A 2,500-year record of environmental change in Highlands Hammock State Park (Central Florida, U.S.A.) inferred from siliceous microfossils
- First Online:
- Cite this article as:
- Pearce, C., Cremer, H., Lammertsma, E. et al. J Paleolimnol (2013) 49: 31. doi:10.1007/s10933-011-9557-2
Analysis of siliceous microfossils of a 79 cm long peat sediment core from Highlands Hammock State Park, Florida, revealed distinct changes in the local hydrology during the past 2,500 years. The coring site is a seasonally inundated forest where water availability is directly influenced by precipitation. Diatoms, chrysophyte statospores, sponge remains and phytoliths were counted in 25 samples throughout the core. Based on the relative abundance of diatom species, the record was subdivided into four diatom assemblage zones, which mainly reflect the hydrological state of the study site. An age-depth relationship based on radiocarbon measurements of eight samples reveals a basal age of the core of approximately 2,500 cal. yrs. BP. Two significant changes of diatom assemblage composition were found that could be linked to both, natural and anthropogenic influences. At 700 cal. yrs. BP, the diatom record documents a shift from tychoplanktonic Aulacoseira species to epiphytic Eunotia species, indicating a shortening of the hydroperiod, i.e. the time period during which a wetland is covered by water. This transition was interpreted as being triggered by natural climate change. In the middle of the twentieth century a second major turnover took place, at that time however, as a result of human impact on the park hydrology through the construction of dams and canals close to the study site.
KeywordsDiatoms Florida Highlands Hammock State Park Aulacoseira coroniformis Eunotia Hydrology
The unique position of Florida between the temperate North American mainland and the tropical Caribbean Islands is reflected in the diverse and specific environments on the peninsula. Water availability is a dominant controlling factor in the mosaic of landscapes (Myers and Ewel 1990). Understanding the response of these environments to hydrological fluctuations forced by climate change as well as human disturbance is crucial for the conservation and restoration plans of the natural areas, as well as regulating the natural resources as groundwater, which is essential for sustaining households and agriculture (Ehrenfeld 2000; Reese 2002).
Long-term records of environmental change in the peninsula’s terrestrial realm are available from pollen analysis of lake deposits from the numerous lakes on the Lake Wales Ridge (Watts 1969, 1975, 1980; Watts et al. 1992; Grimm et al. 1993; Watts and Hansen 1994). The observed major vegetation shifts from dry, oak-dominated landscapes to wetter, pine-dominated landscapes over the past 60,000 years are related to Dansgaard-Oeschger interstadials (Grimm et al. 1993, 2006; Donders et al. 2009). Moreover, hydrological changes occurring in the coastal region are increasingly covered by palynological and geochemical studies on estuarine deposits, reflecting the environmental response to the Holocene sea level rise super-imposed on climatic changes since the last glacial maximum (Willard et al. 2007; Cremer et al. 2007; Van Soelen et al. 2010). While millennial-scale hydrological dynamics are well documented, the minor changes occurring on multi-decadal to multi-centennial time scales are more difficult to deduce from these archives, as the more subtle responses of the environment to these changes are likely smoothed in the regional signal covered in lake and estuarine deposits.
Peat deposits accumulating in the many wetlands of the Florida peninsula are a potentially suitable archive for the reconstruction of short-lived hydrological dynamics of multi-annual to multi-decadal climate variability, of which the El Niño—Southern Oscillation (ENSO) and the Atlantic Multidecadal Oscillation (AMO) are known to strongly affect Florida precipitation patterns (Donders et al. 2005a; Curtis 2008). As the sand-covered limestone subsurface of this region is highly permeable, vegetation and fauna are largely dependent on precipitation for their water supply (Kushlan 1990). Consequently, minor changes in local hydrology, expressed as changes of the water depth or hydroperiod (the time period per year during which a site is flooded) are found to result in distinct environmental changes (Gaiser et al. 1998; Willard et al. 2001; Willard et al. 2004; Donders et al. 2005b).
The analysis of siliceous microfossils has proven to be a highly useful method for reconstructing past changes in aquatic environments. As siliceous algae, like diatoms, chrysophytes, and freshwater sponges have specific habitat preferences, variation in the species assemblage or fossil abundance provides information on changes within the aquatic environment like water depth, hydroperiod and pH (Smol and Stoermer 2010; Wujek 2000; Zeeb and Smol 2001; Frost 2001). Shifts in phytolith assemblages can furthermore be linked to changes in plant groups from which they originate, in response to changes in the local environment (Piperno 2001; Prebble and Shulmeister 2002; Lu et al. 2006).
Diatoms in particular have proven to be a successful tool in reconstructing local hydrological changes. They are characterized by high species diversity and short reproductive rates leading to fast floral alterations under environmental change (Smol and Stoermer 2010). Intensive efforts to deduce water quality changes in Florida surface waters have demonstrated the sensitivity of diatom communities and the applicability of diatom-based transfer functions to determine nutrient status and hydrological conditions in coastal wetlands (Gaiser et al. 2005; Cremer et al. 2007; Wachnicka et al. 2010). Although diatoms have been also successfully applied to deduce past hydrological conditions in freshwater wetlands in the Atlantic coastal plain (Gaiser et al. 2004), diatom-based hydrological reconstructions for central Florida wetlands are so far underrepresented.
Here we present diatom and other siliceous microfossils analyses on a ~2,500 year old peat core from a swamp forest in Highlands Hammock State Park, Florida. The combined analysis of siliceous microfossils allows for detailed and more robust interpretation of the changes in this wetland. Based on these observed changes we aim to reconstruct past environmental dynamics induced by subtle hydrological changes, and thus develop better insight in climate and anthropogenic driven variability in the central Florida hydrology.
The occurrence of natural plant communities in the park strongly reflects the local topography and hydrology. The drier areas at higher elevations are characterized by mesic flatwoods (mainly Florida slash pine, Pinus elliottii, with an undergrowth of predominantly saw palmetto, Serenoarepens), whereas the wetter surroundings of Little Charlie Bowlegs Creek hold mostly basin swamp and marsh vegetation (Baldcypress (Taxodium distichum), swamp laurel oak (Quercus laurifolia), sweetgum (Liquidambar) and red maple (Acer rubrum)). Along the minor waterways, around the hydric hammock, in the central area of the park, so called baygall vegetation is found, consisting mostly of sweetbay (Magnolia virginiana), loblolly bay (Gordonia lasianthus) and swamp bay (Persea palustris) with a diverse undergrowth, e.g. dahoon holly (Ilex cassine), waxmyrtle (Myrica cerifera) and royal ferns (Osmunda spp.). The central feature of the park is the hydric hammock, dominated by a variety of hardwood species like live oak (Quercus virginiana), sweet gum (Liquidambar styraciflua), pignut hickory (Carya glabra), together with cabbage palm (Sabal palmetto) and a dense understory of various shrubs, ferns and epiphytes. Within the hammock a number of ‘domes’ are present due to karstic solution features, where pop ash (Fraxinus caroliniana) and swamp tupelo (Nyssa sylvatica) are common.
Materials and methods
Coring and dating
Radiocarbon dating of samples from core HHA3
Conventional radiocarbon age
Calendar age (year BP)
105.92 ± 0.34 pMC
−55.86 ± 1.01
Leaf fragments, chitin, seed
106.94 ± 0.36 pMC
−53.93 ± 1.14
Leaf fragments, chitin, seeds
100.6 ± 0.35 pMC
−1.31 ± 0.36
230 ± 30 year BP
291 ± 24
Leaf fragments, chitin, seeds
550 ± 40 year BP
539 ± 27
1,360 ± 40 year BP
1,288.5 ± 55.5
1,645 ± 30 year BP
1,551 ± 66
2,460 ± 30 year BP
2,489.5 ± 62.5
Sample processing and diatom analysis
Siliceous microfossil samples were freeze-dried and approximately 1 gram of each sample was used for the extraction of diatom valves. Samples were treated subsequently with hydrogen peroxide (1.5 h at 100°C), hydrochloric acid and nitric acid (2 h at 120°C) for the removal of organic matter and carbonate. Excess acid was removed by seven sedimentation procedures in demineralized water. Microscopic slides were prepared using evaporation trays (Battarbee 1973; Cremer et al. 2001). A known fraction of the cleaned samples in solution was allowed to settle on a cover slip and, after evaporation of the excess water, mounted on slides using the high refraction mountant Naphrax®. On each slide at least 400 diatom valves were counted with a Leica DM2500 microscope equipped with a 63× oil immersion lens and differential interference contrast at a magnification of 945×. Digital images were taken with a Leica DFC320 digital camera. Diatom taxonomy for the most identified species followed Camburn and Charles (2000), Patrick and Reimer (1966), Siver et al. (2005), Gaiser and Johansen (2000), Metzeltin and Lange-Bertalot (2007) and Pearce et al. (2010). A principal component analysis (PCA) was applied to our dataset to create a zonation of the core based on the diatom assemblages (Birks 2010). The PCA was run on the diatom species relative abundance (no data transformation) using the C2 software package (Juggins 2007).
Microfossil abundance determination
The absolute abundance of siliceous microfossils per gram dry sediment was determined by using the methods described in Battarbee (1973). The amount of microfossils per gram dry sediment was calculated considering the mass of dry sediment per sample, the amount of the sample solution that was used to prepare the slide, and the number of siliceous microfossils per surface unit on the cover slip,. Sponge megascleres were broken in most cases; therefore, one count was defined as the occurrence of two spicule endings.
Core lithology and chronology
Core HHA3 entirely consists of dark brown, mostly homogeneous peat (Fig. 2). LOI percentages ranged from 15 to 84%, with a sharp transition at 50 cm depth from relatively lower values in the bottom part of the core to relatively higher organic content in the upper part (Fig. 2). Plant remains become more abundant and less degraded towards the upper part of the core. A well-constrained age model based on eight AMS radiocarbon dates indicates a basal age of the core of approximately 2,500 cal. year BP (Table 1, Fig. 2). For the 79 cm core, this means one centimeter represents on average 32 yrs of sediment accumulation. In the lower parts of the core this goes up to 50 yrs/cm, due to compaction and further degradation of the organic matter.
List of the fifteen most frequent diatom taxa in core HHA3 from Highlands Hammock State Park, including pH and water depth optima and references for each species
Pearce et Cremer (2010)
Pearce et al. (2010, p. 41, Figs. 2–37)
Metzeltin and Lange-Bertalot (2007, Pl. 147, Figs. 6–21)
Patrick and Reimer (1966, p. 209, Fig. 5), Camburn and Charles (2000, p. 18, Pl. 11, Figs. 17–24), Siver et al. (2005, p. 75, Pl. 26, Figs. 24–29; Pl. 27, Figs. 1–5), Gaiser and Johansen (2000, p. 92, Figs. 40–42) (as E. pirla)
W. Smith ex. Gregory (1854)
Gaiser et Johansen (2000)
Gaiser and Johansen (2000, p. 92, Figs. 21–25)
Hustedt ex Patrick (1945)
Eunotia zygodon var. zygodon
Patrick and Reimer (1966, p. 199, Pl. 11, Fig. 8)
Eunotia zygodon var. elongata
Hustedt ex Simonsen (1987)
Gaiser and Johansen (2000, p. 95, Fig. 49)
, p. 144, Figs. 10, 11, 65)
Lange-Bertalot et Krammer (1996)
Diatom zone 1 (35–79 cm), representing the age range from 2,511 to 700 cal. yrs. BP, was dominated by the tychoplanktonic diatom species Aulacoseira coroniformis Pearce et Cremer (>80%) and was characterized by the presence of sponge spicules. The transition between zone 1 and zone 2, ranging from 35–24 cm (700–276 cal. yrs. BP), was gradual. Aulacoseiracoroniformis was gradually replaced by Eunotia zygodon Ehrenberg and Pinnularia viridis Ehrenberg. At the boundary between zone 2 and zone 3, from 24 to 12 cm (276 cal. yrs. BP—AD 1957), the species composition changed within the genus Eunotia. The abundance of Eunotia zygodon and Eunotia zygodon var. elongata Hustedt decreased while Eunotia carolina Patrick and Eunotia tautoniensis Hustedt increased in abundance. The uppermost 11 cm representing the last 50 years (zone 4) were characterized by an increase of the absolute concentrations of both diatoms and phytoliths (Fig. 4). This zone also included Fragilaria javanica Hustedt and Frustulia saxonica Rabenhorst, which were not recorded in older sediment samples of this core.
Diatom zone 1 (2,511–700 cal. year BP)
Environment: permanently flooded acidic wetland.
Diatom zones 2 and 3 (700 cal. year BP to AD 1957)
Environment: seasonally inundated acidic wetland.
Diatom zone 4 (AD 1957–Present)
Environment: wetland pond affected by human impact
The data of the youngest diatom zone reflected a distinctly modified composition of the diatom assemblage and a clear species diversification. A number of previously unobserved species were identified including Fragilaria javanica and species of the genus Frustulia (Fig. 4). The genus Eunotia, which comprised up to 90% of the diatom assemblage in zones 2 and 3, was still present in large numbers, but less dominant. The diatom assemblages of zone 4 consisted of considerable numbers of Fragilaria javanica (up to 40%) and Frustulia saxonica (max. 21%). The latter two species are known to be acidophilous (Hustedt 1938; Gaiser and Johansen 2000), which could mean that there was no major pH change at the transition from zone 3 to zone 4. Water depth optima for Frustulia saxonica and Frustulia crassinervia Lange-Bertalot et Krammer compared to those for the dominant taxa in zone 3 pointed to a slight increase of water depth at the site (Gaiser and Johansen 2000; Siver et al. 2005). Historical records from the twentieth century showed that in the 1930s and 1940s a number of canals were dug throughout the park, altering the sheet flow coming from Haw Branch. These activities certainly had a great effect on the local water system and water availability as well as quality at the coring site (FDEP 2007). Although the main goal of the channel building was to redirect the incoming water in order to permanently flood the hydric hammock in the central area of the park, the construction of an elevated road just north of our coring site most likely has led to increased water availability there. Although it was not possible to link the observed changes of the diatom assemblages at the transition between zone 3 and zone 4 to a single environmental parameter, it is obvious that the various infrastructural measures in the park affected the availability of diatom habitats and species composition of diatom communities.
The studied peat sequence from Highlands Hammock State Park beared well-preserved siliceous microfossil assemblages, including a diverse and abundant diatom flora. Detailed diatom analysis of the peat core revealed changes in the local hydrological state during the past 2,500 years. From 2,500 cal. year BP to 700 cal. year BP, the diatom assemblage were dominated by a tychoplanktonic species of the genus Aulacoseira, indicating a permanently flooded wetland environment. After 700 cal. year BP, the diatom record suggested a shortening of the local hydroperiod. The dominance of acidophilous, epiphytic Eunotia species from 700 cal. year BP onwards was the result of this decreased water level. Since water levels and the hydroperiod at the coring location are directly tied to a precipitation-fed seasonal creek, the observed drying was likely caused by climatic changes, i.e. decreased precipitation. The last turnover in the diatom assemblage occurred in the twentieth century and coincided with intensive human activities in the Highlands Hammock State Park. Through the construction of canals, elevated roads and dams, the hydrography was altered in such a way that the local water availability at our coring location was increased.
The study of diatoms from wetland peat sediments is well suited for the reconstruction of hydrological parameters such as water depth and hydroperiod. A shallow basin that is seasonally inundated is very sensitive to both natural and anthropogenic driven variations in the local hydrology, and the diatom community shows a clear response to these changes. The presence or absence of sponge spicules provides valuable additional palaeoecological information whereas no significant variation in the abundance of chrysophyte cysts and phytoliths was found.
Environmental reconstructions based on sediments from only one coring location are limited to a mainly local signal which makes it difficult to link observed changes to larger scale environmental and climatic phenomena. In order to reconstruct climate variability on a more regional scale, a number of sediment cores from different locations representing a larger area would have to be incorporated. Finally, a problem in the paleoecological interpretation of fossil diatom assemblages is the limited reliable knowledge of the autecological requirements of most of the species present in the studied sediment core. This can only be resolved by extensive taxonomic survey of diatoms and water monitoring in this region combined with the development of modern training sets.
We greatly appreciate the cooperation of State Park District 4 staff and the support of Ken Alvarez (FDEP) during fieldwork. Jan van Tongeren (Utrecht University) is acknowledged for assisting with sample and slide preparation in the lab. Evelyn Gaiser (Florida State University) contributed many fruitful discussions and constructive comments on the manuscript. This study is part of the Hurricanes and Global Change Program funded by Utrecht University and is the Netherlands Research School of Sedimentary Geology publication no 2011.10.02. The paper is part of a special issue on South Florida supported by the Florida Coastal Everglades Long Term Ecological Research Program (National Science Foundation Grant No. DEB-9910514).
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.