} “A DS -A002 OS Be ceun i Spartina sinters for Substrate Stabilization and Salt Marsh Development by W. W. Woodhouse, Jr., E. D. Seneca, and S. W. Broome ~ TECHNICAL MEMORANDUM NO. 46 AUGUST 1974 Prepared for U. S. ARMY, CORPS OF ENGINEERS COASTAL ENGINEERING (8 RESEARCH CENTER YSO Kingman Building ; VY : Fort Belvoir, Va. 22060 yo Te Reprint or republication of any of this material shall give appropriate credit to the U.S. Army Coastal Engineering Research Center. Limited free distribution within the United States of single copies of this publication has been made by this Center. Additional copies are available from: National Technical Information Service ATTN: Operations Division 5285 Port Royal Road Springfield, Virginia 22151 Contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. The findings in this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. ON 4004 ALL 0 0301 00895954 3 eee SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) READ INSTRUCTIONS REPORT DOCUMENTATION PAGE BEFORE COMPLETING FORM T. REPORT NUMBER 2. GOVT ACCESSION NO|| 3. RECIPIENT'S CATALOG NUMBER TM-46 4. TITLE (and Subtitle) 5S. TYPE OF REPORT & PERIOD COVERED PROPAGATION OF SPARTINA ALTERNIFLORA FOR SUBSTRATE STABILIZATION AND SALT MARSH DEVELOPMENT 7. AUTHOR(s) W.W. Woodhouse, Jr. E.D. Seneca S.W. Broome - PERFORMING ORGANIZATION NAME AND ADDRESS Technical Memorandum PERFORMING ORG. REPORT NUMBER 6. 8. CONTRACT OR GRANT NUMBER(s DACW72-70-C-0015 DACW72-72-C-0012 10. PROGRAM ELEMENT, PROJECT, TASK AREA & WORK UNIT NUMBERS North Carolina State University Raleigh, North Carolina 27607 G31167 12. REPORT DATE August 1974 13. NUMBER OF PAGES 155 15. SECURITY CLASS. (of thie report) - CONTROLLING OFFICE NAME AND ADDRESS Department of the Army Coastal Engineering Research Center Kingman Building, Fort Belvoir, Virginia 22060 MONITORING AGENCY NAME & ADDRESS(if different from Controlling Office) 14. Unclassified 15a, DECL ASSIFICATION/ DOWNGRADING SCHEDULE DISTRIBUTION STATEMENT (of this Report) Approved for public release; distribution unlimited DISTRIBUTION STATEMENT (of the abstract entered in Block 20, if different from Report) - SUPPLEMENTARY NOTES - KEY WORDS (Continue on reverse side if necessary and identify by block number) Cordgrass Fertilization Seeding Dredge Spoil Marsh Transplanting Erosion Primary Production Estuaries Salinity ABSTRACT (Continue on reverse side if necesaary and identify by block number) Techniques were developed for propagation of Spartina alternitflora Loisel., smooth cordgrass, in the intertidal zone on dredge spoil and eroding shorelines. Both seeding and transplanting methods were successful. Trans- plants proved to be more tolerant of rigorous conditions such as storm waves and blowing sand, but seeding was more economical and was successful on protected sites. Vegetative development of seeded and transplanted areas was rapid with primary production equal to that of a long established marsh by the second growing season. At the end of the first growing season, FORM DD | jan 73 1473 = EDITION OF 1 Nov 65 1S OBSOLETE UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered) UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE(When Data Entered) 20. Abstract (Continued) more plant cover was produced from seeding at the rate of 100 viable seeds per square meter than from transplanting single-stem plants on a 0.9-meter spacing. The relationship of mineral nutrition to productivity of S. alterntflora was determined. Plants and soils in natural stands were sampled and analyzed for productivity interrelationships using multiple regression techniques. Salinity of the soil solution, plant and soil manganese concentrations, and plant sulfur concentrations were negatively associated with aboveground produc- tion. Variables positively associated with production included phosphorus concentration in the plant tissue and in the soil. Fertilizer experiments showed that the production of a natural stand of S. alterniflora growing on sand was increased significantly by additions of nitrogen and increased three- fold when both nitrogen and phosphorus were added. The production of natural marsh growing on finer-textured sediments doubled when nitrogen was added, but there was no response to phosphorus. Nitrogen and phosphorus fertilizers also enhanced growth of transplants and seedlings on sandy dredge material. UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE(When Data Entered) PREFACE This is a report of research started in November 1969 with support from the U.S. Army, Coastal Engineering Research Center (CERC) through Contract DACW72-70-C-0015, and continued through Contract DACW72-72-C-0012. This effort has also received support from the North Carolina Sea Grant Program, Office of Sea Grant, National Oceanographic and Atmospheric Administration, Department of Commerce, Grant No. GH-78, GH-103, 2-35178 and 04-3-158-40; North Carolina Coastal Research Program; and the North Carolina Agricultural Experiment Station. The authors are W.W. Woodhouse, Jr., Professor of Soil Science, E.D. Seneca, Associate Professor of Botany and Soil Science, and S.W. Broome, Research Associate in Soil Science, North Carolina State University, Raleigh, North Carolina. The authors express appreciation to personnel of the Cape Hatteras and Cape Lookout National Seashores, National Park Service; the U.S. Army Engineer District, Wilmington; and the North Carolina Highway Commission for their cooperation; and to A.W. Cooper for his advice and encourage- ment in the initiation of these studies. Special thanks are due D.M. Bryan, €.L. Campbell, Jr.,.and U.O. Highfill for their assistance in all phases of the field work; and to R.P. Savage, Chief, Research Division, CERC, for his advice and encouragement throughout the course of these studies. NOTE: Comments on this publication are invited. Approved for publication in accordance-with Public Law 166, 79th Congress, approved 31 July 1945, as supplemented by Public Law 172, 88th Congress, approved 7 November 1963. JAMES L. TRAYER Colonel, Corps Engineers Commander and Director an ian 40b i tie a “ ae Hey Ta n » Ey 3 af tite t3 ota ‘ CONTENTS Page PN DRO DW GION) erie, vseercciee tevehewareuee nen tat det lsd! ev itellewaneeunes ier te. leelulew sie es ial eee Generali sinc are Pave Aaey Veer etl co benmonn es ial 2. Biology of Spartina alterntflora mt eeee Be Rh setae 12 3. Characteristics of the North Carolina Goat ash RE ce 14 WIE A DNOLG IDOI Riss 50-5 y) toed i AG Siva Guia: Momrolatol oehOseaabow-biab eucuroulonsd 14 IIE PINORAGNITION o's woh oor eo od so 6 Joe Yoo" fa of bolo 6 6 IS) Iho TSREINS LEI ETN SONS lout G omtic) omomuna oon di delows (oc 19 27, =» OCCA Wo. =. Go Oauo ota bl tomtom sara bo=Gh On OnsOaa Ona Ome 42 Sa Sues Requirements Oe O-oOMB ONG OO GO 2G 6. o O 6 7 IY AIRS SU DENA ILONAWIBINEL = GG 6 On Gud 1d 0 oO ow pio det oro oma “lo 6 76 Vi SHORE SPROTEGLIONPAND! SUBSTRATE STABIEIZATION® = 39s so 6. 90 Pa STNO WAUSEM GUCAarn | erate tenet Met nneot reer Mra Cant Soni Meat tee te watcha. ot 90 Di Cedar sESihAN Cites fs tteutis: Tiss ost het Cnet eee ee ere tien 99 VI THE RELATIONSHIP OF MINERAL NUTRIENTS TO PRODUCTIVITY OF SPARTINA ALTERNIFLORA ... Meee goer aro erase emt a eeNG 102 1. Nutrient Status of Necunall Stands SMU They ou tosses 104 De Sere CES One Gautad llstz Creare tan earn cer men reat Semetl tl sin te esiere: sie 123 VII PLANTING SPECIFICATIONS - SPARTINA ALTERNIFLORA ...... 147 hg.) Wace lembalNe: Bo er tad oS 7606 emia Ao Waco eM eNaNIeNS co 147 ho, WilenESNG- DAKO or Si'te Ss OSS VES a CEN oS SNE! Ge oo. 148 Sane etal 2 atehOMi. wefan cm carinc sw tanger... Sua vsucle, tase aver 6) Zs 148 APES COOLING Wu ecto ca etc ian Waumrsie earaiete cnc um tl Nes: loko ay ie Nua) logeeh weno Ts 148 Wale lelseeOMEIE RGSS Pir CAE Sa tran a mstkceniciatcl aia onsen isch, vouuroMzeWe el Norge casa ou yuciult stamens) ove 149 Xa SO UMMARYAFAN DE CON GIEU STON S wars ares etolmu siren oti tcpurcilircn Comirstit circle) ome ils 150 LUA NOIRE CIM Soe 6 a MMO 6 66 lo loom Bho W8 6 Gg a 76) 6 152 TABLES mse xXperimentalll (SaltCSaeu neck asus Eset 6 GM chases) ciieicniwe 3: toys 16 2 Comparison of Growth of Four Transplant Types S. alternitflora Collected and Transplanted at Drew GINLEE 6 gia Boo ol! eS %e bo a tela (6 dS, Mow ames Mele Matic 21 3 First-Year Aerial Yield of S. alterntflora Transplants from Four Locations Grown in a Flooded Nursery, Gay tonaNomth es Carolina 3 sie icpyirtcs pious) tse / oO B Pe OO @ Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sept Oct Planting date (month) Figure 10. Effect of transplanting date on first-year growth of S. alterntflora. 41 Transplanting By hand - men working in pairs 180 hills per man per hour Machine - tractor; two-row 360 hills per man per hour planter with six-man crew (one on tractor, four on planter, and one supplying plants) Requirements on hectare basis For 0.91 by 0.9l-meter spacing 12,000 hills per hectare or approximately 111 hills pee L OOO Sel ste Hand digging and planting 134 man-hours per hectare Machine digging from 63 man-hours per hectare nursery area and machine planting These estimates may reflect higher productivity per man-hour than would be realistic for large-scale operations. On the other hand, we feel the digging operation could be speeded up appreciably with experience in handling larger volumes, and by further development of harvesting equipment. 2. seeding. There are reports in the literature that S. alterniflora produces very few viable seed (Chapman, 1960; Larimer, 1968), and that seeds are not aS important as rhizomes in spreading this grass. However, early in the studies, seedlings appeared to be the primary means of natural colo- nization of S. alterntflora on freshly deposited sediments in the inter- tidal zone in North Carolina estuaries (Fig. 11). Seed germinate in March and seedlings grow rapidly through the summer, producing flowers and seed by the end of the growing season (October). Seedlings are often numerous in the debris of drift lines at high water and may also be present at lower elevations, particularly in protected areas. If natural seeding is an important means of spreading S. alterniflora, it seems logical that direct seeding would be a suitable method for artificial propagation. This method would have substantial economic and labor-saving advantages over transplanting. To use direct seeding, effi- cient techniques for harvesting, processing, storing and planting seed had to be developed. a. Seed Production. The amount and quality of seed available for harvest varies from location to location and from year to year. As would be expected, the most vigorous plants produce the best seed supply. Such plants are generally found in areas recently colonized by S. alterntflora, 42 ‘yooTq queTdsuer} 1evak-Z 0} JUS De pe SSUT[pse9s JO pueys asuog *¢L61 ABW IT ‘43ND S,mMoUuSs *spoos Aq UOTSBAULT [einjeN II oan3ty and seed production is rather uniform over the entire area. In older marshes, there is little flowering and seed production in thick stands of the short-height zone, but some seed are produced by the tall form along creek banks. In such places, harvesting seed is inefficient because the total area is small. Since the quality and quantity of seed is variable, inspection of potential harvest sites is necessary each year. Variations in rainfall and other climatic conditions apparently affect seed production. Infes- tation by flower beetles (family Moredellidae) may also reduce the seed crop in some areas. b. Harvesting and Processing. An adequate supply of seed for small- scale experiments can be obtained by simply cutting the seed heads by hand. For large field plantings, a more efficient means of harvesting large quantities is desirable. To accomplish this task, a mechanical harvester was developed which consists of a sickle bar blade, a reel, and a canvas bag or tray for catching the seed heads. The apparatus was mounted on a two-wheel garden tractor (Fig. 12). The machine works best in- large areas of seed heads of uniform height. After cutting, seed heads were wrapped in burlap sheets and returned to the laboratory where they were stored temporarily in a cold room (2° to 3° Centigrade) until they could be threshed. A threshing machine, previously used for small- grain plot work, was used to separate the spikelets from the straw. This leaves nearly all seeds still in the spikelets (glumes, lemma, and palea). The threshing procedure reduced the storage space required for keeping the seed over winter (Figs. 13 through 16). Germination studies indicate that seed should be harvested as near maturity as possible, but it is often necessary to compromise on complete maturity since many seed may be lost due to natural shattering if harvesting is delayed too long. c. Storage and Laboratory Testing. A study by Mooring, Cooper and Seneca (1971) showed that 52 percent of S. alterntflora seed germinated when subjected to a 18° to 35° Centigrade diurnal thermoperiod after storage in sea water at 6° Centigrade for 8 months. Seed stored dry during the same period did not remain viable. In the beginning of our studies, we tested the effect of several storage treatments on germination. Seed of S. alterntflora were collected from five locations along the North Carolina coast (Fig. 17) and samples from each location were subjected to the following storage treatments: (1) submerged in estuarine water at 2° to 3° Centigrade, (2) submerged in distilled water at 2° to 3° Centi- grade, (3) suspended over water on screen wire at 2° to 3° Centigrade, (4) frozen dry, (5) frozen in estuarine water, and (6) freeze-dried. Salinity of the estuarine water was between 20 and 25 parts per thousand. Germination was tested in February 1970 by placing 50 seeds, which were disinfected by soaking in a 25 percent Clorox solution for 15 minutes, on moist filter paper in a petri dish. Three replicates of seed from each treatment and location were prepared with a replicate consisting of a petri dish of 50 seeds. Careful selection was made to be reasonably sure a seed was present within each spikelet. The petri dishes containing the 44 45 Cutter bar, reel, tray and motor mounted on two-wheel tractor. Seed harvester. Figure 12. Figure 13. Threshing seed. Thresher in center foreground; seed heads in burlap sheets at left fore- ground and right background. Figure 14. Seed in spikelets mixed with pieces of stems, and seed heads coming from thresher. 46 Figure 16. Seeds being separated from broken stems and unthreshed seed heads using a motorized screening device. Threshed and cleaned seeds in plastic containers; containers ready for filling with salt- water and storing in the cold room. 47 EASTERN NORTH CAROLINA 36° N. Lat. a BEE NOTE By ) 5 aia co N. Lat. A § i BEAUFORT SURF CITY 34° N. Lat. g OAK ISLAND Figure 17. Location of seed collections in North Carolina. seeds were placed in canisters to exclude light, and were subjected to an alternating thermoperiod of 7 hours at 35° Centigrade and 17 hours at 18° Centigrade in a growth chamber. Germinating seed, indicated by the emer- gence of the epicotyl, were counted after 5, 7, 9, 21 and 30 days. Since none of the seeds subjected to freeze-drying germinated, the results of this treatment were deleted from the statistical analysis. Germination failure of freeze-dried seed is consistent with the findings of Mooring, Cooper, and Seneca (1971), and showed S. alterntflora seed must remain moist to retain viability. Results of the effect of storage treatment on germination indicate that freezing, either dry or in estuarine water, was clearly detrimental to germination of seeds from all locations (Fig. 18). Freezing was particularly harmful to seed from Surf City and Oak Island. Germination of seed from Oregon Inlet, Ocracoke and Beaufort, which were stored frozen was fair; however, germination was delayed. This indicates that the after- ripening or development process of the seed was retarded while the seed were frozen. 48 @Estucrine water Distilled water Over woter O Frozen dry 10! Oregon Inlet 4 Frozerin estuarine water LSD .05-7.5 ($0.05-12.2 LSD 05-148 Germination (x) Surf City LSD.05-14.9 804 Ook Island LSD .05-173 5 ‘ 10, 5 20 25 30 Germination period (days) Figure 18. Effect of storage treatments on germination of S. alterni- flora seed from five locations. 49 The effects of storage of seed in estuarine water, distilled water or over water on germination were more variable between locations. Seed from Oregon Inlet showed no statistical difference in percent germination of seed among these three treatments at the end of 30 days. Germination of seed stored over water did lag for the first 10 days of the germination period. Storage over water produced the best germination percentages for seed from Ocracoke and Oak Island, while storage in estuarine water was Significantly better than any other treatment for seed from Beaufort and Suita Grits. The explanation for the variable response to storage treatment of seeds collected from different locations is probably due to the degree of maturity of seed at the time of harvest. Spartina alterniflora seed apparently are never dormant but continue development during an after- ripening period. The degree of seed development at the time of harvest, as well as the environment in which the afterripening proceeds, probably influences viability of the seed. Although there is variation in seed maturity even within a particular stand, flowering and seed maturity occur earlier along the northern coast of North Carolina with about a 3-week span from north to south. The seed collections from the different locations used in the experi- ment were made within 3 days; therefore, seed from the northern coast, e.g., Oregon Inlet, were more mature than those from the southern coast, e.g., Oak Island. Consequently, the difference in the effect of storage in estuarine water, distilled water or over water on germination was least in the seed collected from Oregon Inlet. Storage over water was advantageous for seed collected at Ocracoke and Oak Island. Apparently these seed were less mature at harvest and storage in a saturated atmos- phere, but not submerged, was more conducive to the afterripening process. This study of germination indicates that seed should be harvested as near maturity as possible and that storage in estuarine water or possibly freshwater at 2° to 3° Centigrade is an acceptable and relatively easy way to maintain viability over winter. However, it is often necessary to compromise on complete maturity, since many seed may be lost due to natural shattering if harvesting is delayed too long. At Oregon Inlet the best harvest period has been from about 20 September to 20 October. The best harvest date is later farther south. The effects of date of harvest, length of afterripening period, and storage in distilled or estuarine water were evaluated with seed collected in 1971. Seed harvested on 28 September 1971 at Oregon Inlet had a Significantly higher germination percentage than those harvested 1 week earlier (Fig. 19). However, seed harvested on 21 September were threshed and stored in estuarine water within a few days of harvesting, while those harvested on 28 September were stored in burlap sheets in a cooler for 3 weeks before threshing and storing in estuarine water. Subsequent experience indicates that storing the seed at 2° to 3° Centigrade for several weeks before submerging in water enhances the germination percentage. 50 Germination (%) 100 80 60 40 GHarvested 28 Sept 1971 OHarvested 21 Sept 1971 20 LSD.05-12.5 0 4 8 2 16 20 24 28 Germination period (days) Effect of harvest date on germination of S. alterni- flora seed stored in estuarine water. Seeds were harvested at Oregon Inlet; the germination period began 17 February 1972. 51 Results of a study on the effect of length of storage on germination indicate that immediately after harvest, seed were slow to germinate, requiring about 24 days to reach 50 percent germination (Fig. 20). By February, 50 percent of the seed sample germinated within 4 days. The results suggest that an afterripening process occurs that speeds germi- nation and increases the germination percentage as the length of storage increases. Similar results were obtained by Van Shreven (1958) in a study of S. townsenditt seed. It is difficult to evaluate the effects of length of storage on germination beyond 6 or 7 months since the epicotyl of a large percentage of seed emerges even though stored in salt water at 2° to 3° Centigrade. Experience has indicated that seed stored longer than 1 year do not retain their viability. A study to develop techniques to increase longevity of storage could prove beneficial. Samples of seed harvested 21 September 1971 at Oregon Inlet were used to compare storage in estuarine water with distilled water. After 2 months storage, there was no difference in germination between the storage treatments (Fig. 21). However, after 5 months, when the after- ripening process was apparently complete, seed stored in estuarine water had a significantly higher germination percentage than those stored in distilled water. At least in some cases, storage in salt water enhances seed germination. d. Planting Methods for Field Experiments. Several methods of planting and incorporating seed have been used. In the first field experiment with seed (1970), the plots consisted of three rows of seed planted in furrows 1.0 meter apart and 15.2 meters long. Since the seedlings were confined to narrow rows, little growth per unit area was produced during the first growing season (Tab. 14). The results of this initial experiment indicated that seed could be successfully used to establish S. alterniflora and that seed should be distributed evenly over an area to give a better plant cover during the first growing season. In 1971, comparisons were made between the performance of seed applied to the surface in a clay slurry (attapulgite) and seed broadcast by hand and incorporated 1 to 4 centimeters into the substrate. At Oregon Inlet, seed applied to the surface in the clay slurry produced a greater amount of aboveground growth than those which were covered (Tab. 15). The seed in the clay slurry germinated earlier and got off to a faster start since temperatures on the surface were probably higher and the seedlings did not have to emerge through a layer of sand. However, at Snow's Cut, seed planted in the clay slurry produced almost no seedlings. There is a greater tide range at Snow's Cut, and due to the regular flooding the clay slurry containing the seed did not remain in place long enough for germi- nation and rooting to occur. In adjacent plots, where seed were thoroughly mixed with the substrate by raking to a depth of 1 to 4 centi- meters, there was a good stand of seedlings which produced very good growth (Tab. 15). It appears that broadcasting the seed and covering to a depth of 1 to 4 centimeters is the best method to ensure a good stand of S. alterntflora seedlings over a wide range of conditions. 52 100 Germination (%) Figure 20. cr A 21 Jan 1972 ~ 17 Feb 1972 19 Dec 1971 0 17 Nov 1971 20 Oct 1971 LSD .05-9.8 4 8 12 16 20 24 28 32 36 Germination period (days) Effect of afterripening of S. alterniflora seed stored for various lengths of time. Seed harvested 28 September 1971 at Oregon Inlet and stored in estuarine water until germi- nation was begun on the dates indicated. 53 Table 14. Mean Standing Crops Produced from S. alterniflora seed in Two Growing Seasons.* Dry Weight (g/m*)+ 22 Sept. 70 B Genes 7 Mo® 7.0 WS Plant Component Shoots Rhizomes *Planted in rows 1 meter apart at Oregon Inlet, 21 April 1970. +Means of three samples No data taken Table 15. A Comparison of Growth Measurements of S. alterntflora Seedlings from Two Seeding Methods. Number Flowers/m Clay slurry Raked Snow's Cut (seeded 24 Mar., harvested 15 Sept. 1971) Raked (Plot 1) 1,236.8 260.0 480.0 116.0 Raked (Plot 2) 685.6 120.0 388.0 116.0 *Means of three replications 54 100 DO Estuarine water 80 © Distilled water se 6 60 17 Feb 1972 3 £ E 40 17 Nov 1971 S 20 ) O 4 8 12 16 20 24 28 Si Germination period (days) Figure 21. Comparison of germination of S. alterntflora stored in estuarine water and distilled water. Several methods of incorporating seed were used to replace raking for larger scale plantings. Where access is available, a farm tractor may be used. We have used a standard farm tractor equipped with dual wheels for added flotation and traction. Seed were broadcast by hand on the surface and covered by six narrow sweeps mounted on the tool bar of the tractor, followed by a spiked-tooth harrow (Fig. 22). Seed were also incorporated without the spiked-tooth harrow by going over the area with the sweeps, broadcasting the seed (Fig. 23) and plowing again with the sweeps. Mechanization of distribution of the wet seed would be difficult. However, spreading by hand works very well. Seeds placed in tubs or buckets should be drained of excess water and sufficient dry sand mixed with them to cause the individual seeds to separate (Fig. 24). They are then ready to be broadcast. Many areas are not accessible to a full-sized tractor. Two portable mechanical methods have been utilized in these areas. For small areas, seed may be broadcast (Fig. 25) and incorporated with a rototiller (Fig. 26), 55 AM: RP ALLEN DREAD Figure 22. farm tractor incorporating seed on dredge spoil near Beaufort, North Carolina. Figure 23. Broadcasting seed after preparing a seedbed with tractor and cultivators. 56 Figure 24. Mixing dry sand with wet seed to improve distribution when broadcasting. Figure 25. Broadcasting seed and sand mixture on a small experimental plot on a dredge island near Snow's Cut, North Carolina. Sf *LOT[1IIOJOL & YIM poss Butzerodso0duy "97 o1nsTy 58 but this technique is too slow for larger plantings. A faster method was developed which utilized the same two-wheel garden tractor, equipped with dual wheels, which was used for harvesting seed to pull a cultivator that consisted of. six small sweeps about 25 centimeters apart (Fig. 27). Furrows were opened with the sweeps, seed were broadcast and covered by a second trip over the seeded area. e. Seeding Date. Seed germination under natural conditions begins as early as March on the North Carolina coast. However, when seeding S. alterntflora, delaying the planting date until mid-April lessens the chance of damage due to weather. The probability of storms which produce damaging wave action decreases later in the spring. Earlier plantings produce more growth by the end of the growing season, but a compromise must be made between early planting and increasing the risk of being washed out or buried by wave action. A comparison of two planting dates was made at Beaufort in 1972. About 0.5 hectare was seeded 11 April, but part of the seedlings were washed out by a storm in May. A second seeding was made 21 June. Samples of aboveground and belowground standing crops harvested 5 October 1972 from an area about in the middle of the tide range, indicate that the April seeding was more productive the first growing season (Tab. 16 and Fig. 28). After the second growing season, differences were less striking. Table 16. Mean Aboveground and Belowground Standing Crops Produced from Two Seeding Dates at Beaufort, North Carolina in 1972 Seeding Date Aboveground Standing Belowground Standing Crop (g/m*) by Year Crop (g/m?) by Year 5 Octeelo72 dle Septem lO Simeon Oct. 1972 ia Sept.. 1973 354 644 541 1,077 56 527 176 Three seeding dates were compared in the spring of 1972 at Snow's Cut. Seed were planted on 16 March, 10 April, and 10 May; yields of shoots were 388, 304 and 116 grams per square meter, respectively. These results indicate that there is very little difference in growth between the March and April seeding; therefore, risk of weather damage can be reduced by delaying planting until April without sacrificing growth potential. Delaying until May reduced first season growth greatly. 11 April 1972 21 June 1972 f. Seeding Rate. Since the number of viable seed produced varies from year to year and between locations, it is desirable to estimate the number of viable seed per unit volume for each lot of seed that is avail- able for planting. This can be done by simply measuring a small volume of seed, germinating them, and counting the number of seedlings produced. 59 60 t accessible to the farm tractor. ions no t locat ing a d in see Garden tractor used Figure 27. MS Photo taken Bala Noweats OVD (Gealgalme)).6 S. alterntflora seeded 21 June 1972 (left) Compare with Figure 22. 11 November 1972. Figure 28. Experience has shown the number of viable seed per liter varies from as low as 500 to as high as 27,000. A rule-of-thumb estimate generally used is 10,000 viable seed per liter (10 per milliliter). When planting, the volume of seed applied per unit area can be adjusted according to the results of the germination studies. The rate used as standard is 100 viable seed per square meter. This rate can be adjusted up or down, depending on amount of seed available and the exposure of the site. Fewer seed are necessary in a protected area; heavy seeding rates increase the chances of successfully establishing seedlings in areas exposed to wave action. g. Elevation. The elevation range over which seeding is an effec- tive means of establishing new stands of S. alterntflora is less than the elevation range of transplants. Seedlings are not as hardy as transplants and are not as able to tolerate the rigorous conditions of inundation and wave action characteristic of the lower elevations of the intertidal zone. Observations at Beaufort and Snow's Cut illustrate this point in a quan- titative manner. At Beaufort, an average elevation of several points along the edge of a natural marsh near the seeding experiment, indicated the lower limit of growth to be 0.43 meter above MLW (mean low water). However, seedlings produced by planting 11 April survived only as low as 1.02 meters above MLW. The elevation of the upper edge of the natural marsh was 1.15 meters above MLW. Therefore, the part of the elevation range normally occupied by S. alterntflora which was effectively colonized by seeding was between 1.02 meters and 1.15 meters or a range of 0.13 meter. This amounts to about 19 percent of the 0.72 meter elevation range of the grass at Beaufort. Tide tables list the mean tide range at Beaufort as 0.76 meter. Seedlings produced from seed planted 21 June had an average lower limit of survival of 0.93 meter above MLW. The area colonized represents about 31 percent of the elevation range of S. alterntflora. Apparently the late-planted grass was able to survive at the lower elevation because it was not subjected to storm-induced wave action as was the earlier planting. Seedlings were successfully established over a larger part of the elevation range on a dredge island in the Cape Fear River near Snow's Cut, tide range 1.2 meters. Transplants survived from 0.25 to 1.41 meters above MLW or a range of 1.16 meters. Seedlings survived down to 0.79 meter above MLW which is a range of 0.62 meter or about 54 percent of the elevation range at this location. h. Protection from Blowing Sand. Covering of seedlings by windblown sand appears to be a primary cause of failure of natural S. alterntflora on sandy dredged material. If there is sandy material above the inter- tidal zone, sand can be blown from these higher elevations until it becomes armored with shell or covered with vegetation. When establishing S. alterniflora on these sites by seeding, it is necessary to protect the area with a sand fence, a vegetation strip such as Ammophila breviligulata Fern. (American beachgrass), S. patens (Ait.) Muhl. (saltmeadow cordgrass), Panicum sp., or preferably a combination of fencing and vegetation. 62 The experimental planting site at Beaufort was on the northeast side of a large sandy dredged pile. Consequently, southwest winds periodically moved large amounts of sand onto the planting site. Sand fences were erected to prevent the planting from being smothered (Fig. 29). After the first fence (1.2 meters high) was completely filled, it was necessary to erect a second (Fig. 30) and finally a third fence. A strip of Panicum amarulum Hitchc. §& Chase (silver bunchgrass) was also seeded to intercept the blowing sand. (Seed were provided by the U.S. Department of Agriculture, Soil Conservation Service, Cape May Plant Materials Center, Cape May Court House, New Jersey.) A similar seeding on the southwest side of a large dredged pile on the Ocracoke side of Hatteras Inlet was completely destroyed by drifting sand. This was seeded 13 and 14 April 1972 in the same manner as the Beaufort planting; by mid-May, the seedling stand appeared comparable to that at Beaufort. However, no sand fence was installed on this site before the storm of 24 to 27 May, and the entire seeding was smothered by the deposition of 15 centimeters of sand. Failure to protect seedings from drifting sand can be critical under circumstances such as described for these two sites. This experience emphasizes the potential of wind erosion in transporting sand from dredge material back into the estuary. By stabilizing the dredge material above the tidal zone with vegetation, a Significant amount of refilling of the estuary may be prevented. Sand deposition or erosion by waves or currents are also major threats to seedling survival on exposed sites. If suitable structures could be devised to protect such sites from wave action during the establishment period, seeding might be feasible in many more locations than it is now. i. Introducing Natural Seed Sources. The encouragement of natural invasion of new areas through establishment of small ''seed patches" might be the most economical approach in some situations. Both seedlings and transplants of S. alterniflora almost invariably produce seeds the first year and large numbers of seedlings have been observed among transplants during the spring of the second growing season (Fig. 31). Small plantings were made for this purpose at several locations. The most successful was on a large (2 or 3 hectares) dredge island near Old House Channel. A small area was transplanted in May 1971 on the shore- ward edge of a flat on the southeast shore of the island (Fig. 32). There were no other plants of this species evident on this or the adjoining islands at that time. About 26 months later, S. alterntflora occupied an area of about 0.5 hectare, on each side of and shoreward from the original planting (Fig. 33), which could no longer be identified. This method of colonization may be appropriate where resources are limited or where rapid coverage is not required. j. Large-Scale Seeding. An island, (which we call South Island), comprising in excess of 10 hectares, has been developing for the last 18 months just south of Drum Inlet and several hundred meters west of the barrier island. This island appears to result from sand coming in the 63 “ZL61 ysnsny GT uoyez ooyd $7761 [tady ST poqooug ‘pues 3utmotq WOLF SBUTTpses 399}0Ld 0} JLoFneog Je pojd9LO sadUaF pues 9014} FO YSATF OUL "67 ounsTy 64 'pajder9 useq pey eouez puodes ve pue [[nF Aloqzetdwoo sem oouez YSATF OUI SLET ITtady ZI Ad ‘OS arns Ty 65 *uoseos BUTMOIS YSITF 94} SqUeTdsueIy ay Aq poonpoid o19mM spo9g ‘*Si9q,USD L9ZOUW-T6°O UO ZL6T [tady st poquetdsuer, sem YOTYM JOTU] wNIq TeoU Bote Ue UT YIMOIS SWOZTYI zreoX puoses suowe (¢Z61 [t4dy) sdut[pees "Tg omnsty 66 °1SOM ST MOTA “TZ61T ABW TT Uo punos oot [Weg ‘Tauuey) osnoy PIO seeU Diol fiusez7o *g Butjuetdsueaz]l ‘*z¢ ain3 ty *yanos ST MOTA AR Santee: “(fZ61 AIne ¢Z) syzUOW 97 10}}Ze popods sem vote BUTpUNOAINS dU] shia "es oin3Ty 68 inlet and accumulating around a small (1 hectare or less) spoil island deposited during the course of opening the original channel. An island in this vicinity was seeded in early May 1972, and the resulting seedlings destroyed in the storm of 24 to 27 May of that year. By early April 1973, an estimated 6 to 7 hectares of South Island lay within the upper half of the tide range, the elevation zone in which we have found seeding of S. alterniflora to be feasible. This situation presented a unique opportunity to: (1) undertake field-scale seeding using our portable equipment under the type of conditions for which it was designed, and (2) test the feasibility of seeding a very exposed area where little opportunity for natural invasion by marsh species seemed likely in the immediate future. - The low elevation and closeness to the inlet subjects the area to frequent flooding and strong turbulence. The orobability of seeds drifting onto this site at the appropriate time for germination and subsequent seedling establishment seems rather remote. Seeding was delayed until the week of 16 April to reduce storm hazards. About 4 hectares were seeded on 17 and 18 April using the two-wheel tractor with cultivator. The area to be seeded was cultivated before seeding and again immediately after seeding. Seed were broadcast by hand aS previously described. On 2 May, a second area of about 1 hectare was seeded in the same manner. Germination and emergence were excellent over most of the area; by late May, an adequate stand of seedlings had survived over an area of 3 to 4 hectares. However, the island continued to grow and many seedlings were smothered by sand that moved over the island during the summer. By the end of the growing season a 2-hectare block lying roughly across the westerly one-fourth of the island still retained an adequate stand with scattered plants remaining over another 1 or 2 hectares. Rainfall was below normal for much of the summer in this region which when coupled with the low, flat nature of the island, made the seeded area vulnerable to salt injury (see discussion in Section 3c on Salt Damage). Salt damage was believed to be the cause of both the stand thinning and the slow rate of top growth of this planting. An area near the center of the 2-hectare block (Fig. 34), devoted to a fertilizer test, was sampled 8 November 1973 and the data are presented and discussed later in Section VI and Table 43. Top growth was quite restricted, much less than that of the seeding of 21 June 1972 at Beaufort (Tab. 16). However, root and rhizome production was equal to or better than the Beaufort planting. These results suggest that the periodic salt-induced dieback observed aboveground is not necessarily matched by losses in underground growth. This does not seem too surprising since the large mass of succulent roots and rhizomes underneath established stands probably plays a significant role in their tolerance to salt buildup. This planting was in good condition on 8 March 1974. The island is still growing to the south and east with some erosion along the north- west side. Additional sand has been deposited over most of the seedlings, 69 d “PL6L YdteW g Udye OJOY 61 [tady 8[ popess ‘“pue[s] yInog uo Sutpses ve WOLF sWIND oWOZTYY “PE daIN8TY “CL t RES ATTRA but many of them are already sending new shoots through it. Unless drastic shifts in erosion or deposition rates occur within the next 2 or 3 months, an adequate stand of well established vegetation is expected over the area this spring. This is interpreted to be moderately success- ful and to indicate that seeding under such conditions offers promise of rapid and economical stabilization in some cases. k. Cost. The cost of propagating S. alterntflora from seed is reasonable and not unlike that of agricultural crops, except for diffi- culties in gathering seed, and for access of equipment to some planting sites. The time required to harvest a known amount of seed was determined at Oregon Inlet in September 1972. One man operated the harvester, while two others removed the cut seed heads from it at the end of each round. Three sample areas were harvested (Tab. 17) to quantify yield of seed per unit area and time required for harvesting. The variability of yield is demonstrated by comparing the first sample area to the other two sample areas. The volume of seed per unit area was about five times greater on the first sample area than on the others, even though it was much smaller. Variability of this magnitude in the seed crop of S. alterntflora is common. Consequently, it is difficult to predict the resources necessary to harvest seed at different locations or years or even different areas within the same stand. In the sample, about 5 man-hours were needed to harvest enough seed to plant 1 hectare. The harvested seed were threshed to reduce the space necessary for storage. Threshing required about one-half as many man-hours as harvesting. The cost of storage was negligible, since refrigeration facilities were available. The amount of time required for planting depends on the equipment available. Using a two-wheel garden tractor for preparing the seedbed and covering the seed after they were broadcast by hand, 4 hectares were seeded by 3 men-in about 10 hours. This amounts to 7.5 man-hours per hectare. In addition to the times listed there are other variable costs for transportation to the sites, fuel, etc., which are difficult to estimate. Fixed costs (equipment) are also difficult to estimate. In this case, most of the equipment used was modified from that already owned by the Soil Science Department, North Carolina State University. 3. Site Requirements. a. Elevation and Tide Range. For propagation of salt marshes by either seeding or transplanting, care must be taken in selecting or preparing sites which meet the requirements of the species used. The interactions of such factors as tide range, elevation, slope and salinity determine the species of plants present and influence their vertical zonation in marshes. The vertical range of S. alterntflora is generally stated to occur from about mean sea level to mean high tide. There are many exceptions to this generalization. Variations in vertical 71 *93e107S LOF OEM UT UOTSIoWS 0} LOTId pasos poyserty} TOF o1e SZYsTOM pue SoOUMTOA, TeqOL (9B 2°) Grr (OB GW) S20 000‘ 009 8°ZI 002 0°09 (er S750)" 0S 40 (88 BL) S°O 000‘ 0S 9°IT 061 SITS (ae 2-0)" 8220 0 000°008 9°9T 000‘T 0°08 (98 Z°0) 80°0 (s194TT) (sanoyuen) (3y) (ey/T) (ey) pepess Pees eTqetA aq pr[nos Jo ToqunNn eoLly YoOTYyM eorly azeutxoarddy ouT | ~TYsTtomM PLOTA ,OUNTO/ po seALeYy ZL61 Lequeades ut joeTU] uoseIQ 4e seoty oTdues se1Y] WOLF YSeATeY pees “LT 2TqeL 72 range occur in North Carolina estuaries where tidal fluctuations are dominated by wind direction and velocity and where salinities are low. At our Snow's Cut experimental site the salinity ranges from 7 to 10 parts per thousand. Under this low salinity, freshwater plants have become mixed with the S. alterniflora from about half way in the intertidal zone upward. In completely freshwater, plant species adapted to freshwater would probably become dominant. In several North Carolina estuaries the range of periodic tides is quite low because of few, narrow inlets and the large area of the estua- ries. In such estuaries, surface fluctuations are greatly affected by wind direction and velocity, and consequently, changes in surface levels occur irregularly. This complicates the relationship between tide range and the elevation zone occupied by S. atterntflora. In many locations, S. alternitflora occurs as a narrow fringe at the water's edge. When planning marsh restoration, it is best to take elevation readings of the upper and lower limits of nearby natural marshes and plan to plant within this zone. In preparing a site, the area available for planting can be increased by making the slope as gentle as practicable without ponding of water. The more gentle the slope the larger the area which will be alternately flooded and drained. b. Substrate Texture. Several substrate-related factors affect propagation of S. alterntflora. Texture of the substrate as it affects bearing capacity is an important practical consideration. Substrates with high proportions of silt and clay are not suitable for conventional planting equipment. The Straits test site is the only place where this problem was found. Most of the dredged material in North Carolina estuaries is composed mostly of sand, and consequently has excellent physical properties. Additional opportunities to experiment with propa- gation methods on finer textured materials will be welcome. Sandy substrates are not without limitations. Sandy materials are inherently less fertile than silt and clay since fewer mineral nutrients are adsorbed. Experiments with fertilizers have produced increased growth through applications of nitrogen and phosphorus at locations where the substrate is sandy, such as Drum Inlet and Ocracoke Island. An exception is the Snow's Cut location where the substrate is sand, but nutrients are apparently supplied by the large amount of silt and clay sediments carried by the Cape Fear River and deposited in the marsh on each tidal cycle. At Ocracoke and Drum Inlet little deposition of fine materials has been observed. c. Salt Damage. Another substrate-related factor is salinity. Although S. alterntflora is exceptionally well adapted to growth and sur- vival under saline conditions, it can suffer serious salt damage. This damage we observed in several instances during the field studies. All observations were in the Pamlico and Core Sound region on sites subject to wind setup. Salt damage may occur in such areas any time that extended periods of low water coincide with periods of warm, clear weather. Under these conditions surface salinities build up rapidly and can be particularly 73 severe on seedlings and young transplants. Established plants seem to be able to tolerate much higher salinities at least for short periods of time. Core Sound and the southern part of Pamlico are particularly vulner- able to this phenomena since the southwesterly winds normally prevailing here during the warm part of the year result in low water levels. It occurs also in northern Pamlico where low water results from northeasterly winds. Heavy salt concentrations, even to the extent of a white crust on the soil surface, have been observed. This condition develops at Oregon Inlet, particularly in April and early May when northeasters of several days duration are not uncommon. At such times the effect on young seed- lings is quite severe. While seedling stands are occasionally killed completely by salt damage, the more frequent effect is a temporary dieback, thinning and stunting. The less severe effect may be due to an intermediate salt concentration or to shortening or interruption of the buildup period by rain showers or wind shifts. This probably happened at the field-scale seeding on South Island (Drum Inlet) during the 1973 growing season. Early seedling emergence on most of this site was good, and by early May a very dense stand seemed ensured. However, by early July many seedlings had died and the remaining plants were stunted and exhibited the general appearance we associate with salt damage -- dead leaves and varying degrees of tip burn of living leaves. Spot checks of salinity were made on about every site visit throughout the summer. All readings were approximately sea strength, although plant appearances suggested that some additional salt damage had occurred, probably on more than one occasion. When the planting was sampled 5 November for estimates of aboveground and below- ground production, salinity determinations were made at each sample site. The mean for the 48 samples was 40 parts per thousand (sound water = 35 parts per thousand in this vicinity). Stand losses that appear to be related to salt buildup have also been observed in well established plantings, occurring as small irregular spots. These have been identified at The Straits on recently dredged material, highly variable in texture, and at Drum Inlet on recently dredged material containing a high proportion of sand. The damage always appeared the year after transplanting, and was first observed at The Straits in the summer of 1971. Five sampling stations were established and these were sampled three times (Tab. 18). Wherever salinities of the soil solution exceeded 45 parts per thou- sand, dieback of S. alterniflora leaves was observed; in more severe cases, entire plants were dead. A similar pattern appeared at Drum Inlet in the spring of 1973 on plots transplanted to S. alterntflora in May 1972. Eight sampling stations were established in May and followed through August (Tab. 19). 74 Table 18. Soil Solution and Sound Water Salinity Measurements* in Dieback Spots at The Straits Date Salinityt Soil Solution? 28 July 1971 28 32 to 58 OSepte. 971 35 to 80 BQ) Oees: UUs 5), 160), DS) *Salinity determined in field with hand-held refractometer. +Parts per thousand tRange among five stations §Immediately after period of heavy rains Table 19. Soil Solution Salinity - Drum Inlet, 1973 Station Salinity* Vegetation Condition 16 Aug. Dead, soil bare Normal Dead, soil bare Normal Dead, soil bare Normal Sj. fon Cn SS Ss) TS) Normal Normal *Parts per thousand +Mean low water 75 At The Straits, it was theorized that the high salinity spots were related to the interlayering of very fine and coarse sediments that occurred at that site. This theory was abandoned when similar phenomenon appeared at Drum Inlet, a site containing almost no fine material. At Drum Inlet, the toxic areas were confined to the higher intertidal ele- vation within these plantings. This would be the zone left exposed most often, and it was exposed for days at a time during the summer. This does not account for the localized nature of the high salinity spots within this zone. Station 8 at Drum Inlet is in the "high" elevation zone, but appears very much out-of-place with salinities well below all other stations (Tab. 19). This station is immediately adjacent to a part of the island lying 1 meter or more above the intertidal zone. It is sandy and big enough to develop a freshwater bubble, such as occurs under dunes (Berenyi, 1966), and seepage from this elevated area into the vicinity of Station 8 continually dilutes the soil solution at this location. Freshwater seepage from sandy spoil piles can significantly reduce soil solution salinities in the intertidal zone. This could be important for plant establishment and survival in this zone in regions having substantial wind setup. On South Island the absence of any area high enough to permit the retention of freshwater was a handicap to the initial establishment of a seedling stand. IV. MARSH DEVELOPMENT The ultimate objective of planting S. alterniflora will usually be the initiation of a marsh, which has as one of its functions stabilization of the substrate material. Consequently, the rate at which a planting is able to achieve this objective is of interest. An area at Snow's Cut has been followed since planting 7 April 1971. Planting was by hand, one stem (culm) per hill, 0.91-by 0.91-meter spacing, using transplants dug from nearby natural stands. Their development has been monitored by sampling, counts, and measurements made annually in September, near the end of the period of major aboveground growth. A photographic record has been maintained, and elevation cross sections have been surveyed. The developmental pattern of such a planting can probably be best seen, through the early stages, in the pictorial record. The developmental sequence over the first 12 months for this planting is shown in Figures 35 through 39. Following spring transplanting, there were a few weeks during which the transplants developed new roots and new shoots emerged from the base (Fig. 35). This stage was followed by a period of rapid aboveground growth, as seen by the proliferation of new stems (center culms) around the origi- nal transplant, extension of rhizomes, and emergence of new stems from rhizomes (rhizome culms) at various distances from the transplant (Fig. 36). This stage lasted a little past midsummer after which flowering began and the season's aboveground growth approached maturity (Fig. 37). 76 pue poysttqeise [Tem szueTd mon *YOom YIYSTO 94} UT YMOIZS MoU UO BUTI Nd IZ61 eunc Zz usyez o.oyd ‘TZ61 [tady g poquetdsuesy noe ewer a Sf einsty 77 Sty} BSutanp uotsuedxs prdey “IZ61 Aine 8g uo L9JET SYIOM 78 *speoy poos pue sul[Nd oWOZTYL 9I0N “TL6T toquieydes pT oo a € uosedss SUIMOIS }SITF JO puso IedN "LE eInsty *queTdsuer} wWeqs oTSuTS e WOLF YIMOIS TeOA-ASITY *B8g sAN3TY Rows have disappeared. Starting the second growing season, 10 April 1972. Figure 39, By late September, seeds were mature and aerial growth had decreased. At this point, the stand consisted of large clumps of stems, 20 to 50 or more, centered around the original transplant with rhizomes of various lengths radiating from them and sending up new stems (rhizome culms) (Fig. 38). At this stage, the substrate was partially stabilized by roots and rhizomes but the cover was still quite open, leaving much of the surface bare. Seeds shattered rapidly following maturation in late September through mid-October, and aerial growth took on the color of straw and became susceptible to being broken and exported to the estuary by winter storms. Belowground growth continued as evidenced by new rhizome shoots emerging at the surface through the fall. Although not studied, it is evident from observations that a substantial amount of root and rhizome expansion goes on under these stands during the winter. By March of the second growing season, new shoots so populated the surface that the original hills and rows were no longer identifiable (Fig. 39). Following the flush of spring growth, the site appeared to be fully stabilized, or very close to it, as far as vegetative cover could go. Developmental data of this planting at Snow's Cut over the 3-year period are presented in Table 20. In September 1971, three fairly distinct zones of growth consisting of about the upper 25 percent of the vegetated slope, the center 50 per- cent, and the lower 25 percent were visible. Growth was best in the center of the slope and much poorer at both extremes. Consequently, the planting was divided into these three zones for sampling purposes for 1971. However, at the end of the next growing season, it appeared that four zones would be preferable and sampling afterward was done accordingly. It became evident, after 2 years, that for most measures the two center zones could be combined with little loss of information. Visually there is some difference, but this is primarily in the number of invading plants, and these are numerous enough to sample only in the original upper zone. A large increase in plant cover developed between the end of the first growing season and the end of the second. This is reflected par- ticularly in number of culms and in both aboveground and belowground dry-matter production, with increases ranging from 3- to 10-fold. These results probably indicate a concomitant increase in substrate stabiliza- tion and resistance to wave action of the planted area. By the end of the third growing season, some additional aboveground development was recorded (Fig. 40), but the real change was belowground (Tab. 20). While top growth increased noticeably at the higher elevations, root and rhizome production increased dramatically throughout. Such increase in root and rhizome mass should substantially increase the stability of the area. Although the belowground plant material was not separated by species, it was nearly all S. alterntflora except in the 82 *SOUOZ UOTIEASTO ANOF FO yoes worz so[dues ,W S7°Q ‘e014 ot0m soTdues ¢/ 6] *SOQUOZ UOTJEASTO INOJ Fo yous worz sotdues Ww ¢z7°Q ‘inofz otom sotdues 7/61 atdues 03 s£opeaut mez oops *SQUOZ UOTIBASTO 9914} WOIF sopti sutads Aq poateanod AToreg + SjuetTd [TenplATput 9014} ot0m soTdwes TZ6T soetoeds TIVv4+ eZTS etdures IZ61 Ittdy g pojuetdsues, 9°9T 09 O°ST 02 6°8 6°8 02 9°¢ OG OH EL61 tequeydes gt ZL61 tequeydes 61 (ey/3%) kep/iH wey szopeaut|eutiseds | prota PIOTA punoa3 poze punuy punozsaao0qy -MOTOg POTtog xSUOSP9S BUTMOID 9eLY], TOAO ND S,MoUS 4e SuTJUeT_ D107 f1U10470 *g ue JO JuswdoTeAeqg pue YyIMOIH ‘OZ STIPL 83 upper zone. Based on the first 3 years of stand development on this site, it is evident from the standpoint of vegetative cover, substrate stabili- zation, and primary productivity, that marsh development following trans- planting can be quite rapid. Productivity was 7,000 kilograms per hectare the second year and reached 10,000 kilograms per hectare in the third year which compares with 5,100 and 16,000 for the short and tall height forms in the long established natural marshes of Oak Island (discussed later in Section VI, Tab. 40). Cover, as reflected in number and size of stems, and production increased between the second and third year, but the rate of increase slowed. Belowground growth expanded much more than did top growth during the third year. However, these data may be decep- tive, since this may mean an accumulation of another season's growth added to that of the first 2 years. If this is the correct interpreta- tion, the annual rate of production belowground appears to be slowing also. There may be further increases in vegetative material in the root zone, but it is difficult to comprehend how it can continue to expand at the present rate without much deeper penetration into the substrate. We have seen no evidence of change in this respect from the first year. Distribution of roots and rhizomes by depths was examined at Snow's Cut in 1972. It was feasible to take cores to a depth of about 30 centi- meters, but only at low tide. There was little penetration of roots and rhizomes below this depth, and no adequate method of sampling below 30 centimeters was found. Cores almost invariably broke off at the point of sharp decrease in belowground growth which occurred around 25 to 30 centimeters below surface. Cores collected in 1972 were divided into the 0 to 10- and 10 to 30-centimeter depth segments, and the roots separated from rhizomes. The depth division was selected because belowground plant material was more dense in the upper 10 centimeters. About two-thirds of the roots collected are distributed in the upper 10 centimeters; the remainder occurring in the 10-to 30-centimeter zone (Tab. 21). Rhizomes were more evenly distributed between the two zones. There was a distinct tendency toward less total belowground growth as the period of inundation increased. Sampling variation is high with coefficients of variability of 37.5 to 84.3 percent. Consequently, differences would have to be quite large to be detectable. Estimates of rate of spread were obtained at Snow's Cut at the end of the third growing season. The lateral rate of spread was from 0.9 to 1.5 meters per year (Tab. 22; Fig. 41). Data on downslope spread were avail- able from only one plant source (Ocracoke) at one elevation. Since lateral spread for all plantings at this location was uniform across the four zones of inundation and of the same general magnitude as the single downslope expansion determination, we assume that the latter would be very similar to the lateral spread at all elevations within this range (2 to 12 hours inundation). 85 Table 21. Distribution of Belowground Growth* by Depths, Snow's Cut 1972 Inundation Zones | Dry Weight (kg/ha) Roots|Dry Weight (kg/ha) Rhizomes (hr/day) (cm) *Samples were four core samples (8.5 cm in diameter and 30 cm deep) from each elevation zone. One core was taken from each 0.25 m2 sample area. tLeast significant difference ENot significant §Coefficient of variation Table 22. Rate of Spread (meters per year) of S. alterntflora at Snow's Cut, 8 April 1971 to 27 November 1973 Downslope Spread Lateral Spread Snow's Cut Planting Stock From Single | Along Edge } Planted on Isolated of Large | Old, Very Row Compact Ocracoke Plants in 12 hr/day Inundation Zone Ocracoke Plants Along Edge of Large Block 86 SLOT YIteW S € yn) s i) *uOSBOS BUTMOIS PITY} Fo Buruutsog mous 3e sjzuetd ayoder199 Fo peords [es92e7] Iv 87 Rate of spread of such plantings is useful in estimating the time required for stands, established by seeds in the upper part of the tide range, to advance downslope to the lower limit of growth for this species. For example, direct seeding was successful at Snow's Cut down to 0.79 meters above MLW (inundated 5.7 hours daily); transplants survived down- slope to 0.25 meters above MLW (inundated 16.6 hours daily). The hori- zontal distance between these two points on this site was 39 meters. Assuming an annual rate of spread downslope (1.31 meters) equal to that measured for the transplanted area, 30 years would be required for the seeded stand to spread to the lower limit. This is one of the strong justifications for transplanting where early stabilization of the entire slope is desired. Development of a transplanted stand at Drum Inlet has been followed over a 2-year period (Tab. 23). This stand was one that escaped serious damage from the February 1973 storm and from salt buildup during the summer. Second-year development was equal or somewhat superior to that at Snow's Cut in terms of plant cover, primary production, and belowground growth (Tab. 20). This growth response took place under conditions that varied substantially from those at Snow's Cut. Salinity was much higher (close to sea strength); tide range was lower and erratic due to wind effect, and there appeared to be much less movement of fine grained sedi- ments. The rapid development of S. alterntflora on this site is further evidence of its ability to perform well under a wide range of conditions and suggests the Snow's Cut data may be representative of the development process. Table 23. Development of Transplanted S. alterniflora at Drum Inlet Culms /m2 Height Flowers/m? Basal Area Yield kg/ha | fom) | Come/m?) m?) Center|Rhizome Below- aa Oe ae ee ae Sea 18,038 SAO T2 a were seven individual plants (means are expressed in the table) for aboveground growth. 197.2% TNot sampled £1973 samples were six, 0.25 m* plots. Belowground samples were two cores from each of the aboveground sampling areas. Another aspect of marsh development has been followed at the Drum Inlet and Snow's Cut sites. Invertebrate species were sampled by taking core samples of the substrate material, screening, and identifying those present (Tab. 24). 88 Table 24. Invertebrate Species Found at Drum Inlet and Snow's Cut from March to November 1973 Natural Marsh Dredged Material Drum Inlet Annelida Heteromastus filtiformis Laeonerets culvert Eteone heteropoda Seoloplos fragilis Streblospto benedtett Glycera sp. Nereis succinea Paranais litoralts Henlea ventrtculosa Tubificidae (Ltmnodrilotdes medtoporus?) Annelida Paraonts fulgens Magelona papillicornis Arthropoda Ueca pugtlator Acathohaustortus mtllst Nemertea Tubulanus pelluctdus Arthropoda Dolichopodidae Ephydridae Stratiomyidae Insect adult A Insect adult B Neomysts amertcana Uca pugtlator Ocypode quadrata Mollusca Gemma gemma Mya arenarta Tagelus plebtus Areuatula (= Modtolus) demissa Snow's Cut ' Annelida Mollusca Heteromastus filtformts . Telltna sp. Laeonerets culvert Arcuatula (= Modiolus) Seolocolepides vtridens demissa Nerets succinea Enchytraeus albidus — Nemertea Unidentified Arthropoda Dolichopodidae Ephydridae Leptdactylus dyttscus Cyathura polita Uea pugilator 89 The invading species, in general, appeared to be the most widely adapted of the typical salt marsh species. Common at both Drum Inlet and Snow's Cut were the polychaetes Laeonerets and Heteromastus in the creeks, and Dipteran larvae and Uca pugilator, the fiddler crab, in the marsh itself. The polychaetes are deposit feeders; the others feed mainly on the growth of algae and diatoms at the marsh surface. In time, it is expected that these species will become less important as other less common species invade, but that they will still be among the dominant invertebrate fauna. V. SHORE PROTECTION AND SUBSTRATE STABILIZATION Stabilization was a major objective of the study, but one for which we were unable to develop satisfactory evaluation procedures. The prob- lem is the lack of unaffected or unbiased controls. For example, at the Snow's Cut site, the original experimental plantings were divided into three blocks, 45 to 60 meters wide, that extended roughly from the high spring tide line, downslope to about MLW. These blocks were spaced 30 to 60 meters apart, leaving the intervening strips unplanted and undisturbed. Dominant tidal currents are across the slope (parallel slope contours). Erosion and deposition were monitored in two ways: (1) by cross-sectional surveys started in 1971 through the planted blocks (Figs. 42 through 45) and (2) by cross sections established in early 1972 through the unplanted strips (Figs. 42 and 46 through 49). The latter might seem preferable as controls, compared to following elevation changes of the planted blocks over time. However, the protection afforded these unplanted areas by the adjoining vegetated blocks affects the erosion, the deposition and the revegetation occurring on them. Natural revegetation is gradually elimi- nating these areas as "unvegetated controls." Even the provision of unplanted blocks or strips appears inadvisable on eroding shorelines such as at Cedar Island since they would likely promote erosion of adjacent areas. There appears to be no meaningful way to test stabilization effects directly in small-plot field experiments. However, relative values between - variables based on vegetative growth may be sufficient at this stage. Beyond that, wave tank tests might be the best approach. 1. Snow's Cut. As indicated, certain cross sections were monitored at this site from June 1971 (Figs. 43, 44, and 45) and others (Figs. 46 through 49) from April 1972. All three planted blocks have shown a steady gain in elevation over the 30-month period. The northernmost block (Figs. 42 and 43) has trapped the most sediment, averaging close to 30 centimeters over the entire slope with largest gains in about the upper half of the normal tide range (Fig. 50). It appears that more sediment may be coming from the northern or upstream side and this planting may be intercepting material and reducing the amount available to the areas downstream to it. 90 "4anj) S,MOUS 38 SjOaSUeII FO UOTEIOT BuTMOYS WeIseIq “°7Zp san3TYy Asonjsy oso]uz94/ S [TT aD1S @ y2asups] | ? ! | ! | | | | / | | | | é v S ? ] j | ] | | ] ] j ] j ] ] ] J ] q j ] ] i] ] ] () C Punjs| |!ods 91 € = 1.0 Lower edge of = S. alterniflora Cc 2 i) > 2 w 0.5 0 6) 20 40 60 80 Slope distance (m ) Figure 43. Planted transect, No. 1, Snow's Cut; 2 June 1971 to November 1973. 1.5 £10 Ss = Lower edge of Es S. alterniflora 2 igen, Seueetb aT) = > ZS | WW 26 Nov 1973 e 2 Jun I971 6) 20 40 60 Slope distance (m) Figure 44. Planted transect, No. 3, Snow's Cut; 2 June 1971 to November 1973. 92 e!0 = =) = s 3 926 Nov 1973 loveriedsetat 20.5 © 5 Mar 1973 8. alterniflora 423 Oct 1972 @25 Apr 1972 e2 Jun 1971 0 0 20 40 60 Slope distance (m) Figure 45. Planted transect, No. 7, Snow's Cut; 2 June 1971 to 26 November 1973. 93 Elevation MLW (m) Natural invasion of S. alterniflora Ke Mar Nov 1973 1973 051 026 Nov 1973 #25 Apr 1972 @) 20 40 60 80 Slope distance (m) Figure 46. Unplanted transect, No. 2, Snow's Cut; 25 April 1972 to 26 November 1973. 94 Oct 1972 visual i : Mar 1973 Natural. invasion Nov 1973 Of > alterniflora ro) © oO Elevation MLW (m) 0226 Nov 1973 825 Apr 1972 0) 20 40 60 Slope distance (m) Figure 47. Unplanted transect, No. 4, Snow's Cut; 25 April 1972 to 26 November 1973. 95 rae ere Natural invasion Nov 1973 Of © alterniflora ro) oS on Elevation MLW (m) 9 26 Nov 1973 @ 25 Apr 1973 0) 20 40 60 Slope distance (m) Figure 48. Unplanted transect, No. 5, Snow's Cut; 25 April 1972 to 26 November 1973. 96 Natural invasion of S. alterniflora Oct Nov 1972 1973 ERIEO = ay = Cc 2 rs) > 210!5 WW 2 26 Nov 1973 @25 Apr 1972 O 20 40 60 Slope distance (m) Figure 49. Unplanted transect, No. 6, Snow's Cut, 25 April 1972 to 26 November 1973. There was little difference in elevation gain between the other two planted blocks; the change on these averaged around 15 centimeters. Sediment accumulation definitely decreased in the zone immediately down- slope of the vegetation. Records on the four cross sections through the unplanted strips indi- cate little change in elevation during the period of record (Figs. 46 through 49). Section 2 (Fig. 46) upstream from the others, exhibited the most accumulation. The effect of volunteer seedlings is clearly visible in the most recent survey on all of these unplanted areas. A minor amount of erosion has occurred along these cross sections, largely at the bottom of the slope. There was no way to predict what elevations would exist along the unplanted cross sections if no S. alterniflora were planted in the vicinity. The planted blocks did trap substantial amounts of sediment, and the inter- pretation is that this, along with the behavior of the unplanted strips is reasonable evidence of a strong stabilizing effect of this vegetation. Also, the slowness with which S. alterniflora invaded the unplanted strips appears to indicate that natural colonization of the area would have been extremely slow in the absence of such plantings. No seedlings appeared SN “uoOT Je TNUNDSDe ZUSUT pas BSULMOYS YSOTq paze,o39A FO MOTA Opts “OS 98 on the unplanted strips until the second year following the heavy first- year crop of seed on the planted blocks. It seems, therefore, that these areas were both seeded and protected by the planted blocks. It is possible that without this protection, seeds could not remain in place at this location long enough for germination and establishment, and that revegetation would have to come from gradual expansion of vegetation from a more protected part of the island. 2. Cedar Island. The experimental area is on the former Lola Naval facility situated on a promontory near the southern tip of Cedar Island. It is exposed to a fetch of several miles across Pamlico Sound to the east and northeast and across Core Sound to the south. It appears to be representative of many eroding shorelines in much of the Core and Pamlico Sound region. This area is believed to have a long history of erosion. On more protected shores nearby, Juncus marsh is growing on a thick (0.6 to 1.2 meters) layer of peat well above normal water levels, with the perpendi- cular face of the peat exposed along the water's edge. Apparently the marsh has grown upward on the surface of the accumulating peat and has long since lost contact with the intertidal zone along this edge. Conse- quently, the exposed face is quite vulnerable to erosion. This was probably true in front (east) of the Lola facility some years ago, but if so, the thick peat layer has eroded away. This shoreline is steep (10 to 25 percent slope), and is presently composed of coarse sand interlayered with thin lenses of peat. With only wind setup, a narrow area is available, 2 to 4 meters wide, on which-S. alterniflora could be expected to grow. Before planting in May 1972 there were three or four small eroding patches of S. alterntflora along this shore. About 1,900 S. alterniflora plants, dug near Drum Inlet, were transplanted by hand on about 0.6-by 0.6-meter spacing along a 0.2-kilometer stretch of shoreline south of the Radar Tower (Fig. 51). Following this planting, the area was subjected to strong northeast winds by a subtropical cyclonic disturbance from 24 to 27 May. A few weeks later, 60 percent of the transplants were washed out and the appearance of the 40 percent remaining was not encouraging. Consequently, this trial was written off as a failure and was not checked again until late in the season. At that time, the surviving plants had made a surprising recovery. When growth emerged the following spring (1973), these plantings were well established, and by fall a high propor- tion of this stretch of shoreline was effectively stabilized (Fig. 52). Samples were taken from this area in the fall of 1973 to estimate productivity (Tab. 25). Growth on this site was unusually dense and vigorous. Number of culms and flowers per unit area was quite high, and dry matter production was unusually high. When rechecked in January 1974, the only noticeable change in these plantings was some sporadic sand deposition by storm tides and waves. There was no evidence of stand loss from erosion. 99 “ZZ61 ACW II ‘Sutjzuetdsuer} 1LaqJFe pue[S] IepoD 4e oUTTOZOYS *1S oan3sTy 100 ScLO0 30 Ties Sua UT OUTTOLOYS PUTS] eps) oY} FO MOT/A °@G oLn3 Ty 101 Table 25. Second-Year Development of S. alterniflora Plantings* - Cedar Island, 13 September 1973 Aerial Dry Weight (kg/ha) Number Basal Area Stems /m2 Number Flowers/m2 (cm?/m*) *0.25 m* samples Another planting of 1,900 transplants was made on five small areas north of the Radar Tower on 11 and 17 May 1973. This shoreline is more exposed to the northeast and is eroding more rapidly than the area used for planting in 1972. Plants were closely spaced, about 0.3 by 0.3 meters. The total frontage covered was about 90 meters long. Erosion has con- tinued on this area with little net recession of the shoreline, but sub- stantial movement of material back and forth so that only about 20 percent of the plants could be identified as surviving in September 1973 (Fig. 53). Fewer were in evidence in January 1974, but some of these had grown quite well, some were buried beneath sand and rubble, and the amount of new growth that would appear in the spring was not predictable. Obviously this planting will not be as successful as the 1972 planting nearby, but it could be better than at present. Experience in shoreline stabilization suggests that such plantings do have potential for reducing sound-side shoreline erosion in this region. Many eroding sound shorelines are similar to those in the study area at Cedar Island. Any marsh vegetation remaining on them has lost contact with the sound bottom,separated from it by a scarp. The plants can no longer spread back into the bare zone by rhizomes. Further, since turbu- lence along the shoreline is excessive for seedling establishment, there is little possibility for natural revegetation. It is known from work elsewhere that S. alterntflora can be established by transplanting under conditions that are much too rigorous for natural invasion. If vegetation is to be reestablished on these shores, transplanting is the only possi- bility. VI. THE RELATIONSHIP OF MINERAL NUTRIENTS TO PRODUCTIVITY OF SPARTINA ALTERNIFLORA The mineral nutrition of S. alterniflora was studied to obtain know- ledge of the relationship of fertility of the substrate material to 102 "S/6[T atequieydes uayez ood f¢Z61 ACW TI poquetdsuez} ouTTeLOYS pueTs] IepoD BSG) 103 productivity of the grass. The role of fertility of the substrate material as it affected starting new marshes by artificial propagation had not previously been investigated. The productivity of natural marshes varies greatly from one location to another (Cooper, 1969). There is also variation within a given loca- tion with three distinct height zones (tall, medium, and short) generally recognized (Teal, 1962; Adams, 1963: Cooper, 1969). In the past, attempts were made to relate differences in productivity to such factors as genetic differences, time of inundation, and salinity. The efforts to relate productivity to nutrients in this study were divided into two phases. The nutrient status of plants and soils was sampled in natural stands from seven locations along the North Carolina coast, and related to yields using regression analysis. The second phase measured growth response to fertilizer applications. Plants in natural stands, transplants, and seedlings were included. 1. Nutrient Status of Natural Stands. a. Methods. Stands of S. alterniflora at seven locations along the North Carolina coast were selected for sampling during the summer of 1970 (Fig. 54). These stands are representative of variations in latitude, in tidal range, and in type of substrate occurring along the North Carolina coast. The Oregon Inlet, Hatteras Village and Ocracoke sites are similar in their location on the sound side of barrier islands where the tide range is 30 centimeters or less. Two important damper effects cause the narrow tide range in estuaries north of Cape Lookout: (1) few inlets through the barrier islands, and (2) a large expanse of water behind the islands. Winds greatly influence the surface levels causing a greater but irregular fluctuation. The Oregon Inlet and Ocracoke marshes are young stands on sandy substrate low in organic matter. The Hatteras Village marsh grows on a substrate containing 10 percent organic matter. The North River marsh is also in a location of low tide, growing on sandy substrate, and consists of a narrow fringe (about 20 meters wide) along the shore. The Beaufort, Swansboro and Oak Island marshes are in areas of wide tidal range and are more typical of the regularly flooded southeastern tidal marshes described by Cooper (1969). South of Cape Lookout, inlets are numerous and the estuaries are narrow allowing a greater tide range. Texture of the substrate is quite variable between the locations (Tab. 26). Plant samples were taken from each location and each height zone 17 to 23 June, 5 to 17 August, and 30 September to 2 October. The dis- tinctness of the height zones varied among the locations. The medium height zone was a narrow transition zone which was barely discernable in several locations. Consequently, only samples from the tall and short height zones were included in the statistical analyses. The plant samples, consisting of 0.25-square meter plots, were clipped at ground level from each height zone. The sample plots were selected by establishing a transect across and perpendicular to the height zones. From this line, 104 ; ape pee Ky r \ ) Hatteras ‘y atteras \ sogltts Village eo Ocracoke iy “North River Cape Lookout Beaufort Swansboro Oak Island Figure 54. Locations of sampling sites (indicated by arrows). 105 three points for sampling were randomly selected within each height zone. Sampling was done at low tide when possible. Soil solution salinity was field-determined with a handheld refractometer. Table 26. Soil Texture and Percent Organic Matter of Tall and Short Height Forms of S. LA SLA for Seven Locations Leeseate Sane | Seite || Ponce She | Percent [Boweete he | Percent OM: Location Oregon Inlet] 97.3 Hatteras 10.8 Village Ocracoke OKs North River 2.4 Beaufort 6.0 Swansboro 0.5 Oak Island LSD 0.05+ x ht. zone 10.0 Loc. Mod *Height zone tLeast significant difference (Main effects are not presented because of significant location by height zone interaction) Measurements on the grass included dry weight (dried at 70° Centigrade in a forced-air oven), number of stems and their height (the average of five randomly selected stems in each 0.25-square meter plot). The dried samples were chopped with a silage chopper, and subsamples were ground in a Wiley mill in preparation for nutrient analyses of the plant tissue. Determinations of the concentrations of nitrogen, phosphorus, potassium, sodium, calcium, magnesium, sulfur, iron, zinc, manganese and copper were made by the Department of Soil Science, Analytical Service Laboratory at North Carolina State University. Soil samples of the upper 15 centimeters were taken in each height zone at the time of the first plant sampling. These samples were air dried, and screened, and the following determina- tions were made by the Soil Testing Division of the North Carolina Department of Agriculture using their routine methods: organic matter, calcium, magnesium, phosphorus, potassium, sodium, CEC (cation exchange capacity) and soluble salts. Soil texture was determined by the hydro- meter method (Day, 1956). Regression analysis was the statistical technique used to detect relationships between nutrient levels in the soil, plant tissue, and 106 productivity. After an attempt to use the raw data, yield and height were transformed to their logarithm (base 10) to provide a better fit for the curves, decrease random variation, and normalize the mean square error. There are 48 independent variables in the original data which consisted of the concentration of 11 different mineral nutrients in the plant tissue at each of 3 sampling times, the salinity of the soil solution at each sampling time, and 12 soil chemical and physical proper- ties. The ''best'' regression model was selected using a combination of the maximum R* (multiple correlation coefficient) improvement procedure (Service, 1972), the stepwise regression procedure (Draper and Smith, 1966), and critical examination of the independent variables from an agronomic view. In the stepwise procedure, the single variable model which produces the highest R* is selected; then variables are added one by one to the model according to their significance measured by the F-test (0.1 level of significance for entry). At every stage of regres- sion, the variables incorporated in the previous stages are reexamined for significance. A variable entered at an early stage may lose its Significance because of its relationship to other variables entering the regression. The maximum R* improvement technique developed by J. H. Goodnight (Department of Statistics, North Carolina State University, Raleigh, North Carolina) selects the ''best'' one-variable model, the "best" two-variable model, etc. according to R?%. b. Selecting the Dependent Variable. The independent variables were measurements of soil chemical and physical properties and nutrient con- centrations in the plant tissue at three sampling dates (Tab. 27). Multiple regression procedures were used as variable screening devices to determine which of these independent variables were related to yield and height of S. alterniflora. _It was necessary to select two models (one for logy yield and one for log,g height), because the relationship between height and yield was not as close as might be expected (Fig. 55). esto nertign the relationship between yield and height is highly significant, the R* is only 0.26. The R* was not greatly improved by transforma- tion to logarithms. Stands are thinnest where the tallest grass occurs; consequently, shorter grass may produce higher yields where stands are thicker. An example is seen by comparing the data for yield and height from the tall height zones at Oregon Inlet and Beaufort (Tab. 28). The average height at Beaufort is much greater than at Oregon Inlet, but the yield at Oregon Inlet is greater. This is attributed to there being nearly twice as many stems per unit area at Oregon Inlet to produce the total biomass. This difference in growth habit may be related to tide range or possibly to genetic difference between the grass from different locations. Shorter grass and thick stands occur at Oregon Inlet, Hatteras Village and Ocracoke where the regular tide range is less than 30 centi- meters. At Beaufort, Swansboro; and Oak Island where the tide range is about 1 meter, the grass is taller but the stands are sparse. There is also a significant difference in the number of stems per unit area between the tall and short height zones. The stands are more dense in the short height zones, but the yields are much less than in the tall 107 Table 27. Variables Used in Model Building and Their Simple Correlations with Dependent Variables Height and Yield Variables Abbreviations Simple Correlation Coefficient (r) Organic matter Cation exchange capacity (meq/100 g) Soluble salts (MMho) 0.22 Phosphorus (ppm) -0.09 Potassium (ppm) 0.38* Calcium (ppm) -0.17 Magnesium OR22 Manganese (ppm) -0.08 Sodium 0.29 Sand (%) -0.36* Silt (%) Clay (%) 17 to 23 June 5 to 17 August 30 September to 2 October 17 to 23 June Nitrogen (%) Phosphorus (%) Potassium (%) Sodium (%) Calcium (%) Magnesium (%) Sulfur (%) Iron (ppm) Zinc (ppm) Manganese (ppm) Copper (ppm) ~ *0.01 r > |0.30| +0.05 r > |0.39| 108 Table 27. Variables Used in Model Building and Their Simple Correlations with Dependent Variables Height and Yield- Continued Variables Abbreviations Simple Correlation Coéfficient (r) 5 to 17 August Nitrogen (%) -0.18 Phosphorus (%) 0.09 Potassium (%) -0.33* Sodium (%) -0.08 Calcium (%) 0.09 Magnesium (%) 0.02 Sulfur (%) -0.59T Iron (ppm) 0.43t Zinc (ppm) 0.01 Manganese (ppm) 0.24 Copper (ppm) 0.16 30 September to 2 October Nitrogen (%) -0.27 Phosphorus (%) -0.04 Potassium (%) -0.38* Sodium (%) -0.20 Calcium (%) 0.05 Magnesium (%) -0.05 Sulfur (%) -0.50+ Iron (ppm) 0.40+ Zinc (ppm) -0.02 Manganese (ppm) 0.16 Copper (ppm) | 05 50)| +0.05 r > |0.39| 109 height zones. Height and yield are more closely related within a given marsh or type of marsh (Williams and Murdoch, 1969). That is, at a particular location the taller grass produces higher yields. In this study, the correlation between height and yield was low enough that it seemed appropriate to create separate models for height and yield. 16 = (2 wee 2 R7= 0.26 @he = 4 & e ew?’ C.V. 56.8 =i Y=1298+58.5(X) O O 40 80 120 160 Height (cm) Figure 55. Relationship between height and yield of S. alterntflora. 110 JUBITFTUSTS JON§ QIUSLOFFIP URS OnehE 8 0°S68‘Z 0°008°Z 0°SL6°S 0°00z‘ 7 0°008‘8 0°L98‘ ZI 0°SSz7‘8 “ZW §Z70/sweas “ON | (wo) 2y43TOH —— (ey/34) PTETA SUOT}EDOT UdAES FO sSoUuoZ YSTOH 4LOYS pue Tiel 043 wory setdues DAOTSIULGZ7D °S JO swWoiS FO ToquNN pue USTOH “Pl[OTA LOF SooUSLOTFTG WUeITJTUSTS seo] pue sue TJTUsTS ysevayt S}JOOFFO UTeW YL QUOZ USTOH, auoz *2Yy X *907 auoz “14 "207 >S0°0O £aST +X pueTts]T eo OLOqSUeMS JLoznesg IOATY ULON aYOIeLIO ase TTA sezoi7eH JeTUI uos919 uoOT}eD07 82 eTIeL 111 c. The Model Building Process. There were 47 independent variables in the original model. Silt was omitted since the pefcent sand, silt and clay always sums to 100 percent. This causes any two of these varia- bles to correlate perfectly with the third; therefore, only two of the variables can be used in a regression equation. The number of variables in the original model was reduced with the stepwise regression procedure and the maximum R* improvement procedure. By regression techniques, two sets of variables most likely to be related to yield (Tab. 29) and height (Tab. 30) were selected. Each model contains 11 variables with an R2 of 0.90. Thus, about 90 percent of the variation in yield and height is explained by the independent variables in each regression equation. The variables in these models represent a subset of variables which can be used to explain differences in yield and height; however, they may not be the most important variables affecting plant growth in the salt marsh system. Other subsets may produce similar R2 values, but the two pre- sented were considered the most agronomically feasible. Some models with fewer variables also produced satisfactory R2 values (fabs Si). Regression equations with more independent variables produced higher R* values, but beyond 11 variables there was very little increase in the regression sum of squares or reduction in the error mean square and the coefficient of variation. d. Interpretation of the Yield Model. As would be expected in a natural ecological system, an examination of the correlation matrix for the 11 variables in the model revealed that there was some intercorrela- tion of variables. When there is a correlation between independent variables in a model, the regression coefficients (b's) may not be reliable (Draper and Smith, 1966). However, by considering the regres- sion model in combination with the simple correlations of each independent variable with yield (Tab. 27) and the means and Least Significant Differences obtained by analysis of variance, some insight as to the effect of each variable on yield variation may be obtained. The partial sum of squares are a measure of the relative importance of the variables in a regression equation and the variables may be ranked on this basis (Tab. 29) (Draper and Smith, 1966). The potassium content of the plant tissue at the fall sampling data (CK) is the variable which accounts for the highest amount of yield variation in the regression equation. At five of the seven locations, potassium concentrations were greater in plants taken from the short height zone (Tab. 32). The trend of lower potassium concentrations in higher yielding plants is probably an indication of a greater dilution of potassium in the plant tissue where higher yields occur. Sodium concentration of the plant tissue at the first sampling date (ANa) is positively related to yield. Significant differences in sodium concentrations occur between height zones and locations (Tab. 32). Plants from the tall height zones had higher concentrations than those from the short height zones. The explanation for this is not apparent, particularly since higher salinities of the soil solution (AS-SAL) were clearly related to decreased yields (Tab. 27). High salinity of the soil I Mable 29% Source Regression Deviations Total Source Intercept CK ANat CCat AS-SAL BMnt S-Mn . Analysis of Variance Table, Statistics of Fit for Dependent Variable oe Yield Regression Coefficient and Sum of Squares F-Value GoW (%) pasts tl Bi eeeRi sooo Dian [eee Sel 30 0.3876 | ----- ==- 3.7428 -0.9873 OGIO Ike SOS -0.0161 -0.0095 -0.0046 3.4720 -0.0116 3.2208 0.0005 -0.3368 0. 0. On OF 0. OF 0. 0. 0. 0. 0. *0.01 level of significance +Variables which also appear in the height equation 0.10 level of significance Table 30. Analysis of Variance Table, Regression Coefficients, and Statistics of Fit for Dependent Variable oe Height Source Sum of Squares F-Value GoW. (%) Regression Pia =| ite = aa | (tesa | 54* 3.8 Deviations 0.1420 ---- --- Total histiaar crop! 4722 Coefficient (b) of Squares Intercept ---- BS 0. S28 ANat 0. DS oS BMg 0. Bilea AFe 0. US) 5 7 AMg 0. Wo Abs CN 0. 1WA0F AN 0. 10.2* CCat 0. 8.4* BPt 0. 6.54 BMnt 0. 4.0% ASt 0. Seis *0.01 level of significance tVariables which also appear in the yield equation £0.05 level of significance §0.10 level of significance 114 Table 31. Selected Models for Predicting Log,, of Yield and Height No. in Model R2 Gave Model (%) Yield BS BS, BMn BS, BMn, AS-SAL AS-SAL, BMn, BS, ANa, S-Mn Coins CAs IS ak o wo nN N 7 BMn, BS, AS-SAL, : 4 ANa, BCa, S-Mn, ACu 11 CKwANaaiCGar ; “ll AS-SAL, BMn, S-Mn, AP, AMn, BP, SP, AS Height BS BS, ANa BS, ANa, CCa BS, CCa, ANa, BFe, AS-SAL 7 BS, CCa, ANa, BFe, BMn, S-Ca 11 BS, ANa, BMg, AFe, AMg, CN, AN, CCa, BP, BMn, AS Cn A AS) ES JUBITFIUSTS ONS QOUsLOFFIP JUeITFIUSTS YsvoqTt S}IOFJO UTeW = XI 9UOZ YSTOH, v0'0 § § ouoZ *Y X *907 § S0°0 STO euoz *24 £0°0 60°0 SZ°0 “207 >S0°O $dS7 == = = +X I¢"0 pueTs] eo Z5°0 ozoqsuems Gua) JLOFNeOYg Us>@O TOATY YILON 0s°0 SYOSe.IIOO Sz°0 aBeT ITA setozzeH L220 JOeTUL uos9I09 auoz *3Yy X *907 euoz “4 “307 >$0°O tdS71 = = = +X 9T°0 pueTs]I eo Il‘0 oLoqsuems Sb” 0) JLOFNCSY IT*0O TOATY YIION 60°0 dYOIeIOO OL‘'0 aseLTIIA $sei933eH Z1°0 JeTU] uoserQ 4X | aaoys |. 1TeL (core eric ee en crs eh cre | (ane I8q sqyieg/puesnoy], Lod SRW uoT e907 YYStoaH pue PTOTA OF suotzenby 116 JUBITFTUSTS ONS 61°0O L0°0 vt°O (u/d) aa QOUSLOFFIP JUCOTFIUSTS JSevOTT S}O0TFO uTeU Pee a UOTTTIN tog S}teg/puesnoyuL fog S}ieq/jUooleg poenut}Uuo)-Jystoy pue plat, 1oF suotzenby 9UOZ USTOHy SUOZ 2 exons OF euoz °4H DOF] >S0°0O {ds +X pueTs] eo oLoqsuems JLopnesag LOATY YILON ayooe1IQ ose [IA sesoz7eH JeTUy, uose19 auoz *3Y4 X *907 auoz °*4H S05] >S0°O tdaS7T +X pueTs~T eo .OLOGSUBMS jLojnesg LOATY YZALON dYOIe.IIO oBeLIIA seito7.eH JeTU, uosII09 UuUOTIBIOT uotsseigey ut teodde yoTymM soTqetie, LOF SodUdTOFIIG JYURITFIUSTS Seo] pue sUeoW “ZF STGRL 117 JUROTFIUSTS ONS QUSLOFFIP JURITFIUSTS YSeoTt S}O0JJO UTeWU = XxX} OUuOZ YSTOH, ae LI°6 Leas 9uoz *4Y xX *307 cana V3 § euoz *3H Sie 8b°9 c° 9 BOOT >S0°O fdaSsT +X pue[Ts~T tO OLOqSUeMS JLopneog LOATY YILON aYOIPIIO sseLITA Ses9z3zeY JOeTUl uoseIC9 9UuOZ *Y X *907 9uoz *3H "207 >S$0°O £dS7 —--—--=— +X 87°0 pueTs~ eO IZ °0 oLOqsueMs 87°70 JLOFNvSY 0£*0 LOATY YILON Sr'0 dYORIIQ 97°0 OBET ITA Sesoi.eH Soo JOTU, uossI9 (%) B99 (4) XO UOTT[IW 19g SJteg/puesnoy] Log Si1eg/}UdDIOg uoT}eI07T Ponut}UoO)-YSTOH pue P[OTX OF suotrjzenby uotsseisoy ut teodde ySTYyM SOT qeTreA OF SODUDTOFFIG JUROTFLUSTS }SeoT pue SUeOW “7S OTqeL 118 solution has often been suggested as a cause for reduced growth of 5S. alterntflora (Mooring, Cooper, and Seneca, 1971). The reason for the appearance of calcium concentration in the plant tissue at the fall sampling date (CCa) in the regression equation cannot be readily explained. The means (Tab. 32) and the simple correlations (Tab. 27) show no obvious relationships between yield and calcium con- centrations in the plant. It probably enters the regression equation because of its relationship to other variables. Manganese concentration in the plant tissue at the first sampling date (AMn) and the second sampling date (BMn) and soil manganese (S-Mn) are all negatively related to yield. The solubility of soil manganese is increased under conditions of poor aeration. The importance of manganese in the regression equation may be an indication that the more reduced soils produce lower yields of S. alterntflora. The chemical environment produced by waterlogging of soils produces several toxicity problems which have been studied in relation to reduction of rice yields. Common problems are iron, manganese and sulfide toxicity and toxicity from soluble organic products (Black, 1968). However, manganese toxicity is unlikely because of the concentration in the plant tissue (Tab. 32). Concentrations of 10 times this amount were found with no apparent ill effects to plants in growth chamber studies with S. alterniflora in which the nutrient source was modified Hoagland's solution (Hoagland and Arnon, 1950). Phosphorus concentrations in the plant tissue at the first sampling date (AP) and the second sampling date (BP) and soil phosphorus (S-P) are positively related to yield. The need for adequate supplies of phosphorus for plant growth is well known, and phosphorus is often a limiting factor in the growth of plants. Sulfur at the first sampling date (AS) is negatively correlated with yield. The means in Table 32 clearly show that concentration of sulfur is much less in plants from the tall height zones. It is impossible to determine from the data if this is a dilution effect or if more sulfur is available in the short-height zone. e. Interpretation of the Height Model. Since some variables in this model are also correlated, caution must be observed in interpreting regres- sion coefficients. The sulfur concentration in the plant tissue at the second sampling date (BS) accounts for the highest amount of variation in height in the regression equation. Sulfur concentration at the first sampling date (AS) also appears in the equation. The sulfur concentration at the second sampling time is the single variable with the highest R2 in both the height and yield equations (Tab. 32). There is a clear relationship between increased sulfur concentration and decreased growth, but it is impossible to determine from the data if it is a cause and effect relation- ship. IIS) Sodium concentration in the plants at the first sampling date (ANa) is positively related to height as it was to yield. The concentration is significantly higher in samples from the tall-height zone (Tab. 32). Magnesium concentration in the plant tissue at the first (AMg) and second (BMg) sampling dates are in the regression equation. However, there is no relationship of magnesium with height apparent from the sim- ple correlations (Tab. 27) or the means (Tab. 32). This variable may have entered the regression equation due to its relationship with other variables. The iron concentration in the plant tissue at the first sampling date (AFe) is positively correlated with height. The difference is greater between locations than between height zones. This difference is probably because finer sediments occur at the locations where taller plants are found, and these sites would be expected to contain more iron (associated with clay) than the sandy sites. Nitrogen concentration at the first sampling date (AN) was positively correlated with height, while nitrogen concentration at the third sampling date (CN) was negatively correlated with height. A probable explanation for this is that plant-available nitrogen is scarce in the marsh environ- ment. At the first sampling date, high nitrogen concentrations in the plant are indicative of a high potential for growth. At maturity, the nitrogen concentration is lowest in plants which have achieved maximum growth due to dilution in the greater amount of plant tissue. Calcium concentration at the third sampling date (CCa) probably enters the regression equation because of its relationship to other variables since no relationship with height is apparent from the means (Tab. 32) or the simple correlation with height (Tab. 27). Phosphorus concentration at the second sampling time (BP) was posi- tively correlated with plant height. This indicates that where greater amounts of phosphorus are available for growth, greater heights are attained. The concentration of manganese at the second sampling time (BMn) appears to be negatively related to height according to the regression coefficient (Tab. 30). However, the simple correlation shows a positive relationship of BMn to height. This is an example of sign reversal of b-values when there is correlation between variables in the model. This positive relationship of manganese concentration to height is opposite of the relationship with yield. f. Predicting Standing Crop of Mineral Nutrients from Dry Weight. It is of interest to know the quantities of nutrients conveyed to estu- arine food chains by S. alterntflora (Williams and Murdoch, 1969). From the data in this study, standing crop of nutrients contained in the aboveground part of mature S. alterniflora stands can be calculated (Tab. 33). The data used were three 0.25-square meter samples clipped 120 sotdwes zp jo ueopw4 *(ey/3y) AVTaInjzeu 7e@ sJOOYS puNOTSeAOGe FO JYSTOM AIP = X ‘(ey/3y) ATanZeW Je SszOOYS puUNOTZ9.AOGe UT S}UdTIZNU Fo dots BuTpUeIS = jy (xX) £00000°0 + 8Z00°0 K 79°0 (x) Z0000°0 + 9S00°0- = x 9v°0 (X) T0000°0 + Z000°0- = x £7°0 OOIS0000 = iG 0 == h 82°0 (X)8610°0 + g90°g = k £8°0 (x)Zv00°0 + z6s‘T- = z 06°0 (x)¢00°0 + gz6°0 = A SZ°0 OOSZ0070 2 asecec.: = h ESO (X)1Z00°0 + pI0o°9 = h 88°0 (x)6000°0 + gor'o = A L6°0 (x)6Z00°0 + 9g¢°0- = X : 1G Jee (ey/3y) S}OOYS puNosseAocgy UT S}UsTIyNY Fo doxzyj sutpueys ues oud JO §a'S Ud TIANN plo, f1Udeq7oD *S JO SJOOYS puNorssAoqy ey} UT S}USTIINN [eLouTW Jo doar) ButpueiS LOZ suotzenby uotssers9y pue soT4STIeIS eTdWTS “SE 9TqeL 121 from the tall- and the short-height zones of seven locations when the grass was mature. Since the total amount of nutrients incorporated in the plant tissue is highly dependent on yields, the dry weight is a good predictor for standing crops of the mineral nutrients. The R2 values for standing crop of nitrogen, phosphorus, potassium, sulfur, calcium, magnesium, and sodium are quite high (Tab. 33), and these equations should be reliable and useful for estimating the amount of nutrients contained in the mature shoots of S. alterniflora salt marshes. However, the fate of the nutrients as the plants decompose is more difficult to determine. Because of greater variation, the predictions for iron, zinc, manganese and copper standing crops would be less accurate. g. Summary. The results indicate that tissue concentrations of several nutrients and several soil properties were significantly associated with variations in yield and height of S. alterntflora. It is important to keep in mind the purpose of multiple regression analyses. Predictive models are not necessarily functional, but can lead to insight into a problem. According to Draper and Smith (1966), construction of this type of model from problems where much intercorrelation of data exists is where regression techniques can make their greatest contribution. It provides guidelines for further investigation, pinpoints important variables, and is useful in screening variables. Several variables selected by the multiple regression procedure in this study seem to warrant further investigation to determine their relationship to productivity of S. alterniflora. Such variables negatively associated with yield include salinity of the soil solution, manganese concentration in the plants and soil, and sulfur concentrations in the plants. Important variables positively associated with yield include phosphorus concentration in the plant tissue and in the soil. The reduction of yield of S. alterntflora with increasing soil salinity has long been recognized. This study reconfirms this, but there was not a striking difference in salinity between height zones. This is indicated by the fact that the simple correlation of height with soil salinity is near Ooi@habe 027) The importance of manganese and sulfur in both the yield and height equations and iron in the height equation suggests investigation into the influence of the chemical effects of waterlogging of the marsh soil on S. alterntflora growth. Undoubtedly there are different degrees of aeration both within and between Spartina marshes which affect soil chemi- cal properties. There is an extensive literature on waterlogged soils in connection with rice culture (Redman and Patrick, 1965; Black, 1968). Useful reviews on the chemistry of phosphorus and nitrogen in sediment- water systems are presented by Syers, Harris, and Armstrong (1973) and Keeney (1973). The positive influence of phosphorus concentration in the plant tissue would be expected, since in most natural plant-soil systems phosphorus is second only to nitrogen as a limiting factor in plant growth. Nitrogen 2/2 concentration in the plant tissue was a part of the regression equation for height but not for yield. Perhaps nitrogen does not show up in the equation because it is the limiting factor in growth. If the availability of nitrogen was limiting growth, then growth would proceed whenever nitro- gen became available. Consequently, the concentration of nitrogen in the plant tissue would remain relatively constant due to the increase in biomass. It is possible that if samples had been taken earlier in the growing season, the nitrogen concentration in the plant tissue would have been a better indicator of yield potential. It is interesting to note that there was no significant correlation of yield with the soil properties measured (Tab. 27). There are two factors which contribute to this: (1), the waterlogged conditions of these soils tend to equalize chemical differences, and (2), the methods by which these determinations were made. Standard soil testing procedures were used which probably are not suitable for these soils. An important limita- tion is that North Carolina soil test procedures and extracting solutions are designed for acid soils. The pH (hydrogen-ion concentration) of the soils in this study were between 7 and 8. Developing suitable techniques for studying properties of marsh soils would be an extensive project. In conclusion, it is not within the scope of this study to explain each observed effect, but several relationships were shown to exist between variables which were measured and yield and height of S. alterniflora. Several of these effects may warrant further investigation. 2. Effects of Fertilizer. In the natural marshes, fertilizer was used as an experimental tool to determine if nutrients were limiting factors in growth. Marshes at Ocracoke and Oak Island were selected because of differences in substrate, tide range, and age. On seedlings and transplants, the objective of the fertilizer studies was to determine if adding nutrients would enhance growth to produce cover more rapidly. a. Fertilizer Experiments in Natural Stands. (1) Ocracoke Island. Fertilizer, plots were located on the north end of the island near Hatteras Inlet. This is a relatively young 5S. alterntflora marsh on a sandy substrate with little development of tidal creeks. There is some difference in growth between the different areas within the marsh, which is apparently due to environmental factors. However, zonation of height forms is not as obvious as in many older marshes. The regular lunar tide range at this location is about 30 centimeters, but the added wind effect may extend the range to 1 meter. A fertilizer experiment of factorial design with two rates of phosphorus and four rates of nitrogen was started in 1971. It consisted of three ran- domized complete blocks with 1.22- by 7.62-meter plots and 1.22-meter borders between plots. Phosphorus was supplied by concentrated superphosphate at 123 rates of 0 and 74 kilograms per hectare of phosphorus. Nitrogen rates were 0, 168, 336, and 672 kilograms per hectare of nitrogen supplied by ammonium sulfate. An ammonium form of nitrogen was thought to be more suitable to the marsh environment since it is the form of inorganic nitro- gen found in greatest quantities in reduced soils. Application of nitrate- nitrogen to poorly drained soils is undesirable since it is subject to denitrification and loss to the atmosphere in gaseous forms (Keeney, 1973). The ammonium form has the advantage of being adsorbed by the exchange complex of the soil. It is also possible (considering the flooded condi- tion in which it grows) that S. alterntflora is adapted to utilization of the ammonium form of nitrogen as has been reported for some other plants (Townsend, 1966; Van Den Driessche, 1971). The fertilizer materials were applied in split applications with equal amounts on 12 May, 22 June, and 27 July 1971 by broadcasting evenly over the substrate surface. Samples were harvested 1 September 1971 by cutting a 0.61- by 1.52-meter swath from each plot with a Jari sicklebar mower. Saltecornia spp., dead stems of S. alterniflora from the previous year's growth and other foreign matter, were separated from S. alterntflora plants. The plants were dried at 70° Centigrade and weighed, subsamples were ground in a Wiley mill, and analyzed for nutrient content by the Department of Soil Science, Analytical Service Laboratory of North Carolina State University. The experiment was continued in 1972 with the same rates of nitrogen and phosphorus fertilizers applied in split applications on 13 April, 20 June, and 19 July 1972. The plots were clipped and raked in early spring in 1972 to facilitate harvesting and ensure that all plant material harvested in the fall was produced-in that growing season. Samples were ‘harvested 11 September by clipping a 0.61- by 3.96-meter swath from each plot. The plant samples were dried and processed in the manner previously described. Roots and rhizomes were also sampled in 1972 by taking five cores 8.5 centimeters in diameter and 30 centimeters deep from each plot. In the laboratory the cores were divided into 0 to 10- and 10 to 30-centi- meter layers and washed with tap water to remove the soil material. Ten core samples were selected at random and roots were separated from rhizomes to determine the relative proportions of each. The root and rhizome samples were dried at 70° Centigrade. Combined root and rhizome samples were processed and analyzed in the same manner as the shoots. The experiment was continued in 1973 in the same manner as in 1972, except roots and rhizomes were not separated. The results of the nitrogen-phosphorus factorial experiment at Ocracoke indicate that, although additions of nitrogen alone can increase yields significantly, the availability of phosphorus quickly becomes limiting when nitrogen rates are increased (Fig. 56). Yields at the end of the first growing season were increased only slightly by the addition of nitrogen without phosphorus. The only statistically significant (0.05 level) difference was between the yield of the check plots and those receiving 672 kilograms per hectare of nitrogen. However, when phosphorus was 124 applied at the rate of 74 kilograms per hectare, yield was markedly increased by additions of nitrogen up to 336 kilograms per hectare. The yield produced by 672 kilograms per hectare of nitrogen was not signifi- cantly greater than that where 336 kilograms per hectare were applied during the first growing season. Y=6169.1+17.7X 16 oO < c , a2 Y= 2978.8 + 28.7X-0.024X 2 canscssesesccoscssnsecccoceeeE] R2=.87 = awe” Vx 5555.8 411.37 X- 0.0107 X? , © 3 ° R*=.78 = Y= 4412.1+8.72X-0.009X? o ee aT Va nee (ae | SOR = 40 (ns) ey Ba © No P Supls 0 74 kg “ha P coves 197] — 1972 ) O 168 336 672 Nitrogen (kg/ha/yr) Figure 56. Effect of nitrogen and phosphorus fertilizers on yields during two successive growing seasons. During the second growing season, higher yields and a greater response to fertilization was attained (Figs. 56 and 57). When phosphorus was supplied, the rate of yield increase did not level off beyond 336 kilograms per hectare of nitrogen as in the first year, but was linear over the range of nitrogen rates applied. There was also an increase in the difference between yields of no phosphorus and phosphorus treatments, indicating phosphorus became severely limiting during the second year as yields increased. 125 *szeoXk z 1oF snzoydsoyd Fo reok sod o1e,D90Yy tod sweas3s -O[TY pL pue uss0r1}tu Fo aAvok tod o1e}D9Yy Tod sweISO[TY 719 PaATOIOL (ay8t4)’ qoTd poezt{tqseFZ OYL “ZL6T toquoydes [TT ‘pue[s] eyooe19Q je ysatew [eiInjeu e UT SzOTd poZT{[I}AOF pue pozT{[TJLoFJuN Fo uostaedwuods y “L/S oAnstTy 126 Nitrogen content of the plant tissue was closely related to the rate of nitrogen applied as fertilizer (Fig. 58). During both growing seasons the R* (multiple correlation coefficient, a measure of the percentage variation in the dependent variable which is explained by independent variables in the equation), values were higher where no phosphorus was applied, indicating that a greater proportion of the variation in nitrogen content is accounted for by nitrogen rate when nitrogen alone is applied. Where phosphorus was applied, both the R2 and the actual concentration of nitrogen in the tissue decreased. This decrease was probably due to dilution of nitrogen in a greater amount of dry matter as yield increased when the phosphorus became adequate. The apparent recovery of fertilizer-nitrogen by the aboveground part of the grass and the roots and rhizomes may be calculated by subtracting the uptake of the check plots from that of the treated plots. The calcu- lations reveal a surprisingly high recovery of nitrogen considering the flooding which occurs (Tab. 34). Table 34. Apparent Recovery of Fertilizer Nitrogen in the Shoots at Harvest | Shoots *Rate of P (kg/ha) Concentration of phosphorus in the plant tissue was not significantly affected by nitrogen rate; however, the increase in phosphorus concentra- tion due to phosphorus fertilization was highly significant (0.01 level). In 1971 the mean concentration of phosphorus in the shoots was increased from 0.084 to 0.150 percent, where 74 kilograms per hectare of phosphorus were applied. In 1972 there was a similar increase due to phosphorus fertilization from 0.087 to 0.147 percent in the shoots, and concentration in the roots increased from 0.052 to 0.091 percent. The amount of fertilizer-phosphorus recovered by the shoots for treatments receiving the maximum rate of nitrogen was 15.0 and 26.1 percent of that applied during 1971 and 1972, respectively. Apparent recovery of phosphorus by the roots in 1972 was 12.4 percent. The calculations of recovery of both nitrogen and phosphorus in 1972 may include some carryover from 1971 through storage in the roots and rhizomes or residual in the soil. It is more likely that phosphorus would be retained by the soil than nitrogen. WAT -Jsoarey 3e onsstz quetd Fo JUSWUOD usdoLzTU pue pottdde stoZT[ 134195 uos0i1} TU usemzoq drysuotyep[ot oy] “8S oin3 ty (oy /6y) e405 uabossin GL) gee 891 O Ls) 9C¢e 89] 6) 2000 $100000° - (X) 200° + ya = 4d 04764 940— (X)L00° + SB = A d ON Ove NI (95) uabossiNn "U6 =u OO 9000° + 88° =4 d 947 54 4 O— (Xx) 4000' +Z8° =A gd ON Over cL6\ 1161 128 Analyses for other mineral nutrients included potassium, calcium, magnesium, sulfur, iron, and sodium in both 1971 and 1972. An analysis for manganese was included in 1972. Regression analysis was used to determine if there were significant linear or quadratic relationships between the nutrient concentrations in plant tissue and the nitrogen rate. Of the nutrients checked in 1971, only potassium was found to be signifi- cantly related to nitrogen rate (Tab. 35) and this was true only when phosphorus was applied. In 1972, when no phosphorus was applied, phos- phorus, calcium, and manganese concentrations were significantly affected by nitrogen rate. Calcium and manganese concentrations increased as nitrogen rate increased, while phosphorus concentration decreased as nitrogen rate increased. The decrease in phosphorus concentration indi- cates that there is a limited amount of phosphorus available and the amount in the plant tissue is diluted as growth is increased due to nitro- gen fertilization. Where phosphorus was applied, only iron and manganese concentration was significantly affect by nitrogen rate. Manganese increased and iron decreased as nitrogen rate was increased. Table 35. The Relationship of Nutrient Concentration in the Plant Tissue to Nitrogen Rate* Year | P Rate R2 (kg/ha) 74 SOT Nutrient Regression Equation Vo =) J0n987 + 10500330) P 0 Y= 0.09 - 0.00007(X) + 0.00000007 (Xx) 2 CA O.85 8 0.00036(X) - 0.00000039 (xX) 4% Y= 15.75 + 10.0092(X) ¥i=) 32000 2808219109 Y 0.117(X) *X = Rate of nitrogen fertilizer (kg/ha) +Significant at the 0.05 level tSignificant at the 0.01 level Total uptake of mineral nutrients increased as nitrogen rate was increased (Figs. 59 and 60). This was due mainly to an increase in yield in response to nitrogen and phosphorus fertilization, producing more plant material in which nutrients in adequate supply are incorporated. The standing crop of these nutrients is controlled by yield which is in turn limited by availability of nitrogen and phosphorus to the marsh plants. The amount of nitrogen and phosphorus available is the limiting factor in the total amount of nutrients exported from the marsh in the dead plant material which contribute to the nutrient cycle of the adjacent estuary. An obvious feature of the data for uptake of mineral nutrients is the large difference in sodium uptake between 1971 and 1972. This was probably due to periods of high salt concentration in the substrate in 1972. The level of sodium in the substrate varies with rainfall and frequency of flooding. At Ocracoke, southwest winds during summer often prevent 129 *9re10Y Zod swers0[ty pL FO oFeI 9yi 3e pottddns sem snzoydsoyg ‘untpos pue ‘umtssejod ‘ues01z1u Fo ayeqdn [e202 02 9}2eL UDZOIZTU Jo dtysuoTIe{ser sy, *6S 9ANBTY (By/8y) e2ez UeS0IITN (#y/8y) e3B1 ueSorI TN 719 9£€ g89T 0 719 9€€ 89T 0 Sie ee ee 0 00T 68° = 2u 16" = 74 98° = 24 2 : 00z 18° = py 78° = 7u N ooe 00% 2%) To00'0 - (x) TO + ZLye sk UY 2X) 200070 - (x) ze"0 + 787 = 4 NO CI 4) 204) $000°0 - (x) 8s°0 + 60S = A PN O Ws () 70+7T9 "84 AV 00s () 97°0 + €°8Y =A NO 26) ££000°0 - (x) OT + S°97T = 4 ONG 7L6T IL61 130 (34/34) exeady *ozeqooy zod sweais -O[TY pL FO 932e1 9Yy Ye pottddns sem snzoydsoyg ‘wntsousew pue ‘Singzins ‘umtotes ‘snzoydsoyd Fo oyeydn [e072 07 93eX USSOIZIU Fo dtysuOoTIe[oI oY] (84/34) e3ezr usd0IZIN (qe: 9fE 89T 0 7h (X) 970°0 + 9°6 = A 27 0 000°0 - () oO + TETAS SY S950 (X) 90°0 + 8°87 = 4 3X O (xX) 60°0 + 7°72 = 4 FD O cL61 (ey/34) 29281 ues07T3TN tL9 9EE z(X) $0000°0 - (xX) so°O + T'Y 7%) 60000°0 - (X) OT°0 + €°8 z(X) £0000°0 - (X) 60°0 + °8 2X) £0000°0 - (X) IT°O + 0°2T TL61 °09 9mndTy 89T 0 0 OT 02 og i=] ~ OY & () 0S & ~~ > S 09 OL =k 4®@ =i SV - 08 =i 3H O = 4% 290 06 131 regular flooding of the marsh for a week or more causing an increase in the salinity of the soil solution as evaporation proceeds. Measurements of salinity of the soil solution were made only at harvest time each year. The salinity was 36 parts per thousand on 2 September 1971 and 30 to 34 parts per thousand on 11 September 1972. Although no data are available, it was suspected that there were periods of salt buildup earlier in the 1972 growing season. The standing crop of roots and rhizomes in the upper 30 centimeters was increased significantly by nitrogen fertilization in 1972 (Tab. 36). There was no response to nitrogen rates above 168 kilograms per hectare, indicating that this rate of nitrogen was adequate for maximum growth of roots and rhizomes. When root and rhizome weights are broken down from 0 to 10- and 10 to 30-centimeter layers, a variable response to fertili- zation is noted... In the 0 to 10-centimeter layer there was a significant response to nitrogen and no response to phosphorus similar to the case when the total weight of the upper 30 centimeters is considered. However, in the 10-to 30-centimeter layer there was a significant response to phosphorus but not to nitrogen. The overall means of 120 core samples show that 75 percent of the roots and rhizomes are in the 0-to 10-centi- meter layer and only 25 percent in the 10 to 30-centimeter layer. From field observations, almost all the roots and rhizomes were contained in the upper 30 centimeters and any below this depth would be an insignifi- cant part of the total weight. Results of separating roots and rhizomes from 10 randomly selected core samples indicated that in the 0 to 10-centi- meter layer, the percentages by weight were 49.6 percent roots and 50.4 percent rhizomes. In the 10 to 30-centimeter layer, only 34.6 percent of the dry weight was roots with rhizomes accounting for 65.4 percent. Since there was a higher proportion of rhizomes in the 10 to 30-centimeter layer, it is possible that the response to phosphorus in this layer is due to an increase in rhizome growth -- that is, the response to phosphorus is relatively greater for rhizomes than for roots. The increase in dry weight of the 0 to 10-centimeter layer was probably due to an increase of root growth in response to nitrogen fertilization. The total dry weight of roots and rhizomes in the upper 30 centimeters exceeds the dry weight of shoots (Fig. 61). Since the response of roots and rhizomes to fertilization is less than the response of shoots, the ratio of shoots to roots and rhizomes increases with nitrogen rate from 0.29 for the check to 0.75 at the highest rate of nitrogen. The same fertilization treatments were continued on the plots during the 1973 growing season. The yield of aboveground growth (Tab. 37) was increased significantly by nitrogen applications in combination with phos- phorus; however, yields were not as great as in 1972. Other factors such as rainfall and wind direction (which controls tidal flooding at this location) cause year-to-year variations in shoot yields. Underground growth also showed a significant increase due to fertili- zation, particularly to phosphorus (Tab. 38). 132 UOTJILTIVA JO JUSTITFFIONS ZUBITFTIUBTS ONT QIUILOFFIP JPUCITFIUSTS YsSeoTL S}ZIOFFO uTeU = Xy ted ICI O° LOT‘ IZ | 0° v8z ‘oz Repose OL OVSeSe | Omics Sil *X OwSSScal Oe Apa Sel OG SGe0c 0°S82°9T | O° P98 ‘9T | O° ZOZL‘ST cL9 OHOSSEMG lnOwCLGal Gal) Ome Chalke O°TSZ*9T | O° 00E‘9T | O° Z9T ‘OT 9¢¢ 0°920°7Z | 0°69P‘°ZZ | O'V8S‘IZ 0°O78‘9T | O° SSS‘9T | O° 80S‘ LT 891 O° VS9*ZT | O°SZPS LT | O88 ‘ZI 0°LO08‘°ZI | O° 006‘ IT | O' VIZ‘ €I 0 yideq wo oF 07 0 yideq wo OF 07 OT yidoq wo OT 02 0 ey/s¥ (ey/3) a7ey d ayeyu N (eu/3y) WYystToM AIG puNoIsMOTOg (ZZ61) doij B8utpueiS punorsmoTeg uo snzoydsoyg pue uas0I}IN FO 399FFY “9S OTQeL 133 Table 37. Effect of Nitrogen and Phosphorus Aboveground Growth at Ocracoke (1973) Aerial Dry Weight (kg/ha) N Rate (kg/ha) *X = main effects tLeast significant difference tCoefficient of variation 134 UOTJETALA FO JUSTITFFOODS qUeITFTUSTS IONE QOUSLOFFIP JULITFIUBTS ySeoT} SJOeFFO uUTeU = X, SSS O° LO“ LT OPORCIAG OAS: CG || ae Sees OMG 2 | O Olon 7 Wee = eae os 0° 7Z0*TZ 0°L06‘ rz | O°SSO‘TE | O° 8ZL°8T lo-988‘st | o°6Sr‘7z | O°ZTE ‘ST OCMC OPES SO GOL Se OROL9S Te LORSLS “SCE OR LLL: OL 0°0r8‘sz | 0°8Sz°9Z | O° IZr‘ Tz 0°80E‘8T | 0°20z‘0z | 0°60r“9T OTE “ST IMO NSLS Ve | O° S9G TC 0°6ZE‘LI | 0°8SZ°ST | 0°008°9T yadaq wo 0¢ 02 0 yadeq wo 0¢ OF OT yadeq wo OT 02 0 (ey/3) a27ey d (ey/3y4) Wy8tem AIq punossMOTOg % 8AD S0°0 TO°0 dX N S0°0 LOO al S0°0 KOO RN (ey/3») o1ey N (¢Z61) dotg Sutpueis punossmojtog uo snzoydsoyg pue uaeso1}IN FO WOFFA “BE ITIPL 135 24 [| Roets and rhizomes 33] Shoots 20 Dry weight (metric tons /ha) ~ 168 336 N rate (kg/ha/yr) Figure 61. The effect of nitrogen fertilization on the aboveground and belowground standing crop of S. alterntflora. (2) Oak Island. In 1971 at Oak Island an experiment was begun to test the theory of Adams (1963) that iron deficiency may be the cause of the chlorotic condition often observed in stands of S. alterntflora and to evaluate the effect of nitrogen fertilization. The experiment consisted of three randomized complete blocks with plots that were 1.22 by 15.2 meters with 1.22-meter borders between plots lying perpendicular to a tidal creek so that each plot contained the tall and short forms of S. alternt- flora. Iron and nitrogen fertilizers were applied in split applications 18 May, 30 June and 3 August 1971 (Tab. 39). Ammonium sulphate was the nitrogen source. The iron sources were: (1) a commercial iron chelate applied as a spray to the plants, (2) iron chelate applied in the dry form to the sediment surface and (3) ferrous sulphate applied as a spray. Samples were harvested 4 October 1971 by clipping a 0.25-square meter area from the short S. alterniflora and a 0.25-square meter area from the tall S. alterniflora in each plot. The samples were washed in tap water and distilled water to remove mud from the leaves and stalks, dried at 70° Centigrade, weighed and ground in preparation for nutrient analyses of the plant tissue. 136 JUSULIT} 0} ONP SadUsLOFFIP JPULOTFTUSTS OU 919M 919Y} BdUTS pojUesertd oe sURoW T[eIOAO ATUO|| JUBITFTUSTS ONS QOUILOFFTP JUBITFIUSTS YseoTt UOTJETICA FO JUDTITFFOODL oq eYU/BY B°OT = 91¥I UOT], | wtoy 2YysTOH TTeL § lyo'o| § 62° IT] 9v°8]60°8|8h°6| 19° OT suvoW [T[PI9AQ § €°19¢S‘Z Ov'Sst | 28°02 So°O 25 parts per thousand) and longer periods of inundation (>8 hours). The substrate of a planting site is important in its ability to support equipment, its nutrient supplying capacity, and its effect on salt buildup. Development of transplanted or seeded areas is rapid. After two growing seasons there is little difference in appearance and primary productivity of the vegetation between artificially propagated marshes and long-established natural marshes. The length of time required for a new marsh to achieve a fully functional biological role is unknown. The relationship of mineral nutrition to growth of S. alterntflora was determined by sampling plants and soils in natural stands and by applying fertilizers to natural stands, transplants, and seedlings. Results of the natural marsh sampling and subsequent regression analysis, indicated that tissue concentrations of several nutrients and several soil properties were related to productivity of S. alterniflora. Variables negatively associated with yield were salinity of the soil solution, manganese concentrations in the plants and soil and sulfur concentrations in the plants. Variables positively associated with yield include phosphorus concentrations in the plant tissue and in the soil. Results of fertilizer experiments in natural stands indicate that the productivity of some salt marshes is limited by the supply of nutrients. The standing crop of aboveground growth of S. alterntflora growing on a sandy substrate was increased significantly by additions of nitrogen alone and increased about threefold when phosphorus was also applied. In a marsh developed on finer-textured sediments, nitrogen fertilizer doubled the standing crop of short S. alterniflora, but there was no response to phosphorus. Tall S. alterniflora did not respond to either nitrogen or phosphorus. Nitrogen and phosphorus fertilizers enhance growth of transplants and seedlings artificially established on dredge spoil. These findings suggest that salt marshes may be important in the recycling of nutrients that may otherwise occur as pollutants in the estuary. LITERATURE CITED ADAMS, D.A., ''Factors Influencing Vascular Plant Zonation in North Carolina Salt Marshes,"' Eeology, Vol. 44, No. 3, Summer 1963, pp. 445-456. ANDERSON, C.E., 'A Review of Structure in Several North Carolina Salt Marsh Plants," Ecology of Holophytes, Academic Press, New York, 1974, pp. 303-340. BERENYI, N.M., "'Soil Productivity Factors on The Outer Banks of North Carolina,"’ Ph.D. Thesis, Soil Science Department, North Carolina State University at Raleigh, 1966, Univ. Microfilms, Ann Arbor, Mich. BLACK, C.A., Sotl Plant Relattonshtps, Wiley, New York, 1968. BOURDEAU, P.F., and ADAMS, D.A., ''Factors in Vegetational Zonation of Salt Marshes Near Southport, North Carolina," Bulletin of the Ecological Soctety of America, Vol. 37, 1956, p. 68 (Abstract). BROOME, S.W., "An Investigation of Propagation and the Mineral Nutrition of Spartina alterntflora,'' Ph.D. Thesis, Soil Science Department, North Carolina State University at Raleigh, 1973, Univ. Microfilms, Ann Arbor, Mich. BROOME, S.W., WOODHOUSE, W.W., Jr., and SENECA, E.D., "Propagation of Smooth Cordgrass, Spartina alterntflora from seed in North Carolina," Chesapeake Setence, Vol. 15, No. 4, Dec. 1974 (in press). CHAPMAN, V.J., Salt Marshes and Salt Deserts of the World, Interscience Publishers, New York, 1960. COOPER, A.W., ''Salt Marshes," Coastal Ecologtcal Systems of the United States, Vol. 1, Water Pollution Control Administration, Washington, D.C., 1969. DAY, P.R., "Report of the Committee on Physical Analyses, 1954-1955; Soil Science Society of America," Proceedings of the Sotl Scetence Soctety of America, Vol. 20, No. 2, Apr. 1956, pp. 167-169. DRAPER, N.R., and SMITH, H., Applted Regression Analysts, Wiley, New York, 1966. DURAND, J.B., and NADEAU, R.J., "Water Resources in the Mullica River. Part I. Biological Evaluation of Mullica River, Great Bay Estuary," New Jersey Water Resources Research Institute, Rutgers University, Camden and New Brunswick, N.J., July 1972. GOOD, R.E., "Salt Marsh Vegetation, Cape May, New Jersey," Bulletin of the New Jersey Academy of Setence, Vol. 10, No. 1, Spring, 1965, Dp elle GOSSELINK, J.G., ODUM, E.P., and POPE, R.M., "The Value of the Tidal Marsh,'' LSU-SG-74-03, Center for Wetland Resources, Louisiana State University, Baton Rouge, La., 1974, 30pp. HINDE, H.P., "The Vertical Distribution of Salt Marsh Phanerograms in Relation to Tide Levels," Ecologteal Monographs, Vol. 24, No. 2, Api a 954;, pp. 205-225). HOAGLAND, D.R., and ARNON, D.F., "The Water Culture Method of Growing Plants without Soil,' Circular 347, California Agricultural Experiment Station, University of California at Berkeley, Berkeley, Calif., 1950. JOHNSON, D.S., and YORK, H.H., The Relation of Plants to Tide-Levels, Carnegie Institute Publication, Washington, D.C., 1915. KEEFE, C.W., and BOYNTON, W.R., "Standing Crop of Salt Marshes Surrounding Chincoteague Bay, Maryland-Virginia,"' Chesapeake Setence, Vol. 14, NOkse une MO7Se pps Ll 2 Sr KEENEY, D.R., ''The Nitrogen Cycle in Sediments-Water Systems, Journal of Envtronmental Quality, Vol. 2, No. 1, Jan.-Mar. 1973, pp. 15-28. KIRBY, C.J., Jr., "The Annual Net Primary Production and Decomposition of Salt Marsh Grass Spartina alterntflora Loisel.in the Barotaria Bay Estuary of Louisiana," Ph.D. Thesis, Louisiana State University, Univ. Microfilms, Ann Arbor, Mich, 1971. LARIMER, E.J., ''An Investigation of Possibilities for Creating Saltmarsh in the Estuaries of the Atlantic and Gulf Coasts," Proceedings of the 22nd Conference of Southeastern Game and Fish Commisstoners, 1968, pp. 82-88. MOORING, M.T., COOPER, A.W., and SENECA, E.D., "Seed Germination Response and Evidence for Height Ecophenes in Spartina alterntflora from North Carolina," Amertcan Journal of Botany, Vol. 58, No. 1, Jan. 1971, pp. 48-55. ODUM, E.P., ''The Role of Tidal Marshes in Estuarine Production," Conservationist, Vol. 15, No. 6, June-July 1961, pp. 12-15. ODUM, E.P., and DE LA CRUZ, A.A., "Particulate Organic Detritus in a Georgia Salt Marsh-Estuarine Ecosystem," Estuartes, American Association for the Advancement of Science, Washington, D.C., 1967, pp. 383-388. POMEROY, L.A., et al., "The Phosphorus and Zinc Cycles and the Productivity of a Salt Marsh," Proceedings of the Second National Symposium on Radioecology, USAEC Doc. CONF-670503, 1969, pp. 412-449. RANWELL, D.S., "World Resources of Spartina townsendit (Sensu Lato) and Economic Use of Spartina Marshland," Journal of Applted Ecology, Vol. 4, Nowa. May, 19.67%) opp 259-256). 153 REDMAN, F.H., and PATRICK, W.H., Jr., "Effect of Submergence on Several Biological and Chemical Soil Properties," Bull. 592, Louisiana Agricultural Experiment Station, Louisiana State University, Baton Rouge, La., 1965. REED, J.F., "The Relation of the Spartinetum glabrae Near Beaufort, N.C. to Certain Edaphic Factors," Amertcan Midland Naturalist, Vol. 38, No. 3, Nov. 1947, pp. 605-614. SENECA, E.D., "Germination and Seedling Response of Atlantic and Gulf Coasts Populations of Spartina alterntflora,"" Amertcan Journal of Botany, (in press). SERVICE, J., "A User's Guide to the Statistical Analysis System," Student Supply Stores, North Carolina State University at Raleigh, N.C., 1972. SHEA, M.L., WARREN, R.S, and NIERING, W.A., "The Ecotype-Ecophene Status of Varying Height Forms of Spartina alterntflora," Bulletin of the Ecologteal Soctety of America, Vol. 53, No. 2, June 1972, pp. 15-16. SMALLEY, A.E., "The Growth Cycle of Spartina and its Relation to the Invertebrate Populations in the Marsh," Proceeding Conference Salt Marshes, Marine Institute, University of Georgia, Athens, Ga., Apr. 1959, pp. 96-100. STALTER, R., and BATSON, W.T., "Transplantation of Salt Marsh Vegetation, Georgetown, South Carolina," Feology, Vol. 50, No. 6, Aug. 1969, pp- 1087-1089. STROUD, L.M., and COOPER, A.W., "Color Infrared Aerial Photographic Interpretation and Net Primary Productivity of a Regularly-Flooded North Carolina Salt Marsh," Water Resources Research Institute, North Carolina State University at Raleigh, Raleigh, N.C., Nov. 1968. SYERS, J.K., HARRIS, R.F., and ARMSTRONG, D.E., ''Phosphate Chemistry in Lake Sediments," Journal of Environmental Quality, Vol. 2, No. 1, Jan.-Mar. 1973, pp. 1-14. TEAL, J.M., "Energy Flow in the Salt Marsh Ecosystem of Georgia," Ecology, Vol. 43, No. 4, Autumn 1962, pp. 614-624. TEAL, J.M., and KANWISHER, "Gas Transport in the Marsh Grass, Spartina alterntflora," Journal of Experimental Botany, Vol. 17, No. 51, May 1966, pp. 355-361. TOWNSEND, L.R., "Effect of Nitrate and Ammonium Nitrogen on the Growth of the Lowbush Blueberry," Canadian Journal of Plant Science, Vol. 46, No. 2, Mar. 1966, pp. 209-210. 154 VAN DEN DRIESSCHE, R., "Response of Conifer Seedlings to Nitrate and Ammonium Sources of Nitrogen,'' Plant and Sotl, Vol. 34, No. 2, Apr. 1971, pp. 421-439. VAN SCHREVEN, A.C., '"Kiem-en Kweekproeven met Spartina townsendit Groves,"' Acta Botantea Neerlandtca, Vol. 7, 1958, pp. 205-216. WILLIAMS, R.B., and MURDOCH, M.B., ''The Potential Importance of Spartina alterniflora in Conveying Zinc, Manganese, and Iron into Estuarine Food Chains," Proceedings of the Second National Symposium on Radtoecology, USAEC, CONF-670503, Ann Arbor, Mich, 1969. WOODHOUSE, W.W., Jr., SENECA, E.D., and BROOME, S.W., "Marsh Building with Dredge Spoil in North Carolina," Bull. 445, North Carolina Agricultural Experiment Station, North Carolina State University at Raleigh, Raleigh, N.C., July 1972 (available from NTIS as CERC Reprint R2-72). WOODHOUSE, W.W., Jr., SENECA, E.D., and BROOME, S.W., "Establishing Salt Marsh on Dredge Spoil," Proceedings of Fifth World Dredging Conference, Hamburg, Ger., June 1973, pp. 51-78. 155 97 “ou 8 wa1gcn’ €79 97°ou waLgsn* €0720L *(J0e1}uU0D) (SeTtas) *zoygne qufof *°q*g ‘eoeueg “IT ‘“eTATL ‘1 ‘“TFods *¢ uot Ae8edorg - PLOTJTUIEZTe euTJIeds *Z ‘uoTJeRaSaA - SaysieW TepTyL °[ *poutuiezep ostTe sen Duo7zfimtaz7D “9 Jo ARTATIONpord 07 uoTATAINu TerouTM Jo drysuotjepTea ayyZ *[TnJsseoons eteM spoyjeu SufjueTdsuei} pue Zufpess yjog *‘seuTTeLoYys B3uTpoze pue Ttods espeip jo auoz TepTjzejUT ayy uy ‘(ssea3px10d yoous) DLT {Ute 1D purqapdg yo uot e8edord sy, 103 pedoteaep saenbtuyoej, saeqt1zoseq “ccl-zc, *d :kydeasoz{qrg “(TL00-O-ZL-ZLMOVE PUB CL00-9-0L-ZLMOVG 29P1QUOD 19}Uap yoreessy SufzrseuTzug [TeIseoD *S*n) (9H ‘OU wWNpUeTOWeW TeOTUYDeL *19}Ue8D YOIeesoy BuTiseutTsug [eqyseog *s*n) “*sqzeyo ‘*snTTT °d cect “7Z6l ‘zaqUeD YoIeesey ZuTIeeuT3ug [eqseoD *s*p ‘*eA SATOATAag 310g *(szeyzO0 pue) eoeues “q*g ‘ar ‘asnoypoos, *M*mM Aq SjueudoTeaep ysaeu q[es pue uotqezTTTqeq4s e,ez4sqns Toy exroTJTurejTe eupqyzeds jo uot ,esedo1rg “MN WETTTEM *esnoypoon 97 ‘ou wa,gcn* £069 97-0U watgcn* £02OL *(30821}uU09) (SaTras) *zoyqne qufof **q°g ‘eoeueg “IT ‘eTaTL ‘1 ‘TFods *¢ -‘uozieSedorg - BLOT JTuIERTe eutqaeds *Z ‘uoTIeJeseA - SoysreW TePpTL “| *pouptutejep osTe sem nuo7fimiaq 7D “9 jo ARFATIONpord oF uoTZTAqnu [eazouzw jo drysuotiepeaz sayz “[nyJssevons alan spoyjeu SutjueTdsuerz} pue Suptpees yjog *‘seutTeaoys Z3utpoxre pue {[tods a8perp Jo auoz [eptjisjuT eyi ut ‘(sseaBp109 yjoous) nL07f2U1eq7D puiqzaindg yo uotqe8edord ay, 103 pedoteaep senbtuyoeq, saqtsoseq “cCl-cSL ‘d :Ayudeasortqta “(ZL00-0-ZL-ZZNOVA PUB S100-0-0L-ZLMOVG JOeAIUOD 1Aa}UeD yoiresey SufzieeuTZuq [eqseoD *S‘n) (9y “OU wNpuUeAOWE_ TeOTUYoe] *1eque) Yyoteesoy BufyitseuTsuq Tejseop *s*n) “*sqaaeyo ‘**snT{TT *d cc, “LOL ‘zequep yoreesey ZuTiseuTZuq Teqseog “sn ‘eA SATOATEgG 310g *(S1ayIO pue) eooues *q'g **ar Sasnoypooy *M*mM Aq ‘jueudoTeaep ysiew 4Tes pue uoFIeZTTTqGeys eJeAAsqns TOF eAOTJTUIeITe euT reds Jo uotIesedo1g “M WETTTTM ‘esnoypoon 97 “ou walLgcn’ €79 HPCs woLgsn* €0ZOL °(90813u09) (seftes) *zoyjne jutof S*q*y ‘edeues ‘II “eTIFL ‘1 “*Trods °¢ cuozqe3edo1g = PlOTJTuIeITe euTIIeds *Z ‘“uoTIeJaBaA - SoysreW TepTL “[ *peuptutejep ostTe sem nto7zfiumisq1D *¢9 yo ARTATIONpord OF uoTATAQNU TereuTW jo dfysuotjzepter sup *[nyJsseoons ateM spoyjeu suTjueTdsueiq pue 3utpees yj0g ‘SauUTTeL1oYs B3uTpore pue TTods aspeip JO euoz TeptTjizejUyT ayi ut ‘(ssei3p109 yjoows) DL0OZ{[1Uu18470 purqzapdg yo uot,esedoid ayy 103 padotTeasp saenbtuyoeq saqtazssaq “GGt=zc, *d :Ayder80tT qT “(ZLO0-9-7L-ZZNOVE PUP GL00-0-04=Z/NOVA 39P1TWUOD Aa}UED yoreesey BuTiseutTZuq Teqseop °*S°n) (97 ‘OU wNpuUerTOWA TedTUYde, *‘Zajuep) yoteesey ButTzrseuTsugq Teqseog) *s°n) ‘sqzeyo S*snqTTr *d cc, “PL6l *laqUeD yOTePesey SuTTeeutTsuq [Te seo) *Ss*n ‘S*eA SATOATEeG 3IOq *(S19ayI0 pue) eosues “q’g ‘*ar ‘asnoypoom “N° Aq ‘jueudoTenep ysrew Tes pue uoTIeZTTTqGeIs e,eARSqns ATOZ eLOTFTuIeATe eutqszeds jo uot Iesedorg “M WETTTTM ‘esnoypoom 97 “ou 8 wa1esn’ €729 97°ou wargcn* £070L *(30e1QuU09) (Sseties) *zoyqne jupfof ‘°q*g Seoeues “Il ‘eTITL ‘I ‘*TFods *¢ ‘uopje3edo1g - P1OTJTUTSRTe euTJIeds *Z ‘UOoTIeIaeBaA - SoYsTeW TepTL “| *peupTuiejep osTe sem Dt07f1Uleq7D °¢ Jo AjzTAFIONpoad of UOTAFARANU TerseuTW jo drysuotTzeTer syL *[nJssaoons 31am spoy jeu SuTjueTdsueijz pue 3utTpees yjOg “seuf-Tez0ys B3utTpoze pue T[Tods aspeip Jo auoz TepTjzAejUT 9y i UT ‘(Sssea3p109 YyJooUsS) DLO7{[1U1e471D purqzupdg jo uot,e8edoid ey3 103 pedoteaep senbfuyseq seqtazoseq “cSt-zot *d :hydeaotT qT *(ZL00"0"72L-ZLMOVE PUP CL00-0-02=Z7ZMOVG 39BAIUOD 19}UaD yoieesey SutT1eeuT¥ug Te3seoD *S°n) (9H *OU uMpueIOWEW TeoTuUYyoe] ‘ZajUeD) YoTeesey BuTleeuTsuq Teqseop *s*n) “sqzaeYyo ‘*°snTTT *d cL “7L6l *zaque) YyYoreeSey BuTAVeUTSUy TeIseoD “Sn S*eA SATOATSG 310g *(S19Y4R0 pue) eoaueg ‘q*g **ar ‘asnoypoom *M*mM Aq ‘jueudoTeAep ysiew Tes pue uoTJeZTTTqeqs e,eaqsqns A1ojF eAOTJTUIeqTe euTaeds jo uoTzesedo1g ‘mM WETTTEM ‘esnoypoom 947 *ou w31gcn° £79 97 °Ou wIigcn’ €0720L *"(J9e81}u00) (SeTires) *Zoyzne jutof *°q°g ‘eoeueg “II “eTITL ‘I ‘“TFOds *¢ ~*uotTqe8edozrg - PioTjTuteszTe euTqjieds *Z ‘uot }EJeseA - SaysiewW TepTL °f *poututejzep osTe sem Duo7fiusleaq7D *¢ Jo ARTATIONpord oF uoTITAz4nu TereuTu jo dtysuotjepez ayy *[NhJsseoons viem spoyjeu SutjueTdsuei} pue Sutpeess yjog ‘seut[Tezoys 3utpora pue [tods aspeip JO ouoZ Tept}IezUT ay. uz ‘(ssea3pz09 yqoous) Dio7 f2mraq7D puzzapdg jo uot e8edord ayy 103 pedoteaap senbruysay Saqtiosaq “SSt-7G, *d tAydeasotT qT "(ZL00-0-ZL-ZLMOVA PUP $L00-0-0L-ZLMOVG 39P41QUOD 1aqUED yoreesey Suyiseutsug [e3seoD *s*n) (gy ‘OU wNpUeTOWA_ TeOTUYDeL *IeqUe) YyOLeesoy ButTiseuTsug Teqyseog *s*n) *sqzeyo ‘**snqTT *d cel “PL6L ‘Zequep yoieesey BuTiseuTsuq [eqseog *s*n **eA SATOATeg 10g aaeaee pue) eoeues “qd*g ‘*ar Sasnoypooy *M*mM Aq SjueudoTeaap ysaeu 4Tes pue uoTIeZTTTqeIS BJeIASGNS TOF eAOTFTULERTe eutyizeds jo uot zesedo1rg “M WETTTIM ‘esnoypoom 97 ‘ou witiecn’ £79 97° ou waitecn* €07OL *“(909B1}uU09) (SaTi9s) *zoyane jutof ‘°q°g ‘eoeueg “IT ‘eTatL ‘1 ‘Ttods ‘¢ *uoT zesedoirg - B10TJTUIaITe euTqjaeds *Z ‘uoT}JeJedaQ - SayszeU [ePTL ‘| *poututelap osTe sem nio7fiuile4z7D “g jo Ajz~ATIONpord oF uoTZyTAynu Terzoutw Fo drysuotjepTea ayy "[n}Jsseoons alan spoyjeu sutjueTdsuez} pue Zutpaes yjOog ‘seutT[etoys 3utpoze pue Ttods espeip Jo auoz TepTjiejuyT oy up ‘(ssea8pi0. yqoous) vio7s1Uutsez7D purqzapdg jo uot eZedo1rd ay, 103 pedoteaep senbtuyoej seqtiosaq “GGL-zSL “d :Aydeasotq{qtg “(@L00-9-ZL-ZZMOVA PUB C100-0-0L-ZZNOVG }9eARUOD 12qUAaD yorresey Suyiesurtsug Te3seoD *S‘n) (9H “OU wNpUeAOWAW TeOTUYDeZ *1eqUe) YoTeesoy BupyiseuTZuq [eqjseoy “s‘n) *sqaeyo ‘*snqT{TT *d coy “LOL ‘lequeD) Yyoieesey BuyieeuTsugq Teqseog ‘sn ‘*eA ‘ATOATeg J40q *(s19y30 pue) eos0ues *q*g ‘*ar ‘Saesnoypoony *M*mM Aq ‘juUeudoTeAep ysieu Tes pue uoTIeZTT Tqe4s ejeiq4sqns 10JF eAOTJTuISITe euTAzeds jo uotzesedo1g “M WETTTIM ‘esnoypoon 97 “ou waLgcn’ £79 97e0u wa},gcn* £0ZOL *(90P12uU0D) (SeTiIes) ‘zoygne qutof S°q*y ‘eoeueg “II ‘eTITL *I ‘Ttodg *¢ *uofzesedorg - PLOTJTUIERTeS euTIIedS *Z ‘UOTIeIeZaA - SoYsAeU [TepT] “[ ‘peutwtejep oste sen nio7zfiusez1D “9 Jo AQTATJONpord 07 uoTITAANU [TereuTW jo dfysuotzeTer sys *[njgsseoons 31am spoyjou suTjueTdsuer} pue 3uTpees yjog *SauTTetoys Z3utTpoare pue {tods adpeip Jo suoz TeptjtequT 9yQ UT ‘(Ssei3pi09 yJoous) DL07zf1U1841D purqapdsg Jo uotqesedoid ay 10g padoteAep senbtuyseq seqtaosaq “GGl-zs_ “d :AydersotTqTg “(ZL00-9-7L£-Z2MOVG PUP CLO0-9-0L-Z/MOVd 39¥4IU0D I1aqUaD yoreesey BuTLseuTsuq TeIseog *S*n) (97 ‘OU WNpUeIOWA TeoTUYyoa]L *Zajuep yoteesay BuTiseuTsug TeAseoD *S*n) *sqzeyo S*snqTTt *d cc, “Plot ‘ZeaqUeD YDteessy ButresutTsugq TeIseoD “s*n **eA fATOATAgG 3A0g *(S19y}0 pue) eossues “q’*g ‘*ar Sasnoypoom *y*mM Aq SjueudoTenap ysreu Tes pue uoTJeZTTTqeq3s eJe1ASqnNS OF eALOTFTUISRTe eu zedsg jo uoTWesedo1rg “M WETTTTM Sesnoypoom OV SOUS Sw eG Nts €729 9750u warssn* €07OL *(q0P1JU0D) (SeT1aS) *zoyjne jutol S°q°gq Seoeues “II “eTITL “I ‘“Ttods *¢ uot zesedorzg - B1OTjJTureqzTe eutqieds *Z ‘uoTRIeJaZaA - SaysTeW TepTL “1 “pouTutejep osTe sem Doz fiUMloq1D “¢ yo Ajtatqonpoad 09 uoTzFAANU TeLouTw jo drysuoTzeTer ayy *[njsseoons 310m spoyjeou sutjueTdsueizj} pue Sutpses yjOg “‘seutTetoys BSutTpozs pue TtTods a3peip Jo auoz [eptzi1sjuT 9y} UT *(SSei8p10D YQOOUS) DuO7Zf1UuIAa71D pu1qzapdg jo uotjesedoid ay} 103 padoTeaap senbtuyoeq saqTtazoseq “sci-zct *d :Aydeasortqta “(ZL00-9-Z22-ZZMOVE PUP CL00-0-0L=ZLMOVG 39BAZUOD 1ejUED yoleesey BSuTiseuTsuq TeqseoD *s°N) (9y “OU uMpUeTOWEW TeOTUYDE], *ZoqueDd yoTeesey BSuTiseutT3uq Te seoD *S*n) *szateYyD ‘*snqTTT °d GCL “PL6L ‘leque) yoreesey BuTisauT3uq Teqyseog “s'n S*eA SATOATOG 340g *(S19Y470 pue) eosues “"q’y S*ar fasnoypoom *M*m Aq ‘SjueudoTeAep ysieu [es pue uoTJeZTTTqeqs eIeIASGNS 1OF eAOTJTUADRTe euTJzeds jo uot zesedo1g ‘M WETTTTM ‘esnoypoom Ate \ i iy i a iy Son