveground biomass (g). Each month rhizomes were dug from the sample plots, measured (length (cm), diameter (mm), # nodes, and # live buds), and transplanted into pots and placed in the greenhouse for approximately 94 days. At that time, measurements were taken of the number of live and dead culms produced per pot and the height of the tallest culm (cm). Results indicate that rhizomes collected and transplanted in the greenhouse during winter and spring months (December-May), had significantly greater culm production success and produced significantly taller culms. Also, rhizomes intermediate in length (18-30 cm) that contained 5 to 12 nodes and 4 to 9 live buds tended to have higher than average culm production. Finally, a positive correlation existed between the amount of aboveground biomass and the number of rhizome nodes, the number of live rhizome buds, and the length of rhizomes found in sample plots. Aboveground biomass can predict the amount of belowground rhizome propagules that can be used for canebrake restoration. Study 2 is a two year continuation of work initiated in 2011 and reported on by Margaret Anderson in 2014 on the effects of fire, fertilization, and fire and fertilization combined on the growth and expansion (culm density, height and diameter) of canebrakes within the Cache River Watershed in southern Illinois. Results showed considerable year to year variability among treatments within the canebrake for some growth parameters. However, three years after disturbance, all treatments tended to have similar culm density and growth values and canebrake expansion occurred for all treatments into exterior plots. Fire alone was similar to controls in growth parameters and did not produce any long-term negative effects. Thus, fire can be used as an effective tool to reduce competition from other species, allowing managed canebrakes to persist longer than those that remain undisturbed. Fertilization used alone and in combination with fire, produced slight growth and density increases, but added costs may not warrant its use in canebrake management. Study 3 compared fire and mowing disturbances on the growth and expansion of remnant canebrakes located in southern Illinois. Eleven replications in remnant canebrakes were established throughout the Cache River watershed. Each replication contained a fire only treatment, a mowing only treatment, and a control. Measurements were taken in the dormant season early in 2012 prior to a single mowing and a single fire in March, and after each growing season thru 2014. Measurements including live and dead culm density (#m2), culm height (cm), and culm diameter (0.01 mm), were taken within subplots in the canebrake interior and exterior. Results from this study showed that mowing can be used as an effective alternative to fire for the management of remnant canebrakes through reductions in competition. Neither fire nor mowing produced negative effects of growth within the measured canebrakes or in adjacent areas. Like the previous study, all treatments including the control experienced an outward expansion of cane culms throughout the course of the study. This study shows that both fire and mowing can be used as effective tools to reduce competition within canebrakes to aid in their continued vigor.

AN ABSTRACT OF THE THESIS OF MARGARET MARZIYE ANDERSON, for the Masters of Science degree in Forestry, presented on September 12th, at Southern Illinois University Carbondale. FIRE AND FERTILIZATION EFFECTS ON THE GROWTH AND EXPANSION OF EXISTING NATIVE CANEBRAKES [ARUNDINARIA GIGANTEA (WALT) MUHL] IN SOUTHERN ILLINOIS MAJOR PROFESSOR: Dr. Jon Schoonover Giant cane [Arundinaria gigantea (Walt) Muhl.], a native bamboo, is an integral component of bottomland forests in the southeastern United States. Cane occurs as monodominant stands, also known as canebrakes, which historically covered vast areas of land. As a result of land conversion, overgrazing and altered fire regimes, an alarming 98% reduction of canebrakes has occurred. Due to the ecological significance of giant cane as wildlife habitat, a riparian buffer, its role in soil stabilization and potential as woody biomass, restoration interest has increased. Research with planted cane indicated fertilization and burning had interacting effects on cane growth, however in remnant natural stands, the influence of burning and fertilization on canebrake growth and spread is unknown. This study examined the survival and growth response of cane to burning and fertilization in remnant stands to provide guidance for rehabilitation, restoration and management. Four treatment plots were replicated eight times across seven sites in canebrakes growing in riparian zones adjacent to agricultural fields in the Cache River Watershed, Illinois. The four treatments were randomized factorial design of: 1) burning, 2) fertilization, 3) burning/fertilization, or 4) control. Within treatment plots, two interior and three exterior 1-m² subplots were randomly established to measure culm density (stems/ha), height (cm), diameter (mm), and spread (increase in live culm density by the outward movement from interior subplots into exterior subplots) prior to treatment and after one and two growing seasons. Fertilized and fertilized/burned plots were treated in summers of 2011 and 2012 with a half corn rate of nitrogen (56 kg ha-1), phosphorus (22 kg ha-1), and potassium (37 kg ha-1). Prescribed burning took place in March 2012. Data were analyzed using a three way analysis of variance (fire, fertilization and subplot) (α = 0.05). At year 0 (2011), culm density, height and diameter were not significantly different among treatments. By year 2, live culm density in interior plots slightly increased, however density in exterior plots tended to more than double, indicating canebrake expansion over time. Fertilization tended to increase height and had little effect on cane diameter. Research suggests that cane typically increases in both height and diameter simultaneously, suggesting that fertilization only partially provides the resources needed to stimulate growth. Further analysis on fertilization application rates and timing may be necessary to ascertain the efficiency of its role in culm growth and development. Giant cane responded to prescribed burning through a decrease in height and culm diameter. However, fire increased culm density through stimulation of the growth of new culms. In addition, though fire consumed a portion of existing culms, the canebrake emerged vigorously, demonstrating prescribed fire's utility as a tool for land managers to reduce competition and increase canebrake health and expansion.

Riparian buffers are a commonly utilized best management practice (BMP) for mitigating non-point source pollution from agricultural fields. This practice is most effective when runoff enters the buffer as sheet flow. However runoff from becomes concentrated and forms concentrated flow paths (CFPs) breaking through the buffer in critical erosional areas. These critical areas can be determined prior to the implementation of a BMP such as a riparian buffer through incorporation of the Revised Universal Soil Loss Equation (RUSLE) with Geographic Information Systems (GIS). This approach utilizes field characteristics with hydrological processes in order to determine the erosion risk areas from observed field data throughout an entire watershed. In order to re-establish a riparian buffer in critical areas managers need a method that can prevent erosion and the loss of top soil. This study evaluates an innovative method, "Buffer in a Bag", used to restore riparian buffer benefits and to prevent further erosion in CFPs. There were three different treatments initiated for this study the first two treatments the "Buffer in a Bag", and Hand-Planted were implemented prior to the broadcasting of switchgrass seed. The final treatment was the Control, which required no planting within the treatment plot. The "Buffer in a Bag" treatment utilized the placement of giant cane rhizomes within burlap bags along the channel of concentrated flow. The hand planted treatment utilized the placement of giant cane rhizomes planted in the agricultural field adjacent to the path of concentrated flow. The presence of these CFPs breaking through riparian buffers has been noted, but the actual significant soil erosion occurring within the CFP and in the adjacent field has yet to be quantified. Through the use of topographical surveys and geostatistical analysis to produce digital elevation models (DEM) of in-field elevation changes the hydrological processes and volumetric difference for CFPs can be determined. The goal was to define the variability and the spatial pattern from the data as a means of providing comparisons between subsequent survey periods over time and space to estimate the difference and movement of sediment within CFPs. The volumetric differences found when comparing the DEMs between survey periods provide a quantifiable method for evaluating the implementation of a treatment, such as the "Buffer in a Bag". Using geostatistical analysis of volumetric difference between the treatment plots determined that that there was no significant difference between the CFP's over the study period. The mean percentage volumetric difference within the treatment plots between "Buffer in a Bag", Hand-Planted, and Control treatments were -2.49%, -3.72%, and -2.37%, respectively. There was overall net erosion across all treatments with the erosion and accumulation being significantly different across the survey periods for all the treatment CFPs. The mean percentage volume loss across treatments and survey periods varied from positive 10% accumulation to negative 15% erosion for overall loss when looking at each individual survey and CFP. The displayed loss of valuable topsoil and continued expansion of these concentrated flow paths emphasis the need for research in further management applications. There needs to be an effective method developed to address the volumetric difference within CFPs forming in these high risk erosion areas. Using a GIS/RUSLE approach to indicate these high risk erosion areas will hopefully facilitate the prevention of the formation of concentrated flow paths prior to the implementation of a BMP, such as riparian buffers.