The expansion of black mangrove Avicennia germinans into historically smooth cordgrass Spartina alterniflora-dominated marshes with warming temperatures heralds the migration of the marsh-mangrove ecotone northward in the northern Gulf of Mexico. With this shift, A. germinans is expected to outcompete S. alterniflora where it is able to establish, offering another prevalent food source to first order consumers. In this study, we find A. germinans leaves to be preferable to chewing herbivores, but simultaneously, chewing herbivores cause more damage to S. alterniflora leaves. Despite higher nitrogen content, A. germinans leaves decomposed slower than S. alterniflora leaves, perhaps due to other leaf constituents or a different microbial community. Other studies have found the opposite in decomposition rates of the two species’ leaf tissue. This study provides insights into basic trophic process, herbivory and decomposition, at the initial stages of black mangrove colonization into S. alterniflora salt marsh.
Citation: Macy A, Sharma S, Sparks E, Goff J, Heck KL, Johnson MW, et al. (2019) Tropicalization of the barrier islands of the northern Gulf of Mexico: A comparison of herbivory and decomposition rates between smooth cordgrass (Spartina alterniflora) and black mangrove (Avicennia germinans). PLoS ONE 14(1): e0210144. https://doi.org/10.1371/journal.pone.0210144
Editor: Suzannah Rutherford, Fred Hutchinson Cancer Research Center, UNITED STATES
Received: October 23, 2017; Accepted: December 18, 2018; Published: January 7, 2019
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: This project was funded by the United States Fish and Wildlife Service, grant #F11AC01390 to JC and KLH. Support was also provided by the U.S. Department of Commerce’s National Oceanic and Atmospheric Administration under NOAA award NA14OAR4170098 and the Mississippi-Alabama Sea Grant Consortium. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: Just Cebrian as an editor for PLOS ONE and Patric Harper as an employee for the Fish & Wildlife Service in the Competing Interests section. We confirm that this does not alter our adherence to PLOS ONE policies on sharing data and materials.
As climates warm, tropical species are expanding their population ranges into historically subtropical environments , and lack of severe freeze events as the climate warms is expected to allow for more permanent expansion of mangrove populations in the northern Gulf of Mexico (nGOM)  and around the world [3, 4], rather than the alternating expansion/contraction cycles witnessed in the past. Coastal wetlands of the nGOM are largely dominated by herbaceous marsh comprised of smooth cordgrass Spartina alterniflora and/or black needlerush Juncus roemerianus, though some small regions of warmer climate within the nGOM contain stunted populations of the black mangrove Avicennia germinans. The Intergovernmental Panel on Climate Change predicts fewer severe cold events in the coming decades , which would facilitate continued poleward expansion of A. germinans and a gradual conversion of coastal wetlands along the nGOM from herbaceous marsh to woody mangrove swamps .
Along with temperature, other factors also limit the establishment and growth of A. germinans, such as wave action , propagule dispersal , propagule viability , microclimate , and hydrology [10, 11]. A. germinans often colonizes shorelines populated with S. alterniflora through herbaceous marsh entrapment of mangrove propagules , and this mangrove colonization comes at the expense of some salt marsh . This shift from salt marsh to mangrove is associated with substantial shifts in ecosystems structure and function [13, 14]. This mangrove succession over salt marsh may lead to shifts in energy transfer through consumer preference and detrital pathways. Despite the significance of energy transfer on ecosystems, it can be difficult to quantify in the field.
Studies of herbivory and decomposition rates on mangroves are limited to areas where mangroves are dominant, but these rates likely differ near the edges of their geographic extent due to colder climates and different faunal and microbial assemblages. Duarte & Cebrian  conducted an extensive review of carbon budgets in coastal systems, and concluded herbivory amounted to 7–11% of carbon transfer in mangroves and 24–38% carbon transfer in marshes, though more recent work suggests herbivory on mangroves has been underestimated [16, 17]. Decomposition rates vary with environmental conditions, including temperature , decomposer community , redox potential , and nutrient availability . No study to the authors’ knowledge has compared herbivory and decomposition rates of marsh and mangrove at the poleward extent of A. germinans.
In this study, we focus on a barrier island located at the most northern extent of A. germinans occurrence in the nGOM . Three A. germinans individuals were discovered in 2012 to have colonized Horn Island. Though the exact time of colonization is unknown, the trees were 0.72–1.08m tall in April 2012. We carried out a two-year study comparing leaf herbivory, leaf nutrient content, and decomposition rates between S. alterniflora and A. germinans at this early stage of A. germinans colonization. Our results demonstrate a dichotomy in the levels of leaf herbivory, based upon mass or areal removal analysis, and offer novel findings on differences in decomposition rates between A. germinans and S. alterniflora, illustrating important consideration of region-specificity in generalizations concerning expanding or migrating species ranges.
Horn Island is a barrier island off the coast of Mississippi (30°14'27" N, 88°40'43"W) that marks the northern limit of A. germinans in the Gulf of Mexico . The entire known mangrove population on the island, composed of three individuals at the time of this study (2012–2013), resided along the shoreline of a bayside inlet to a lagoon on the interior of Horn Island (Fig 1). The shoreline was dominated by S. alterniflora interspersed with patches of black needlerush J. romerianus and other salt marsh species.
Study region indicated on inset of Horn Island in top right. Each station contained a mixed and reference plot, 5m apart. Mixed (2012–2013) and reference (2013) plot sizes were scaled to the size of the A. germinans tree in the corresponding station (Stations 1 and 3: 1.0m2; Station 2: 0.25m2). Imagery of the study site and the Horn Island insert were obtained from USGS National Map Viewer (public domain).
For the first year of the study (2012), we established three stations, each comprised of a single “mixed” plot centered on one of the three surveyed A. germinans trees, thus containing the tree and underlying marsh. Each station was expanded in the second year (2013) to include a second, marsh-only (“reference”), plot, 5m away from its paired mixed plot. S. alterniflora was the dominant species of marsh plant and referred to as “S. alterniflora-mixed” in mixed plots and “S. alterniflora-ref” in reference plots. Mixed and reference plot sizes were scaled to the size of the A. germinans tree (0.25–1.0m2) in the corresponding station (Fig 1).
Floral communities, A. germinans and S. alterniflora morphometrics, and herbivory were measured bimonthly over the 2012 and 2013 growing seasons (June-October). Leaf decomposition rates using the conventional litterbag method [23, 24] were measured in 2012; in 2013, many of the litterbags were washed away, precluding analysis of 2013 data, and leaf carbon and nitrogen content were measured bimonthly in 2013.
With the permission of the United States Fish and Wildlife Service, thirty green A. germinans leaves were collected haphazardly from each tree on each sampling date. Thirty live shoots of S. alterniflora were collected haphazardly from just outside each plot around or adjacent to the tree. Photos were taken of all A. germinans leaves. S. alterniflora leaves were peeled off the shoot and their basal terminus defined by a distinct line where the leaf diverges from the stem. Height and basal width were measured for all S. alterniflora leaves. Photos were taken of the leaf sections with bite marks. A. germinans leaf area, and the area of the bite marks on A. germinans and S. alterniflora leaves were measured with ImageJ software.
To estimate leaf area for S. alterniflora leaves with tip present, we derived an allometric equation with fifteen S. alterniflora shoots collected from a marsh on Dauphin Island (30°15'3.56"N, 88° 4'39.91"W). Many S. alterniflora leaves were missing the tip, particularly those with many bite marks. This was not the case for A. germinans leaves since grazed leaves maintained their tip and oval shape despite having grazing marks. Missing tips could have endured substantial grazing; thus, disregard of these tips could lead to severe grazing underestimates . We used a two-step approach to derive a possible range of herbivory intensity for leaves where the tip was missing.
Four possible situations were encountered for S. alterniflora leaves collected (Fig 2). To estimate area of missing tips, we took photos of leaves with tip missing and extended both edges of each leaf using ImageJ  until they intersected. We measured the taping angle (mean value ± standard error: 89.5° ± 0.2°). Then, assuming an isosceles triangle shaped-leaf, we calculated leaf area as tan(Θ)*(0.5b)2, where θ is the mean tapering angle and b is the basal width. From this we derived the width of the top edge at which the leaf was cut off and the tip missing (i.e., width of the isosceles triangle at the height of the leaf), using the relationship: top width = 2Δh/tan(Θ), where Δh is height difference between the reconstructed isosceles triangle and leaf height. In turn, the width of the top leaf edge was taken as the basal width for the isosceles triangle representing the missing tip, and the area of the missing tip was calculated from this width measurement and the mean tapering angle value. These estimates were included in the S. alterniflora leaf area allometric regression equation, Leaf Areacalculated = 0.3478 * Leaf Areameasured + 0.1383, (r2 = 0.91), to obtain final values of missing tip areas. The missing tip reconstruction approach was tested by severing 20 complete leaves at haphazardly-chosen cross sections along each leaf blade to determine an adjustment factor for our calculated missing tip areas: Missing Tipcalculated = 2.6498 * Missing Tipmeasured + 0.2265, (r2 = 0.92).
Four possible situations encountered for S. alterniflora leaves: a) leaf tip intact, no grazing (76.1% of leaves), b) leaf tip intact, grazed (6.4% of leaves), c) leaf tip missing, no grazing on remaining portion (3.2% of leaves), d) leaf tip missing, grazed on remaining portion (14.2% of leaves).
Upon estimation of the area of the missing tip, the second step involved derivation of herbivory values assuming that 25%, 50%, or 75% of the missing tip had been consumed to illustrate a wide range of possible consumption scenarios. These calculations were done separately for each missing tip and the area of the bite marks added to the grazing estimation for the missing tip. Herbivory estimates were derived as percent leaf area consumed (cm2grazed cm-2leaf), absolute leaf area consumed (cm2grazed leaf-1), and absolute leaf mass consumed (grams dry weight leaf-1).
To assess leaf mass consumption, leaf area (cm2) to dry weight (g) conversion ratios were derived from additional subsamples: S. alterniflora (r2 = 0.88): Oven − Dry Weight = 0.0193 * Leaf Area − 0.1657; A. germinans: (r2 = 0.80): Oven-Dry Weight = 0.0197 * Leaf Area + 0.0252.
Cumulative leaf damage from grazers is a common proxy for herbivory [27, 28], but when compared between species, leaf lifespans should be taken into consideration. Leaves of A. germinans and S. alterniflora in the studied area have similar lifespans (100–150 days, [29, 30]); thus, our estimates should represent similar time scales for cumulative leaf damage.
Leaf nutrient content.
Between 10–20 of the green A. germians or S. alterniflora leaves used to measure herbivory were oven dried (60°C to a constant weight) then ground into powder using a Thomas-Wiley plant tissue grinder and analyzed for percent carbon (%C) and nitrogen (%N) by mass using a CNS analyzer .
Leaf litter decomposition.
The conventional litterbag method [23, 24] was used to estimate leaf decomposition rates. Nylon mesh bags (1.0 mm mesh) containing a known mass (air-dried) of green leaf detritus of A. germinans or S. alterniflora were fastened to rebar at the sediment surface within the plot. Leaves of both species begin to senesce while still attached to the plant, such that plucked senescent leaves would have been at different stages of decomposition. To normalize the stage of decomposition, we used green leaves for comparison. For measuring litterbag mass remaining, a conversion between oven- and air-dried material was necessary: A. germinans (r2 = 0.98): Air-dry weight = 1.8143 * Oven-Dry weight + 0.0477; S. alterniflora (r2 = 0.92): Air-dry weight = 1.3959 * Oven-Dry weight + 0.0186.
The following equation was fit for each vegetation type in each of the three stations: where DMretrieval and DMdeployment correspond to the detritus mass (air-dry weight) remaining at retrieval and deployed at the beginning of the incubations, respectively; C is a constant, k is the decomposition rate, in days-1; and Δt is the time elapsed from deployment to retrieval, in days. The fits were done with least-square regression in SigmaPlot 12.3.
We used the following model for the analysis of leaf nutrient content, herbivory, and decomposition: where Station is a random blocking factor corresponding to each of the three locations where the plots were situated, Subject is a random factor, nested within Station, accounting for repeated measurements on individual organisms, Plant Type is a fixed factor corresponding to A. germinans, S. alterniflora-mixed or S. alterniflora-ref, and Time is a random, within-subject factor corresponding to collection date. For nutrient content, the response variable was percent nitrogen; for herbivory, the response variable was amount grazed (percent (%), absolute, or mass); and for decomposition, the response variable corresponded to ln(DMretrievalDMdeployment). All values obtained from a given plot on each sampling date were averaged into one true replicate. We used R version 3.2.4  with package “lme4”  for analysis. A likelihood ratio test was used to determine significance of each factor: (1) two models were constructed for each comparison, the full model and a model lacking the factor of interest; (2) a likelihood ratio test using the “anova ()” function provided a comparison between the models; and (3) any differences between the models are attributed to the factor of interest . When comparing models for a fixed factor, maximum likelihood estimators were used in the full and partial models; for random factors, restricted maximum likelihood estimators were used for both models being compared [34, 35]. Post-hoc comparisons were computed for specific factor levels following the same approach. Recent, robust statistical analysis developments have allowed for inclusion of several factors to be accounted for smaller sample sizes (e.g., n = 3), though the statistical power may still be reduced [33–35].
Herbivores removed more leaf area of S. alterniflora but consumed more leaf mass of A. germinans, per leaf. Fig 3 shows consumption (S. alterniflora: portion of overall leaf removal; A. germinans: equal to leaf removal), but overall leaf removal would account for 100% of missing tips (Yearly Leaf Removal Averages; Percent Area: S. alterniflora-mixed: 3.66%, S. alterniflora-ref: 2.22%, A. germinans: 0.75% and Absolute Leaf Area: S. alterniflora-mixed: 0.60 cm2 leaf-1, S. alterniflora-ref: 0.68 cm2 leaf-1, A. germinans: 0.06cm2 leaf-1). Significantly more area was removed from each S. alterniflora leaf than each A. germinans leaf. On a mass basis, even if 75% of missing S. alterniflora tips were consumed, more mass per leaf would be entering herbivores via A. germinans than via S. alterniflora (Fig 3 and Table 1). Post-hoc analysis did not reveal any significant differences in herbivory between S. alterniflora-mixed and S. alterniflora-ref. A. germinans leaves were more frequently damaged (63.1% of leaves intact; 36.9% of leaves grazed) than S. alterniflora leaves (See Fig 2; 76.1% intact, no grazing; 6.4% intact, grazed; 3.2% not intact, no grazing; 14.2% not intact, grazed).
Grazing on leaf tissue was compared on the bases of percentage of total leaf area (left column plots), absolute leaf area (center column), and absolute leaf mass (right column). Note y-axes’ scales vary in left and center columns. Values between top, middle, and bottom rows vary only for S. alterniflora, with different assumptions of missing tip consumption (top: 25%, middle: 50%, bottom: 75%).
Leaf nutrient content
Leaf nitrogen content was higher in A. germinans leaves than in S. alterniflora leaves (p < 0.05, Fig 4 and Table 1), with %N in S. alterniflora leaves not varying significantly between S. alteriflora-mixed and S. alterniflora-ref. These differences were consistent over time (Plant Type x Time: p > 0.05, Table 1).
Leaf decomposition was faster for S. alterniflora than for A. germinans leaves (Plant Type x Time: p < 0.001, Fig 5 and Table 1). Similar leaf detritus weights for the two plant types remained in the bags over the first collection times, but 328 days after deployment only ca. 2% of the initial weight remained for S. alterniflora as opposed to ca. 15% for A. germinans leaf detritus. We used the single exponential decay model because other models (such as the linear or double exponential) did not improve the quality of the fit (S1 and S2 Figs).
Y values, ln(DMt/DMo), are the natural logarithms of the proportion of detrital mass remaining at time, t. Duplicate bags of each plant type were collected 6, 12, 25, 68, 113, and 328 days after deployment. Upon retrieval, the samples were rinsed off with fresh water, and the contents within the bag carefully removed and sieved through a 0.5mm mesh. Samples were oven-dried (60°C to constant mass) and converted to an air-dried weight using regressions obtained with additional samples. Least square regression for single exponential models (p < 0.05, A. germinans: r2 = 0.90; S. alterniflora r2 = 0.95). A. germinans leaf detritus decomposed slower than S. alterniflora leaf detritus (p < 0.05, Table 1).
There was a greater mass consumed from A. germinans leaves than S. alterniflora leaves, even when assuming 75% of each missing S. alterniflora tip was consumed (A. germinans: 0.018 g leaf-1, S. alterniflora-mixed: 0.007 g leaf-1, S. alterniflora-ref: 0.004 g leaf-1, Fig 3 and Table 1), and these differences were consistent over time (Plant Type x Time: p > 0.35). Based on previous degrees of herbivory in other systems , 0% and 75% consumptions are unlikely; we chose to include analysis of a conservative 75% scenario to highlight the consumption relative to A. germinans. In addition to the greater mass consumption per leaf, mangrove leaves contained 48–63% greater nitrogen concentrations than S. alterniflora leaves (A. germinans: 1.86%, S. alterniflora-mixed: 1.25%, S. alterniflora-ref: 1.14%, Fig 4 and Table 1), suggesting herbivores may be targeting A. germinans for their higher nutritional value. Typical leaf nitrogen content ranges from 1–4% , indicating S. alterniflora is relatively nitrogen poor. Onuf et al.  fertilized red mangroves to the extent of increasing leaf nitrogen concentrations by 33%, resulting in a 400% increase in herbivory. The leaf nitrogen concentration discrepancy of 48–63% observed in our study may influenced grazer preference of leaves.
Other factors such as lignin concentration, tannins, and silica content [38–40] can affect palatability. Relative to other mangrove genuses, A. germinans leaves contain fewer tannins (0.8% DW, ) but still more than S. alterniflora (0.4% DW, ). The most likely dominant grazers of A. germinans in our system are grasshoppers, based upon identification of nearby marsh grazing populations . The most likely dominant grazer of S. alterniflora is the marsh periwinkle Littoraria irrorate , who incidentally consumes S. alterniflora as it consumes fungus that infects S. alterniflora . The results demonstrate A. germinans may be a preferable option to certain grazers in the nGOM. Despite this apparent grazer preference of live leaf tissue, the morphology differences between A. germinans and S. alterniflora leaves lend to a greater impact on S. alterniflora, the less consumed plant.
Primary production is a function of leaf area, and leafy plants have exhibited several adaptations to exposure of leaf surface area to sunlight [46, 47]. Grazing incidentally decreases leaf area, but to sustain the same level of secondary production, more leaf area of one species of plant may need to be consumed (or removed through sloppy feeding, wherein material is removed from the plant but not consumed by the grazer) than the leaf area of another plant species. There is a greater mass per unit area in A. germinans leaves than in S. alterniflora leaves, but the overall effect of herbivory on either plant species is small. If herbivores are 100% responsible for the missing areas, they are still only removing <5% of leaf area in cordgrass and <1% of leaf area in mangroves, but this does not account for entire leaves that are removed from the plants. No published studies quantifying herbivory on A. germinans along the nGOM could be found, but in warmer climates, higher rates of herbivory were observed on A. germinans (0.83–4.5% of leaf area in Guadeloupe ; 8–36% in Belize ). Alongi  asserts that usually <10% of mangrove defoliation is attributed to insect herbivory. As such, most leaf biomass ends up as detrital matter.
Carbon, nitrogen, and energy also make their way into the food chain via detrital pathways: detritivores and decomposers extract a portion of this sustenance before it is either exported out of the system or sequestered in the wetland peat. Higher leaf nutrient content may result in faster decomposition [49–51], yet we found slower decomposition of the nitrogen-rich mangrove leaves than the nitrogen-poor cordgrass leaves. This counter-intuitive result could be influenced by potentially stronger factors, such as higher lignin concentrations or waxier cuticle [50, 52], tannins, alternative nutrient limitation (i.e., phosphorus), or different microbial communities . While marine systems are often nitrogen-limited, Johnson et al.  documented phosphorus limitation in seagrass beds nearby, suggesting detritivores may have preferred phosphorus-rich material. Phosphorus levels in leaves may have varied significantly but were not measured. Studies in this region indicate leaf phosphorus concentrations are similar between A. germinans (0.11–0.15%, ) and S. alterniflora (0.08–0.15%, ). Perry and Mendelssohn  found A. germinans leaves decomposed faster than S. alterniflora leaves in an area with a long-established population of A. germinans, though microbial communities can vary substantially with relative mangrove influence . Microclimate conditions between litterbags of each study were unlikely to have a significantly different effect, despite the variable dimensions and mesh sizes .
In this study, we found contrasting perspectives by investigation of first order consumption. Between A. germinans and S. alterniflora leaves, chewing herbivores consumed more A. germinans leaf material while causing less damage to the leaf, perhaps enticed by the higher nitrogen content. Chewing herbivores removed more leaf area from S. alterniflora but ingested less plant material. S. alterniflora leaves’ greater area to mass ratio likely also contributed to its more rapid decomposition. The introduction of new producers like A. germinans within a system presents herbivores (and detritivores) another food choice. The magnitude of subsequent shifts in trophic processes will align with the extent of the new producer’s dominance within the native system; however, those magnitudes will also align with the relative preference of first order consumers to a producer growing in that location, which may differ from that producer’s growth in its native environment. By understanding dichotomies of herbivory and decomposition at range extremes, we can better identify and explain the impending changes to shifts associated with economically and ecologically valuable systems, like S. alterniflora-dominated salt marsh ecosystems of the nGOM. We look forward to future research identifying changes in trophic processes with mangrove expansion and/or changes in the marsh communities alongside other climate-motivated species range expansions.
S1 Fig. Linear decomposition model.
Y-axis: dry mass at time t divided by initial dry mass. A. germians: r2 = 0.89; S. alterniflora: r2 = 0.95.
We thank D. Byron, T. Ferrero, R. Gamble, J. Hulsey, A. McDonald, B. Tuttle, and C. Wessel for their help in the field and lab. We thank J. Alberti for his statistical software advice. This publication (MASGP-18-045) was partially supported by the U.S. Department of Commerce’s National Oceanic and Atmospheric Administration under NOAA Award NA14OAR4170098 and the Mississippi-Alabama Sea Grant Consortium. The views expressed herein do not necessarily reflect the views of any of these organizations.
- 1. Walther G-R, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, et al. Ecological responses to recent climate change. Nature. 2002;416(6879):389–95. pmid:11919621
- 2. Osland MJ, Enwright N, Day RH, Doyle TW. Winter climate change and coastal wetland foundation species: salt marshes vs. mangrove forests in the southeastern United States. Glob Chang Biol. 2013 May;19(5):1482–94. pmid:23504931
- 3. Osland MJ, Enwright NM, Day RH, Gabler CA, Stagg CL, Grace JB. Beyond just sea-level rise: considering macroclimatic drivers within coastal wetland vulnerability assessments to climate change. Glob Chang Biol. 2016;22(1):1–11. pmid:26342186
- 4. Osland MJ, Day RH, Hall CT, Brumfield MD, Dugas JL, Jones WR. Mangrove expansion and contraction at a poleward range limit: climate extremes and land‐ocean temperature gradients. Ecology. 2017 Jan 1;98(1):125–37. pmid:27935029
- 5. IPCC 2013. IPCC 2013 Summary. In: Stocker TF , Qin D., Plattner G.-K., Tignor M., S.K., Boschung J., et al., editors. Climate Change 2013: The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press; 2013.
- 6. Zhang Y, Huang G, Wang W, Chen L, Lin G. Interactions between mangroves and exotic Spartina in an anthropogenically disturbed estuary in southern China. Ecology. 2012 Mar;93(3):588–97. pmid:22624213
- 7. Stevens PW, Fox SL, Montague CL. The interplay between mangroves and saltmarshes at the transition between temperate and subtropical climate in Florida. Wetlands Ecol Manage. 2006 Oct;14(5):435–44.
- 8. Alleman LK, Hester MW. Reproductive ecology of black mangrove (Avicennia germinans) along the Louisiana Coast: propagule production cycles, dispersal limitations, and establishment elevations. Estuaries Coasts. 2011;34(5):1068–77.
- 9. Devaney JL, Lehmann M, Feller IC, Parker JD. Mangrove microclimates alter seedling dynamics at the range edge. Ecology. 2017 Oct;98(10):2513–20. pmid:28779524
- 10. McKee KL. Seedling recruitment patterns in a Belizean mangrove forest: effects of establishment ability and physico-chemical factors. Oecologia. 1995;101(4):448–60. pmid:28306959
- 11. Lewis RR. Ecological engineering for successful management and restoration of mangrove forests. Ecol Eng. 2005 Apr;24(4):403–18.
- 12. Armitage AR, Highfield WE, Brody SD, Louchouarn P. The contribution of mangrove expansion to salt marsh loss on the Texas Gulf Coast. PLoS One. 2015 May 6;10(5):e0125404. pmid:25946132
- 13. Osland MJ, Spivak AC, Nestlerode J a., Lessmann JM, Almario AE, Heitmuller PT, et al. Ecosystem Development After Mangrove Wetland Creation: Plant–Soil Change Across a 20-Year Chronosequence. Ecosystems. 2012 May 19;15(5):848–66.
- 14. Guo H, Weaver C, Charles SP, Whitt A, Dastidar S, D’Odorico P, et al. Coastal regime shifts: rapid responses of coastal wetlands to changes in mangrove cover. Ecology. 2017 Mar;98(3):762–72. pmid:27984665
- 15. Duarte CM, Cebrian J. The fate of marine autotrophic production. Limnol Oceanogr. 1996;41(8):1758–66.
- 16. Burrows D. The role of insect leaf herbivory on the mangroves Avicennia marina and Rhizophora stylosa. 2003;238.
- 17. Cannicci S, Burrows D, Fratini S, Smith TJ, Offenberg J, Dahdouh-guebas F. Faunal impact on vegetation structure and ecosystem function in mangrove forests: a review. Aquat Bot. 2008;89:186–200.
- 18. Davidson EA, Janssens IA. Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature. 2006 Mar 9;440(7081):165–73. pmid:16525463
- 19. Mcguire KL, Treseder KT. Microbial communities and their relevance for ecosystem models: decomposition as a case study. Soil Biol Biochem. 2010;42:529–35.
- 20. Reddy KR, Patrick WH. Effect of alternate aerobic and anaerobic conditions on redox potential, organic matter decomposition and nitrogen loss in a flooded soil. Soil Biol Biochem. 1975 Mar 1;7(2):87–94.
- 21. Rejmánková E, Sirová D. Wetland macrophyte decomposition under different nutrient conditions: Relationships between decomposition rate, enzyme activities and microbial biomass. Soil Biol Biochem. 2007 Feb 1;39(2):526–38.
- 22. Scheffel W, Heck K, Byron D. Range expansion of black mangrove (Avicennia germinans) to the Mississippi Barrier Islands. Gulf Mex Sci. 2013;2013(1–2):79–82.
- 23. Hackney CT, Cruz AADL. In situ decomposition of roots and rhizomes of two tidal marsh plants. Ecology. 1980;61(2):226–31.
- 24. Hunter A, Cebrian J, Stutes JP, Patterson D, Christiaen B, Lafabrie C, et al. Magnitude and trophic fate of black needlerush (Juncus roemerianus) productivity: Does nutrient addition matter? Wetlands. 2015;35(2):401–17.
- 25. Prado P, Tomas F, Alcoverro T, Romero J. Extensive direct measurements of Posidonia oceanica defoliation confirm the importance of herbivory in temperate seagrass meadows. Mar Ecol Prog Ser. 2007;340:63–71.
- 26. Schneider Caroline A., Rasband Wayne S., and Eliceiri Kevin W. 2012. “NIH Image to ImageJ: 25 Years of Image Analysis.” Nature Methods 9 (7): 671–75. pmid:22930834
- 27. Farnsworth EJ, Ellison AM. Patterns of herbivory in Belizean mangrove swamps. Biotropica. 1991;23(4b):555.
- 28. Saur E, Imbert D, Etienne J, Mian D. Insect herbivory on mangrove leaves in Guadeloupe: effects on biomass and mineral content. Hydrobiologia. 1999;413:89–93.
- 29. Gallagher JL. Seasonal patterns in recoverable underground reserves in Spartina alterniflora Loisel. Am J Bot. 1983;70(2):212–5.
- 30. Suárez N, Medina E. Salinity effect on plant growth and leaf demography of the mangrove, Avicennia germinans L. Trees. 2005 Aug 25;19(6):722–8.
- 31. Pennock JR, Cowan JLW. C&N analytical instrumentation methods manual. 2001.
- 32. R Core Team (2016). R: A language and environment for statistical computing. R Foundation for Statistical Computing [Internet]. R Foundation for Statistical Computing; Available from: https://www.R-project.org
- 33. Bates D, Mächler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, Articles. 2015;67(1):1–48.
- 34. Winter B. Linear models and linear mixed effects models in R with linguistic applications. 2013; Available from: http://arxiv.org/abs/1308.5499.pdf
- 35. Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. Statistics for Biology and Health. Springer Science+Business Media, LLC; 2009.
- 36. Zhang J, Wang X-J, Wang J-P, Wang W-X. Carbon and nitrogen contents in typical plants and soil profiles in Yanqi Basin of Northwest China. J Integr Agric. 2014 Mar 1;13(3):648–56.
- 37. Onuf CP, Teal JM, Valiela I. Interactions of nutrients, plant growth and herbivory in a mangrove ecosystem. Ecology. 1977;58(3):514–26.
- 38. Herms DA, Mattson WJ. The dilemma of plants: To grow or defend. Q Rev Biol. 1992;67(3):283–335.
- 39. Massey FP, Hartley SE. Physical defences wear you down: progressive and irreversible impacts of silica on insect herbivores. J Anim Ecol. 2009;78:281–91. pmid:18771503
- 40. Barbehenn RV, Constabel CP. Tannins in plant–herbivore interactions. Phytochemistry. 2011;72:1551–65. pmid:21354580
- 41. Alongi DM. Oecologia on intertidal meiobenthos in tropical estuaries. Oecologia. 1987;71:537–40. pmid:28312223
- 42. Filip Z, Alberts JJ. Humic substances isolated from Spartina alterniflora (Loisel.) following long-term decomposition in sea water. Sci Total Environ. 1989;83:273–85.
- 43. Sparks EL, Cebrian J. Effects of fertilization on grasshopper grazing of Northern Gulf of Mexico salt marshes. Estuaries Coasts. 2014 Jul 22;38(3):988–99.
- 44. Silliman BR, Koppel JVD, Bertness MD, Stanton LE, Mendelssohn IA. Drought, snails, and large-scale die-off of southern U.S. salt marshes. Science. 2005;310(5755):1803–6. pmid:16357258
- 45. Silliman BR, Bertness MD. A trophic cascade regulates salt marsh primary production. Proc Natl Acad Sci U S A. 2002 Aug 6;99(16):10500–5. pmid:12149475
- 46. Kikuzawa K. Leaf phenology as an optimal strategy for carbon gain in plants. Can J Bot. 1995;163(August 1993):158–63.
- 47. Enríquez S, Pantoja-Reyes NI. Form-function analysis of the effect of canopy morphology on leaf self-shading in the seagrass Thalassia testudinum. Oecologia. 2005 Sep;145(2):235–43. pmid:15942763
- 48. Alongi DM. Ecosystem Dynamics. In: The energetics of mangrove forests. 2009. p. 129–62.
- 49. Enriquez S, Duarte C, Sand-Jensen K. Patterns in decomposition rates among photosynthetic organisms: the importance of detritus C:N:P content. Oecologia. 1993;94(4):457–71. pmid:28313985
- 50. Hobbie SE. Interactions between litter lignin and soil nitrogen availability during leaf litter decomposition in a Hawaiian montane forest. Ecosystems. 2000;3(5):484–94.
- 51. Cebrian J, Lartigue J. Patterns of herbivory and decomposition in aquatic and terrestrial ecosystems. Ecol Monogr. 2004;74(2):237–59.
- 52. Millar CS. Decomposition of coniferous leaf litter. Biology of plant litter decomposition. 1974;1:105–28.
- 53. Johnson MW, Heck KL, Fourqurean JW. Nutrient content of seagrasses and epiphytes in the northern Gulf of Mexico: Evidence of phosphorus and nitrogen limitation. Aquat Bot. 2006;85:103–11.
- 54. Méndez-Alonzo R, López-Portillo J, Rivera-Monroy VH. Latitudinal variation in leaf and tree traits of the mangrove Avicennia germinans (Avicenniaceae) in the central region of the Gulf of Mexico. Biotropica. 2008;40(4):449–56.
- 55. Patrick WH, Delaune RD. Nitrogen and phosphorus utilization by Spartina alterniflora in a salt marsh in Barataria Bay, Louisiana. Estuarine Coastal Mar Sci. 1976 Jan;4(1):59–64.
- 56. Perry CL, Mendelssohn I a. Ecosystem effects of expanding populations of Avicennia germinans in a Louisiana salt marsh. Wetlands. 2009 Mar;29(1):396–406.
- 57. Barreto CR, Morrissey EM, Wykoff DD, Chapman SK. Co-occurring Mangroves and Salt Marshes Differ in Microbial Community Composition. Wetlands. 2018 Jun 1;38(3):497–508.
- 58. Bokhorst S, Wardle DA. Microclimate within litter bags of different mesh size: Implications for the “arthropod effect” on litter decomposition. Soil Biol Biochem. 2013;58:147–52.