Accepted Manuscript
Are seaward pneumatophore fringes transitional between mangrove and lower-shore system compartments?
R.S.K. Barnes
PII: S0141-1136(16)30310-5 DOI: 10.1016/j.marenvres.2017.01.008 Reference: MERE 4269
To appear in: Marine Environmental Research
Received Date: 4 December 2016 Revised Date: 24 January 2017 Accepted Date: 30 January 2017
Please cite this article as: Barnes, R.S.K., Are seaward pneumatophore fringes transitional between mangrove and lower-shore system compartments?, Marine Environmental Research (2017), doi: 10.1016/j.marenvres.2017.01.008.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 ACCEPTED MANUSCRIPT 1 Are seaward pneumatophore fringes transitional between mangrove and
2 lower-shore system compartments?
3
4
5
6
7
8 R.S.K. Barnes
9
10
11 School of Biological Sciences & Centre for Marine Science, University of Queensland,
12 Brisbane 4072, Queensland, Australia MANUSCRIPT 13 Biodiversity Program, Queensland Museum, Brisbane 4101, Queensland, Australia 14
15 E-mail address: [email protected]
16
ACCEPTED 2 ACCEPTED MANUSCRIPT 17 ABSTRACT
18 Work in temperate New Zealand has concluded that seaward fringes of Avicennia
19 pneumatophores (P) form an 'important ecological transitional environment' between
20 seagrass (Z) and mangrove (M), supporting intermediate macrofaunal numbers and
21 biodiversity (Alfaro, 2006). This study re-examined that hypothesis in subtropical Moreton
22 Bay, Queensland, and investigated its dependence on the nature of the lower-shore habitat;
23 i.e. whether seagrass or sandflat (S). Adjacent macrobenthic assemblages across 45 m deep
24 Z:P:M and S:P:M interfaces were compared uni- and multivariately and via various
25 assemblage metrics. Here, system compartment P was not intermediate. In Z:P:M
26 interfaces it was essentially an extension of the lower-shore assemblage and supported peak
27 biodiversity. In contrast, P in S:P:M interfaces was partly an extension of the upper-shore
28 assemblage with unchanged biodiversity but minimum abundance. Several species 29 spanned the whole interface zone, and assemblageMANUSCRIPT structure and several metrics remained 30 unchanged across it. These findings are discussed in relation to ecotones in general. Like
31 other such zones the characteristics of pneumatophore-fringe ecotones are context
32 dependent.
33
34 Keywords: biodiversity, ecotone, macrobenthos, mangrove, Moreton Bay,
35 pneumatophores, sandflat, seagrass ACCEPTED 36
37
38 3 ACCEPTED MANUSCRIPT 39 1. Introduction
40 Undisturbed soft-sediment shores in sheltered tropical and subtropical bays generally
41 take the form of a lower-shore band of seagrass extending up from the sublittoral and a
42 higher-shore belt of mangrove that merges into fully-terrestrial coastal vegetation inland
43 (Davies, 1972; Woodroffe, 2002; Gillis et al., 2014). Lower-shore and higher-shore systems
44 then meet in the vicinity of mean sea level (MSL). In some areas, patches of sediment
45 lacking macrophytes also occur within the seagrass beds or between them and the mangrove
46 fringe (McGlathery et al., 2013; Moffett et al., 2015).
47 Inter-connectivity, complementarity, and passage of materials and organisms between
48 these system compartments have received considerable recent attention (Bouillon et al.,
49 2007; Unsworth et al., 2009; Lee et al., 2014), much of it motivated by the perceived general
50 importance of the ecological and economic roles played by coastal interface zones (Levin et 51 al., 2001; Waycott et al., 2011; Boström et al., MANUSCRIPT 2011). Patterns of change in biodiversity and 52 in relative abundance of faunal assemblages across mosaics of mangrove, seagrass and
53 intertidal-flat habitats, however, constitute an important gap in current knowledge (Waycott
54 et al., 2011), although the interface zones between seagrass and adjacent unvegetated-
55 sediment systems have recently been investigated in some detail (Ollivier et al., 2015;
56 Barnes & Hamylton, 2016). Seagrass-associated studies have been conducted within the
57 context of its recent worldwide decline and fragmentation (Waycott et al., 2009; Fourqurean 58 et al., 2012) ACCEPTEDand resultant concern that the alternative denuded habitat state may support 59 lesser animal abundance and biodiversity (Pillay et al., 2010; Barnes & Barnes, 2012)
60 thereby impacting significantly on the ecosystem services otherwise provided by seagrass
61 beds (Costanza et al., 1997; Barbier et al., 2011). 4 ACCEPTED MANUSCRIPT 62 Mangrove is also a globally threatened habitat (Alongi, 2002; Duke et al., 2007) and
63 it also contributes significantly to the ecosystem services of coastal macrophytes (Barbier et
64 al., 2011); mangrove-associated crustaceans and molluscs alone, for example, are worth >4
65 billion US$ yr -1 (Ellison, 2008). But much less attention has been paid to the nature of
66 changes across the interface between mangroves and adjacent regions of the lower shore, be
67 these regions seagrass or unvegetated sediment. Indeed the role of the mangrove
68 compartment as a whole is still relatively poorly understood (Lee, 2008; Nagelkerken et al.,
69 2008; Lee et al., 2014). Intertidal interfaces with mangroves are often represented by an
70 extensive seaward 'fringe' of Avicennia or Sonneratia pneumatophores that can extend >25
71 m metres away from the parent trees (Macnae, 1978; pers.obs.) (Fig. 1A). In a much cited
72 study, Alfaro (2006) specifically investigated this type of boundary between seagrass and
73 mangrove in the Matapouri Estuary in New Zealand, finding that the seagrass beds were the
74 most biodiverse, mangrove was the least and the pneumatophore-fringe interface was 75 intermediate, and concluding that this boundary MANUSCRIPT formed an 'important ecological transitional 76 environment between seagrasses and mangrove' (p. 107). She suggested that the sandier
77 sediment within the pneumatophore interface (relative to the muddier mangrove), and the
78 entrapment of drift algae by the aerial roots, might be amongst the factors promoting the
79 greater faunal diversity and density observed there than within the mangrove proper.
80 Greater light intensity in the relatively open marginal habitat could also result in more
81 microphytobenthic food for consumers (Morrisey et al., 2010), as could the characteristically 82 higher rates ACCEPTEDof decomposition (Imgraben & Dittman, 2008). Wells (1986) also studied a 83 three-compartment mangrove-boundary system in the tropical Bay of Rest, Western
84 Australia although here the lower-shore habitat was mudflat rather than seagrass. He
85 sampled the larger molluscs present in transects extending 50 m on either side of the
86 mangrove/bare-sediment interface and found that the boundary pneumatophore fringe 5 ACCEPTED MANUSCRIPT 87 supported higher densities than either of the adjacent habitats, not intermediate levels,
88 although it did contain an intermediate number of species, of which approximately equal
89 numbers were shared with each of its two larger neighbouring habitat blocks. Several other
90 studies have compared the biodiversity of local expanses of seagrass or intertidal flat with
91 that of neighbouring mangrove (e.g. Wells, 1983; Sheridan, 1997; Dittmann, 2001;
92 Bloomfield & Gillanders, 2005; Armenteros et al., 2007; and Dalia et al., 2014) but have not
93 examined the nature of their interfaces.
94 A transitional state, however, is but one of several theoretically possible outcomes
95 within boundary ecotones (Turner, 2005; Matias et al., 2013; Langhans and Tockner, 2014)
96 and, as indeed indicated by Alfaro (2006), further comparative studies are necessary in other
97 areas dominated by Avicennia before its status can be assured, not least because her study
98 area being temperate may differ in important respects from equivalent habitats in lower 99 latitudes (Alfaro, 2006; Morrisey et al., 2010).MANUSCRIPT It was therefore the purpose of the present 100 work to investigate the biodiversity, abundance and nature of faunal assemblages and their
101 metrics across seaward pneumatophore-fringe interfaces with replacing lower-shore habitats
102 in another (and non-temperate) geographical area. The objects were to test the wider
103 applicability of the transitional-role hypothesis and to determine whether the outcome might
104 be dependent on characteristics of the lower-shore habitat in question, i.e. whether the
105 sediment was vegetated or not. It should be stressed that the present study did not seek to
106 compare characteristic local mangrove faunal assemblages with those further downshore but 107 was solely concernedACCEPTED with their interfaces: seagrass assemblages near their landward 108 margin are known to be relatively species-poor and not representative of those across the
109 intertidal zone in general (Barnes and Ellwood, 2011; Barnes and Barnes, 2012) and
110 likewise those in the seaward margins of mangroves may not typify that habitat either
111 (Morgan and Hailstone, 1986; Metcalfe and Glasby, 2008). 6 ACCEPTED MANUSCRIPT 112
113 2. Materials and Methods
114 2.1 Study area, sample collection and processing
115 Macrofaunal sampling was conducted over a period of 10 weeks during the austral
116 spring (late September – end November) along the Rainbow Channel coast of North
117 Stradbroke Island (aka Minjerribah) within the relatively pristine Eastern Banks region of
118 the oligohaline, mesotidal, subtropical Moreton Bay Marine Park, Queensland (Dennison
119 and Abal, 1999; Gibbes et al., 2014). This is the same area that supported earlier studies of
120 faunal differences between adjacent seagrass and unvegetated sand (Barnes & Barnes, 2012)
121 and of the nature of faunal transitions across the interface between the two (Barnes &
122 Hamylton, 2013, 2016). This coast also provides ecotones between a landward belt of 123 Avicennia marina var australasica and RhizophoraMANUSCRIPT stylosa mangrove and two alternative 124 lower-shore habitats: (a) a stretch of unvegetated, Trypaea - and Mictyris - structured, fine- to
125 medium-grained, tidal-delta quartz sand [i.e. a bare muddy sand (S) – seaward
126 pneumatophore fringe (P) – mangrove (M) transition]; or (b) more rarely a bed of intertidal
127 seagrass dominated by Zostera (Zosterella ) capricorni extending over the lower shore and
128 up into the pneumatophore fringe (Fig. 1) [i.e. a seagrass (Z) – seagrass + pneumatophore
129 fringe (ZP) – mangrove (M) transition]. This mangrove belt has been stable over the last
130 50+ years (Accad et al., 2016), and extends along (and beyond) the entire 35 km long 131 western coastlineACCEPTED of the island except in the vicinity of the Dunwich township.
132 In order to compare strictly equivalent habitats across the c. MSL interfaces, the study
133 was confined to benthic epi- and infaunal assemblages of the surface layers of the
134 substratum, so that although projecting pneumatophores and seagrass plants were included, 7 ACCEPTED MANUSCRIPT 135 the extensive hard substrata provided within the mangrove by tree trunks were not. In all
136 cases, data were collected from two-dimensional grids of samples set across and along the
137 seaward pneumatophore fringe and its interfacing habitats, each grid being a 9 x 8 lattice of
138 core samples, each type of transition being sampled at two different sites, the S-P-M
139 transition at Capembah (27º28'03"S,153º25'25"E) and Yerrol (27º29'08"S,153º24'28"E) and
140 both Z-ZP-M transitions in Deanbilla Bay but at sites 125 m apart separated by an area of
141 unvegetated sand (27º30'31"S,153º24'39"E and 27º30'34"S,153º24'42"E). The Capembah
142 site was located well away from the discharge of freshwater from Capembah Stream (Arnold
143 et al., 2014). Rows of 8 core samples, each core of 8.3 cm diameter (0.0054 m 2) with
144 intervals between them of 2 m, along lines parallel to the interface formed sample
145 ‘horizons’, and 3 such horizons were sampled in each of the 3 compartments forming the
146 boundary zones. Horizons in compartment P or ZP above were positioned so as to represent
147 the lower, middle and upper zones of the fringe, which was taken to be the region between 148 the general limit of the most seaward tree trunks MANUSCRIPT of Avicennia and that of the most seaward 149 pneumatophores (outliers being excluded; i.e. equivalent to the 'higher pneumatophore zone'
150 of Penha-Lopes et al., 2009). Horizons within S, Z and M were located at intervals of 5, 10
151 and 15 m into the compartments concerned from their interfaces with P or ZP (Fig. 2). The
152 width of the pneumatophore fringe was 15 ± 4 m at the four sites, thus forming, at least
153 potentially, a very gradual or soft interface sensu Strayer et al. (2003). Each grid covered
154 some 670 m2.
155 Since ACCEPTEDmost soft-sediment macrofauna occurs in the top few mm of sediment (e.g. 98%
156 in the top ≤50 mm in the studies of Lewis & Stoner, 1981 and Klumpp & Kwak, 2005), each
157 core was taken to a depth of 100 mm. This procedure collects the smaller and most
158 numerous species that constitute the large majority of invertebrate biodiversity (Dittmann,
159 2001; Gaudêncio & Cabral, 2007; Albano et al., 2011), but not the scarcer megafauna or 8 ACCEPTED MANUSCRIPT 160 deeply-burrowing forms. Collection and treatment of core samples followed the same
161 procedure as in the earlier studies of the intertidal North Stradbroke macrofaunal
162 assemblages cited above. Cores were taken soon after tidal ebb from the sites concerned,
163 and were gently sieved through 710 µm mesh on site. Retained material from each sample
164 (i) was placed in a large polythene bag of seawater within which all seagrass or mangrove
165 root material was disassociated and shaken vigorously to dislodge all but sessile animals and
166 then discarded; (ii) was then re-sieved and transported immediately to the laboratory, and
167 (iii) was there placed in a 30 x 25 cm white tray on an A3 LED board in which the living
168 fauna was located by visual examination using magnifying glasses, continuing until no
169 further animal could be seen during a 3 min search period. Animals were identified to
170 species level wherever possible, with all nomenclature as listed in the World Register of
171 Marine Species (www.marinespecies.org), accessed December 2016, except, following
172 Ozawa et al. (2009), for the separation of Velacumantus from Batillaria . Several taxa, 173 however, although relatively important in Moreton MANUSCRIPT Bay have not yet been investigated 174 systematically in southern Queensland (Davie & Phillips, 2008) and their identification to
175 named species poses ‘enormous problems’ (Davie et al., 2011: p. 8). Such animals were
176 treated as morphospecies, an operationally appropriate procedure to detect spatial patterns of
177 numbers of species and their differential abundance (Dethier & Schoch, 2006; Albano et al.,
178 2011).
179 2.2 Analyses ACCEPTED 180 Numbers of each component species within the various sampling horizons and
181 system compartments were subjected to similarity analysis, and assemblage metrics were
182 derived and compared within and across the compartments and habitat types. 9 ACCEPTED MANUSCRIPT 183 Univariate and multivariate assessments of faunal assemblage similarity and
184 calculation of biodiversity metrics were carried out via EstimateS 9.1.0 (Colwell, 2013) and
185 PAST 3.14 (Hammer et al., 2016). Univariate metrics assessed were: (i) faunal abundance
186 per unit area, (ii) observed (N0) and estimated (Nest ) numbers of species per unit area [i.e.
187 Gotelli and Colwell's (2001) 'species density'], (iii) observed number of species per unit
188 faunal numbers [i.e. 'species richness', Nn], (iv) α-diversity, and (v) relative evenness. Nest
189 was assessed as the bias-corrected Chao1 metric (Chao, 2005; Glowacki, 2011); and Nn was
190 obtained by rarefaction or extrapolation using 100 randomizations. Alpha-diversities were
191 measured as Hill's (1973) N2 which is generally insensitive to both spatial grain/extent and
192 size of species pool, and for those reasons is recommended by Dauby & Hardy (2012) and
193 Chase & Knight (2013). Relative evenness was measured by Pielou's J (Beisel et al., 2003;
194 Jost, 2010). Metrics assessed to compare adjacent pairs of horizons and compartments were:
195 (vi) observed (Vobs ) and estimated (Vest ) percentages of shared species and (vii) Bray-Curtis 196 MANUSCRIPT compositional similarities [Legendre and Legendre's (1998) S 17 ]. Vest was based on 197 Abundance Coverage-based Estimates of total species present and on Chao et al.'s (2000)
198 estimator of species shared by such estimates. Comparison of univariate assemblage metrics
199 used one-way ANOVA after ln[x+1] transformation of data, followed by post-hoc Tukey
200 HSD tests where applicable. Faunal patchiness was assessed via Morista's Iδ (see Barnes,
201 2016).
202 MultivariateACCEPTED comparison of macrofaunal assemblage composition used hierarchical 203 clustering analysis of Bray-Curtis similarity, ANOSIM, SIMPER, indices of taxonomic
204 diversity ( ∆) and distinctness ( ∆*) (Clarke and Warwick, 1998), and ordination by non-
205 metric multidimensional scaling (nMDS) carried out on square-root transformed species
206 abundances, all with 9999 permutations. 10 ACCEPTED MANUSCRIPT 207 In order to detect underlying patterns of faunal affinity, two widespread species were
208 excluded from the study; in each case because of extremely patchy distributions coupled
209 with very high density within patches (>50 ind. 0.01m -2). These were: Mictyris longicarpus ,
210 a wandering crab found, especially in the form of juveniles, across the whole
211 mangrove/sandflat boundary zone but in locations changing with each tide (see Barnes and
212 Hamylton, 2016), and Velacumantus australis , a relatively large gastropod that likewise
213 occurs across, and beyond, the whole seagrass/mangrove belt but in the form of dense
214 aggregations separated by areas from which it is absent (see, e.g., Barnes and Hamylton,
215 2015). In each case, randomly encountering a dense patch in some samples and missing
216 them in others had dramatically confounding effects on the results.
217 Relative importance of individual species within the various macrofaunal
218 assemblages was assessed by Barnes's (2014) index of numerical importance which 219 combines data from both abundance and occupancyMANUSCRIPT (= frequency of occurrence). As the 220 pattern of relative importance of individual specie s in an assemblage (whether measured
221 as abundance, occupancy or a composite of the two) can yield information on factors
222 structuring ecological assemblages when displayed in rank order as a Whittaker Plot (e.g.
223 Magurran, 2004; Foster & Dunstan, 2010; Jenkins, 2011), such ranked Species
224 Importance Distributions (SIDs) were constructed for each system compartment. All data
225 sets were standardised for overall species density (by dividing all ranks by the total
226 number of species in the set) and for assemblage sample size (by dividing each species 227 total by the overallACCEPTED number of individuals in the set) to permit comparison of curve slopes 228 (Passy, 2016), a measure of equitability in individual species contribution to the total
229 (Whittaker, 1972).
230 11 ACCEPTED MANUSCRIPT 231 3. Results
232 The North Stradbroke intertidal samples yielded a subset of 115 of the species
233 known to occur in intertidal soft sediments in Moreton Bay, at densities within the ranges of
234 those previously recorded from the habitats concerned (e.g. Morgan and Hailstone, 1986;
235 Camilleri, 1992; Barnes and Barnes, 2012; Barnes and Hamylton, 2016). 40% of species
236 occurred only as singletons or doubletons, and a mere 8% comprised >50% of the total
237 individuals within the whole 45 m deep boundary zone; in descending order of abundance,
238 these were: Calopia imitata (truncatelloid gastropod), Enigmaplax littoralis (ocypodoid
239 brachyuran), Cerithium coralium (cerithioid gastropod), Iravadia goliath (truncatelloid
240 gastropod), Indoaustriella lamprelli (lucinoid bivalve), Dasybranchus caducus (scolecidan
241 polychaete), Aglaophamus australiensis (phyllodocidan polychaete), Urohaustorius
242 mertungi (phoxocephaloid amphipod) and Nassarius burchardi (buccinoid gastropod). 243 Besides Mictyris and Velacumantus , some 15%MANUSCRIPT of the species occurred in all three system 244 compartments, including various infaunal polychaete s (particularly Aglaophamus
245 australiensis but also Capitella sp., Dasybranchus caducus and Scoloplos cf normalis ), a
246 few amphipods ( Victoriopisa australiensis and an ischyrocerid) and the pyramidellid
247 gastropod ?Pyrgulina sp., that appears associated with Velacumantus . Considering only
248 those species present as >4 individuals in total, no species was restricted to the
249 pneumatophore-fringe zone; five appeared effectively to be confined to the sand (including
250 the dominant Urohaustorius and Koloonella sp.); two to the mangrove proper 251 (IndoaustriellaACCEPTED and Boccardia sp.); and 12 to seagrass (i.e. across both zones Z and ZP), 252 including the dominant Calopia . C. imitata has not been recorded from other intertidal
253 habitats in Moreton Bay, but it is widespread and abundant in most types of sublittoral soft
254 substrata there (Rachello-Dolmen, 2013). 12 ACCEPTED MANUSCRIPT 255 Except for compartment P, the assemblages of each habitat type at different sampling
256 sites displayed significant compositional differences (ANOSIM P <0.01; R = >0.13), to
257 which the dominant species listed in Table 1 contributed >40% (SIMPER). Notwithstanding
258 these differences, however, ordination of faunal assemblages produced distinct
259 compartment-based clusters, with the interfacing pneumatophore fringes of the two types of
260 transition clustering with system compartments at opposite tidal heights; those of P with the
261 adjacent upper-shore M, and those of ZP with Z (Fig. 3). This was very clearly so at
262 Z:ZP:M transitions where ZP was largely an extension of Z and there was no significant
263 variation in assemblage composition across the six component seagrassed horizons at each
264 site (Figs 4 & 5). Nevertheless, there were consistent changes in some faunal components
265 across the seagrass. One dominant seagrass microgastropod, Calopia imitata , for example,
266 declined in importance upshore (Pearson R = 0.86; P = 0.03), where it was increasingly
267 replaced by Iravadia goliath (Table 1); and less abundant related truncatelloids such as the 268 Pseudoliotia spp. were also replaced by what canMANUSCRIPT be presumed to be the functionally 269 equivalent truncatelloids Stenothyra australis , Wakauraia fukudai and Tatea huonensis as
270 well as by the small pulmonate Salinator (particularly S. rhamphidia ), all more
271 characteristic of mangrove. The clustering of the fauna of P with that of M was less close,
272 however, and significant changes in assemblage composition occurred across their boundary
273 (Figs 4 & 5). The proportions of species shared between P and both S and M were also
274 more similar than their clustering pattern would suggest (Table 2).
275 BesidesACCEPTED the change in assemblage composition across the interface between P/ZP 276 and M seen at all four sites, and that between S and P occurring at both S:P:M transitions
277 (Figs 4 & 5), significant change also occurred within compartment P at Yerrol and within M
278 at all sites except Yerrol. SIMPER analysis showed that change across M/P or M/ZP
279 interfaces was largely consequent on the presence of Indoaustriella and the greater 13 ACCEPTED MANUSCRIPT 280 abundance of Iravadia and Cerithium in the landwards habitat, and 30-35% of the
281 significant change within M was due to fluctuation in the relative abundance of these same
2 282 species, all three of which were patchily distributed (Morisita's Iδ >2; χ >180; P <<0.001).
283 35% of the significant change within P at Yerrol was due to the presence to landwards of
284 Cerithium and a switch in the polychaete fauna from Scoloplos to Simplisetia (SIMPER).
285 Biodiversity measures (observed and estimated species density, species richness and
286 species diversity per horizon) changed significantly across the Z:ZP:M boundary zone (all
287 one-way ANOVA F2,15 >6, P <0.02), with ZP consistently supporting the greatest values of
288 the three compartments for all these metrics, but abundance, evenness and taxonomically-
289 derived indices showed no significant differences (all ANOVA F2,15 <2.4, P >0.1) (see
290 Table 2). The results for evenness and taxonomic indices were the same across the S:P:M
291 boundary, but otherwise the converse was the case for this transition; abundance per horizon
292 showing significant change (ANOVA F2,15 = 8.3;MANUSCRIPT P <0.004), with the P horizons supporting 293 the least numbers, whilst the four biodiversity indices showed no significant differences (all
294 ANOVA F2,15 <2.0; P >0.17). Unlike in some reef/seagrass ecotones (Tuya et al., 2011),
295 similarities or contrasts in number of species in the various system compartments were not
296 consequent on their patterns of abundance (see Gotelli and Colwell, 2001) (see Fig. 6).
297 Indeed, for the Z:ZP:M transition, the two system compartments with the least assemblage
298 abundance (1200 and 1270 ind m 2) supported the most and the least species (38 and 56 spp),
2 299 and for the S:P:MACCEPTED transition, the two with the greatest numbers (810 and 842 ind m ) also 300 supported the most and least species (28 and 40) (Table 2).
301 Irrespective of the widely varying numbers of species (observed 14-56 and expected
302 24-86) and numbers of individuals (33-465) in each individual system compartment across
303 the two types of transition, the ranked Species Importance Distributions of compartmental 14 ACCEPTED MANUSCRIPT 304 data were extremely similar, with exponential decay constants ranging only from -0.07 to -
305 0.13 across both types of transition (Fig. 7). The type of habitat sampled, intertidal soft –
306 basically sandy – sediments at MLW, was clearly comparable across all compartments
307 sampled, and it would appear that whatever structures the differential importance of the
308 various species in each assemblage acts irrespective of the presence of seagrass or mangrove
309 trees.
310
311 4. Discussion
312 There is no doubt that mangrove forms part of a single interconnected biological
313 system that extends well offshore to adjacent reefs and coastal waters (Meynecke et al.,
314 2008), and hence it would be very surprising if some species did not range across the 315 interfaces between that habitat and immediatelyMANUSCRIPT adjacent seagrass or sediment flat. 316 Epifaunal gastropods such as Velacumantus australis and infaunal polychaetes such as
317 Aglaophamus australiensis demonstrate the essential unity of the sheltered North
318 Stradbroke soft-sediment intertidal ecosystem despite the increasing presence of a
319 vegetation cover. Other species are likely to extend across the interface between mangrove
320 and lower-shore systems if only because of 'mass effects' (Kunin, 1998). To that extent,
321 any spatially intermediate system compartment can be expected to be transitional in the
322 broad sense of containing, permanently or temporarily, some elements more characteristic 323 of each contiguousACCEPTED habitat type. But usage of the term 'transitional' in the present ecotonal 324 context implies either (or both) having values of important assemblage metrics intermediate
325 between those of the two larger interfacing systems (the usage of Alfaro, 2006) or having a
326 faunal composition part-way between those two (as in, e.g., Van Hoey et al., 2004). 15 ACCEPTED MANUSCRIPT 327 In terms of its assemblage composition, any spatially intermediate zone might
328 potentially support either: (a) essentially the same assemblage as one of the interfacing
329 habitats but not that of the other, and the faunal change could occur elsewhere; (b) both
330 assemblage types, and therefore be more rich and diverse than either; (c) neither of them,
331 and be either (c 1) a relatively barren ecological 'no-man's-land' or else (c 2) a region with its
332 own distinctive assemblage, unrelated to those of either contiguous habitat; or indeed (d) a
333 state or states intermediate between those of the adjacent two. Features of all such
334 outcomes except c 2 were to some extent represented along the 6 km stretch of Australian
335 coast along which the sites were located. Secondly, besides being intermediate in
336 magnitude, ecotonal assemblage metrics could potentially also show other states relative to
337 those of the habitats on either side, and there seem no a priori reasons why any value
338 applicable to the mangrove/seagrass interface should be the same as that to the
339 mangrove/sandflat one, nor why the response of different assemblage metrics (e.g. of 340 abundance, evenness and biodiversity) should MANUSCRIPTtake the same form. In practice, relative 341 values of assemblage metrics of the North Stradbroke pneumatophore fringe did (a) in
342 many cases vary with the nature of the lower-shore habitat and (b) not show consistent
343 pattern from metric to metric. All possible relative magnitudes were observed, and in only
344 two out of 16 instances was the pneumatophore-fringe metric intermediate between those
345 of the mangrove and lower-shore habitats. One of these was indeed the species density
346 shown to be intermediate in value by Alfaro (2006) (albeit in a seagrass-mangrove 347 transition whereasACCEPTED the present case was in a sandflat-mangrove one), but in neither case was 348 the difference between the three compartments statistically significant.
349 Insofar as the present results permit generalisation (see below), North Stradbroke
350 pneumatophore-fringe interfaces with seagrass showed domination by the down-shore
351 seagrass fauna although biodiversity was there somewhat higher because of the additional 16 ACCEPTED MANUSCRIPT 352 presence of some more characteristically mangrove animals. Overall abundance, however,
353 although relatively low in comparison with the seagrass, was effectively the same as in the
354 mangrove and abundance did not vary significantly across the whole zone. In contrast,
355 fringe interfaces with sandflat showed more evenly mixed faunas although overall
356 assemblage densities were significantly less than (only half) those in the adjacent
357 compartments and because of possessing equivalent species density and richness were
358 accordingly of high species diversity. A wide range of ecotonal responses are therefore
359 represented, and the same has been demonstrated to occur across other equivalent habitat
360 interfaces. Faunal abundance across seagrass/bare-sand boundaries, for example, is known
361 on occasion or at some sites to increase towards the interface because of larval settlement
362 patterns and/or of active accumulation of adults at the margins (Bologna and Heck, 2002;
363 Godbold et al., 2011; Wong and Dowd, 2015). In contrast, other instances show either an
364 abrupt c1 response (Barnes and Hamylton, 2013, 2016) or very little change at all (Barnes 365 and Barnes, 2014). Indeed, in the latter case seagMANUSCRIPTrass assemblages at given points 366 displayed more similarity with those of adjacent unvegetated sand than they did with other
367 local areas of seagrass, and likewise in respect of the sandflat assemblages. Context does
368 indeed appear everything in explaining ecotonal form, rendering attempted generalisation
369 across sites problematic (Warman et al., 2013).
370 With reference to relative abundance at habitat interfaces, Ries and Sisk (2004),
371 however, have proposed a contextual model essentially for individual species that can 372 neverthelessACCEPTED be extended to multi-species assemblages if responses of the various species to 373 the interface are additive (Ries et al., 2004). This does seems generally to be the case in
374 respect of the reactions of abundance and diversity of local seagrass and sandflat
375 assemblages to their interface, albeit that this boundary can be abrupt rather than gradual
376 (Barnes and Hamylton, 2016). In the Ries and Sisk (2004) model, only where one habitat 17 ACCEPTED MANUSCRIPT 377 supports more of the same types of resource (in the widest sense of that word) than the
378 adjacent habitat, i.e. shows a quantitative not qualitative difference, will abundance take a
379 transitional, intermediate form at the interface. In contrast, in all cases where resources are
380 qualitatively different across the interface, regardless of their quantitative values, the
381 interface should support greater abundance than either compartment; and where resources
382 in the two habitats are quantitatively and qualitatively similar, no change in abundance
383 would be predicted. A comparable argument could be put forward for resource diversity
384 (or more generally habitat diversity) and species density and/or diversity. Only where one
385 habitat type supports only a limited subset of resources, microhabitats, etc., available in the
386 other, should species density or richness decline across the interface zone. Qualitatively
387 different potential niches in the two habitats, regardless of their relative frequencies, should
388 lead to elevated numbers of species in the boundary region. Because it lay outside the
389 scope of their model, none of the Ries and Sisk (2004) scenarios predicted a situation in 390 which abundances (and, in parallel, species density MANUSCRIPT/richness) are least at the interface, as in
391 case c1 above. Such was the outcome, however, not only at the seagrass/unvegetated-sand
392 boundary along the North Stradbroke intertidal (Barnes and Hamylton, 2016) but also in
393 the experimental intertidal turf habitats of Matias et al. (2013). The cause of this state was
394 not specifically investigated but, for example, it is likely to occur when a key structural
395 habitat resource in each compartment declines to zero at that point to be replaced by an
396 incompatible alternative. Thus although the presence of seagrass is self-evidently the key 397 feature of a seagrassACCEPTED bed, minimal abundance at an interface with sandflat is likely also to 398 require the presence of powerful bioturbators in the unvegetated compartment, e.g. Trypaea
399 and Mictyris on Stradbroke, in that without such bioturbation there may be no c1-type
400 outcome (Barnes and Barnes, 2014). Failure of larval settlement near the interface for
401 whatever reason is another possible cause of the effect (Matias et al., 2013). 18 ACCEPTED MANUSCRIPT 402 Hence the nature of relative values of faunal abundance and diversity across
403 boundary zones may be suggestive of relative resource availability in the interfacing
404 habitats, although the basis of such theory is still very far from complete. If considered in
405 such a light, however, the presence of seagrass in the spatially intermediate
406 pneumatophore-fringe ecotone in this subtropical Queensland locality would appear to
407 enhance the variety of the local resource base, as might be expected. It would not,
408 however, appear at the same time to increase its overall quantity, which seems somewhat
409 counter intuitive, although obviously not all consumers are equal even with respect to the
410 same resource base. In contrast, numbers declined across the interface. Although the
411 variety of the resource base may be enhanced, the limit to consumer assemblage abundance
412 would appear to be set by factors other than overall resource quantity; for example by
413 predators for whom those consumers are the resource, including the suites of juvenile and
414 adult prawns and fish entering the intertidal system at high tide (Lewis and Anderson, 415 2012; Whitfield, 2016). The densities of seagrass MANUSCRIPT invertebrates in the North Stradbroke 416 beds, and in equivalent beds elsewhere, appear likely to be well below potential carrying
417 capacity which would accord with top-down control (Barnes, 2016). The gradient in the
418 essentially sandy substratum at Capembah and Yerrol from the unvegetated condition at
419 one extreme to a shaded and leaf-litter rich sediment at the other also creates a situation of
420 habitat change beyond local resource availability. Here the intervening pneumatophore
421 fringe is neither the relatively clean, unstructured expanse of sand that can support, for 422 example, specialistACCEPTED mobile burrowers such as Urohaustorius (not least because of the dense 423 meshwork of fibrous rootlets of Avicennia just beneath the surface in which Indoaustriella
424 nestles), nor as sheltered a detrital habitat for epifaunal gastropods. To that extent it
425 approximates a c1 system, although a few burrowing polychaetes and a few epifaunal 19 ACCEPTED MANUSCRIPT 426 gastropods can clearly be accommodated at low densities, possibly because of dispersal
427 pressure.
428 A further factor reducing the efficacy of generalisation is that not all fringing
429 expanses of pneumatophores are the same; amongst other environmental variables likely to
430 impinge on invertebrate assemblages, some pneumatophore fields will be more sheltered
431 and have muddier substrata than others, and pneumatophore density also varies widely
432 (Dahdouh-Guebas et al., 2007). The pneumatophore fringes under consideration here are on
433 the open shores of a large bay and possess relatively dense, short pneumatophores (up to
434 >500 m -2). Crabs such as Uca (sensu lato ), however, prefer open areas with sparse
435 pneumatophores (Skilleter and Warren, 2000; Nobbs and Blamires, 2015), e.g. along creek
436 banks, and accordingly were scarce in the current samples. The substratum studied by
437 Alfaro (2006) was soft mud, not sand, and appears to support a different type of fauna to 438 that on North Stradbroke, with for example a greateMANUSCRIPTr range and density of bivalve molluscs 439 and larger gastropods. The effects of such habitat variables on pneumatophore ecotonal
440 biodiversity, however, have not been studied. Much of the work on mangrove-related
441 invertebrates that has been carried out has been restricted to the larger infaunal crabs
442 (Kristensen, 2008; Mokhtari et al., 2015) and larger epifaunal molluscs (Vermeij, 1974;
443 Pape et al., 2008). The smaller, though more numerous, elements of the macrofauna
444 documented by this and related studies along the North Stradbroke intertidal zone fall into a
445 category of largely ignored species (Albano et al., 2011; Middelfart et al., 2016), being 446 completely omittedACCEPTED from many mangrove faunal inventories and from reviews such as that 447 of Kathiresan and Bingham (2001), for example. Hence considerable further work will be
448 necessary before the full range of ecotonal responses to the interfaces of mangroves with
449 lower-shore habitats can be documented, let alone understood in resource – or any other -
450 terms. Is the situation described by Alfaro (2006) more common in temperate areas, whilst 20 ACCEPTED MANUSCRIPT 451 the present results are more characteristic of the subtropics? Do interfaces with Rhizophora
452 prop-roots differ significantly from those involving cable-roots and pneumatophores? Does
453 interface form vary with salinity? At least at the sites sampled, however, it is clear that the
454 richest area of biodiversity is located in those peripheral mangrove zones that also support
455 seagrass. It is also clear from historical data on Google Earth that seagrass has been
456 retreating downshore at localities such as Capembah for at least the last 12 years. Whilst
457 the whole mangrove ecosystem is in urgent need of conservation (Duke et al., 2007), there
458 may therefore be a particular need for biodiversity conservation effort in their downshore
459 interface regions (Araujo, 2002).
460
461 Acknowledgements
462 I am most grateful to the Quandamooka YoolooburrabeMANUSCRIPTe Aboriginal Corporation (QYAC), 463 the Quandamooka Aboriginal Land and Sea Management Agency, and the Queensland
464 Parks and Wildlife Service for permission to conduct research within the native title area of
465 the Quandamooka People, and within Habitat Protection and Marine National Park zones
466 of the Moreton Bay Marine Park, under permit QS2014/CVL588. My gratitude is also due
467 to Ian Tibbetts and the staff of the Moreton Bay Research Station for their help and support,
468 to Andrea Alfaro for commenting on a draft of the paper, and to Adrian Friday for advice
469 on use of PAST3 . This research did not receive any specific grant from funding agencies in 470 the public, commercial,ACCEPTED or not-for-profit sectors.
471
472 References 21 ACCEPTED MANUSCRIPT 473 Accad, A., Li, J., Dowling, R., Guymer, G.P., 2016. Mangrove and associated communities
474 of Moreton Bay, Queensland, Australia: change in extent 1955-1997-2012. Queensland
475 Herbarium, Dept of Science, Information Technology and Innovation, Brisbane.
476 Albano, P.G., Sabelli, B., Bouchet, P., 2011. The challenge of small and rare species in
477 marine biodiversity surveys: microgastropod diversity in a complex tropical coastal
478 environment. Biodiversity and Conservation , 20 : 3223-3237
479 Alfaro, A.C., 2006. Benthic macro-invertebrate community composition within a
480 mangrove/seagrass estuary in northern New Zealand. Estuarine, Coastal and Shelf
481 Science , 66 : 97-110
482 Alongi, D.M., 2002. Present state and future of the world's mangrove forests. Environmental
483 Conservation , 29 : 331-349
484 Araujo, M.B., 2002. Biodiversity hotspots and MANUSCRIPT zones of ecological transition. Conservation 485 Biology , 16 : 1662-1663
486 Armenteros, M., Williams, J.P., Hidalgo, G., Gonzalez-Sanson, G., 2007. Community
487 structure of meio- and macrofauna in seagrass meadows and mangroves from NW shelf
488 of Cuba (Gulf of Mexico). Revista de Investigaciones Marina , 28 : 139-150
489 Arnold, T., Freundlich, G., Weilnau, T., Verdi, A., Tibbetts, I.R., 2014. Impacts of 490 groundwaterACCEPTED discharge at Myora Springs (North Stradbroke Island, Australia) on the 491 phenolic metabolism of eelgrass, Zostera muelleri , and grazing by the juvenile rabbitfish,
492 Siganus fuscescens . PLoS ONE , 9(8): e104738. doi:101371/journal.pone.0104738 22 ACCEPTED MANUSCRIPT 493 Barbier, E.B., Hacker, S.D., Kennedy, C., Koch, E.W., Stier, A.C., Silliman. B.R., 2011.
494 The value of estuarine and coastal ecosystem services. Ecological Monographs , 81 :
495 169-193
496 Barnes, R.S.K., 2014. The nature and location of spatial change in species assemblages: a
497 new approach illustrated by the seagrass macrofauna of the Knysna estuarine bay, South
498 Africa. Transactions of the Royal Society of South Africa , 69 : 75-80
499 Barnes, R.S.K., 2016. Spatial homogeneity of benthic macrofaunal biodiversity across
500 small spatial scales. Marine Environmental Research , 122 : 148-157
501 Barnes, R.S.K., Barnes, M.K.S., 2012. Shore height and differentials between
502 macrobenthic assemblages in vegetated and unvegetated areas of an intertidal sandflat.
503 Estuarine, Coastal and Shelf Science , 106 : 112-120
504 Barnes, R.S.K., Barnes, M.K.S., 2014. Biodiversity MANUSCRIPT differentials between the numerically- 505 dominant macrobenthos of seagrass and adjacent unvegetated sediment in the absence of
506 sandflat bioturbation. Marine Environmental Research , 99 : 34-43
507 Barnes, R.S.K., Ellwood, M.D.F., 2011. The significance of shore height in intertidal
508 macrobenthic seagrass ecology and conservation. Aquatic Conservation , 21 : 614-624
509 Barnes, R.S.K., Hamylton, S., 2013. Abrupt transitions between macrobenthic faunal 510 assemblagesACCEPTED across seagrass bed margins. Estuarine, Coastal and Shelf Science , 131 : 511 213-223
512 Barnes, R.S.K., Hamylton, S., 2015. Uniform functional structure across spatial scales in
513 an intertidal benthic assemblage. Marine Environmental Research , 106 : 82-91 23 ACCEPTED MANUSCRIPT 514 Barnes, R.S.K., Hamylton, S., 2016. On the very edge: faunal and functional responses to
515 the interface between benthic seagrass and unvegetated-sand assemblages. Marine
516 Ecology Progress Series , 553 : 33-48
517 Beisel, J-N., Usseglio-Polatera, P., Bachmann, P., Moreteau, J-C., 2003. A comparative
518 analysis of evenness index sensitivity. International Revue of Hydrobiology , 88 : 3-15
519 Bloomfield, A.L., Gillanders, B.M., 2005. Fish and invertebrate assemblages in seagrass,
520 mangrove, saltmarsh, and nonvegetated habitats. Estuaries , 28 : 63-77
521 Bologna, P.A.X., Heck, K.L.Jr, 2002. Impact of habitat edges on density and secondary
522 production of seagrass-associated fauna. Estuaries , 25 : 1033-1044
523 Boström, C., Pittman, S.J., Simenstad, C., Kneib, R.T., 2011. Seascape ecology of coastal
524 biogenic habitats: advances, gaps, and challenges. Marine Ecology Progress Series , 427 : 525 191-217 MANUSCRIPT
526 Bouillon, S., Dehairs, F., Velimirov, V., Abril G., Borges, A.V., 2007. Dynamics of organic
527 and inorganic carbon across contiguous mangrove and seagrass systems (Gazi Bay,
528 Kenya). Journal of Geophysical Research , 112 : G02018, DOI:10.1029/2006JG000325
529 Camilleri, J.C., 1992. Leaf-litter processing by invertebrates in a mangrove forest in
530 Queensland. Marine Biology , 114 : 139-145
531 Chao, A., 2005.ACCEPTED Species estimation and applications. In: Balakrishnan, N., Read, C.B.,
532 Vidakovic, B. (eds) Encyclopedia of Statistical Sciences, 2 nd Edn , 12 : 7907-7918. Wiley,
533 New York
534 Chao, A., Hwang, W-H., Chen, Y-C., Kuo, C-Y., 2000. Estimating the number of shared 24 ACCEPTED MANUSCRIPT 535 species in two communities. Statistica Sinica , 10 : 227-246
536 Chase, J.M., Knight, T.M., 2013. Scale-dependent effect sizes of ecological drivers on
537 biodiversity: why standardised sampling is not enough. Ecology Letters , 16 : 17-26
538 Clarke, K.R., Warwick, R.M., 1998. A taxonomic distinctiveness index and its statistical
539 properties. Journal of Applied Ecology , 35 : 523-531
540 Colwell, R.K., 2013. EstimateS : Statistical estimation of species richness and shared
541 species from samples, version 9.1.0Mac. Users guide and application available at
542 http://purl.oclc.org/estimates
543 Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K.,
544 Naeem, S., O'Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., van den Belt, M., 1997.
545 The value of the world's ecosystem services and natural capital. Nature , 387 : 253-260 MANUSCRIPT 546 Dahdouh-Guebas, F., Kairo, J.G., de Bondt, R., Koedam, N., 2007. Pneumatophore height
547 and density in relation to micro-topography in the grey mangrove Avicennia marina .
548 Belgian Journal of Botany , 140 : 213-221
549 Dalia, S.V., Satheesh, K.P., Pillai, N.G.K., 2014. Biodiversity and seasonal variation of
550 benthic macrofauna in Minicoy Island, Lakshadweep, India. Acta Oceanologica Sinica ,
551 33 : 58-73
552 Dauby, G., Hardy,ACCEPTED O.J., 2012. Sample-based estimation of diversity sensu stricto by
553 transforming Hurlbert diversities into effective number of species. Ecography , 35 : 661-
554 672
555 Davie, P.J.F. and others (2011) Wild Guide to Moreton Bay, 2 nd Edn . Queensland Museum, 25 ACCEPTED MANUSCRIPT 556 Brisbane
557 Davie, P.J.F., Phillips, J.A., 2008. The Marine Fauna and Flora of Moreton Bay Workshop:
558 a focus on species. In: Davie, P.J.F., Phillips, J.A. (eds) Proceedings of the Thirteenth
559 International Marine Biological Workshop. The Marine Fauna and Flora of Moreton
560 Bay, Queensland. Memoirs of the Queensland Museum – Nature , 54 (1): xi-xvii
561 Davies, J.L., 1972. Geographical Variation in Coastal Development , Edinburgh, Oliver &
562 Boyd.
563 Dennison, W.C., Abal, E.G., 1999. Moreton Bay Study: A Scientific Basis for the Healthy
564 Waterways Campaign. SE Queensland Regional Water Quality Management Strategy,
565 Brisbane.
566 Dethier, M.N., Schoch, G.C. 2006. Taxonomic sufficiency in distinguishing natural spatial 567 patterns on an estuarine shoreline. Marine EcologyMANUSCRIPT Progress Series , 306 : 41-49
568 Dittmann, S., 2001. Abundance and distribution of small infauna in mangroves of
569 Missionary Bay, North Queensland, Australia. Revista de Biologia Tropical , 49 : 535-
570 544
571 Duke, N.C., Meynecke, J-O., Dittmann, S., Ellison, A.M., Anger, K., Berger, U., Cannicci,
572 S., Diele, K., Ewel, K.C., Field, C.D., Koedam, N., Lee, S.Y., Marchand, C., Nordhaus, 573 I., Dahdouh-Guebas,ACCEPTED F., 2007. A world without mangroves? Science , 317 : 41-42
574 Ellison, A.M., 2008. Managing mangroves with benthic biodiversity in mind: moving
575 beyond roving banditry. Journal of Sea Research , 59 : 2-15 26 ACCEPTED MANUSCRIPT 576 Foster, S.D., Dunstan, P.K., 2010. The analysis of biodiversity using rank abundance
577 distributions. Biometrics , 66 : 186-195
578 Fourqurean, J.W., Duarte, C.M., Kennedy, H., Marba, N., Holmer, M., Mateo, M.A.,
579 Apostolaki, E.T., Kendrick, G.A., Krause-Jensen, D., McGlathery, K.J., Serrano, 0.,
580 2012. Seagrass ecosystems as a globally significant carbon stock. Nature Geoscience , 5:
581 505-509
582 Gaudêncio, M.J., Cabral, H.N., 2007. Trophic structure of macrobenthos in the Tagus
583 estuary and adjacent coastal shelf. Hydrobiologia , 587 : 241-251
584 Gibbes, B., Grinham. A., Neil, D., Olds, A., Maxwell, P., Connolly, R., Weber, T., Udy, N.,
585 Udy, J., 2014. Moreton Bay and its estuaries: a sub-tropical system under pressure from
586 rapid population growth. In: Wolanski, E. (ed.), Estuaries of Australia in 2050 and
587 beyond. Springer, Dordrecht, pp. 203-222 MANUSCRIPT 588 Gillis, L.G., Ziegler, A.D., van Oevelen, D., Cathalot, C., Herman, P.M.J., Wolters, J.W.,
589 Bouma, T.J., 2014. Tiny is mighty: seagrass beds have a large role in the export of
590 organic material in the tropical coastal zone. PloS One , 9(11): e111847
591 doi:10.1371/journal.pone.0111847
592 Glowacki, L., 2011. Accuracy of species richness estimators applied to fish in small and
593 large temperate lowland rivers. Biodiversity and Conservation , 20 , 1365-1384 ACCEPTED 594 Godbold, J.A., Bulling, M.T., Solan, M., 2011. Habitat structure mediates biodiversity
595 effects on ecosystem properties. Proceedings of the Royal Society (B) , 278 : 2510-2518
596 Gotelli, N.J., Colwell, R.K., 2001. Quantifying biodiversity: procedures and pitfalls in the
597 measurement and comparison of species richness. Ecology Letters , 4: 379-391 27 ACCEPTED MANUSCRIPT 598 Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2016. PAST: Paleontological statistics software
599 package for education and data analysis, Version 3.14.
600 Hill, M.O., 1973. Diversity and evenness: a unifying notation and its consequences.
601 Ecology , 54 : 427-432
602 Imgraben, S., Dittmann, S., 2008. Leaf litter dynamics and litter consumption in two
603 temperate South Australian mangrove forests. Journal of Sea Research , 59 : 83-93
604 Jenkins, D.G., 2011. Ranked species occupancy curves reveal common patterns among
605 diverse communities. Global Ecology and Biogeography , 20 : 486-497
606 Jost, L., 2010. The relation between evenness and diversity. Diversity , 2, 207-232
607 Kathiresan, K., Bingham, B.L., 2001. Biology of mangroves and mangrove ecosystems.
608 Advances in Marine Biology, 40 : 81-251 MANUSCRIPT 609 Klumpp, D.W., Kwak, S.N., 2005. Composition and abundance of benthic macrofauna of a
610 tropical sea-grass bed in north Queensland, Australia. Pacific Science , 59 : 541-560
611 Kristensen, E., 2008. Mangrove crabs as ecosystem engineers; with emphasis on sediment
612 processes. Journal of Sea Research , 59 : 30-43
613 Kunin, W.E., 1998. Biodiversity at the edge: a test of the importance of spatial 'mass 614 effects' inACCEPTED the Rothamsted Park grass experiments. Proceedings of the National 615 Academy of Sciences of the USA , 95 : 207-212
616 Langhans, S.D., Tockner, K., 2014. Edge effects are important in supporting beetle
617 biodiversity in a gravel-bed river floodplain. PloS ONE , 9(12): e114415.
618 doi:10.1371/journal.pone.0114415 28 ACCEPTED MANUSCRIPT 619 Lee, S.Y., 2008. Mangrove macrobenthos: assemblages, services and linkages. Journal of
620 Sea Research , 59 : 16-29
621 Lee, S.Y., Primavera, J.H., Dahdouh-Guebas, F., McKee, K., Bosire, J.O., Cannicci, S.,
622 Diele, K., Fromard, F., Koedam, N., Marchant, C., Mendelssohn, I., Mukherjee, N.,
623 Record, S., 2014. Ecological role and services of tropical mangrove ecosystems: a
624 reassessment. Golbal Ecology and Biogeography , 23 : 726-743
625 Legendre, P., Legendre, L., 1998. Numerical Ecology, 2 nd English Edn. Elsevier,
626 Amsterdam.
627 Levin, L.A., Boesch, D.F., Covich, A., Dahm, C., Erséus, C., Ewel, K.C., Kneib, R.T.,
628 Moldenke, A., Palmer, M.A., Snelgrove, P., Strayer, D., Weslawski, J.M., 2001. The
629 function of marine critical transition zones and the importance of sediment biodiversity.
630 Ecosystems , 4: 430-451 MANUSCRIPT 631 Lewis, F.G., Stoner, A.W., 1981. An examination of methods for sampling macrobenthos
632 in seagrass meadows. Bulletin of Marine Science , 31 : 116-124
633 Lewis, L.S., Anderson, T.W. , 2012. Top-down control of epifauna by fishes enhances
634 seagrass production. Ecology , 93 : 2746-2757
635 Macnae, W., 1968. A general account of the fauna and flora of mangrove swamps and 636 forests inACCEPTED the Indo-West-Pacific region. Advances in Marine Biology , 6: 73-270
637 Magurran, A.E., 2004. Measuring Biological Diversity . Blackwell, Oxford
638 Matias, M.G., Coleman, R.A., Hochuli, D.F., Underwood, A.J., 2013. Macrofaunal
639 responses to edges are independent of habitat-heterogeneity in experimental landscapes. 29 ACCEPTED MANUSCRIPT 640 PloS ONE , 8(4): e61349. doi 10.1371/journal.pone.0061349
641 McGlathery, K.J., Reidenbach, M.A., D'Odorico, P., Fagherazzi, S., Pace, M.L., Porter,
642 J.H., 2013. Nonlinear dynamics and alternative stable states in shallow coastal systems.
643 Oceanography , 26 : 221-231
644 Metcalfe, K.N., Glasby, C.J., 2008. Diversity of Polychaeta (Annelida) and other worm
645 taxa in mangrove habitats of Darwin Harbour, northern Australia. Journal of Sea
646 Research , 59 : 70-82
647 Meynecke, J.O., Lee, S.Y., Duke, N.C., 2008. Linking spatial metrics and fish catch
648 reveals the importance of coastal wetland connectivity to inshore fisheries in
649 Queensland, Australia. Biological Conservation , 141 : 981-996
650 Middelfart, P.U., Kirkendale, L.A., Wilson, N.G., 2016. Australian tropical micromolluscs: 651 An overwhelming bias . Diversity , 8(17); doi:10.3390/d8030017MANUSCRIPT
652 Moffett, K.B., Nardin, W., Silvestri, S., Wang, C., Temmerman, S., 2015. Multiple stable
653 states and catastrophic shifts in coastal wetlands: progress, challenges, and opportunities
654 in validating theory using remote sensing and other methods. Remote Sensing , 7: 10184-
655 10226
656 Mokhtari, M., Ghaffar, M.A., Usup, G., Cob, Z.C., 2015. Determination of key 657 environmentalACCEPTED factors responsible for distribution patterns of fiddler crabs in a tropical 658 mangrove ecosystem. PloS ONE , 10 (1): e0117467. doi:10.1371/journal.pone.0117467
659 Morgan, G.J., Hailstone, T.S., 1986. Distribution of gastropods in a mangrove habitat in
660 south-east Queensland. Journal of the Malacological Society of Australia , 7: 131-140 30 ACCEPTED MANUSCRIPT 661 Morrisey, D.J., Swales, A., Dittmann, S., Morrison, M.A., Lovelock, C.E., Beard, C.M.,
662 2010. The ecology and management of temperate mangroves. Oceanography and Marine
663 Biology: An Annual Review , 48 : 43-160
664 Nagelkerken, I., Blaber, S.J.M., Bouillon, S., Green, P., Haywood, M., Kirton, L.G.,
665 Meynecke, J-O., Pawlik, J., Penrose, H.M., Sasekumar, A., Somerfield, P.J., 2008. The
666 habitat function of mangroves for terrestrial and marine fauna: a review. Aquatic Botany ,
667 89 : 155-185
668 Nobbs, M., Blamires, S.J., 2015. Spatiotemporal distribution and abundance of mangrove
669 ecosystem engineers: burrowing crabs around canopy gaps. Ecosphere , 6(5); 84.
670 doi:10/1890/ES14-00498.1
671 Ollivier, Q.R., Bramwell, N.A., Hammill, E., Foster-Thorpe, C., Booth, D.J., 2015. Are the 672 effects of adjacent habitat type on seagrass MANUSCRIPTgastropod communities being masked by 673 previous focus on habitat dyads? Australian Journal of Zoology , 63 : 357-363
674 Ozawa, T., Köhler, F., Reid, D.G., Glaubrecht, M., 2009. Tethyan relics on continental
675 coastlines of the northwestern Pacific Ocean and Australasia: molecular phylogeny and
676 fossil record of batillariid gastropods (Caenogastropoda, Cerithioidea). Zoologica Scripta,
677 38: 503-525
678 Pape, E., Muthumbi, A., Kamanu, C.P., Vanreusel, A., 2008. Size-dependent distribution
679 and feedingACCEPTED habits of Terebralia palustris in mangrove habitats of Gazi Bay, Kenya.
680 Estuarine, Coastal and Shelf Science , 76 : 797-808
681 Passy, S.I., 2016. Abundance inequality in freshwater communities has an ecological origin.
682 American Naturalist , 187 : 505-516 31 ACCEPTED MANUSCRIPT 683 Penha-Lopes, G., Bouillon, S., Mangion, P., Macia, A., Paula, J., 2009. Population structure,
684 density and food sources of Terebralia palustris (Potamididae: Gastropoda) in a low
685 intertidal Avicennia marina mangrove stand (Inhaca Island, Mozambique). Estuarine,
686 Coastal and Shelf Science , 84 : 318-325
687 Pillay, D., Branch, G.M., Griffiths, C.L., Williams, C., Prinsloo, A., 2010. Ecosystem
688 change in a South African marine reserve (1960-2009): role of seagrass loss and
689 anthropogenic disturbance. Marine Ecology Progress Series , 415 : 35-48
690 Rachello-Dolmen, P.G., 2013. Biodiversity and historical ecology of marine gastropod
691 assemblages from subtropical Moreton Bay, Queensland, Australia. PhD thesis,
692 University of Queensland, Brisbane.
693 Ries, L., Fletcher, R.J.Jr, Battin, J., Sisk, T.D., 2004. Ecological responses to habitat edges: 694 mechanisms, models, and variability explained.MANUSCRIPT Annual Review of Ecology, Evolution 695 and Systematics , 35 : 491-522
696 Ries, L., Sisk, T.D., 2004. A predictive model of edge effects. Ecology , 85 : 2917-2926
697 Sheridan, P., 1997. Benthos of adjacent mangrove, seagrass, and non-vegetated habitats in
698 Rookery Bay, Florida, U.S.A. Estuarine, Coastal and Shelf Science , 44 : 455-469
699 Skilleter, G.A., Warren, S., 2000. Effects of habitat modification in mangroves on the 700 structure ACCEPTEDof mollusc and crab assemblages. Journal of Experimental Marine Biology and 701 Ecology , 244 : 107-129
702 Strayer D.L., Power, M.E., Fagan, W.F., Pickett, S.T.A., Belnap, J., 2003. A classification of
703 ecological boundaries. Bioscience , 53 : 723-729 32 ACCEPTED MANUSCRIPT 704 Turner, M.G., 2005. Landscape ecology. What is the state of the science? Annual Revue of
705 Ecology, Systematics and Evolution , 36 : 319-344
706 Tuya, F., Vanderklift, M.A., Wernberg, T., Thomsen, M.S., 2011. Gradients in the number
707 of species at reef-seagrass ecotones explained by gradients in abundance. PloS ONE ,
708 6(5); e20190. doi:10.1371/journal.pone.0020190
709 Unsworth, R.K.F., Garrard, S.L., Salinas de Leon, P., Cullen, L.C., Smith, D.J., Sloman,
710 K.A., Bell, J.J., 2009. Structuring of Indo-Pacific fish assemblages along the mangrove-
711 seagrass continuum. Aquatic Biology , 5: 85-95
712 Van Hoey, G., Degraer, S., Vincx, M., 2004. Macrobenthic community structure of soft-
713 bottom sediments at the Belgian continental shelf. Estuarine, Coastal and Shelf Science ,
714 59 : 599-613
715 Vermeij, G.J., 1974. Molluscs in mangrove swamps: MANUSCRIPT physiognomy, diversity, and regional 716 differences. Systematic Zoology , 22 : 609-624
717 Warman, L., Bradford, M.G., Moles, A.T., 2013. A broad approach to abrupt boundaries:
718 looking beyond the boundary at soil attributes within and across tropical vegetation
719 types. PloS ONE , 8(4):e60789. doi:10.1371/journal.pone.0060789.
720 Waycott, M., Duarte, C.M., Carruthers, T.J.B., Orth, R.J., Dennison, W.C., Olyarnik, S., 721 Calladine,ACCEPTED A., Fourqurean, J.W., Heck, K.L. Jr, Hughes, A.R., Kendrick, G.A., 722 Kenworthy, W.J., Short, F.T., Williams, S.L., 2009. Accelerating loss of seagrass across
723 the globe threatens coastal ecosystems. Proceedings of the National Academy of
724 Science of the USA, 106 : 12377-12381 33 ACCEPTED MANUSCRIPT 725 Waycott, M., McKenzie, L.J., Mellors, J.E., Ellison, J.C., Sheaves, M.T., Collier, C.,
726 Schwarz, A-M., Webb, A., Johnson, J.E., Payri, C.E., 2011. Vulnerability of
727 mangroves, seagrasses and intertidal flats in the tropical Pacific to climate change. In:
728 Bell, J.D., Johnson, J.E., Hobday, A.J. (eds), Vulnerability of Tropical Pacific Fisheries
729 and Aquaculture to Climate Change, 297-368. Noumea: Secretariat of the Pacific
730 Community.
731 Wells, F.E., 1983. An analysis of marine invertebrate distributions in a mangrove swamp
732 in northwestern Australia. Bulletin of Marine Science , 33 : 736-744
733 Wells, F.E., 1986. Distribution of molluscs across a pneumatophore boundary in a small bay
734 in northwestern Australia. Journal of Molluscan Studies , 52 : 83-90
735 Whitfield, A.K., 2016. The role of seagrass meadows, mangrove forests, salt marshes and 736 reed beds as nursery areas and food sourcesMANUSCRIPT for fishies in estuaries. Reviews in Fish 737 Biology and Fisheries , doi:10.1007/s11160-016-9454-x.
738 Whittaker, R.H., 1972. Evolution and measurement of species diversity. Science , 147 : 250-
739 260
740 Wong, M.C., Dowd, M., 2015. Patterns in taxonomic and functional diversity of
741 macrobenthic invertebrates across seagrass habitats: a case study in Atlantic Canada.
742 Estuaries and Coasts , 38 : 2323-2336 ACCEPTED 743 Woodroffe, C.D., 2002. Coasts. Form, Process and Evolution, Cambridge, Cambridge
744 University Press.
745
746 34 ACCEPTED MANUSCRIPT 747 Table 1. The four most numerically dominant members of the macrofaunal assemblages
748 (excluding Mictyris and Velacumantus ) and their percentage importance values in each of
749 the three compartments spanning the MSL interface between mangrove and the adjacent
750 lower-shore habitat at each locality.
751
752 CAPEMBAH YERROL DEANBILLA 1 DEANBILLA 2
753 S S Z Z
754 Urohaustorius Mysella Calopia Calopia
755 mertungi (39.9) vitrea (15.2) imitata (19.8) imitata (12.5)
756 Scoloplos Aglaophamus Enigmaplax Enigmaplax
757 cf normalis (22.7) australiensis (9.6) littoralis (11.1) littoralis (12.4)
758 Mysella Scoloplos Monilea Nassarius
759 vitrea (14.4) cf normalis (8.7) callifera (10.1) burchardi (9.1) 760 Koloonella Dasybranchus ?MANUSCRIPTPyrgulina Dasybranchus 761 sp. (4.0) caducus (7.6) sp. (6.6) caducus (8.2)
762 P P ZP ZP
763 Aglaophamus Cerithium Enigmaplax Enigmaplax
764 australiensis (19.4) coralium (12.2) littoralis (13.8) littoralis (17.7)
765 Cerithium Scoloplos Calopia Aglaophamus
766 coralium (16.0) cf normalis (11.3) imitata (9.2) australiensis (8.4) 767 Scoloplos ACCEPTED Iravadia Monilea Iravadia 768 cf normalis (13.0) goliath (9.7) callifera (8.5) goliath (7.7)
769 Dasybranchus Salinator Dasybranchus Dasybranchus
770 caducus (9.7) rhamphidia (8.0) caducus (7.1) caducus (7.1)
771 M M M M 35 ACCEPTED MANUSCRIPT 772 Cerithium Iravadia Cerithium Iravadia
773 coralium (22.9) goliath (15.7) coralium (20.0) goliath (19.0)
774 Indoaustriella Indoaustriella Indoaustriella Indoaustriella
775 lamprelli (17.9) lamprelli (12.9) lamprelli (18.7) lamprelli (10.9)
776 Aglaophamus Cerithium Simplisetia Simplisetia
777 australiensis (8.6) coralium (12.0) aequisetis (9.4) aequisetis (8.3)
778 Iravadia Stenothyra Iravadia Cerithium
779 goliath (8.0) australis (10.7) goliath (6.9) coralium (8.0)
780
MANUSCRIPT
ACCEPTED 36 ACCEPTED MANUSCRIPT Table 2. Summary magnitudes of compartmental biodiversity metrics and similarity coefficients across the two types of interface (overall data per 0.13 m 2). Compartments P or
ZP showing significantly larger or smaller values than one or both of the adjacent
compartments in their values of per-horizon metrics are indicated in bold (one-way ANOVA
F2,15 >6; P <0.02).
S : P : M Z : ZP : M
interface: interface:
Faunal abundance 218 : 103 : 210 465 : 329 : 330
N0 species density 28 : 32 : 40 52 : 56 : 38
Nest species density 29 : 44 : 49 77 : 86 : 46
N300 species richness 30 : 48 : 47 42 : 56 : 39
N2 species diversity 7.8 : 13.9 : 10.5 9.2 : 16.1 : 9.1
J species evenness 0.77 : 0.86 MANUSCRIPT : 0.78 0.72 : 0.83 : 0.75 ∆ taxonomic diversity 4.1 : 4.4 : 4.1 4.1 : 4.4 : 4.2
∆* taxonomic distinctness 4.7 : 4.7 : 4.6 4.6 : 4.7 : 4.7
S17 with lower-shore habitat 0.29 0.55
S17 with mangrove 0.33 0.22
† Vobs with lower-shore habitat 19% 54%
† Vobs with mangrove 21% 18% Vest with lower-shoreACCEPTED habitat 33% 33% Vest with mangrove 26% 25%
†Excluding singleton species and those occurring throughout all three system compartments.
37 ACCEPTED MANUSCRIPT
Legends for Figures
1. Seaward pneumatophore fringes. (A) Representative MSL fringe on North
Stradbroke Island, Queensland (here between an Avicennia marina + Rhizophora stylosa
mangrove and a largely unvegetated sandflat at the Yerrol sampling locality). (B) Detail of
the surface of a fringe supporting seagrass.
2. Diagrammatic representation of the sampling grid across the seaward interface
between mangrove and seagrass or sandflat at each site, showing the horizon numbering
system used in the text.
3. Compositional similarities between faunalMANUSCRIPT assemblages of the various individual system compartments forming the MSL interface between mangrove and adjacent lower-
shore habitats, as indicated by non-metric multidimensional scaling of superimposed group-
average S17 Bray-Curtis values using square-root transformed abundances of the component species. Compartments are clustered at the indicated levels of S17 similarity. (C,
Capembah; Y, Yerrol; D, Deanbilla)
4. Compositional similarities between faunal assemblages of the different sampling horizons at eachACCEPTED locality across MSL interfaces between mangrove and adjacent lower-
shore habitats, as indicated by non-metric multidimensional scaling of superimposed S17
Bray-Curtis clusters using square-root transformed abundances of the component species.
Horizons not differing significantly (ANOSIM P >0.06; R <0.22) are enclosed by 38 ACCEPTED MANUSCRIPT envelopes at the indicated levels of S17 similarity; horizon numbering system is as per Fig.
2.
5. Change in faunal-assemblage composition between adjacent sampling horizons across MSL interfaces between mangrove and lower-shore habitats at the four localities, as measured by S17 Bray-Curtis similarities. The horizon numbering system is as per Fig. 2.
✴✴✴ symbols indicate significant change in assemblage composition between the adjacent horizons concerned (ANOSIM P <0.03; R = 0.11-0.36).
6. Representative curves of species-richness vs assemblage density in the three compartments of MSL (A) mangrove/sandflat (Capembah) and (B) mangrove/seagrass
(Deanbilla 1) interfaces. Note the sampling artefact of apparently homogeneous species richness across all three compartments in samples of < c. 20 individuals (see Barnes, 2016). The equivalent artefact also applies to species MANUSCRIPTdensity in samples of < c. 0.07 m 2 in the pneumatophore-fringe and the lower-shore compartments.
7. Lack of change in basic compositional structure of faunal assemblages across the three compartments of MSL (A) mangrove/seagrass and (B) mangrove/sandflat interfaces, as indicated by standardized Species Importance Distributions. Decay constants ( c) and
coefficients of determination ( R2) of the fitted exponential decay curves are indicated for each distribution.ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED
ACCEPTED MANUSCRIPT • 'Transitional status' of seaward Avicennia pneumatophore fringes re-investigated
• Mangrove : seagrass interface zones basically extensions of the lower-shore seagrass
• But show higher biodiversity, though not abundance, than either adjacent habitat
• Mangrove : sandflat interfaces show most faunal affinity with higher-shore mangrove
• But have lower abundance, though not biodiversity, than either adjacent habitat
• Characteristics of pneumatophore-fringe assemblages are context dependent
MANUSCRIPT
ACCEPTED