This item is only available as the following downloads:
1 RESTORATION STRATEGIES FOR IMPROVING SURVIVAL AND COMPOSITION OF PLANT SPECIES NATIVE TO COASTAL DUNES IN THE FLORIDA PANHANDLE By NATALIE N. HOOTON A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011
2 2011 Natalie N. Hooton
3 To my two best friends and biggest supporters, LSH and MJH
4 ACKNOWLEDGMENTS I thank my advisors Dr. Debbie Miller and Dr. Mack Thetford for their time and expertise as well as my committee members Dr. Sandra Wilson and Dr. Carrie R einhar d t Adams for their guidance and encouragement throughout this pro cess. I greatly appreciate the opportunities provided by the Florida Department of Environmental Protection and the support of my supervisor, Amy Baldwin Moss and co worker Beth Fugate. I thank the Gulf of Mexico Foundation who helped to fund this study. I would also like to thank the following people for their contributions: Pat Frey and Keona Muller for their extensive assistance with la b and greenhouse work; Tova Spec initi ating this whole process; Dr. Jack Putz for allowing me to find my way and follow my mentor and friend when I needed it the most. I am f orever grateful to my familial support system.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Stresses on Coastal Dune Ecosystems ................................ ................................ .. 11 Importance of Coastal Dune Ecosystems ................................ ............................... 11 Ecological Restoration of Coastal Dunes ................................ ................................ 12 2 THE EFFECTIVENESS OF SURROGATE WRACK ON PROMOTING THE SURVIVAL AND GROWTH OF PLANTED COASTAL DUNE VEGETATION ........ 16 Background ................................ ................................ ................................ ............. 16 Methods ................................ ................................ ................................ .................. 22 Field Site ................................ ................................ ................................ .......... 22 Species ................................ ................................ ................................ ............. 23 Study Design ................................ ................................ ................................ .... 24 Surrogate wrack effects on sea oats planting ................................ ............ 24 Surrogate wrack and interplanting with herbaceous species ..................... 2 4 Data collection ................................ ................................ ................................ .. 25 Surrogate wrack effect on sea oat planting ................................ ................ 25 Survival of herbaceous plantings ................................ ............................... 26 Dune profiling ................................ ................................ ............................. 26 Data Analysis ................................ ................................ ................................ ... 26 Surrogate wrack effects on sea oats plantings ................................ ........... 26 Herbaceous interplanting ................................ ................................ ........... 27 Results ................................ ................................ ................................ .................... 27 Sea Oats Growth ................................ ................................ .............................. 27 Survival ................................ ................................ ................................ ............. 28 Sand Accumulation ................................ ................................ .......................... 29 Discussion ................................ ................................ ................................ .............. 29 Concluding Remarks ................................ ................................ ............................... 36 3 THE IMPORTANCE OF RECOGNIZING DISTINCT ECOTYPES FOR RESTORATION USING TWO SPEC IES OF GOLDENASTER ( Chrysopsis godfreyi AND Chrysopsis gossypina spp. cruiseana ) AS THE MODEL ORGANISMS ................................ ................................ ................................ .......... 49
6 Background ................................ ................................ ................................ ............. 49 Methods ................................ ................................ ................................ .................. 53 Collection Areas ................................ ................................ ............................... 53 Collection Metho ds ................................ ................................ ........................... 54 Initial Germination and Viability test ................................ ................................ 54 Germination Trials ................................ ................................ ............................ 55 Data Analysis ................................ ................................ ................................ ... 57 Results ................................ ................................ ................................ .................... 58 Discussion ................................ ................................ ................................ .............. 59 Concluding Remarks ................................ ................................ ............................... 62 4 IMPLICATIONS FOR PROMOTING VEGETATION ESTABLISHMENT AND RECREATING NATURAL DIVERSITY IN COASTAL DUNE RESTORATION ....... 70 LIST OF REFERENCES ................................ ................................ ............................... 74 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 81
7 LIST OF TABLES Table page 2 1 Mean height, width, number of tillers, and aboveground biomass of U. paniculata samples collected from Johnson Beach, Gulf Islands National Seashore in response to surrogate wrack. ................................ ......................... 38 2 2 Overall mean survival (%) of the herbaceous dune species in response to planting with Uniola paniculata and surrogate wrack. ................................ ......... 39 2 3 Ove rall mean survival (%) of three dune species planted in late July 2010 at Johnson Beach, Gulf Islands National Seashore. ................................ ............... 39 2 4 Mean survival (%) of three dune species in response to planting with Uniola paniculata (sea oats) and surrogate wrack. ................................ ........................ 40 2 5 Mean difference in sand accumulation in Uniola paniculata plots by treatment .. 40 2 6 Mean difference in sand accumulation in herbaceous only plots with and without surrogate wrack ................................ ................................ ...................... 40 3 1 Flower and seed production in two forms of Chrysopsis godfreyi and one subspecies of Chrysopsis gossypina. ................................ ................................ 63 3 2 Characteristic morphology, germination and viability percentages of the three Chrysopsis spp. as determined by studies conducted by MidWest seeds starting in June 2010 (n=100). ................................ ................................ ............ 64 3 3 Mean germination of all three Chrysopsis species after 55 days in a growth chamber in response to photoperiod and temperature. ................................ ...... 64
8 LIST OF FIGURES Figure page 2 1 Mean monthly maximum and minimum temperatures for Perdido Key, Florida during the time of plant installation and monitoring in 2010 compared to historic mean monthly temperatures ................................ ................................ .. 41 2 2 Total monthly precipitation at Perdido Key, Florida during the time of plant installation and monitoring in 2010 compared to historic mean total monthly precipitation ................................ ................................ ................................ ....... 42 2 3 Location of the six replicate sites within Gulf Islands National Seashore Johnson Beach, Perdido Key, Florida. ................................ ............................... 43 2 4 Photographs, descriptions and geographic origins of the six species included in the herbaceous plot designs. ................................ ................................ .......... 44 2 5 Block and planting design ................................ ................................ ................... 46 2 6 Mean change in sand height by 61 cm segments across sea oats plantings with and without surrogate wrack that occurred during a three month time period ................................ ................................ ................................ ................. 48 2 7 Mean change in sand height by 61 cm segments across herbaceous species only plantings with and without surrogate wrack that occurred during a three month time period. ................................ ................................ .............................. 48 3 1 Location of 13 individual Chrysopsis spp. plants at the Naval Air Station in Pensacola, FL that had seed heads harvested in December 2009. ................ 65 3 2 Location of 20 individual Chrysopsis spp. plants at Perdido Key State Park in Escambia County, Florida that had seed heads harvested in December 2009 .. 65 3 3 Photos and distribution maps of the three Chrysopsis spp. collected ................. 66 3 4 Mean germination (%) of three Chrysopsis spp. over eight weeks in a greenhouse in Fort Pierce, Florida from September 2010 November 2010 ....... 67 3 5 Total mean germination (%) of all seeds of Chrysopsis spp. under different temperature regiments representing Florida seasons ................................ ........ 68 3 6 Total mean germination (%) of all Chrysopsis spp. seeds in response to temperature and photoperiod ................................ ................................ ............. 68 3 7 Mean germination (%) of three Chrysopsis spp. over eight weeks in under two photo periods ................................ ................................ ................................ 69
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the R equirements for the Degree of Master of Science RESTORATION STRATEGIES FOR IMPROVING SURVIVAL AND COMPOSITION OF PLANT SPECIES NATIVE TO COASTAL DUNES IN THE FLORIDA PANHANDLE By Natalie N. Hooton August 2011 Chair: Debbie Miller Cochair: Mack Thetford Major: Interdisci plinary Ecology The increasing number of natural and anthropogenic stresses on coastal dune ecosystems requires the use of more effective restoration strategies to enhance dune building, increase veg etati on r e e stablishment and maintain plant diversity. In this study, t he use of a surrogate wrack was an effective method to improve growth of spring planted Uniola paniculata Mean aboveground biomass of U. paniculata 6 month s after planting in plots with surrogate wrack was 9.25 g 1.00 g compared to 2.18 g 0.24 g for those without surrogate wrack Number of tillers, tiller height, and basal width w ere also greater at the end of the first growing season for plants treated with the surrogate wrack ( p <0.05 ). Survival of three Chrysopsis species, Physalis angustifolia, and Oenothera humifusa was not improved by the presence of a surrogate wrack or when interplanted among U. paniculata Revegetation strategies for these species should focus on relative placem ent on the dunes instead of the addition of organic matter. The difference in sand accumulation was marginally significant between U. paniculata plots with the surrogate wrack and those plots with U. paniculata but no wrack ( P = 0.1 093 ) The increased sa nd accumulation suggests that a surrogate wrack can either directly or
10 indirectly trap more sand by creating an additional obstacle or promoting the growth of dune grasses which create additional barriers which hold greater sand. Maintaining natural diversity of plant species influences th e success of habitat reestablishment and the restoration of autogenic processes. Three morphologically distinct Chrysopsis have been observed in the coastal dunes of the western portion of the Florida pa nhandle. These three goldenasters presumably Chrysopsis godfreyi f. viridis (CHGOV), Chrysopsis godfreyi f. godfreyi ( CHGOG ) and Chrysopsis gossypina spp. crusieana (CHGOC) differ in terms of increasing pube scence on the l eaves and stems. Seed morphologies, seed viability, and germination also diffe red between the three ( P <0.05). Mean germination of all seeds was highest in the incubator studies with altern ating temperature regimes of 25 C/15 C and 20 C/10 C, most similar to mean fall and winter seasonal temperatures Whil e germination between the three sample populations did not differ due to an interaction with temperature regimes and photoperiod, overall germination under all tested temperature regimes and photoperiods was lowest for Chrysopsis godfreyi f. godfreyi These findi ngs indicate that two distinct ecotypes of C.godfreyi as well as C. gossypina spp. crusieana may exist and will help determine the number of seeds that must be collected and the best time to germinate them to reach target numbers for restoration outplanting. In addition to morphological and physiological differences, our field observations s uggest that CHGOG prefers a different type of microsite than the CHGOC and CHGOV With these characteristics considered, maintaining separate stock for morphologically distinct Chrysopsis spp. during production would be most appropriate.
11 CHAPTER 1 INTRODUCTION Stresses on Coastal Dune Ecosystems Coastal dunes are threatened throughout the world by various natural agents and human activities (Defeo et al. 2009). They are dynamic systems that experience high degrees of disturbance even under natural conditions from strong winds, high tides, and storm surges. On the Gulf of Mexico, dunes on barrier islands can take 5 to 10 years to recover from hurricane impacts (Perrow and Davy 2002). Plant species that persist in this environment must be capable of enduring high evaporation rates, limited access to macronutrients, blowing sands, and sa lt stress (Snyder and Boss 2002). In addition to these natural stresses, coastal dune ecosystems are subject to several anthropogenic factors. With 30% of the population in the United States living in coastal counties and up to 75% predicted to live th ere within the next 15 years, development and engineering will continue to be a major concern for coastal dunes (Hinrichsen 1999, Crowell et al. 2007). Additional stresses come from recreational use and the often associated beach nourishment and grooming practices, both terrestrial and marine pollution, introduction of invasive species, and coastline retreat related to sea level rise from global climate change (Defeo et al. 2009). With so many factors confronting the coastal dune ecosystem, one must consi der the importance of its continued existence. Importance of Coastal Dune Ecosystems Coastal dunes which cover 20% of all coastal landscape areas provide irreplaceable ecosystem services and natural capital stocks as well as offer socially valued opportun ities (Acosta et al. 2005, Wilson et al. 2005, Martinez et al. 2007). In
12 terms of ecological values, coastal dunes support high biodiversity and a large number of endemic vegetative species; provide habitats for threatened and endangered plant species tha t are critical for nesting and foraging activities; filter water and recycle nutrients (Martinez and Psuty 2004, Schlacher et al. 2007, Lomba et al. 2008). The economic benefits of the sand shore and dune ecosystem are substantial. It provides protection and reduces impacts associated with disturbances such as hurricanes, flooding, and storm surge on coastal communities and can also regulate coastal erosion (Wilson et al. 2005, Martinez et al. 2007, Schacher et al. 2007). Additionally, in many countries the collection and use of renewable biotic resources such as wood from coastal dunes provide a source of income. Opportunities to capitalize on natural products for genetic, medical, and ornamental purposes have also been explored (Wilson et al. 2005). Tourism is also a major source of revenue for coastal communities. People are willing to travel great distances and spend their earnings to enjoy the aesthetic beauty that only the sandy shores can provide. During their time at the coast, people often en gage in recreational activities such as driving, beachcomb ing, and camping that typically have damaging effects on the coastal environment. Ecological Restoration of Coastal Dunes Unfortunately, many of the practices that are economically and socially ben eficial to humans often degrade the sandy shore and coastal dune environment altering important ecosystem functions and processes (Wilson et al. 2005, Martinez et al. 2007, Schlacher et al. 2007). Dune ecosystems are dynamic and subject to natural stocha stic events, but their resilience to these disturbances is reduced when they are undermined by human activities (Adger et al. 2005, Pries et al. 2009). Preservation and restoration of these habitats are crucial to sustaining the ecosystem services and nat ural capital
13 stocks that are only provided by this system and are vital to the survival of coastal communities around the world. While the opportunity for man and nature to have a synergistic relationship exists, demands for natural resources and the perc eived need for ongoing development limit the possibilities for preserving land in a pristine state. Ecological restoration requires human intervention to assist and accelerate the recovery of an ecosystem that has been misused or mismanaged and/or degrade d by natural disturbances. Restoration of coastal dunes is often challenging due to the dynamic nature of that environment, changing shape and extent and therefore, shifting environmental gradients, in response to disturbances (Schlacher et al. 2007, Mill er et al. 2008). Initiating autogenic processes through the installation of appropriate native vegetation can be widely effective in restoring ecosystem functions and processes allowing for a self sustaining recovery (Whisenant 1999). Treating the soil s urface can optimize allogenic conditions to promote survival and growth of installed vegetation. Along with abiotic factors that are capable of supporting the successful establishment of primary producers, species selection can aid in realizing management objectives. The success of a coastal dune restoration project is determined based on the proper implementation of these two factors (Whisenant 1999). To better understand the interactions between restoration actions such as soil amendments and species selection and the natural ecology of the site and therefore, improve the process of assisting the recovery of natural systems, restoration should be science based, guided by sound theories and experimentation. To enhance nutrient availability, dune vegeta tion is often installed with water adsorbing polymer gels.
14 However, the effectiveness of these gels in aiding nutrient uptake is unclear. According to coastal natural resource managers in Florida, there has been limited or no increased su rvival in plants observed when installed with the gel versus those planted without it (personal communication with Florida Department of Environmental Protection (FDEP) scientists ). Therefore, alternative methods for providing nutrients and retaining water should be cons idered to encourage higher survival and increased growth rates of plants used in coastal dune restoration projects. Additionally, plant selection should continue to be evaluated on a site by site basis. Species such as U. paniculata and Panicum amarum sh ould be considered in early stages of coastal dune restoration because their root structures support sand accumulation and stabilization (Miller et al. 2008). Coastal dunes also have relatively high endemicity and are biologically diverse especially in du ne slacks (Ward et al. 2003, Martinez and Psuty 2004 ). Species that are endemic and/or endangered should be considered a critical component of restoration efforts. In 1995, two hurricanes severely impacted the barrier islands that make up the Florida portion of Gulf Islands National Seashore. To protect communities along the Gulf coast as well as provide necessary habitat for wildlife that depend on the barrier islands for shelter and food, coastal dune restoration in those areas i s needed. Addressi ng two important aspects of ecosystem restoration, this study investigated the use of a surrogate vegetative wrack layer to improve survival and promote growth of plant species installed as well as increase the rate of dune formation at the site Our seco nd focus was to further examine the presence and habits of two species of an important secondary dune genus Chrysopsis both of which are en dangered and found
15 only in northwestern coastal counties in Florida, to improve propagation methods and survival wh en included in restoration.
16 CHAPTER 2 THE EFFECTIVENESS OF SURROGATE WRACK ON P ROMOTING THE SURVIVA L AND GROWTH OF PLANTE D COASTAL DUNE VEGET ATION Background Despite being a highly dynamic environment, coastal dunes contribute substantially to both marine and terrestrial ecological functions, support unique biological communities, and provide important ecosystem services that are not offered by other systems (N ordstrom et al. 2000, Wilson et al. 2005, Dugan and Hubbard 2010). Healthy coastal dunes offer nesting and foraging habitat to various endangered species, act as a buffer for inland sites against storm events, allow for nutrient exchange and cycling betw een marine and terrestrial environments, and offer a site for many recreational activities (Wilson et al. 2003, Defeo et al. 2009). Coastal dunes which were number of environment al and anthropogenic disturbances including hurricanes, storm surge, overwash, global climate change and sea level rise, coastal development, and beach grooming all of which contribute to erosion and loss of biological diversity (Snyder and Boss 2002, Mill er et al. 2003, Wilson et al. 2003, Defeo et al. 2009, Khalil and Finkl 2009, Nordstrom et al. 2010). km of sandy shores make its coastal dunes especially vulnerable to human impacts. In 2009, over 80 million people vacationed in Florida with tourist spending estimated to have reached 60.9 billion dollars and the tourism industry employing nearly one million people. More than one third (>26.6 million) of those tourists visit beaches ( VISIT FLORIDA Research 2011) suggesting that coastal dune environments are not only ecologically important but economically essential as well.
17 Because this ecosystem has such a high value, it is necessary to not only conserve these areas, but also to attempt to restore coastal dunes that have been damaged by environmental stresses or degraded by human activities. There a re several strategies that have been considered for stabilizing eroding dunes and rebuilding severely damag ed ones all of which focus on creating biological and/or structural barriers to trap sand transported by aeolian processes (Wilcock and Carter 1977, Surrency 1992, Olafson 1997, Nordstrom et al. 2010). For sites that are easily accessible and time is extremely limited due to project requirements and/or restrictions on funding personnel, mechanical grading of coastal sand dunes has been recommended. However, the newly established slopes are still subject to erosion by wind and wave action (Wilcock 1977, Olafson 1997, Khalil and Finkl 2009) and the use of heavy machinery may have a negative impact on vegetation and existing fragmented dunes. Wilcock and Carter (1977) suggest that dune regrading should be res tricted to sites that are in early successional stages and recreation pressures are high. Wooden or fabric sand fences can decrease sand movement and increase sand accumulation in the immediate area (Wilcox and Carter 1977, Surrency 1992, Olafson 1997, Mil ler et al. 2001). They are immediately effective and can be installed at any time of the year. Once sand has accumulated to the top of the fences or the material has deteriorated (in the case of fences made of biodegradable materials such as geojute), th ey no longer impact sand movement. Additional fencing must be installed to further sand accretion in those areas (Surrency 1992, Olafson 1997, Miller et al. 2001). Dune cross over structures such as boardwalks can be used to limit foot traffic and to al low for natural re establishment of dune vegetation by providing some
18 protection from erosion (Surrency 1992). These structures can be expensive, require maintenance, and are limited in terms of appropriate placement and design. As an alternative to usin g mechanical and structural measures to encourage dune stabilization, soft engineering strategies such as planting vegetation and/or installing vegetative mats or netting can be effective. Vegetative mats or netting can protect bare sand surfaces such as t hose found in dune blowouts following severe storm events. They can be used in combination with seedling plantings or transplants propagated from cuttings, but do little to increase sand accretion rates (Olafson 1997). To promote dune building or maintai n dune structure, revegetation of the site is a relatively inexpensive, effective technique that encourages the restoration of autogenic processes which decreases the need for ongoing maintenance following successful installation and establishment (Surrenc y 1992, Olafson 1997, Snyder and Boss 2002, Miller et al. 2003, Mountney and Russell 2006, Williams and Feagin 2010). When selecting plant species for coastal dune restoration projects, several factors need to be considered. Coastal dunes are harsh envir onments and the vegetation must be adapted to tolerate considerable salt spray, widely fluctuating temperatures resulting in high evapotranspiration rates in the summer months, and fluvial and aeolian processes (Olafson 1997, Miller et al. 2001, Snyder and Boss 2002, Miller et al. 2003). Ideal vegetation for stabilizing coastal dunes are those that can withstand the afore mentioned environmental stressors and are perennial, have high survival rates, have extensive root structures, have aboveground biomass that can slow sand movement, and that can grow quickly enough to exceed sand accumulation levels (Surrency 1992, van der Laan et al. 1997, Snyder and Boss 2002, Miller et al.
19 2003). Oftentimes, perennial grasses such as sea oats ( Uniola paniculata ) beach grass ( Ammophila spp ), and bitter panicum ( Panicum amarum ) are installed to facilitate dune building after existing dunes have been fragmented or destroyed (Surrency 1992, van der Laan 1997, Miller et al. 2003). Because coastal dune environments are typ ically nutrient poor, vegetative recolonization may be slow with several coastal dune species needing three or more growing seasons to establish (van der Laan et al. 1997, Miller et al. 2001, Hannan et al. 2007). To increase growth rates and therefore inc rease sand holding capacity, several studies and land management ma nuals suggest the use of fertilizers or other soil amendments in a limited capacity for specific restoration goals such as increased first season growth or accelerated root spread (Surrency 1992, Seliskar 1995, Olafson 1997). The majority of coastal dune vegetation is adapted to the naturally low nutrient regime common in this type of habitat. However, in damaged dune systems, fertilizer assisted growth rates may be one of the determining factors of revegetation/restoration success. Seliskar (1995) found that the application of nitrogen, phosphorus, or potassium fertilizers along with limestone allowed for successful recolonizat ion of Ammophila spp. after an initial die out while sites with no amendments took up to five months to revegetate. Using hydrated hydrogel along with slow release fertilizer to accelerate the growth of planted sea oats in coastal dune restoration projects has been a common approach (Surrency 1992, Olafson 1997). For U. paniculata S. patens P. amarum the Americus Plant Materials Center for the USDA recommends adding a slow release fertilizer (ex. Osmocote ) in each hole at the time of planting or applying between 200
20 400 pounds/acre of 10 10 10 fertilizer a few week s after planting and then once annually (Surrency 1992). These strategies can be time consuming and expensive making them potentially unachievable given the financial limitations of many restoration projects. However, there are promising alternatives tha t have been considered more recently such as the use of vegetative wrack and other organic materials that can In an attempt to develop a study that could answer questions that were relevant to coastal land ma nagers I contacted Tova Spector, an environmental scientist with the Northwest District of the F lorida D epartment of E nvironmental P rotection (FDEP) She suggested that we consider th e role of the natural wrack, how it might be simulated, and if that surrogate could be usef ul for coastal revegetation. high tide line or as a result of overwash during storm surges is often considered an important marine subsidy for terrestrial coastal communities (Behnehani and Croker 1982, Orr et al. 2005, Nordstrom et al. 2010). In addition to marine originating debris, wra ck found higher on the beach and/or closer to pre existing dunes often contain s vegetative reproductive structures (culms, rhizomes, seeds) from terrestrial dune species (Nordstrom et al. 2010). It has been considered as not only a nutrient supply, but als o as a potential source of protection, as a method to retain soil moisture, as an obstacle to promote sand accumulation, and as a highly heterogeneous surface to catch seeds and therefore increase species diversity and richness (Dugan et al. 2003, Orr et a l. 2005, Cardon and Garcia 2008, Nordstrom et al. 2010, Williams and Feagin 2010).
21 The potential ecological significance of vegetative wrack on the beaches has been dismissed in the past due to its unattractive appearance which many beach goers find unfavo rable (Behnke 2010), but further consideration should be given. Recent studies have investigated variability in wrack composition (Orr et al. 2003), the effects of wrack abundance on various macrofauna communities and shorebirds (Dugan et al. 2003), and t he importance of this organic matter on enhancing coastal dune plant growth (Cardona and Garcia 2008, Williams and Feagin 2010) as well as the impacts of wrack removal through beach grooming such as a raking (Dugan and Hubbard 2010, Nordstrom et al. 2010). Orr et al. (2005 ) found that wrack deposition does not occur in a predicta ble pattern and that substratum grain size influences the amount of wrack that can be trapped both of which determine the species composition that can be supported by this material More specifically, Cardon and Garcia (2008) determined that the seagrass component of vegetative wrack acts as a nitrogen source for coastal foredune vegetation. Brown alga, another component found in wrack, can also be an important supplier of nitroge n and phosphorus (Williams and Feagin 2010). During a meeting with Tova Spector and John Bente, a FDEP biologist the two proposed the use of wheat straw as a surrogate wrack. It is an appealing option to managers because it is readily available and relat ively inexpensive. While the species composition varies substantially, there are many similarities between the vegetative wrack that washes ashore and wheat straw which is commonly applied during various restoration projects to reduce surface erosion an d to promote vegetative growth. Wheat straw contains moderate amounts of potassium and nitrogen as well as some phosphorus. It also acts as protective surface cover which will reduce
22 sand movement under it as well as retain moisture at the surface of the soil. Additionally, its uneven surface acts as a topographic obstacle to aeolian transport and for seed dispersal, encouraging recruitment of coastal plant species in that area. To determine the feasibility of using vegetative wrack and/or similar cost effective organic substances in coastal dune revegetation/restoration projects to increase the survival, accelerate the growth of coastal dune plant species, and to enhance dune building through sand accumulation, we applied a layer of wheat straw to our p lanting sit es. Using the wheat straw as a surrogate wrack, we expected to have increases in survival, growth, and sand accumulation in plots were it had been applied. Methods Field Site This study was conducted on Johnson Beach, Gulf Islands National Seas hore ( located on Perdido Key, Florida a barrier island about 24 km southwest of Pensacola, Florida. The dominant species of the study site were U. paniculata P. amarum beach elder ( Iva imbricata ), gulf bluestem ( Sc hizachyrium maritimum ), and seabeach evening primrose ( Oenothera humifusa ). Mean monthly maximum and mini mum temperatures varied from 33C (max) and 24.9C (min) in the summer to 14.2C (max) and 2.8 C (min) in the winter (Figure 2 1). Total monthly precipitation during the study period ranged from a high of 30.15 cm in August 2010 to a low of 0.97 cm in October 2010 (Figure 2 2 ). In 2004, Hurricane Ivan caused substantial coastal erosion both in terms of beach profile lowering and dune erosion on Pe rdido Key. Additionally, within the Johnson Beach Park, extensive overwash of sand occurred across the barrier island (FDEP 2004). Large overwash fans extended from the Gulf of Mexico to the back side
23 of the island (Perdido Bay) in several locations thro ughout the approximately 8 km of undeveloped land beyond Johnson Beach Road (Figure 2 3). Species Six native dune species that are native to the site and are important in dun e building and stabilization or s er v e as important food sources for nativ e dun e animals were used in this study: U. paniculata, O. humifusa, coastal groundcherry ( Physalis angustifolia ) Gulf coast searocket ( Cakile edentula ) two forms of ( Chrysopsis godfreyi f. godfreyi and C. godfreyi f. viridis ) and cottony goldenaster ( Chrysopsis gossypina ssp. cruiseana ) (Figure 2 4 A and Figure 2 4B ) U. paniculata is a native, dune stabilizing grass that occurs on primary dunes and barrier islands in the southeastern United States. It typically has a naturally fra gmented, linear distribution and reproduction occurs by wind pollination as well as clonally. O. humifusa is a native perennial that is common in frontal and interdune troughs throughout Perdido Key. They can form large clumps and promote sand accumulatio n. P. angustifolia is a native, rhizomatous perennial that produces a small, round fruit enclosed in a papery husk. On Perdido Key, it is common along roadways and can be found on the sound side of Johnson Beach in small patches. C. edentula is an annu al and is often one of the first species to recolonize a beach after a major disturbance. It is common on foredunes especially in the wrack that forms at the high tide line. C. cruiseana and both forms of C godfreyi are endangered perennial s found only i n the coastal region of Alabama and the western portion of the Florida panhandle. In Perdido Key, they are found in secondary dunes or more established areas on the sound side of the barrier island. In addition to assisting with dune forma tion and stabil ization, the six species
24 considered in this study serve as critical components of the endangered Perdido Key beach mouse ( Peromyscus polionotus trissyllepsis ) diet and/or habitat. Study Design Surrogate wrack effects on s ea oats planting Six replicate interdunal sites were selected along the roadless portion of Johnson Beach on Perdido Key (Figure 2 1). In April 2010, approximately 1000 U. paniculata plugs propagated from seed of local stock were planted within a 21.34 m by 3.96 m plot ran domly selected at each of the six replicate sites. Plants were spaced approximately 30.5 cm apart in straight lines. Two weeks later, at each replicate site, the U. paniculata plantings were divided into subplots (10.67 meters by 3.96 meters; Figure 2 5 A) with half of the suplots randomly selected to receive five bales of wheat straw. The remaining subplots did not receive wheat straw. Wheat straw was used as a surrogate for wrack. This surrogate wrack layer measured approximately 10 cm in depth. Th e straw was carefully placed around U. paniculata to avoid burial, but allow the entire surface of the treatment area to be covered. Surrogate wrack and interplanting with herbaceous species At each replicate site at least 25 m away from the plots planted with sea oats, an additional plot consisting of bare sand (10.67 meters by 3.96 meters; Figure 2 2A) and not planted with U paniculata was established. In early July 2010, five bales of wheat straw were applied to randomly selected subplots within the ba re sand plots (Figure 2 5 A). The wheat straw layer measured approximately 10 cm in depth. The remaining subplot of bare sand (10.67 m by 3.96 m) received no straw and no U. paniculata Each of the four sea oats/ surrogate wrack treatments (sea oats no st raw ; sea oats with straw ; straw no sea oats, and bare sand) within each site were further divided
25 into seven, 1.52 m by 3.96 m sub plots (Figure 2 5 B). Five species of herbaceous plants P. angustifolia, C. edentula C. godfreyi [ wit h two forms, f. godfre yi (CHGOG) and f. viridis (CHGOV ) ] C. gossypina spp. cruiseana (CHGOC) and O. humifusa were randomly installed in one of the seven areas within each treatment. Plants were planted between U. paniculata using 61 cm spacing (Figure 2 5 C). The original design included 21 individuals of each species for a select 1.52 m by 3.96 m subplot. H owever, due to variable survival rates of planted stock that number was not avai lable for all species/forms The final design included 21 P. angustifolia, 21 O. humif usa, 20 CHGOV 10 C. edentula 9 CHGOC, or 9 CHGOG randomly selected for one of the seven subplots within each of the four sea oats/ surrogate wrack treatment areas within a block. Species that had less than 21 individuals available were planted at the cen ter of the specified sub plot. Data collection Surrogate wrack effect on s ea oat planting Sea oats survival was determined one month and six months after planting. The aboveground biomass of U. paniculata was harvested from five individuals in each of th e two surrogate wrack treatments ( straw no straw ) for all six sites four and six months after planting. Individuals were randomly selected, and the coordinates were recorded to avoid harvesting the same plant during the second collection. The three oute rmost rows and columns were avoided to reduce inclusion of edge effects. Prior to harvesting, the number of tillers, the maximum height of each tiller, and the basal width of each tiller was measured and recorded. Biomass from each plant was placed in a p aper bag and closed with two staples. Sealed bags were load ed into a drying oven set at 65 C for 2.5
26 days. Masses were immediately taken and recorded after removal from the drying oven. Survival of herbaceous plantings Survival of herbaceous species wa s determined two (08/30/2010) and four (10/30/2010) months after planting. Individuals of each species were counted and that number was compared against the known number of planted individuals. Dune profiling Six, 1.52 m PVC poles were installed within e ach block. Three poles were installed at each plot two on either side, 1.07 m beyond the plot area in the center and one in between the two treatment areas within that plot also in the center. A string was tied to one of the end poles and extended acros s the plot to the other pole. A string level was used to ensure the line was straight. Measurements were taken every 61 cm using a meter stick measuring from the line down to the sand surface. This data was collected two and five months after the last wheat straw was applied and compared to calculate sand accumulation or loss Data Analysis Surrogate wrack effects on s ea oats plantings The experimental design followed a randomized complete block design with a repeated measure of survival. Each interdun al site represented a block and a replicate. Main effects were surrogate wrack treatment (presence or absence of straw ) and date (two sample dates within the growing season). Response variables were survival (%), above ground biomass (g), mean height (c m), tillers produced (no.), and plant width (cm). Residual analysis was performed and assumptions of equal variance and normality were met (skewedness was 0). Transformations were made for October data
27 as follows: width (log 10), mass (log), and tiller (l og2). To normalize August data, tiller (log), width (log2), and mass (log + 1) were used. Mixed models (PROC MIXED) repeated measures analysis was used to test main effects and interaction of wrack treatment and date. Means separation of dependant varia bles was conducted with the LSMeans procedure with a bonferonni correction. Herbaceous i nterplanting The experimental design followed a balanced split plot design with the whole plots arranged in a randomized complete block design and included a repeated measure of the response variable. Each interdunal site represented a block and a replicate. Four sur rogate wrack treatments (sea oats no straw ; sea oats with straw ; straw no sea oats and bare sand) were the whole plots and each whole plot was separated into 7 sub plots ( 6 herbaceous species and a control either non planted or with U paniculata only ). The response variable (survival) was measured on multiple dates necessitating repeated measures analysis of the data. Residual analysis was performed and assumptions of equal variance and normality were met (skewedness was 0). The transformation log+1 fo r survival was performed. Mixed models (PROC MIXED) repeated measures analysis was used to test main effects of treatment and interaction of treatment and date. Means separation of the dependant variable was conducted with the LSMeans procedure with a bo nferonni correction. Results Sea Oats Growth Four months after planting, mean tiller height (cm) and mean aboveground biomass (g) of U. paniculata planted with surrogate vegetative wra ck ( straw ) differed significantly from those planted without wrack ( F (1 ,5)=7.27, P =0.04 0.05 ;
28 F (1,5)=13.28, P =0.01 0.05 respectively; Table 2 1). Height and aboveground biomass were greater in individuals planted with surrogate wrack. Six months after planting, mean tiller height (cm) and mean aboveground biomass remained greater for U. paniculata installed with surrogate wrack ( P =0.002 0.05 ; P = 0.0002 0.05 respectively; Table 1). Additionally, the mean number of tille rs and mean plant width (cm) also varied between U. paniculata planted with and without surrogate wrack ( P =0.039 0.05 ; P = 0.003 0.05 respectively; Table 2 1). In all four growth indicator measures, the mean values were greater for U. paniculata w ith the application of wrack. The change in the number of tillers and the aboveground biomass between August and October differed significantly between wrack treatments as well ( P = 0.036 = 0.05 ; P = 0.0007 0.05 respectively; Table 2 1) with greater mean changes measured in U. paniculata plots with wrack. Survival Six months after installation, the mean U. paniculata survival in plots with surrogate wrack was 98.9 0.35% and 97.1 0.35% in plots without During late July 2010, Escamb ia County was experiencing drought conditions and many C. edentula did not transition well from greenhouse culture to the coastal environment despite a transitional period on an outdoor irrigation pad ; survival was extremely low following the transport fro m the greenhouse to the beach. Because of this stress, planted individuals were not considered in overall survival analyses. Overall survival of the herbaceous species was the lowest, 15.6 4.0% and significantly different in the wrack only plots compar ed to the other three treatments, both wrack and U. paniculata interplanted with U. paniculata and no wrack, and with bare sand ( no U. paniculata and no wrack ) (Table 2 2). Survival of the three dune species planted with both wrack and U. paniculata
29 int erplanted with U. paniculata and no wrack, and planted with in bare sand were very similar (Table 2 3) The species with the highest mean survival overall was P. angustifolia with 51.2 6.0%. P. angustifolia survival varied significantly from the overall survival of C. godfreyi f. viridis (14.4 4.0% ; Table 2 3 ) and O. humifusa (23.8 5.9 % ; Table 2 3). Mean survival of the three dune species did not vary significantly in response to U. paniculata and surrogate wrack (p=0.6894; Table 2 4). Sand Accumulation The mean change in sand height (cm) between plots with surrogate wrack and U. paniculata plots without surrogate wrack did not vary significantly ( P =0.1093), but did indicate a notable difference in sand accumulation with a mean increase of 1 1.16 1.71 cm in surrogate wrack plots and an increase of 7.78 0.78 cm in U. paniculata plots without surrogate wrack (Table 2 5). Sand height changes in herbaceous planting plots with and without surrogate wrack were highly variable and were not stat istically significant (Table 2 6). Discussion U. paniculata survival in restoration projects is variable and often differs from site to site and with environmental conditions at time of planting. Overall U. paniculata survival in our study at over 95% regardless of surrogate wrack treatment was higher than many other documented coastal dune restoration studies (Finch and Bauer 1986, Feagin et al. 2009). For example, a revegetation effort of a manmade spoil island in Clearwater Harbor, Florida showed between 46% and 75% survival of U. paniculata planted in January after six months at two sites (Finch and Bauer 1986). Three months after planting, U. paniculata survival was 22.91% for a restoration project on Galveston Island, Texas that experienced a mild drought following plant installation (Feagin et al.
30 2009). The nearly 80% mortality observed in this study suggests that while U. paniculata is adapted to relatively harsh conditions, it has a threshold for environmen tal factors such as moisture availability. Higher survival in our study could be the result of a number of factors. Our plants were installed in the first half of April 2010 when temperatures and precipitation were typical for the region. Cooler spri ng temperatures compared to summer temperatures would limit evapotranspiration and on average precipitation in April is higher than May in this region. Survival of U. paniculata can be inhibit ed due to factors associated with relative location; dune plan tings closer to the Gulf of Mexico are subjected to higher wind energy, increased exposure to salt spray, and higher levels of soil salinity than those planted behind foredunes (Thetford et al. 2005). Our U. paniculata were planted on the back side of the island where wind speed is reduced and seedlings are less likely to be excavated by shifting sands. When p lanted behind the protection of foredunes of at least 1m in height survival of Schizachyrium maritimum was nearly 100% and did not vary with distanc e from the Gulf 15 months after installation in an earlier study on Santa Rosa Island, Florida (Miller et al 2008). There is some evidence that suggests facilitative interactions can result in increased survival when plant spacing for dune species is red uced because environmental extremes are weakened by neighboring plants (Franks 2003). The spacing in our study may have facilitated survival. On Santa Rosa Island (SRI), located approximately 30 km to the wes t of the plantings in our study Miller et al. (2001) measured survival of U. paniculata seven months after a spring planting where plants were placed 100 cm apart and found a somewhat higher mortality rate of 15.8%
31 compared to our study where plants were spaced 30 cm apart. The U. paniculata planted on SRI were larger at the time of planting compared to the seedlings planted on Perdido Key, but excavation by wind was the primary cause of mortality in the SRI study which may have been prevented by the back dune location and the closer spacing of U. pa niculata in our study. U. paniculata may improve the survival of neighboring U. paniculata or other dune species, but perennial grasses have also been shown to have adverse effects on other dune species establishment and production of tillers so competit ion can also occur and therefore facilitation within U. paniculata plantings is not absolute (Keddy 1982; Franks 2003). Facilitation as an adaptation to the harsh, high energy environment of coastal dune systems seems plausible when the presence of surro unding plants improve environmental conditions for the target individual (Franks and Peterson 2003). The main constraints impacting vegetative growth in coastal sand dunes are deficiencies of water, nitrogen, phosphorus, potassium, and organic matter (M aun 1994). The addition of macronutrients such as nitrogen, phosphorus, and potassium has been shown to result in increased total biomass production, higher leaf elongation rates, and a greater number of stems in U. paniculata and Panicum amarum (Hester a nd Mendelssohn 1990). Other studies have shown that in addition to an increase in available nitrogen and soil organic matter, wheat straw can increase soil moisture content on the soil surface and in the first few centimeters of soil (Wilson et al. 2004, Pervaiz et al. 2009). In coastal dune environments, natural recruitment of seedlings has been shown to correspond with seasonal moisture availability when precipitation occurs uniformly throughout the system. In this study, the wheat straw was saturated from a
32 recent rain event when it was applied to the planted sea oat plots. The presence of the wheat straw may have kept the water from additional precipitation near the surface where the young seedlings with relatively shallow roots could access it for a longer period. Whereas infiltration rates in bare sandy soils is high and percolation through the soil profile is rapid. There is evidence that the application of straw can increase the soil moisture content in sandy soils down to 60 cm with the greate st increase observed in the top 15 cm of soil (McDonald 1934). U. paniculata in our study had greater above ground biomass and tiller number with the application of surrogate wrack. Sea oats can grow from 30 cm to 60 cm in the first year after establish ment under normal growing conditions (Hazell et al. 2010). Gormally and Donovan (2010) found that sea oats growing closer to the shoreline where soil K content was highest were taller and had higher N and K content. The mean change in height of U. panic ulata within the surrogate wrack plots in our study was almost 20 cm in only two months. Nutrient analysis was not conducted on the wheat straw used in this study, but typical values have been established for nitrogen, phosphorus, and potassium and are 6. 1 7.2 g N kg 1 0.64 1.0 g P kg 1 and 10.1 11.7 g K kg 1 respectively (Tarkalson et al. 2009, NRCS 2011). While the total amount of N, P, and K are likely not available for plant uptake, the presence of these macronutrients in the wheat straw suggest that increased nutrients may have contributed to the greater increase in growth variables of U. paniculata planted in plots where the surrogate wrack was applied. The overall adverse effect of surrogate wrack on the herbaceous species was not antic ipated. While we are unclear as to the actual mechanism that was detrimental to
33 plant survival, there are a number of possibilities. At the time of planting in July, mean daily temperatures on Perdido Key were slightly higher than average and precipitati on was substantially lower than average. While the plants were installed with soil that had been saturated, a rain event at the study site did not occur until five days after planting. The presence of dry wheat straw on the soil surface surrounding the p lants may have inhibited small amounts of moisture from reaching the root zone during a critical period for establishment as the moisture may have been absorbed by the wheat straw. Additionally, with less wind energy at the site during that time of year, the wrack remained exposed and was not compacted by the sand at the time of the initial height measurements Overall low survival among species in all treatment groups included in the study may have been a result of unfavorable temperatures and moisture av ailability for establishment for the week following planting, but there may be species specific explanations for these results. Species may have been influenced differentially by aeolian transport and sand burial at some of the sites by the time the second sand measure was collected. Previous studies suggest tha t coastal dune species respond differently to sand burial which may have been a factor affecting survival (Maun 1998). While there is evidence that suggests that interplanting with U. paniculata c an adversely affect survival of other herbaceous dune species, published accounts of this phenomena are limited to O. humifusa Although not significant, general trends between treatments indicate that perennial, clumping coastal species such as O. humifu sa may not need or benefit from organic matter additions or the presence of primary successional grasses such as U. paniculata to stabilize surrounding sand. This
34 may be due to the fact that species like O. humifusa are often some of the first to colonize bare coastal sands and are relatively tolerant of burial and aeolian sand transport (Gibson and Looney 1994, Stallins and Parker 2003). While not significant, sur vival of the two forms of C. godfreyi and C. gossypina spp. cruiseana was generally higher in plots where individuals were interplanted with U. paniculata Unlike P. angustifolia, C. edentula, and O.humifusa C. godfreyi and C. gossypina are more sensitive to salt spra y and water overflow and appear to be more abundant in swales and secondary dunes at the sites where seeds and vegetative cuttings were collected for propagation (Stopp, Jr. Because interior and sound side dunes typically have more dense vegetative cover than frontal dunes, the Chrysopsis species may have benefited from the presence of sea oats that had already been growing at the site for almost four months prior to the herbaceous species installation. Although C. edentula is a cool season annual, it was originally included in the experimental design because of its natural distribution in coastal dune environments and the potential for creating a seed bank for this species in highly disturbed dune systems. Absence of a seed bank for annual plants may consequently re sult in the species removal from that ecosystem (Maun 1998). Because this species is a potential food sources for the endangered Perdido Key beach mouse ( Peromyscus polionotus trissyllepsis ) and it can be found under various environmental conditions from open sandy areas such as blowouts to dunes established with abundant vegetation (Keddy 1982) it was an ideal representative for our study. Additionally, C. edentula fruit can float or may be carried by the wind and is often a component of the vegetative w rack
35 (Keddy 1982). Unfavorable greenhouse conditions due to various irrigation malfunctions stressed the C. edentula population which was already near the end of its annual life cycle. In addition to providing nutrients to terrestrial plant species that c an trap sand, natural wrack in and of itself can assist in dune building near the high tide line by acting as an obstacle that can reduce wind and wave energy at the surface of the soil and trap sediment (Dugan and Hubbard 2010, Nordstrom et al. 2010, Will iams and Feagin 2010). Because the wheat straw used in this study also creates an obstacle that can cause sand to accumulate around it, increased accumulation was expected in the two plots that were treated with this surrogate wrack In the two sub plo ts containing U. paniculata there was a marginally significant difference between sand accumulation in the surrogate wrack plot versus the no surrogate wrack plot supporting the hypothesis of increased accumulation with wrack (Figure 2 6) However, much of the influence of surrogate wrack on sand accumulation may be the effect surrogate wrack had on U. paniculata growth. U. paniculata capture wind blown beach sand and stabilize it with an extensive, fibrous root system (Miller et al. 2003). The extent of U. paniculata by comparing the sand accumulation of four plot types: U. paniculata only, wrack only, U. paniculata and wrack, and bare sand. Sand accumulation is greater in fall, winter and early spring in Northwest Florida than summer because of higher wind speeds and sand movement in associated with cold front passage (Miller et al. 2003). We measured sand accumulation in late summer and fa ll. Additional measurements taken prior to surrogate wrack application,
36 following complete wrack coverage as well as additional fall and winter measures of sand accumulation may more greatly accentuate effects of surrogate wrack on sand accumulation. T here was no evidence to support the hypothesis that wrack increases accumulation in the herbaceous only plots. Similar to explanations related to decreased survival in those herbaceous wrack plots, timing of application (summer) as well as precipitation m ay have influenced the changes in sand accumulation observed across the plot profile. Additionally, sand that was caught by the surrogate wrack or by the increased number of tillers produced in surrogate wrack plots may have created additional obstruction s that slowed or stopped the transport of sand from moving into the plots that did not have U. paniculata and/or the surrogate wrack. Concl uding Remarks Our findings suggest that the application of a surrogate wrack such as wheat straw can potentially bene fit coastal dune restoration projects by increasing the rate of growth of dune grasses such as U. paniculata Additionally, the presence of surrogate wrack along with U. paniculata generally improves sand accumulation in the immediate area. However, it a ppears that wheat straw can have a negative impact on planted herbaceous species when plants are installed when environmental conditions are not optimal for survival (higher temperatures, lower precipitation). Because of this potential adverse effect, tim e of planting, time of wheat straw application, and current, local conditions should be considered before using this method to promote establishment and survival of herbaceous dune species. Additional studies investigating plant response to wheat straw ap plications along the environmental gradient of a barrier island as well as those examining the similarities
37 between natural vegetative wrack and the surrogate wrack would help to advance the understanding and application of this strategy in coastal dune r estoration projects.
38 Table 2 1. Mean height, width, number of tillers, and aboveground biomass of U. paniculata samples collected from Johnson Beach, Gulf Islands National Seashore in response to surrogate wrack.
39 Table 2 2 Overal l mean survival (%) of the herbaceous dune species in re sponse to planting with Uniola paniculata and surrogate wrack Table 2 3 Overall mean survival (%) of three dune species planted in late July 2010 at Johnson Beach, Gulf Islands National Seashore.
40 Table 2 4 Mean survival (%) of three dune species in response to planting with Uniola paniculata (sea oats) and surrogate wrack Table 2 5 Mean difference in sand accumulation in Uniola paniculata plots by treatment Table 2 6 Mean difference in sand accumulation in h erbaceous only plots with and without surrogate wrack
41 Figure 2 1. Mean monthly maximum and minim um temperatures for Perdido Key, Florida during the time of plant installation and monitoring in 2010 compared to historic mean monthly temperatures Values obtained for 1985 2010 are from the National Weather For ecast Office http://www.weather.gov/climate/index.php?wfo=mob 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 Apr May June July Aug Sept Oct Nov Dec Temperature ( C) Mean Max. Temp 2010 Mean Min. Temp 2010 Historic Mean Max Temp Historic Mean Min Temp
42 Figure 2 2. Total monthly precipitation at Perdido Key, Florida during the time of plant installation and monitoring in 2010 compared to historic mean total monthly precipitation Values obtained for 1985 2010 are from the National Weather Forecast Office http://www.weather.gov/climate/index.php?wfo=mob 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Apr May June July Aug Sept Oct Nov Dec Total Precipitation (cm) 2010 Historic
43 Figure 2 3. Location of the six replicate sites within Gulf Islands National Seashore Johnson Beach, Perdido Key, Florida.
44 Figure 2 4 Photographs, descriptions and geographic origins of the six species included in the herbaceous plot designs. (A) U. paniculata, O. humifusa, P. angustifolia, and C. edentula. (B) C. godfreyi f. godfreyi, C. godfreyi f. viridis, and C. gossypina spp. cruiseana. Map credits: Wunderlin and Hansen 2008 Botanical Name Uniola paniculata Oenothera humifusa Physalis angustifolia Cakile edentula Plant Form Geographic Distribution A
45 Botanical Name Chrysopsis godfreyi f. godfreyi Chrysopsis godfreyi f. viridis Chrysopsis gossypina spp. cruiseana Plant Form Geographic Distribution Figure 2 4 continued B
46 Figure 2 5 Block and planting design (A) Block (Total of 6 blocks at Gulf Islands National Seashore) presence and absence of wheat straw is randomized. U. paniculata plantings approximately 30.5 cm (12 in) apart. (B) One of four subplots within a block location of each species and/or phenotype were randomized. In plots where U. paniculata was not planted, there will just be wheat straw or bare sand depending on the assigned treatment. (C) Planting arrangement wi thin one of seven species sub plots ( U. paniculata will not be present in half of the treatment areas). U. paniculata = 30.5 cm (12 in) spacing; Assigned species (either C. godfreyi f. viridis, C. godfreyi f. godfreyi, C. gossypina spp. cruiseana, C. eden tula, O. humifusa, or P. angustifolia ) 61 cm (24 in) spacing
47 Figure 2 5. Continued
48 Figure 2 6 Mean change in sand height by 61 cm segments across sea oats plantings with and withou t surrogate wrack that occurred during a three month time period. (n=6) Error bars represent 1 standard error of the mean Figure 2 7 Mean change in sand height by 61 cm segments across herbaceous species only planting s with and without surrogate wrack that occurred du ring a three month time period. (n=6) Error bars represent 1 standard error of the mean. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Difference in sand height (cm) Segment Number in Profile Surrogate Wrack No Wrack 3.00 2.00 1.00 0.00 1.00 2.00 3.00 4.00 5.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Difference in Sand Height (cm) Segment Number in Profile Surrogate Wrack No Wrack
49 CHAPTER 3 THE IMPORTANCE OF RECOGNIZING DISTINCT ECOTYPES FOR RESTORATION USING TWO SPECIES OF GOLDENASTER ( Chrysopsis godfreyi AND C hrysopsis gossypina spp c ruis eana ) AS THE MODEL ORGANISMS Background Coastal dune ecosystems are subject to both natural and anthropogenic disturbances. Because coastal dunes provide critical ecosystem services such as erosion and flooding control, it is important to consider how res ilient these systems are to such stresses. Typically, coastal dunes that have less human development and are in a healthy, natural state are less vulnerable to disturbances and can reduce adverse impacts on surrounding areas (Mart nez et al. 2006). In fa ct, the use of assessments determining the plant diversity for coastal dunes has been deemed a suitable tool for measuring the sensitivity of dunes to natural and anthropogenic pressures (Garcia Mora et al. 2000). In an attempt to regain many of the ecosystem services, the foundation of coastal dune restoration is the re establishment of native vegetation. Plants can contribute to the dune building process and/or assist in the rehabilitation of autogenic processes that can support animal species that depend on their presence and proper function (Fant et al. 2008). Three functional groups of coastal dune vegetation have been identified: dune builders, burial tolerant stabilizers, and burial intolerant stabilizers a s well as a fourth group of passenger species with no direct influence on topographic modification (Walker 1995, Stallins 2002). Maintaining vegetative diversity is a necessary component of restoration because in addition to their role as producers, coasta l dune plant species influence the structure of their habitat, alter the topography, and support the establishment of other species (Stallins 2002; Corenblit and Steiger 2009). In addition
50 to species diversity, genetic variation of native species must als o be preserved when revegetating coastal dunes. determined by its genetic coding as well as phenotypic plasticity (Gray 1996, Falk et al. 2001). Various genotypes and/or phenot ypes of each species are often found within a population (Gray 1996). This genetic diversity is responsible for producing traits that allow tolerance of environmental variations related to moisture availability, temperature, nutrient availability, light a vailability, and herbivory (Falk et al. 2001). Survival and fitness between and within populations as well as within individuals is influenced by the genetic variation that exists at each of those levels. Additionally, the ability of a species and/or pop ulation to persist over time when their physical and/or biological environments have shifted to a different state is determined by the genetic composition of that species (Freeman and Herron 1998, Falk et al. 2001). Because of the influence genetic variab ility and phenotypic variations of local populations can have on the success of re establishment, diversity is an important consideration when collecting plant materials for propagation in the field or purchasing them from native plant suppliers (Gray 1996 Hufford and Mazer 2003, Carney 2005, Smith et al. 2009). Vegetat ion collection protocols have recently come under scrutiny because of the potential consequence of altering the genetic composition and possibly fitness of local ecotypic populations through outbreeding and inbreeding depression, and genetic swamping (Gray 1996, Hufford and Mazer 2003, Fant et al. 2008, Smith et al. 2009). For restoration projects, limited sampling geographically and ecologically can have long term consequences related to sp ecies establishment and persistence if the range of
51 phenotypic variation between ecotypes is not represented at the sites of revegetation (Smith et al. 2009). Installation of non local ecotypes can cause the remnant population to fail due to competition w ith the new forms and/or intraspecific hybridization (Holmstrom et al. 2010). While the genetic structure of many species considered for diversity without having that informa tion available. Traditionally, researchers advocated the collection of a large number of seeds or individual plants to optimize sample diversity (Allard 1970). This approach is unreasonable for many restoration projects due to limited time for collection and resources for storing seeds and/or propagating. Fortunately, more recent studies suggest that as few as 10 34 random individuals from a minimum of five populations could be sampled and still maintain the necessary ecotypic diversity (Brown and Biggs 1991; Lawrence et al. 1995). However, the species distribution across the be influenced by their individual surroundings. Therefore, a habitat based sampling strategy wh ere a comparatively small number of individuals are collected from populations that exist in different habitats across the environmental gradient (Gray 1996). An approach that incorporates specific knowledge of a species biology and habit as it relates t o diversity within the species can further optimize sampling protocols for maintaining the natural genetic composition for a species included in a restoration project. The Center for Plant Conservation uses this strategy for sampling threatened
52 and endang ered species and places a high priority on site variations within a species to determine the number of populations to sample (Falk 1991, Gray 1996). The inclusion of endangered and threatened plant species as a component of coastal dune restoration is imp ortant for multiple reasons: they serve functional roles within their environment; coastal dunes are being degraded by natural and anthropogenic activities resulting in an overall reduction in the system type and an increased risk of species extinction; an d, increased biodiversity in coastal dunes reduces their vulnerability to disturbances. To create a sustainable population of an endangered plant species, the genetic composition and structure of other local populations should be reconstructed as differen ces in morphology and fecundity have been observed in different habitat types across a species geographical range (Fant et al. 2008, Smith et al. 2009). Chrysopsis godfreyi ), first identified in 1978, is an endangered coastal dune pl ant endemic to a few counties in the central and western panhandle of Florida (Semple 1978, DeLaney and Wunderlin 2002). It is most common on barrier islands, sometimes forming large colonies, and assisting in sand stabilization on secondary dunes (Semple of the species exist, C. godfreyi f. godfreyi and C. godfreyi f. viridis and can be distinguished by leaf pubescence with form godfreyi form viridis lacking the pu bescence. However, field observations at Perdido Key State Park and at the Naval Air Station in Pensacol a suggest that three similar morphologies exist which may indicate the presence of Chrysopsis gossypina spp. cruiseana which is native to the area, but has not been documented at the two collection sites (Figure 3 3).
53 C. gossypina spp. cruiseana displays thin pubescence on the basal leaves, the tips of some leaves, and/or on the youngest leaves, but not through out the entire stem. C. godfreyi f. viridi s lack s pubescence, appears green in color, and leaves exude a sticky subst ance. C. godfreyi f. godfreyi is densely covered with pubescence on all stems and leaves, and has a silvery white like appearance because of the large amount of hairs present. The intent of this study was to determine if the se three physically distinct types of C hrysopsis also differed in terms of seed production as well as seed viability, and germination in response to differing photoperiods and seasonal temperatures. This data co uld indicate whether it was necessary to include all apparent phenotypes in restoration efforts to maintain the morphological (potentially genetic) diversity within the local population. Additionally, the determination of germination success for the three phenotypes in response to various temperature regimes can provide important details to support the propagation of this endangered species for inclusion in future coastal dune restoration projects. Methods Collection Areas Pensacola. Escambia County, Fl elevation 9 m), (Figure 3 1). The Naval Air Station is located west of downtown Pensacola. The dominant plant species present are Chrysoma pauciflosculosa (woody goldenrod), Chrysopsis spp. (goldenasters), Conradina canescens (false rosemary), Physalis angusti folia (ground cherry), Polygonella spp. (jointweeds), Quercus geminata (sand live oak) and Quercus myrtifolia (myrtle oak).
54 Perdido Key. 872 3 2). Perdido Key State Park is located on a barrier island about 24 km southwest of Pensacola, Florida. The dominant vegetative species present are Chrysoma pauciflosculosa (woody goldenrod), Chrysopsis spp. (goldenasters), Quercus geminata (sand live oak) and Quercus myrtifolia (myrtle oak). Collection Methods In December 2009, a total of 150 seed heads from 33 individuals of Chrysopsis species were collected from the Naval Air Station and Perdido Key State Park (DAC permi t no. 888; DEP permit no. 09111111). Individual plants were selected based on the form s Plants that lacked living leaf tissue or that had both wooly and densely stipit ate glandular leaves were not selected. Once a plant was recognized for collection, a photo of the individual was taken, its identity was recorded as C. godfreyi f. viridis (green leaves with little to no pubesc ence and sticky leaf surface), C. gossypina spp. cruiseana (gre en leaves with pubescence), or C. godfreyi f. godfreyi (heavily pubescent with a silvery appearance) (Figure 3 3), and its location marked using a Garmin GPS receiver (Garmin Ltd., Olathe, Kansas). Up to five non fractured, fruiting heads were collected from selected individuals. Heads were placed in envelopes, sealed, and stored in a gro wth chamber at approximately 23 C. Total number of seeds collected from an individual plant was counted and recorded. Initial Germination and Viabi lity test Mature inflorescences were removed from each plant at each site and cleaned by hand. Immature seeds or seeds with visible indication of pathogen or insect damage were discarded. Cleaned see ds were gravity air dried at 22 C for 48 to 72 h before
55 analysis. In accordance with the TetrazoliumTesting Handbook, Contribution No. 29 Associationof Official Seed Analysts rules (Peters 2007), pre germination viability tests were replicated twice on a subset of 100 seeds per s pecies (two forms of C. godfreyi and C. gossypina spp. cruiseana ) Seeds were pretreated by allowing them to imbibe between moist blotter pa per overnight at at 20 25 C Seeds were then cut longitudinally and stai ned for 18 to 24 h at 30 to 35 C in 1.0% tetrazolium (2, 3, 5 triphenyl chlo ride)solution with positive staining patterns confirming seed viability (Mid West SeedService Inc., Brookings, SD). An additional 400 seeds per species/form were subjected to germination tests (four replications o f 100 seeds per test) at 30/20 C (8 h photoperiod at 30 C followed by 16 h darkness at 20 C) for 28 d (Mid West SeedService Inc.). Seeds were arranged in germination boxes (containing two layers of moistened blue blotter paper) that were placed in incubators equipped with cool white fluorescen t lamps. Germination readings were taken at Day 14 with a final count at Day 28.Ungerminated seed were subjected to post germination viability tests (as described previously) and used to report percent germination of viable seeds. Germination Trials Seed p reparation : Seeds were collected from 33 individuals that represented the three phenotypes/ species of interest. All collected seeds were graded and counted. Class 1 seeds were defined as fully developed, light to dark tan in color, and no signs of herbiv ory. Class 2 seeds were defined as underdeveloped or had evidence of herbivory. Those seeds likely to germinate (class 1) were used for the germination trials.
56 Four replications per treatment of 50 seeds (n=1600) of C. godfreyi f. viridis C. godfreyi f. godfreyi and C. gossypina spp. cruiseana were soaked in a Physan 20 (Marlin Products, Inc., Tustin, California) solution (1.0 mL Physan 20 per 50 0 mL of deionized water) for five minutes to reduce the amount of surface contaminants. Physan was tested pr ior and was determined to have no effect o n germination rates. Seeds were placed in 32 10.9 x 10.9 cm transparent polystyrene germination boxes (Hoffman Manufacturing, Inc., Albany, OR) containing two sheets of germination paper (Hoffman Manufacturing Inc.) that were saturated with 15 mL of nanopure water. An additional 5 to 10 mL of deionized water was added to germination boxes as needed Temper ature and photoperiod: The prepared germination b oxes were placed in temperature and light controlled chamb ers equipped with cool white fluorescent lamps (Model818; Precision Scientific, Winchester, VA). The seeds were exposed to four, 12 hour alternating temperatures 20/10 C 25/15C, 30/20 C, and 35/25 C to simulat e day and night temperatures that are simil ar to mean temperatures common to winter, fall, spring, and summer, respectively. Four replicates of each phenotype experienced a 12 hour daily photope riod for each temperature treatment (photosynthet ic photonflux was 22 to 30 mmol m 2 s 1 at sh elf level) Aluminum foil was placed around each of the remaining four replicates for each phenotype to allow for continuous darkness within those petri dishes throughout the trial. These were not opened until the end of the experiment. Treatments for the germination trials were arranged in a 3 (phenotype) x 2 (photoperiod) x 4 (temperature) factorial. A randomized complete block design was used for this study. Germination counts for seeds in the light treatment were taken every other day for 8 weeks and at 8 weeks for those in the dark treatment.
57 Differences in Seedl ing Emergence : To determine seedling emergence under typical growing conditions, additional germination tests were conducted at the greenhouse during late summer through early fall in Fort Pierce, Florida. Four replications of 50 seeds (n=200) of C. godfreyi f. viridis C. godfreyi f. godfreyi and C. g ossypina spp. cruiseana were soaked in a Physan 20 solution (1.0 mL Physan 20 per 500 mL of deionized water) for five minutes to reduc e the amount of surface contaminants. After decontamination, seeds were placed on the surface of a peat based germination media and covered lightly with vermiculite. The trays were then placed in the greenhouse and watered by seep irrigation as needed. G ermination data was collected every other day for 8 weeks. Data A nalysis Mean flower and seed production for each of the Chrysopsis spp. was calculated and analysis of variance was performed using the PROC GLM procedure in SAS software, v. 9.1 (SAS Inst itute, Cary, North Carolina). Mean germination percentage data (MidWest Seeds, Inc.) were transformed using the arcsine of the square root and analysis of variance was performed using the SAS PROC GLM procedure with a Tukey correction. Non transformed d ata are presented. For germination studies under greenhouse conditions and those with varying temperature regimes and photoperiods, mean percent germination over time was calculated. Overall mean data were transformed using the arcsine of the square roo t. Analysis of variance was performed using the SAS PROC GLM procedure and means separation was by LSMeans with a Bonferroni correction in SAS software, v. 9.2. Significance of main effects and in teractions was determined with PROC MIXED procedure in S AS
58 Results The mean number of inflorescence, the number of flower heads per inflorescence, the number of flower heads per plant, and the estimated number of seeds per plant did not differ significa ntly between the two forms of C. godfreyi and C. cruiseana (Table 3 1) The number of seeds produced per flow er head by C. godfreyi f. godfreyi (CHGOG) (118 7) was lower t han the number produced by the C. godfreyi f. viridis (CHGOV) (137 6) and C. gossypina spp. cruiseana (CHGOC) (137 6) (Table 3 1). Seeds tested by MidWest Seeds Lab from all three phenotypes indicated that the mean percentage of normal seeds was lower (40.50% 1.76%) for CHGOG than for the other t wo ( P = 0.0036; Table 3 2). CHGOC and CHGOG had significantly different percentages of dead seeds, 35.00% 2.8% and 47.25% 3.35%, respectively ( P = 0 .0342; Table 3 2). The proportion of viable seeds w as significantly lower for CHGOG (51.00% 2.91%) than for the other two phenotypes ( P =0.0145; Table 3 2) while the proportion of viable seeds that germin ated differed between the CHGOV (89.75% 1.65%) and CHGOG (79.75% 1.69%) phenotypes ( P =0.0461; Table 3 2). The percentage of abnormal, dormant, and those non germinating seeds deemed to be viable according to the TZ test results did not differ significan tly between the three Chrysopsis tested (Table 3 2). The TZ test were highly variable but generally suggest that the mean percentage of seeds in natural populations of the CHGOV, CHGOC, and CHGOG are 73% 21%, 75 22, and 48 14, respectively. Overall mean ger mination in the greenhouse did not differ after eight weeks. There was a significant difference in germination in the greenhouse at six we eks between all three Chrysopsis spp ( Figure 3 4). C. godfreyi f. viridis had the highest
59 mean germination (27% 3%) followed by C. gossypina spp. cruiseana (22% 2%) with C. godfreyi f. godfreyi having the lowest germination (18% 0.8%) at six weeks. Overall mean germination differed bet ween photoperiods with 13% 2% of all seeds exposed to 24 hours of darkness and 49% 3 % subject to a 12 hour photoperiod germinating after 55 days (n=48; P < 0.0001). Cumulative mean percentages from all photoperiod and temperature combinations differed between phenotypes as well with 38% 5% of CHGOC 23% 4% of CHGOG a nd 33% 4% of CHGOV germinating after 55 days (n=32; P < 0.0001). Total germination across all three phenotypes varied according t o temperature regimes (seasons ) with the highest val ues in winter (44% 4%) and fall (40% 5%) exhibiting significant differences between summer (11% 3%) and spring (30% 6%) germination percentages with the two warmer temperature regimes differing significantly as well ( P <0.0001; Figure 5). The respo nse to the interaction between photoperiod and season treatments also varied significantly (Figure 3 6). While the spring, fall, and winter seeds with a 12 hr photoperiod responded similarly and having mean germination percentages above 55% 3%, they dif fered from the seeds with the same photoperiod, but exposed to the summer alternating temperature regime (21% 3%). However, the light summer treatment did not differ from the dark winter treatment (30% 4%) or the dark fall treatment (18% 3%). The w inter alternating temperature regime had the highest mean germination percentage under zero light conditions. (Table 3 3; Figure 3 6). Discussion Differences in the number of seeds produced per flower head can be caused by genetic variation, differing env ironmental conditions such as resource availability, and or
60 various biological stressors such as competition for the parent plant at the time of seed development. Some studies have shown that increased pollen levels resulted in a greater number of seeds (Galen and Weger 1986, Lalonde and Roitb erg 1989). C. godfreyi f. godfreyi which had fewer seeds per flower head, was most commonly found near the tree line or next to large shrubby vegetation while the other two Chrysopsis were found in swale like depressions or in less densely vegetated areas of the backdunes. The form godfreyi reduced the number of pollinators visiting parent plants. Another possibili ty is that those plants had fewer resourc es available. A study by McGinley and Charnov (1988) found that the relationship between seed number and size was influenced by the amount of carbon and nitrogen available for seed production. Characteristics of germination can vary within a species due to genetics, the environment, or an interaction of genetics and the environmental conditions at the time of seed maturation (Baskin and Baskin 2001). While there was no difference in dormancy between the three sample populat ions there were v ariations in the proportion of normal, dead, viable, and germination of viable seeds. The viability of seeds collected is an important consideration for restoration projects. If only a small percentage of seeds are viable for a species of interest, more seeds must be collected to reach the target number of individuals. Because all of the seeds tested by MidWest Seeds were collected at the same time and were subjected to the same conditions, the length of day and the growing season did not differ. Also, the seeds wer e collected from random individuals across the entire habitat so it is unlikely age of the parent pla nt differed between the three Chrysopsis species However, several studies have
61 indicated that environmental conditions required for germination may vary between populations of a species (Groves et al. 1982, Fady 1992, Baskin and Baskin 2001). Varying individual sensitivities of salinity, soil pH, photoperiod, and temperatures have been observed (Baskin and Baskin 2001). In the greenhou se trials, all three performed similarly over the eight week period suggesting that the differences in abundance of the three se en in the field may be due to conditions that did not differ in the greenhouse such as moisture and nutrient availability, photoperiod and temperatu re regimes, or interactions with other organisms in terms of herbivory or competition. Additionally, the greenhouse study took place during Fall 2010; the growth chamber germination study indicated that all three phenotypes had the highest germination wit h a normal photoperiod (12 h ) and alternati ng temperatures that mimic temperatures most similar to It may be possible that differences in germinat ion between the three would become apparent in the greenhouse during less favorable g rowing seasons. Germination speed is considered to be a heritable trait in some plant species, but has also been shown to differ by cultivar at lower temperatures (Farris 1988, Otubo et al. 1996, Baskin and Baskin 2001). Under all four seasonal temperatur e regimes used in the gro wth chamber, C. gossypina spp. cruiseana had the highest initial percent germination. This suggests that the germination speed for C hrysopsis species may be genetic because the relative speed of germination was consistent for all seasonal temperatures tested. Van Loenhoud and Duyts (1981) found that ideal light and temperature requirements for 11 microspecies of Taxraxacum germination differed significantly. Of
62 the temperature regimes and photoperiods included in this study, it appears that all three phenotypes favor a 12 hour photoperiod with alternating temperatures consistent winter likely allow seeds to experience less direct sunlight and more exposure to moisture due to reduced evaporation at those times of year. Conclu ding Remarks This stud y indicates that the two forms of Chrysopis godfreyi and the subspecies of Chrysopsis gossypina respond differently to seasonal temperature regimes an d to differing photoperiods which may indicate that they represent di fferent ecotypes of that genus At minimum, they represent two species with three distinct morphologies that naturally occur; therefore, efforts should be made to include all of them in revegetation of coastal dune habitats within their native range. Our results suggest that optimal seasons for ger mination of all three in Florida are fall and winter. Seed s can be collected in the winter and immediately sown for greatest germination. Additionall y, because C. godfreyi f. godfreyi has a lower proportion of viable seeds that germinate, it may be necessary to collect additional seed heads of this type compared to the other two to get the same number of plants. Because of the importance of maintaining species diversity and genetic diversity, future studies are needed on C. godfreyi and C. gossypina as well as on other coastal dune species to determine the sou rce of variation. If multiple, distinct ecotypes or subspecies exist, it will be necessary to incorporate all of them in management plans and restoration efforts.
63 Table 3 1. Flower and seed production in two forms of Chrysopsis godfreyi and one subspecie s of Chrysopsis gossypina
64 Table 3 2 Characteristic morphology, germination and viability percentages of the three Chrysopsis spp. as determined by studies conducted by MidWest seeds starting in June 2010 (n=100 ). Table 3 3 Mean germination of all three Chrysopsis species after 55 days in a growth chamber in response to photoperiod and temperature.
65 Figure 3 1. Location of 13 individual Chrysopsis spp. plants at the Naval Air Station in Pensacola, FL that had seed heads harvested in December 2009. Individuals are labeled N01 N13. Figure 3 2. Location of 20 individual Chrysopsis spp. plants at Perdido Key State Park in Escambia County, Florida t hat had seed h eads harvested in December 2009 Individuals are labeled P01 P20.
66 Figure 3 3. Photos and distribution maps of the three Chrysopsis spp. co llected. (A) Chrysopsis godfreyi f. godfreyi represents individuals with heavy pubescence on leaves as well as the stems and give it a silvery white appearance. (B) Chrysopsis godfreyi f. viridis represents individuals with a sticky leaf surface and little to no pubescence on any leaves (C) Chrysop sis gossypina spp. cruiseana represents individuals with moderate pubescence on leaves with increased coverage on basa l leaves Map credits: Wunderlin and Hansen 2008. A B C
67 Figure 3 4. Mean germination (%) of three Chrysopsis spp. over eight weeks in a greenhouse in Fort Pierce, Florida from September 2010 November 2010. CHGOV= Chrysopsis godfreyi f. vir i dis represents individuals with a sticky leaf surface and little to no pubescence on any leaves; CHGOC= Chrysopsis gossypina spp. cruiseana represents i ndividuals with moderate pubescence on leaves with increased coverage on basal leaves; CHGOG= Chrysopsis godfreyi f. godfreyi represents individuals with heavy pubescence on leaves as well as the stems and give it a silvery white appearance. Error bars rep resent 1 standard error of the mean. *= week where cumulative germinat ion between the three species was statistically different. 0 10 20 30 40 50 60 1 2 3 4 5 6 7 8 Mean Germination (%) Weeks After Sowing CHGOC CHGOV CHGOG
68 Figure 3 5 Total mean germination (%) of all seeds of Chrysopsi s spp. under different temperature regiments representing Florida seasons. Error bars represent 1 standard error. Means separation by LSD (alpha= 0.05) with a Bonferroni correction. Bars containing the same letter do not differ. Figure 3 6 Total mean ge rmination (%) of all Chrysopsis spp. seeds in response to temperature and photoperiod. Error bars represent 1 standard error of the mean. Light = a 12 hour photoperiod; dark= 24 hours of darkness. b c a a 0 10 20 30 40 50 60 70 80 90 100 SPRING (30/20) SUMMER (35/25) FALL (25/15) WINTER (20/10) Germination (%) Season (Day Temp. C/Night Temp. C) n=24; p < 0.0001 0 10 20 30 40 50 60 70 80 90 100 SPRING (30/20) SUMMER (35/25) FALL (25/15) WINTER (20/10) Germination (%) Season (Day Temp. C/Night Temp. C) LIGHT DARK n=12; p < 0.0001
69 Figure 3 7 Mean germina tion (%) of three Chrysopsis spp. over eight weeks in under four different temperature regimes CHGOC= Chrysopsis gossypina spp. cruiseana represents individuals with moderate pubescence on leaves with increased coverage on basal leaves; CHGOG= Chrysopsis godfreyi f. godfreyi represents individuals with heavy pubescence on leaves as well as the stems and give it a silvery white appearance L ight 12 hr light/12 hr dark. Dark 24 hr of darkness
70 CHAPTER 4 IMPLICATIONS FOR PROMOTING VEGETATION ESTABLISHMENT AND RECREATING NATURAL DIVERSITY IN COASTAL DUNE RESTORATION As a result of overexploitation and inc reases in natural disturbances, coastal dune s have experienced a net loss worldwide (Mart nez et al. 2008). These ecosystems continue to experience stress from major storm events as well as structural development, terrestrial and marine pollution, encroachment by non native species and coastl ine retreat related to sea level rise from global climate change (Perrow and Davy 2002 Crowell et al. 2007, Defeo et al. 2009). Restoration of coastal dune ecosystems in response to these natural and anthropogenic stressors is necessary to preserve the u nique biodiversity and ecosystem services that only they provide. Re vegetation of coastal dunes i s a necessary step in the restoration process and studies related to improved establishment of planted vegetation as well as those that can identify local eco types will help practitioners restore the autogenic processes that occur when natural plant diversity is present at a site. The addition of organic matter to planting sites can help stabilize the sand, add moisture to the microenvironment surrounding the installed plants, and provide nutrients that could increase the growth and spread of dune grasses (Hester and Mendelssohn 1990, Seliskar 1995 Nordstrom et al. 2010, Williams and Feagin 2010 ) Another component of successfully restoring coastal dunes is restoring the natural diversity that occurred at the site prior to the disturbance or that occurs at the reference site. Because ecological systems are complex and all interactions and/or func tions have not been identified, it is important to maintain the genetic diversity even within a species to preserve the integrity of the community in which it belongs.
71 At Gulf Islands National Seashore in Escambia County, Florida, we applied a surrogate wr ack in the form of wheat straw to Uniola paniculata a native dune grass at six replicate sites and compared them with U. paniculata that did not receive the surrogate wrack treatment. Six months after the application, we found that the mean aboveground biomass, the height, the basal width, and the number of tillers per plant were all greater for U. paniculata that were in plots with wrack than those that were grown in bare sand ( P <0.05 ). On average, U. paniculata grown in the wrack plots had 2 more till ers, were 7 grams heavier, and 20 centimeters taller than those plants grown without the surrogate wrack. Additionally, wrack plots accumulated 11.16 cm 1.71cm of sand over three months compared to only 7.78 cm 0.78 cm in plots with no wrack. While t his difference was only marginally significant ( P = 0.1093), it suggests that either the presence of a surrogate wrack or the additional grass growth as a result of the wrack influences sand accumulation in the immediate area. We also collected seeds from three phenotypes representing the two endangered Chrysopsis species endemic to the coastal dunes of the Florida panhandle to de termine if the three phenotypes observed in the field were expressions of the species phenotypic pla sticity or if they are in fact distinct ecotypes or subspecies. The three Chrysopsis differ in the amount of pubescence present on the leaves and stems with Chrysopsis godfrey i f. viridis having a sticky leaf surface and little to no p ubescence on any lea ves, Chrysopsis gossypina spp. cruiseana present moderate pubescence on leaves with increased c overage on basal leaves, and Chrysopsis godfreyi f. godfreyi include individuals with heavy pubescence on leaves as well as the stems and give it a silvery white appearance.
72 We found no overall difference in germination in greenhouse studies that were conducted in the fall of 2010 in Fort Pierce, Florida but significant differences in germination between phenotypes did occur in growth chambe r studies. Chrysopsi s godfreyi f. godfreyi had the lowest germination over eight weeks with a 12 hour photoperiod under four different alternating temperature regimes: 20/10 C 25/15C, 30/20 C, and 35/25 C which in essence ditionally, C. godfreyi f. gofreyi had the least number of seeds per flower head, the fewest seeds with normal morphology, and the lowest viability as measured by the TZ test. Our results show that the highest percen tage of all three Chrysopsis spp. seeds will germinat e at temperatures that are most similar Our study suggests that the application of a surrogate wrack can enhance the growth of dune grasses such as U. paniculata Wheat s traw is generally inexpensive and readily available to managers. A more favorable approach to a surrogate wrack may be the use of the natural vegetative wrack that occurs along the high tide line. Other studies have shown that this wrack material can fertilize coastal vegetation and can assist in initial stages of dune building (Cardona and Garcia 2008, Nordstrom et al. 2010, Williams and Feagin 2010). Using the natural wrack especially at sites where it is typically removed could reduce the cost of restoration by replacing the hydro gel and fertilizers t hat are often used and eliminating the need for off site transport of wrack from beach grooming. Conversely, we found that wheat straw can have a neutral or negative effect for herbaceous species naturally occurring in open dune habitat s. This indicates that the use of wheat straw as a surrogate wrack could be applied in restorations that use a
73 successional approach to vegetation installation where the wrack would be applied in the initial stage along with the dune building grasses. Herbaceous species co uld then be added after the grasses have established and the straw is covered in sand. Our investig ation on the three C hrysopsis spp. s uggests that these physical traits represent different genotypes within the population. While more data need s to be collected to confirm classification, it is apparent that to recreate the natural diversity of these species, all three phenotypes should be included in restoration projects. Additionally, because germination success varies, it is important to coll ect the appropriate number of seeds for propagation to reach target numbers. Wrack has long been considered a n important marine subsidy for invertebrates, but has only recently been investigated for its role as a natural fertilizer. Dr. Semple identified two forms of C. godfreyi in the late 1970s, but despite their and C. gossypina spp. s, endangered status little research has been done to determine effective ways to recreate their natural diversity on restoration projects. The willingness of r estoration ecologists and practitioners to consider new approaches and/or re evaluate current strategies is a crucial component to the future success of restoration projects.
74 LIST OF REFERENCES Acosta A., C. Carranza and F. Izzi 2005. Combining land cover mapping of coastal dunes with vegetation analysis. Applied Vegetation Science 8:133 138. Adger W. N., T. P. Hughes C. Folke, S. R. Carpenter and J. Rockstrm. 2005. Social ecological resilience to coastal disasters. Science 309:1036 1039. Allard, R.W. 1970. Population structure and sampling methods. Pages 97 107 In O. H. Frankel and E. Bennett, editors. Genetic resources in plants their exploration and conservation Blackwell Scientific Publications, Oxford England Baskin, C. C., J. M. B askin. 2001. Causes of within species variations in seed dormancy and germination characteristics. Pages 181 237 in C. C. Baskin and J. M. Baskin. Seeds: ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego, CA, U SA. Behbehani, M. I. and R. A. Croker 1982. Ecology of beach wrack in northern New England with special reference to Orchestria platensis Estuarine, Coastal, and Shelf Science 15:611 620. The Wildlife Forecast. Florida Fish and Wildlife Commission, Tallahassee, FL, USA. Brown, A. D. and J. D. Briggs. 1991. Sampling strategies for genetic variation in ex situ collections of endangered plant species. Pages 99 119 In Genetics and conservatio n of rare plants. D.A. Falk and K. E. Holsinger, editors. Oxford University Press, Oxford England Cardona, L. and M. Garcia 2008. Beach cast seagrass material fertilizes the foredune vegetation of Mediterranean coastal dunes. Oecologica 34:97 103. Carney, S. E. 2005. Population genetic issues associated with revegetation using national park collected plant materials: Final Report. Corenblit, D. and J. Steiger 2009. Vegetation as a major conductor of geomorphic changes on the Earth surface: toward e volutionary geomorphology. Earth Surface Processes and Landforms 34:891 896. Crowell M., S. Edelman K. Coulton and S. McAfee 2007. How many people live in coastal areas? Journal of Coastal Research 23:iii vi. Davidson Arnott, R. D., Y. Yang J. Ollerhead P. A. Hesp and I. J. Walker. 2008. The effects of surface moisture on Aeolian sediment transport threshold and mass flux on a beach. Earth Surface Processes and Landforms 33:55 74.
75 Defeo O., A. McLachlan D. S. Schoeman T. A. Schlacher J. E. Dugan A. Jones M. Lastra, and F. Scapini 2009. Threats to sandy beach ecosystems: a review Estuarine, Coastal and Shelf Science 81:1 12. DeLaney, K. R. and R. P. Wunderlin 2002. A new species of Chrysopsis (Asteraceae, Astereae) from Central Florida. The Botanical Explorer 2:1 20. Dugan, J. E. and D. M. Hubbard. 2010. Loss of coastal stand habitat in southern California: the role of beach grooming. Estuaries and Coasts 33:67 77. Dugan, J. E., D. M. Hubbard M. D. McCrary and M. O. Pierson 2003. The response of macrofauna communities and shorebirds to macrophyte wrack subsidies on exposed sandy beaches of southern California. Estuarine, Coastal and Shelf Science 58S:25 40. Ehrenfeld, J.G. 2000. Defining the limits of restoration: the need for realistic goals. Restoration Ecology 8:2 9. Falk, D. A. 1991. Joining biological and economic models for conserving plant genetic diversity. Pages 209 223 In Genetics and conservation of rare plants D. A. Falk and K. E. Holsinger, editors. Oxford Univ ersity Press, Oxford England Falk, D. A., E. Knapp and E. O. Guerrant 2001. An introduction to restoration genetics. Society for Ecological Restoration 1 33. Fant, J.B., R. M. Holmstrom E. Sirkin, J. R. Etterson and S. Masi 2008. Genetic structure o f threatened native populations and propagules used for restoration in a clonal species, American beachgrass ( Ammophila breviligulata Fern.). Restoration Ecology 16:594 603. Farris, M. A. 1988. Quantitative genetic variation and natural selection in Cleome serrulata growing along a mild soil moisture gradient. Canadian Journal of Botany 66:1870 1876. Feagin, R.A., R. E. Koske, J. N. Gemma and A. M. Williams 2009. Restoration of sea oats ( Uniola paniculata ) with mycorrhizae on Galveston Island: Final Repor t. Finch, T., and G. Bauer 1986. Spoil Island 25 Restoration and Stabilization: Final Report. Franks, S. J. 2003. Competitive and facilitative interactions within and between two species of coastal dune perennials. Canadian Journal of Botany 81:330 337. F reeman, S., and J. C. Herron. 1998. Evolutionary analysis. Prentice Hall, Upper Saddle River, New Jersey, USA Galen, C., and Weger, H. G. 1986. Re evaluating the significance of correlations between seed number and size: evidence from a natural population of lily, Clintonia borealis American Journal of Botany 73:346 352.
76 Garca Mora, M. R., J. B. Gallego Fernndez and F. Garca Novo 2000. Plant diversity as a suitable tool for coastal dune vulnerability assessment. Journal of Coastal Research 16:990 995 Gibson, D. J. and P. B. Looney 1994. Vegetation colonization of dredge spoil on Perdido Key, Florida. Journal of Coastal Research 10:133 143. Gormally, C. L. and L. A. Donovan 2010. Responses of Uniola paniculata L. (Poaceae), an essential dune build ing grass, to complex changing environmental gradients on the coastal dunes. Estuaries and Coasts 33:1237 1246. Gray, A. 1996. Genetic diversity and its conservation in natural populations of plants. Biodiversity Letters 3:71 80. Hannan, L. B., J. D. Roth L. M. Ehrhart and J. F. Weishampel 2007. Dune vegetation fertilization by nesting sea turtles. Ecology 88:1053 1058. Hazell, J., S. H. Brown and K. Cooprid er 2010. Uniola paniculata. IFAS Extension University of Florida Press. Hester, M.W. and I. A. Mendelssohn. 1990. Effects of macronutrient and micronutrient additions on photosynthesis, growth parameters, and leaf nutrient concentrations of Uniola paniculata and Panicum amarum. Botanical Gazette 151:21 29. Hinrichsen D. 1999 The coastal population explosion. Pages 27 29 In Trends and Future Challenges for U.S. National Ocean and Coastal Policy: Proceedings of a Workshop Washington, D.C. USA Holmstrom, R. M., J. R. Etterson and D. J. Schimpf. 2010. Dune restoration introduces genetically distinct American beachgrass, Ammophila breviligulata into a threatened local population. Restoration Ecology 18:426 437. Hufford, K. M. and S. J. Mazer 2003. Plant ecotypes: genetic differentiation in the age of ecological restoration. TRENDS in Ecology and Evolution 18 :147 155. Lalonde, R. G. 1989. Resource limitation and offspring size and number of trade offs in Cirsium arvense (Asteraceae). American Journal of Botany 76:1107 1113. Lawrence, M.J., D. F. Marshall and P. Davies 1995. Genetics of genetic conservation. I. Sample size when collecting germplasm. Euphytica 84:89 99. Lomba, A., P. Alves and J. Honrado 2008. Endemic sand dune vegetation from northwest Iberian peninsula : diversity, dynamics and significance for bioindication and monitoring of coastal landscapes. Journal of Coastal Research 24:113 121. Keddy, P. A. 1982. Population ecology on an environmental gradient: Cakile edentula on a sand dune. Oceologia 52:348 355.
77 Khalil, S. M. and C. W. Fink. 2009. Regional sediment management strategies for coastal restoration in Louisiana, USA. Journal of Coastal Research 56:1320 1324. Martnez, M. L., J. B. Gallego Fernndez J. G. Garca Franco C. Moctezuma, and C. D. Jimn ez 2006. Assessment of coastal dune vulnerability to natural and anthropogenic disturbances along the Gulf of Mexico. Environmental Conservation 33:109 117. Martnez M. L ., A. Intralawan G. Vzquez O. Prez Masqueo P. Sutton and R. Landgrave. 2007. The coasts of our world: ecological, economic and social importance. Ecological Economics 63:254 272. Martnez, M. L., M.A. Maun, and N. P. Psuty. 2008. The fragility and conservation of the world's coastal dunes: geomorphological, ecological and s o cioeconomic Perspectives Ecological Studies 171:355 369. Martnez, M. L. and N. P. Psuty, editors. 2004 Coastal Dunes Ecology and Conservation Series : Ecological Studie s 171. Maun, M. A. 1994. Adaptations enhancing survival and establishment of seed lings on coastal dune systems. Plant Ecology 111:59 70. Maun, M. A. 1998. Adaptations of plants to burial in coastal sand dunes. Canadian Journal of Botany 76:713 738. McDonald, J. A. 1934. Mulch in experiments on cacao. Fourth Annual Report on Cacao Research for 1934 1935 :64 74. McGinley, M. A., and Charnov, E. L. 1988. Multiple resources and the optimal balance between size and the number of offspring. Evolutionary Ecology 2:77 84. McGwynne, L. E, A. McLachlan and J. P. F urstenberg 1988. Wrack bre akdown on sandy beaches its impact on interstitial meiofauna. Marine Environmental Research 25:213 232. Miller D. L., M. Thetford and M. Schneider 2008. Distance from the Gulf influences survival and growth of three barrier island dune plants. Journal o f Coastal Research 24:261 266. Miller, D. L., M. Thetford and L. Yager 2001. Evaluation of sand fence and vegetation for dune building following overwash by Hurricane Opal on Santa Rosa Island, Florida. Journal of Coastal Research 17:936 948. Miller, D. L., L. Yager M. Thetford and M. Schneider 2003. Potential use of Uniola paniculata rhizome fragments for dune restoration. Restoration Ecology 11:359 369. Minchinton, T. E. 2006. Rafting on wrack as a mode of dispersal for plants in coastal marshes. Aquatic Botany 84:372 376.
78 Mountney, N. G. and A. J. Russell 2006. Coastal aeolian dune development, Slheimasandur, southern Ireland. Sedimentary Geology 192: 167 181. Natural Resources Conservation Service (NRCS). 2011. Plant Nutrient Content database. United S tates Department of Agriculture, Washington, D.C., USA. Nordstrom, K. F., N. L. Jackson, K. H. Korotky, and J. A. Puleo 2010. Aeolian transport rates across raked and unraked beaches on a developed coast. Earth Surface Processes and Landforms DOI: 10.1002/esp.2105 Nordstrom, K. F., R. Lampe and L. M. Vandemark. 2000. Reestablishing naturally functioning dunes on developed coasts. Environmental Management 25:37 51. Olafson, A. 1997. Stabilization of coastal dunes with vegetation. Restoration and Re clamation Review 2:1 7. Orr, M., M. Zimmer D. E. Jelinski and M. Mews 2005. Wrack deposition on different beach types: spatial and temporal variation in the pattern of subsidy. Ecology 86:1496 1507. Otubo, S. T., M. P. Ramalho, A. B. Abreu, and J. B. d os Santos. 1996. Genetic control of low temperature tolerance in germination of common bean ( Phaseolus vulgaris L.) Euphytica 89:313 317. Perrow M. R. and A. J. Davy 2002. Principles of Restoration. Pages 257 278 In Handbook of Ecological Restoration, Vol ume 1 Princ iples of Restoration. Cambridg e University Press, Cambridge, England. Pervaiz, M.A., M. Iqbal K. Shahzad and A. Ul Hassan 2009. Effect of mulch on soil physical properties and N, P, K concentration in maize ( Zea mays L.) shoots under two tillage systems. International Journal of Agriculture and Biology 11:119 124. Pries A.J., L. C. Branch and D. L. Miller 2009. Impact of hurricanes on habitat occupancy and spatial distribution of beach mice. Journal of Mammalogy 90:8 41 850. Schlacher T. A., J. Dugan D. S. Schoeman M. Lastra A. Jones F. Scapini, A. McLachlan, and O. Defeo 2007. Sandy beaches at the brink. Diversity and Distributions 13:556 560. Seliskar, D. M. 1995. Coastal dune restoration: a strategy for alleviating the dieout of Ammophila breviligulata Restoration Ecology 3:54 60. Semple, J. C. 1978. A new species endemic to West Florida: Chrysopsis godfreyi (Compositae Astereae). Canadian Journal of Bota ny 56:2092 2096. Smith, B.M., A. Diaz R. D aniels L. Winder and J. M. Holland 2009. Regional and ecotype traits in Lotus corniculatus L., with reference to restoration ecology. Restoration Ecology 17:12 23.
79 Snyder R. A., and C. L. Boss. 2002. Recovery a nd stability in barrier island plant comm unities. Journal of Coastal Research 18 :530 536. Stallins, J. A. 2003. Dune plant species diversity and function in two barrier island biogeomorphic systems. Plant Ecology 165:183 196. Stallins, J.A., and A. J. Parker 2003. The influence of complex system interactions on barrier island dune vegetation pattern and process. Annals of the Association of American Geographers 93:13 29. Stopp, Jr., G. H., and 1996. Initial effects of a hurricane storm s urge on barrier island vegetation. Technical Pa per 81. Florida Sea Grant College, Gainesville, FL, USA. Surrency, D. 1992. Measures for stabilizing coastal dunes: Alabama, Georgia Americus Plant Materials Center Americus, Georgia, USA. Tarkalson, D., B. Brown H. Kok and D. L. Bjorneberg. 2009. Irrigated small grain residue management effects on soil properties and nutrient cycling. Western Nutrient Management Conference. Salt Lake City, Utah, USA. Thetford, M D. L. Miller L. Smith and M. Schneider 2005. Container size and planting zone influence on transplant survival and growth of two coastal plants. Horticultural Technology 15:554 559. v an der Laan, D., O. R. van Tongeren W. H. van der Putten and G. Veenbaas 1997. Vegetation development in co astal foredunes in relation to methods of establishing marram grass ( Ammophila arenaria ). Journal of Coastal Conservation 3:179 190. van Loenhound, P. J., and H. Duyts. 1981. A comparative study of the germination ecology of some microspecies of Taraxacum Wigg. Acta Botanica Neerlandica 30:161 182. Velander, K., and M. method? Marine Pollution Bulletin 38:1134 1140. VISIT FLORIDA Research. 2011. Research. http://media.visitflorida.org/research.php accessed 11 Mar 2011 Walker, B. 1995. Conserving biological diversity through ecosystem resilience. Conservation Biology 9:747 752. Ward D. B., D. F. Austin and N. C. Coile 2003. Endangered and threatened plants of Florida, ranked in order of rarity. Southern Appalachian Botanical Society 68:160 174. Whisenant S.G. 1999. Repairing Damaged Wildlands, a process orientated, lands cape scale approach. C ambridge University Press, Cambridge, England
80 Wilcock, F. A. and R. G. Carter 1977. An environmental approach to the restoration of badly eroded sand dunes. Biological Conservation 11:279 291. Williams, A. and R. Feagin 2010. Sargassum as a natural solution to enhance dune plant growth. Environmental Management 46:738 747. Wilson M. A., R. Costanz a R. Boumans and S. Liu 2005. Integrated assessment and valuation of ecosystem goods and services provided by coastal systems. Pages 1 23 In J. G. Wilson, editor. Roy al Irish Academy, Dublin, Ireland. Wilson, S. D., J. D. Bakker J. M. Christian X. Li L. G. Ambrose J. Waddington. 2004. Semiarid Old field restoration: is neighbor control needed ? Ecological Applications 14:476 484. Wunderlin, R. P., and B. F. Hansen. 2008. Atlas of Florida Vascular Plants (http://www.plantatlas.usf.edu/).[S. M. Landry and K. N. Campbell (application development), Florida Center for Community Design and Research.] Institute for Systematic Botany, University of South Florida, Tampa
81 BIOGRAPHICAL SKETCH Natalie graduated from Stetson University (DeLand, FL) in 2002 with a Bachelor of Science degree in Biology. She then spent five years promoting inquiry based science as a middle school science teach er of the gifted and science fair advocate and coordinator. She obtained her Master of Science degree in August of 2011 in Interdisciplinary Ecology from the School of Natural Resources and the Environment at the University of Florida.