American Journal of Botany 90(12): 1720–1728. 2003.
RESPONSES
OF COMMON AND SUCCESSIONAL
HEATHLAND SPECIES TO MANIPULATED SALT SPRAY AND WATER AVAILABILITY1
MEGAN E. GRIFFITHS2
AND
COLIN M. ORIANS
Department of Biology, Tufts University, Medford, Massachusetts 02155 USA Coastal sandplain heathlands are a rare plant community in the northeastern United States. Salt spray and water availability are likely important factors determining heathland distribution. Field surveys and manipulative experiments were performed to examine heathland species’ responses to salt spray and water availability. We surveyed field distributions of four typical heathland species: Solidago puberula, Solidago rugosa, Gaylussacia baccata, and Myrica pensylvanica. The distributions of two native tree species, Pinus rigida and Quercus ilicifolia, were also surveyed because they succeed into coastal heathlands with low disturbance frequency. We then manipulated salt spray and water in the field and measured species’ water status, necrosis, and growth responses to the treatments. Predawn xylem pressure potential and necrosis were strongly affected by high salt spray and low water availability. Shoot elongation was also limited in S. puberula and S. rugosa grown in high salt, low water treatments. Gaylussacia baccata and Q. ilicifolia were particularly sensitive to high salt spray and low water, suggesting that they might excluded be from areas with those conditions. The interaction between salt spray and water availability could affect the landscape scale and should be incorporated into conservation management plans. Key words:
coastal sandplain heathlands; disturbances; heathland species; northeastern USA; salt spray; water availability.
Many ecosystems require abiotic and biotic disturbances to maintain plant community composition (White, 1979). One such ecosystem is sandplain heathland, which is a dwarf shrub community that occurs in coastal regions in the northeastern United States. Fire and domestic livestock grazing are two anthropogenic disturbances that have maintained and caused the historical expansion of heathlands by preventing succession of native tree species such as Pinus rigida and Quercus spp. (Godfrey and Alpert, 1985; Dunwiddie et al., 1997). In recent decades the combined effects of fire suppression, grazing cessation, and residential development have caused a decline of coastal heathlands, and these ecosystems are now considered to be endangered (Godfrey and Alpert, 1985; Noss et al., 1995). While much research has focused on the effects of anthropogenic disturbances on the composition coastal heathland communities (Zaremba et al., 1983; Dunwiddie and Caljouw, 1990; Dunwiddie, 1991; Dunwiddie et al., 1997), very little attention has been given to the possible role of natural abiotic conditions. Salt spray is an abiotic disturbance that plays a critical ecological role in many coastal plant communities. Research has demonstrated that salt spray disrupts the water balance of plants (Munns, 1993), causes necrosis or loss of leaves (Karschon, 1958), and leads to growth reduction (Tominaga et al., 1991). As a result of these physiological, morphological, and growth effects, salt spray causes dwarfing and asymmetric Manuscript received 4 March 2003; revision accepted 6 June 2003. The authors thank R. Keith for field assistance and R. Johnson, L. Raleigh, C. Egan, and T. Chase for logistical support. The following organizations permitted this research to take place on their properties: Sheriff’s Meadow Foundation (Priscilla Hancock Meadow and Quansoo Beach), Quansoo Beach Association (Quansoo Beach), and The Trustees of Reservations (Long Point Wildlife Refuge). Housing for M. G. was provided by The Nature Conservancy. Funding came from Tufts Institute for the Environment, the Massachusetts Environmental Trust, the Biology Department of Tufts University, and the Andrew Mellon Foundation. We thank E. Farnsworth and two anonymous reviewers for comments on the manuscript. 2 E-mail:
[email protected]. 1
growth in coastal woody plants (Wells and Shunk, 1937; Roper-Lindsay and Say, 1986). In addition to affecting plant stature, salt spray can also greatly influence coastal plant community composition. Zonation and succession in coastal systems are often the result of variation in species’ tolerance to salt spray (Oosting and Billings, 1942; Oosting, 1945; Boyce, 1954; Randall, 1970; van der Valk, 1974; Barbour and DeJong, 1977; Barbour, 1978; Tyndall et al., 1987; Sykes and Wilson, 1988). We provide evidence that salt spray might be an important abiotic disturbance in coastal heathlands. Through field surveys, we found that salt spray accumulation on heathland plants is closely correlated with increased water stress, increased leaf necrosis, reduced height, and an absence of trees (Griffiths, 2003). Greenhouse studies have demonstrated that these responses are the direct result of salt spray (Griffiths and Orians, 2003). This could be one mechanism through which the characteristic dwarf stature of heathlands is maintained. Because salt spray primarily inhibits plant growth by causing water stress (Pammenter and Smith, 1983; Munns, 1993), we hypothesized that the effects of salt spray will be more pronounced when there is low water availability. In fact, heathlands frequently occupy areas with low water availability and many heathland species can tolerate high levels of water stress (Roper-Lindsay and Say, 1986). Thus, in areas close to the ocean with high salt spray and low water availability, nonheathland species may be excluded by the abiotic conditions. One goal of this study was to explain within- and among-site variation in plant community composition as a function of water availability and salt spray accumulation patterns. Our second objective was to carry out manipulative field experiments on common and successional heathland plants to determine interspecific differences in salt spray tolerance. The physiological, morphological, and growth responses of six species to different salt spray and water availability were measured. We predicted that differences in species’ responses to salt spray and water availability would be reflected in the distribution of the species relative to the ocean and soil water availability.
1720
December 2003]
GRIFFITHS
AND
ORIANS—SALT
SPRAY AND WATER EFFECTS ON HEATHLANDS
TABLE 1. Salt spray accumulation on the windward side of Myrica pensylvanica at Quansoo Beach and Long Point Wildlife Refuge during the summer growing season (means 6 1 SE). At Quansoo Beach, no Myrica pensylvanica plants grew at 25 and 75 m (ND 5 no data). Salt spray accumulation (mg · dm22 · d21)
Distance from dune crest (m)
Quansoo Beach
25 50 75 100 125 150 175 200
ND 3.07 6 1.12 ND 1.83 6 0.75 2.14 6 1.18 2.00 6 0.89 1.72 6 0.87 1.62 6 0.99
Long Point
4.49 3.32 3.26 2.46 1.94 1.67 1.70 1.10
6 6 6 6 6 6 6 6
1.90 1.44 1.29 1.55 1.04 0.95 1.11 0.59
MATERIALS AND METHODS Field surveys—We performed field surveys on Martha’s Vineyard, Massachusetts, USA (418229 N, 708409 W) during the summer of 1999. Two conservation areas were used for field surveys: Quansoo Beach (QB) and Long Point Wildlife Refuge (LP). The two field sites used in this study differ in topography. Quansoo Beach (QB) has a high single dune that extends approximately 75 m inland from the dune crest, while the single dune at Long Point Wildlife Refuge (LP) is low and extends 15 m inland. Coastal areas of QB have been actively managed with mowing in the last 20 yr, while LP has not been managed for at least 50 yr. These sites are 2 km apart on the southern shore of Martha’s Vineyard and have similar exposure to the ocean. We had previously measured salt spray accumulation on windward-facing leaves of Myrica pensylvanica plants and found that salt spray accumulation decreases as distance from the dune crest increases (Table 1). Along the distance gradient, plants growing 25 m from the dune crest accumulate approximately 4.5 mg NaCl · dm22 · d21 on windward leaves, while plants at 125 m accumulate approximately 2 mg NaCl · dm22 · d21. Precipitation data were collected at 15-min intervals by a Davis Weather Monitor II (Davis Instruments, Hayward, California, USA) at Long Point Wildlife Refuge. We measured soil moisture in both sites at eight distances from the dune crest: 25, 50, 75, 100, 125, 150, 175, and 200 m. At each distance, two replicate soil cores were taken from the top 15 cm of the soil, excluding the litter layer. The soil samples were mixed and a 10-g soil subsample was placed in an aluminum tin and dried in a 1108C oven for 24 h. Soil water content was calculated by subtracting dry mass from wet mass to give the amount of water expressed in grams of water per 100-g sample (Brower et al., 1997). We measured species presence/absence in 1 3 50 m quadrats running parallel to the ocean. At both field sites, two quadrats were placed at each of eight distances from the dune crest. In addition, plant stem density was sampled per 1 m2. At each distance, we measured the number of stems of all species in six randomly placed 1-m2 plots. Here we report stem density only for the focal species used in the manipulative experiment: Solidago puberula Nutt., Solidago rugosa Miller, Gaylussacia baccata (Wangenh.) K. Koch, Myrica pensylvanica Mirbel, and Quercus ilicifolia Wangenh. Pinus rigida Miller, although present in adjacent inland forests, was not present in any of the coastal heathlands surveyed. Manipulative experiment—Our manipulative study was carried out at Priscilla Hancock Meadow, a conservation area owned by the Sheriff’s Meadow Foundation. The study site was located approximately 1 km from the ocean, just inland from a stand of tall trees. Such stands can effectively block salt spray and wind from areas leeward of the windbreak (Clayton, 1972). We measured salt spray and found that it did not accumulate in the study area during the experiment, which allowed us to control for natural wind and salt spray effects while manipulating salt spray levels. Background soil moisture was also measured; the levels around focal plants were not significantly dif-
1721
ferent among treatment plots, with percentage soil moisture of 9.99 6 0.48 (mean 6 1 SE). The study used six focal species: four typical heathland species and two native tree species that are successional in heathlands. The heathland species included the forbs S. puberula and S. rugosa, the shrubs G. baccata and M. pensylvanica, and the successional trees P. rigida and Q. ilicifolia, which are frequently found in heathlands with low disturbance levels. Eighteen individuals of each species were tagged and randomly assigned into salt spray and water treatments (n 5 3 plants per treatment combination). Because G. baccata is clonal (Harper, 1995), individuals of this species were used only if they were separated from other individuals by a distance greater than 10 m. All species were standardized by height, and only reproductive stalks were used for Solidago spp. Three levels of salt spray treatment and two levels of water treatment were used for each species. Filtered seawater (31 ppt), collected off the southern shore of Martha’s Vineyard, was applied with a handheld plant mister at either 0 mg NaCl · dm22 · d21, 2 mg NaCl · dm22 · d21 (equivalent to windward exposure at 125 m from the dune crest), or 4 mg NaCl · dm22 · d21 (equivalent to windward exposure at 25 m from the dune crest). For the water treatment, either no water or supplemental fresh water was applied at the level of 1 L/ d to the base of the plant using a backpack water sprayer. Treatments began on 28 June 1999 and lasted for 8 wk. The precipitation during the first four weeks of the experiment was 3.23 cm, lower than the normal average of 7.42 cm, which could have resulted in high water stress for plants that did not receive supplemental water. Following the treatment period, water status, necrosis, and growth measurements were taken. Two aspects of water status were tested: predawn xylem pressure potential and leaf water content. Predawn xylem pressure potential of stems was measured in the field at 0400 using a pressure chamber (PMS Instrument Company, Corvalis, Oregon, USA). Leaf water content was measured on the first three fully expanded leaves or needles at the top of each plant. Leaves were oven-dried at 608C for 48 h to determine dry mass. Leaf water content was calculated using the following equation: water content 5 (leaf wet mass 2 leaf dry mass)/leaf dry mass. No mortality occurred during the course of the experiment. Leaf necrosis, which results from chloride toxicity (Parsons and Gill, 1968), was measured as an index for salt spray damage. Leaf necrosis was assessed on the same leaf used for water content measurements and was measured using a grid and expressed as a proportion of the total leaf area with necrotic damage. Growth was measured using leaf area and shoot elongation. We sampled the area of the first fully expanded leaf on all six species tested. Shoot elongation was determined by measuring the height of plants before and after the treatment period. Solidago puberula and S. rugosa were the only species that exhibited vertical growth during the experimental period, so only data for these two species are reported. Statistical analyses—All statistical analyses were run using general linear model procedures on SAS (SAS Institute, 1990). Soil water content was analyzed with a two-factor analysis of variance. For the manipulative experiments, three-factor analyses of variance using the main effects of species, salt spray treatment, and water treatment were run on all data. Additional twofactor analyses of variance using the main effects of salt spray treatment and water treatment were run for each individual species to determine the degree of response for each species to manipulated salt spray and water levels.
RESULTS Field surveys—The two field sites differed in their soil water content. On average, soil water content was higher at QB than at LP, and the two sites differed in soil water content across distances (Fig. 1a). At QB, soil moisture was low at 25 m and 50 m where sampling stations were on the sand dune and soil moisture peaked at 100 m in an area of salt marsh. Sampling areas inland of 100 m had relatively constant soil water content of approximately 15%. Soil water content at LP was approximately 10% across the distance gradient.
1722
AMERICAN JOURNAL
OF
BOTANY
[Vol. 90
all quadrats inland of 125 m from the dune crest, while it was present in all sampling quadrats at LP. At QB, M. pensylvanica was present in more quadrats as distance from the dune crest increased, but the opposite trend was found for the same species at LP. The stem density of different species varied by site (Fig. 1b–c). Solidago puberula occurred infrequently, with only two stems at 175 m at QB. Solidago rugosa was present at both QB and LP, although it did not occur at 25 m from the dune crest at either site. Solidago rugosa stem density peaked at 175 m at QB and at 100 m at LP. Gaylussacia baccata was present at both field sites, but patterns in stem density differed between the sites. At QB, G. baccata was not present until 125 m, but then it was present at relatively high density at sampling points farther inland. At LP, G. baccata appeared at 25 m and maintained a very high stem density until 200 m. Myrica pensylvanica was also present at both field sites. At QB, M. pensylvanica appeared in sampling plots at 125 m and had a relatively constant density until 200 m. At LP, M. pensylvanica occurred at 25 m, but the density was relatively low and, although present in quadrats (Table 2), no individuals of this species were captured in 1-m2 sampling plots inland of 75 m. While present in quadrats at both field sites, the successional tree species Q. ilicifolia was only captured by sampling plots at LP, where it had low stem density from 150 to 200 m. Pinus rigida, although present in inland forests adjacent to the study areas, was not present in the heathlands surveyed. Manipulative experiment—Many of the plant traits measured had physiological, morphological, and growth responses to manipulated salt spray and water availability. The degree of response to salt spray and water availability varied among species. The most significant effects of treatments were detected for predawn xylem pressure potential and necrosis. Water status—All six of the species tested showed a significant decrease in predawn xylem pressure potential in response to salt spray, indicating that the plants became more waterstressed with increasing salt spray (Table 3, Fig. 2). There were significant species effects, significant water treatment effects, and an interactive effect between salt spray and water treatment, indicating that additional water availability ameliorated the effects of high salt spray (Table 3). The other measure of water status, leaf water content, varied by species but did not change significantly under salt spray and water treatment (Tables 3, 4).
Fig. 1. (a) Percentage soil water content and (b–c) stem density at Quansoo Beach (QB) and Long Point Wildlife Refuge (LP). Species surveyed were Solidago puberula, Solidago rugosa, Gaylussacia baccata, Myrica pensylvanica, and Quercus ilicifolia. Bars represent the mean of six replicate 1-m2 samples taken at each site for each distance from the dune crest. ND 5 no data.
All species used in the manipulative experiment except for P. rigida were found in the 50-m2 sampling quadrats at QB and LP (Table 2). The forb species S. puberula was the least common species, occurring only in one quadrat at 175 m from the dune crest. Solidago rugosa (at QB) and Q. ilicifolia (at QB and LP) were present in more quadrats as distance from the dune crest increased. At QB, G. baccata was present in
Necrosis—Necrosis increased with increasing levels of salt spray in four of the five species measured (Table 3, Fig. 3). Necrosis was not measured in P. rigida due to the difficulty of measuring necrosis on needles. In the remaining five species examined, there was a significant three-way interaction (Table 3). Further analysis by species revealed significant salt spray effects for S. puberula, S. rugosa, and G. baccata (Fig. 3), but the native heathland shrub M. pensylvanica showed no necrosis (data not shown). The shrub species G. baccata had the most severe levels of necrosis. There were no interactive effects between salt spray and water treatments, but there was a slight decrease in salt spray damage in supplemental water treatments. While Q. ilicifolia was not significantly affected by salt spray, necrosis increased in this species under increasing salt spray levels and had an increased salt spray response in the high water treatment.
December 2003]
GRIFFITHS
AND
ORIANS—SALT
1723
SPRAY AND WATER EFFECTS ON HEATHLANDS
TABLE 2. Number of 50-m2 quadrats at Quansoo Beach and Long Point in which study species occurred. Two quadrats were censused at each distance. Distance from dune crest (m)
Quansoo Beach 25 50 100 125 150 175 200 Long Point 25 50 75 100 125 150 175 200
Solidago puberula
Solidago rugosa
Gaylussacia baccata
Myrica pensylvanica
Quercus ilicifolia
0 0
0 0
0 1
0 0
0 0 1 0
2 2 2 2
0 0 Salt Marsh—None Present 2 2 2 2
2 2 2 2
1 1 0 1
0 0 0 0 0 0 0 0
1 2 2 2 2 2 1 2
2 2 2 2 2 2 2 2
2 2 2 2 0 1 2 1
1 1 1 1 2 2 2 2
Growth responses—Leaf area varied by species but not by salt spray and water treatments because no species except Solidago spp. flushed leaves during the experiment (Tables 5 and 6). Only the two forb species, S. puberula and S. rugosa, grew during the growth period. Shoot elongation of both was strongly reduced with increasing salt spray, although these changes were not statistically significant (Table 6, Fig. 4). Solidago puberula grew more in the high water treatment (Fig. 4). Growth was inhibited for these two species under high salt spray treatments, and the effect was ameliorated by adding water.
In addition, heathland species grew only in areas with soil water content between 10 and 20%. Previous studies have determined that complex interactions between multiple abiotic factors shape coastal plant community composition (Barbour and DeJong, 1977; Houle, 1997). Our manipulative experiments demonstrated an interactive effect among salt spray, water availability, and species. Salt spray and water, both individually and in combination with one another, might play an important role in determining which species grow in high salt spray heathlands.
DISCUSSION
Field surveys—Soil water content was constant across the distance gradient at LP, and the levels were lower than at QB. Because QB also had low soil water content close to the ocean, heathland plant species might be limited by high salt and low water in areas close to the ocean at QB. The higher water availability might allow heathland plants to grow in high salt spray areas at LP. Based on the soil water availability profile, we expected that species distributions in the QB plant community close to the ocean would differ along a distance gradient, while LP distributions would be more consistent across the site. Indeed, in the field surveys at QB, heathland species, with the exception of M. pensylvanica, did not grow until 125 m from the dune crest. Therefore the extremes in water avail-
Species distributions are determined by intrinsic factors such as recruitment and by extrinsic factors such as disturbances and abiotic conditions. In coastal areas, salt spray often influences the distributions of species, and plant population density can increase along a gradient of decreasing salt spray (Cheplick and Demetri, 2000). Our field surveys identified distinctive patterns of species distribution that reflected differences in salt spray accumulation and water availability at different distances from the ocean and different field sites. Because some of the heathland species, including the successional tree species Q. ilicifolia, were rare close to the dune crest, they might have been limited by high salt spray in those areas.
TABLE 3. ANOVA table of water status (xylem pressure potential and leaf water content) and necrosis of Solidago puberula, Solidago rugosa, Gaylussacia baccata, Myrica pensylvanica, Pinus rigida, and Quercus ilicifolia in response to three levels of salt spray and two levels of water. Pinus rigida was not included in the analysis for necrosis. Boldface type indicates significance at the 0.05 level. Xylem potential
Leaf water content
Necrosis
Source of variation
df
F
P
df
F
P
df
F
P
Species Salt Water Species 3 salt Species 3 water Salt 3 water Species 3 salt 3 water Error
5 2 1 10 5 2 10 286
12.35 49.29 19.07 1.73 1.41 5.03 1.29
,0.01 ,0.01 ,0.01 0.09 0.23 0.01 0.25
5 2 1 10 5 2 10 286
4.51 0.58 0.01 0.36 0.01 1.34 1.38
,0.01 0.57 0.94 0.96 1.00 0.27 0.21
4 2 1 8 4 2 8 240
36.61 38.03 9.54 0.11 2.06 0.33 5.06
,0.01 ,0.01 ,0.01 0.74 0.09 0.72 ,0.01
1724
AMERICAN JOURNAL
OF
BOTANY
[Vol. 90
Fig. 2. Predawn xylem pressure potential at 0400 for (a) Solidago puberula, (b) Solidago rugosa, (c) Gaylussacia baccata, (d) Myrica pensylvanica, (e) Pinus rigida, and (f) Quercus ilicifolia in manipulated salt spray and water treatments (means and 1 SE). No salt 5 0 mg · dm22 · d21; medium 5 2 mg · dm22 · d21; and high 5 4 mg · dm22 · d21; not watered 5 no supplemental water; watered 5 1 L/d.
ability from 25 to 100 m might have limited most heathland species. Unlike QB, soil water content at LP was constant at approximately 10% across the distance gradient. The survey of plant community structure at LP revealed that all heathland species except S. puberula were found throughout the distance gradient, but the patterns of distribution varied among species. Solidago puberula was relatively infrequent in the study areas, although it is more commonly found at other coastal sites.
Given its observed presence close to the ocean at other sites, the rarity of S. puberula at QB and LP is likely due to an intrinsic factor such as recruitment limitation rather than abiotic conditions alone. At LP, the forb species S. rugosa was present at all distances from the ocean, but stem density was highest at intermediate distances from the ocean. Given a low abundance at 25 m, S. rugosa might be limited by salt spray or some other factor close to the ocean. The shrub species M.
December 2003]
GRIFFITHS
AND
ORIANS—SALT
1725
SPRAY AND WATER EFFECTS ON HEATHLANDS
TABLE 4. Leaf water content in six focal species at three levels of salt spray (no salt 5 0 mg · dm22 · d21, medium 5 2 mg · dm22 · d21, and high 5 4 mg · dm22 · d21) and two levels of water. Not watered Species
Solidago puberula Solidago rugosa Gaylussacia baccata Myrica pensylvanica Pinus rigida Quercus ilicifolia
No salt
2.07 1.64 1.30 1.31 1.93 1.13
6 6 6 6 6 6
Watered (1 L/d)
Medium
0.02 0.09 0.05 0.02 0.03 0.03
2.30 1.73 1.32 1.30 2.05 1.13
6 6 6 6 6 6
0.08 0.11 0.05 0.04 0.03 0.03
High
2.20 1.51 1.27 1.30 4.54 1.13
pensylvanica was most common close to the dune crest, while G. baccata was common throughout the distance gradient. Gaylussacia baccata and S. rugosa are both clonal perennial species, which may explain why they have such high density when they are present in the community. The successional tree Q. ilicifolia was rare except at distances farther from the dune crest, suggesting that it may be limited by salt spray close to the ocean. Pinus rigida did not grow in any of the study areas, but previous studies have described this species as being saltintolerant (Boyce, 1954). In other studies, P. taeda was eliminated from high salt spray areas unless it was sheltered from wind (Wells and Shunk, 1938; Johnson and Young, 1993). Because P. rigida was absent close to the ocean, salt spray or low water availability may be limiting the growth of this species. Manipulative experiment—Water status was affected by salt spray and water treatments. All species had decreasing predawn xylem pressure potential with increasing salt spray accumulation. This salt spray effect was ameliorated in all species when they were given supplemental water, particularly for S. puberula and P. rigida, suggesting that both common and successional trees would be better able to survive in high salt spray areas with high water availability. Gaylussacia baccata and Q. ilicifolia exhibited the lowest predawn xylem pressure potential measurements in low water, high salt spray treatments. Of the species tested, these two species are the most likely to be limited from growing in high salt spray, low water areas. We did not find a significant increase in leaf water content, which would be expected if plants altered leaf succulence as an adaptation to avoid salt spray stress (Boyce, 1951; Cartica and Quinn, 1980; Boyd and Barbour, 1986; Alpha et al., 1996). Necrotic damage increased with increasing salt spray levels in low water S. puberula, S. rugosa, G. baccata, and Q. ilicifolia. The necrotic response was slightly lower for high wa-
6 6 6 6 6 6
No salt
0.13 0.07 0.02 0.02 3.27 0.09
1.99 1.66 1.28 1.23 2.00 1.09
6 6 6 6 6 6
0.05 0.06 0.05 0.01 0.01 0.03
Medium
2.16 1.65 1.48 1.26 5.06 1.15
6 6 6 6 6 6
0.06 0.05 0.04 0.02 3.09 0.02
High
2.08 1.59 1.34 1.30 1.47 1.07
6 6 6 6 6 6
0.16 0.07 0.08 0.02 0.24 0.04
ter plants of all species except for Q. ilicifolia, which showed an increase. Necrosis was not measured in P. rigida, but there is evidence from previous studies that salt spray does accumulate in needles of Pinus spp. and cause necrosis, even at low levels (Bedunah and Trlica, 1979; Zobel and Nighswander, 1990). Myrica pensylvanica had no necrosis, demonstrating that this species is extremely salt tolerant and might not be limited in high salt spray areas, even if there is extremely low water availability. Growth is often affected by salt spray, although the type and degree of growth response differs among species. Growth in some species is stimulated when sprayed with salt (Rozema et al., 1982; Maze and Whalley, 1992), while growth is suppressed in others (Cheplick and Demetri, 1999). In our experiment, growth responses were difficult to detect due to the short duration of the experimental period. Leaf areas did not differ among treatments in any of the species tested even though some species, most notably S. puberula and S. rugosa, were producing new leaves during the treatment period. While leaf area was not reduced in salt spray treatments, photosynthetic leaf area was substantially reduced by necrosis, up to 50% in some species. Such a reduction could have long-term consequences on how much the plants grow, how much energy they put into storage and reproduction, and how much they can grow and reproduce in subsequent years (Parker and Patton, 1975). Despite the reduction, changes in shoot elongation were only detected in S. puberula and S. rugosa, although the differences among treatments were not significant. However, the general trend found for the Solidago spp. was that salt spray decreased shoot growth in low water treatments. A reduction in growth was not present in the high water treatment, suggesting that these species would be able to survive in high salt spray areas that also had high water availability. Scaling up from physiology to distribution—In the context of results from our manipulative experiment, the field distri-
TABLE 5. ANOVA table of final leaf area and shoot elongation of Solidago puberula, Solidago rugosa, Gaylussacia baccata, Myrica pensylvanica, Pinus rigida, and Quercus ilicifolia in response to three levels of salt spray and two levels of water. Pinus rigida was not included in the analysis for leaf area. Boldface type indicates significance at the 0.05 level. Leaf area
Shoot elongation
Source of variation
df
F
P
df
F
P
Species Salt Water Species 3 salt Species 3 water Salt 3 water Species 3 salt 3 water Error
4 2 1 8 4 2 8 240
144.51 0.21 0.10 0.18 1.15 0.68 0.44
,0.01 0.81 0.76 0.99 0.34 0.51 0.923
5 2 1 10 5 2 10 286
24.57 29.68 0.81 6.01 1.19 1.14 0.94
,0.01 ,0.01 0.37 ,0.01 0.32 0.33 0.50
1726
AMERICAN JOURNAL
OF
BOTANY
[Vol. 90
Fig. 3. Percentage leaf area with necrosis in (a) Solidago puberula, (b) Solidago rugosa, (c) Gaylussacia baccata, and (d) Quecus ilicifolia (means 1 1 SE). No salt 5 0 mg · dm22 · d21; medium 5 2 mg · dm22 · d21; and high 5 4 mg · dm22 · d21; not watered 5 no supplemental water; watered 5 1 L/d.
bution of common and successional heathland species can be explained by the independent and combined effects of salt spray and water availability. The field distributions of heathland plants at QB appear to be strongly controlled by low water availability close to the ocean. At LP, where water availability is more constant across the distance gradient, the distributions of plants are likely to be more strongly controlled by salt spray. In such areas, many heathland plants grow close to the ocean, and only those that are salt-intolerant are limited from the coastal areas. The two forb species, S. puberula and S. rugosa, had similar physiological responses to salt spray and water treatments. Under a high salt spray regime, both had more water stress and necrotic damage, as well as a slight but nonsignificant growth inhibition, suggesting that they might be limited from growing
in high salt spray areas. These effects were ameliorated when supplemental water was added, indicating that these species might be able to survive in high salt spray areas if there is sufficient water. The distributions of these two species are difficult to generalize because S. puberula was so rare. However, S. rugosa distributions followed the predictions made from the manipulative study. At QB, this species was absent from the high salt spray and low water areas at 25 and 50 m from the dune crest and appeared in the community only at 125 m. Solidago rugosa at LP occurred at all sampling points, although it was most abundant in the intermediate distances. The areas close to the ocean at LP have high salt spray inputs but much higher water availability than at QB. Of the two common heathland shrubs tested, G. baccata was more susceptible to salt spray treatments. Water stress and
TABLE 6. Leaf area of five focal species at three levels of salt spray (no salt 5 0 mg · dm22 · d21, medium 5 2 mg · dm22 · d21, and high 5 4 mg · dm22 · d21) and two levels of water. Not watered Species
Solidago puberula Solidago rugosa Gaylussacia baccata Myrica pensylvanica Quercus ilicifolia
No salt
16.4 10.1 5.29 7.09 13.5
6 6 6 6 6
2.10 1.69 0.77 0.52 0.89
Medium
15.5 9.20 3.52 6.34 13.5
6 6 6 6 6
0.41 1.05 0.29 0.31 2.30
Watered (1 L/d) High
16.3 10.7 4.68 6.72 14.6
6 6 6 6 6
1.59 1.67 0.63 0.25 2.18
No salt
14.1 10.9 4.10 6.39 16.5
6 6 6 6 6
1.20 0.85 0.53 0.50 1.56
Medium
15.5 12.3 4.10 6.42 14.4
6 6 6 6 6
1.47 1.41 0.36 0.11 1.14
High
14.1 9.64 4.74 7.03 15.6
6 6 6 6 6
2.35 0.18 0.08 0.71 3.23
December 2003]
GRIFFITHS
AND
ORIANS—SALT
SPRAY AND WATER EFFECTS ON HEATHLANDS
1727
Fig. 4. Shoot elongation of (a) Solidago puberula and (b) Solidago rugosa in manipulated salt spray and water treatments (means 1 1 SE). No salt 5 0 mg · dm22 · d21; medium 5 2 mg · dm22 · d21; and high 5 4 mg · dm22 · d21; not watered 5 no supplemental water; watered 5 1 L/d.
necrotic damage increased under manipulated salt spray, but supplemental water lessened these effects. At QB, this species was absent at 25 and 50 m from the dune crest, but it was common in the community beyond 125 m and abundant at all distances at LP. As with S. rugosa, the presence of G. baccata in the high salt spray zone at LP suggests that the species was able to grow close to the ocean because high water availability ameliorated the effects of salt spray. The second heathland shrub tested, M. pensylvanica, was the least susceptible to salt spray. Although water stress slightly increased under salt spray, M. pensylvanica had no necrotic damage, indicating that salt spray does not physiologically damage this species. Myrica pensylvanica grows quite close to the ocean, even in high salt spray, low water areas. Quercus ilicifolia showed the most extreme water stress response to salt spray treatment, and water only slightly ameliorated the effect, suggesting that, of the species tested, it is the most likely to be inhibited in high salt spray areas. Field observations support this prediction. At QB, Q. ilicifolia was present but not common in areas far from the ocean with low salt spray and high water. Similarly, at LP, this species was present across the distance gradient, but it was not abundant in areas close to the ocean. The other successional tree species tested, P. rigida, showed less severe water stress under salt spray manipulations, which would lead us to predict that it potentially could grow in heathlands with lower salt spray. Despite this prediction, we did not find any P. rigida growing in our survey areas within 200 m from the dune crest. The possible explanations for this absence are discussed in Griffiths (2003). Conservation implications—There are complex interactions between salt spray, water availability, and plant stress. In the biphasic model of plant responses to salinity (Munns, 1993), the first stage of growth inhibition may occur because of a disruption in water relations, leading to water stress very similar to that caused by drought. If plants are growing with high water availability, however, the plant will be able to maintain normal water relations even under high salt spray regimes, perhaps explaining why we were able to detect changes in water status but not growth after 8 wk. While salt spray and water availability do not entirely account for the zonation in
coastal areas, they are likely to have a profound impact on the community close to the ocean. When designing and implementing management plans, we need to consider the many different abiotic factors that influence plant community composition. Salt spray may limit the growth of some species in heathlands growing close to the ocean. However, it might be more likely that successional species will occur even in the high salt spray zone if the sites have high water availability. Managers should seek to understand the profile of salinity, water availability, and salt spray at sites that are managed for conservation purposes. Sites with higher water availability may support a more diverse heath assemblage, but will be susceptible to succession by P. rigida and Q. ilicifolia. Management activities to facilitate persistence of indicator heathland species should focus on areas where both salt spray and water availability are intermediate, which is optimized to promote growth of salt-tolerant heathland species and to discourage growth of salt-intolerant successional species. Managers should also be aware that hydrological changes (i.e., fresh- or saltwater intrusion) will affect community structure, probably within a few growing seasons. Disturbance events, particularly tropical storms and hurricanes may ‘‘reset’’ the successional clock by eliminating more saltintolerant species from heath (Griffiths, 2003). LITERATURE CITED ALPHA, C. G., D. R. DRAKE, AND G. GOLDSTEIN. 1996. Morphological and physiological responses of Scaevola sericea (Goodeniaceae) seedlings to salt spray and substrate salinity. American Journal of Botany 83: 86–92. BARBOUR, M. G. 1978. Salt spray as a microenvironmental factor in the distribution of beach plants at Point Reyes, California. Oecologia 32: 213–224. BARBOUR, M. G., AND T. M. DEJONG. 1977. Response of West Coast beach taxa to salt spray, seawater inundation, and soil salinity. Bulletin of the Torrey Botanical Club 104: 29–34. BEDUNAH, D., AND M. J. TRLICA. 1979. Sodium chloride effects on carbon dioxide exchange rates and other plant and soil variables of ponderosa pine. Canadian Journal of Forestry Research 9: 349–353. BOYCE, S. G. 1951. Salt hypertrophy in succulent dune plants. Science 114: 544–545. BOYCE, S. G. 1954. The salt spray community. Ecological Monographs 24: 29–67. BOYD, R. S., AND M. G. BARBOUR. 1986. Relative salt tolerance of Caikile
1728
AMERICAN JOURNAL
edentula (Brassicaceae) from lacustrine and marine beaches. American Journal of Botany 73: 236–241. BROWER, J. E., J. H. ZAR, AND C. N. VON ENDE. 1997. Field and laboratory methods for general ecology. WCB McGraw-Hill, Boston, Massachusetts, USA. CARTICA, R. J., AND J. A. QUINN. 1980. Responses of populations of Solidago sempervirens (Compositae) to salt spray across a barrier beach. Bulletin of the Torrey Botanical Club 104: 29–34. CHEPLICK, G. P., AND H. DEMETRI. 1999. Impact of saltwater spray and sand deposition on the coastal annual Triplasis purpurea (Poaceae). American Journal of Botany 86: 703–710. CHEPLICK, G. P., AND H. DEMETRI. 2000. Population biology of the annual grass Triplasis purpurea in relation to distance from shore on Staten Island, New York. Journal of Coastal Conservation 5: 145–154. CLAYTON, J. E. 1972. Salt spray and mineral cycling in two California coastal ecosystems. Ecology 53: 74–81. DUNWIDDIE, P. W. 1991. Comparisons of aboveground arthropods in burned, mowed and untreated sites in sandplain grasslands on Nantucket Island. American Midland Naturalist 125: 206–212. DUNWIDDIE, P. W., AND C. CALJOUW. 1990. Prescribed burning and mowing of coastal heathlands and grasslands in Massachusetts. In R. S. Mitchell, C. J. Sheviak, and D. J. Leopold [eds.], Ecosystem management: rare species and significant habitats, 271–275. New York Museum State Bulletin 471. DUNWIDDIE, P. W., W. A. I. PATTERSON, J. L. RUDNICKY, AND R. E. ZAREMBA. 1997. Vegetation management in coastal grasslands on Nantucket Island, Massachusetts: effects of burning and mowing from 1982 to 1993. In P. D. Vickery and P. W. Dunwiddie [eds.], Grasslands of northeastern North America, 85–98. Massachusetts Audubon Society, Lincoln, Massachusetts, USA. GODFREY, P. J., AND P. ALPERT. 1985. Racing to save the coastal heaths. Nature Conservancy News 35: 11–13. GRIFFITHS, M. E. 2003. Salt spray effects on rare New England coastal sandplain heathland plant communities. Ph.D. dissertation, Tufts University, Medford, Massachusetts, USA. GRIFFITHS, M. E., AND C. M. ORIANS. 2003. Salt spray differentially affects water status, necrosis, and growth in coastal sandplain heathland species. American Journal of Botany 90: 1188–1196. HARPER, K. A. 1995. Effect of expanding clones of Gaylussacia baccata (black huckleberry) on species composition in sandplain grassland on Nantucket Island, Massachusetts. Bulletin of the Torrey Botanical Club 122: 124–133. HOULE, G. 1997. Interactions between resources and abiotic conditions control plant performance on subarctic coastal dunes. American Journal of Botany 84: 1729–1737. JOHNSON, S. R., AND D. R. YOUNG. 1993. Factors contributing to the decline of Pinus taeda on a Virginia barrier island. Bulletin of the Torrey Botanical Club 120: 431–438. KARSCHON, R. 1958. Leaf absorption of wind-borne salt and leaf scorch in Eucalyptus camaldulensis Dehn. Ilanoth 4: 5–25. MAZE, K. M., AND R. D. B. WHALLEY. 1992. Effects of salt spray and sand burial on Spinifex sericeus R. Br. Australian Journal of Ecology 17: 9– 19.
OF
BOTANY
[Vol. 90
MUNNS, R. 1993. Physiological processes limiting plant growth in saline soils: some dogmas and hypotheses. Plant, Cell and Environment 16: 15–24. NOSS, R. F., E. T. LAROE, III, AND J. M. SCOTT. 1995. Endangered ecosystems of the United States: a preliminary assessment of loss and degradation. U.S. Department of the Interior, National Biological Service, report no. 28. Washington, D.C., USA. OOSTING, H. J. 1945. Tolerance of salt spray of plants of coastal dunes. Ecology 26: 85–89. OOSTING, H. J., AND W. D. BILLINGS. 1942. Factors affecting vegetational zonation on coastal dunes. Ecology 23: 131–142. PAMMENTER, N. W., AND V. R. SMITH. 1983. The effect of salinity on leaf water relations and chemical composition in the sub-Antarctic tussock grass Poa cookii Hook F. New Phytologist 94: 585–594. PARKER, J., AND R. L. PATTON. 1975. Effects of drought and defoliation on some metabolites in roots of black oak seedlings. Canadian Journal of Forestry Research 5: 457–463. PARSONS, R. F., AND A. M. GILL. 1968. The effects of salt spray on coastal vegetation at Wilson’s Promontory, Victoria, Australia. Proceedings of the Royal Society of Victoria 81: 1–10. RANDALL, R. E. 1970. Vegetation and environment on the Barbados coast. Journal of Ecology 58: 155–172. ROPER-LINDSAY, J., AND A. M. SAY. 1986. Plant communities of the Shetland Islands. Journal of Ecology 74: 1013–1030. ROZEMA, J., F. BIJL, T. DUECK, AND H. WESSELMAN. 1982. Salt-spray stimulated growth in strand-line species. Physiologia Plantarum 56: 204– 210. SAS INSTITUTE. 1990. SAS user’s guide, vers. 6, 4th ed. SAS Institute, Cary, North Carolina, USA. SYKES, M. T., AND J. B. WILSON. 1988. An experimental investigation into the response of some New Zealand sand dune species to salt spray. Annals of Botany 62: 159–166. TOMINAGA, T., H. KOBAYASHI, AND K. UEKI. 1991. Clonal variation in salt tolerance of Imperata cylindrica (L.) Beauv. var. koenigii (Retz.) et Schinz. Journal of Japanese Grassland Science 37: 69–75. TYNDALL, R. W., A. H. TERAMURA, C. L. MULCHI, AND L. W. DOUGLASS. 1987. Effects of salt spray upon seedling survival, biomass, and distribution on Currituck Bank, North Carolina. Castanea 52: 77–86. VAN DER VALK, A. G. 1974. Environmental factors controlling the distribution of forbs on coastal foredunes in Cape Hatteras National Seashore. Canadian Journal of Botany 52: 1057–1073. WELLS, B. W., AND I. V. SHUNK. 1937. Seaside shrubs: wind forms vs. salt spray forms. Science 85: 499. WELLS, B. W., AND I. V. SHUNK. 1938. Salt spray: an important factor in coastal ecology. Bulletin of the Torrey Botanical Club 65: 485–492. WHITE, P. S. 1979. Pattern, process, and natural disturbances in vegetation. Botanical Review 45: 229–299. ZAREMBA, R. E., W. A. I. PATTERSON, AND J. L. RUDNICKY. 1983. Prescribed burning experiments in Nantucket heathlands. Cape Naturalist 12: 49– 55. ZOBEL, A., AND J. E. NIGHSWANDER. 1990. Accumulation of phenolic compounds in the necrotic areas of Austrian and red pine needles due to salt spray. Annals of Botany 66: 629–640.
