Effects of Nutrient Loading and Mercury Chemical Speciation on the

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Effects of nutrient loading and mercury chemical speciation on the formation and degradation of methylmercury in estuarine sediment Van Liem-Nguyen, Sofi Jonsson, Ulf Skyllberg, Mats B. Nilsson, Agneta Andersson, Erik Lundberg, and Erik Björn Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b01567 • Publication Date (Web): 03 Jun 2016 Downloaded from http://pubs.acs.org on June 7, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19

Environmental Science & Technology

1 2

Effects of nutrient loading and mercury chemical speciation on the formation and degradation of methylmercury in estuarine sediment

3 4 5

Van Liem-Nguyen1, Sofi Jonsson1,2,3, Ulf Skyllberg4, Mats B. Nilsson4, Agneta Andersson2,5, Erik Lundberg2, Erik Björn1,*

6 7 8

1

9 10 11 12 13 14

Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden Umeå Marine Sciences Centre, Umeå University, SE-910 20 Hörnefors, Sweden 3 current address: Department of Marine Sciences, University of Connecticut, 1080 Shennecossett Road, Groton, CT06340, USA 4 Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, SE-901 83 Umeå, Sweden 5 Department of Ecology and Environmental Science, Umeå University, SE-901 87 Umeå, Sweden 2

AUTHOR INFORMATION

15

Corresponding author

16

*e-mail: [email protected]; phone +46 90 7865189

17 18

ABSTRACT

19

Net formation of methylmercury (MeHg) in sediments is known to be affected by the availability of

20

inorganic divalent mercury (HgII) and by the activities of HgII methylating and MeHg demethylating

21

bacteria. Enhanced autochthonous organic matter deposition to the benthic zone, following

22

increased loading of nutrients to the pelagic zone, has been suggested to increase the activity of HgII

23

methylating bacteria and thus the rate of net methylation. However, the impact of increased nutrient

24

loading on the biogeochemistry of mercury (Hg) is challenging to predict as different geochemical

25

pools of Hg may respond differently to enhanced bacterial activities. Here, we investigate the

26

combined effects of nutrient (N and P) supply to the pelagic zone and the chemical speciation of HgII

27

and of MeHg on MeHg formation and degradation in a brackish sediment-water mesocosm model

28

ecosystem. By use of Hg isotope tracers added in situ to the mesocosms or ex situ in incubation

29

experiments, we show that the MeHg formation rate increased with nutrient loading only for HgII

30

tracers with a high availability for methylation. Tracers with low availability did not respond

31

significantly to nutrient loading. Thus, both microbial activity (stimulated indirectly through plankton

32

biomass production by nutrient loading) and HgII chemical speciation were found to control the

33

MeHg formation rate in marine sediments.

34

INTRODUCTION

1

ACS Paragon Plus Environment


Page 2 of 19

35

Methylmercury (MeHg) is a neurotoxic compound which biomagnifies in aquatic food-webs causing

36

adverse effects to wildlife and humans. Formation of MeHg from inorganic, divalent Hg (HgII) is

37

mediated by phylogenetically diverse anaerobic microorganisms,1-5 carrying the hgcA and hgcB

38

genes.6 Bacteria are also responsible for degradation of MeHg in environmental compartments with

39

low sunlight exposure, such as in many sedimentary systems.7 Net formation of MeHg under such

40

conditions is driven by multiple factors such as the availability of HgII and MeHg for uptake by

41

methylating/demethylating bacteria8-13 and factors in control of the activity of these bacteria.2,14-15

42

The current perception is that bacteria methylate HgII to MeHg in their cell, after up-take of aqueous

43

forms such as HgII‒sulfide complexes,8 and low molecular mass HgII‒thiol complexes.11-12 The major

44

pools of HgII and MeHg in sediments and soils however, are represented by different solid/adsorbed

45

HgII and MeHg forms with different solubilities and dissolution/desorption rates. Under equilibrium

46

or close to equilibrium conditions the chemical speciation of Hg in solid/adsorbed phases controls

47

aqueous concentrations of HgII and MeHg, and will potentially limit their availabilities for bacterial

48

uptake.13,16 Indeed, Jonsson et al.16 demonstrated that the MeHg concentration in estuarine

49

sediments and biota is controlled by the bioavailability of different geochemical Hg pools, which in

50

turn is largely controlled by the solid/adsorbed phase chemical speciation and vertical localization of

51

Hg in the sediment.

52

The rate of nutrient supply, e.g. via catchment runoff, as well as temperature and light conditions are

53

principal factors controlling primary production in aquatic ecosystems and thus the rate of formation

54

and deposition rate of autochthonous natural organic matter (NOM) to the sediment. The deposition

55

of autochthonous NOM will in turn control the activity of saprotrophic microbial organisms,17-18

56

including bacteria with the capability to methylate HgII. Observations made in the field have in

57

several studies been interpreted as a direct or indirect link between autochthonous production of

58

biomass and MeHg net formation in marine and estuarine ecosystems. Two hypotheses have

59

emerged from these studies: 1) The sediment and pore water partitioning coefficient, Kd (L kg-1), for

60

HgII (and MeHg) is controlled by, and increases with, the amount of NOM in the sediment.19-22 With

61

increased Kd the pore water concentration of HgII decreases and thus the presumed forms available

62

for uptake by methylating bacteria also decrease. The chemical speciation of HgII in solid/adsorbed

63

phases was however not established in these studies and recent reports have highlighted that the

64

type and concentration of dissolved organic matter strongly influences the relationship between

65

sediment NOM content and Kd of Hg.22-24 It therefore remains to develop a comprehensive chemical

66

speciation model that quantitatively describes the solubility of HgII (and MeHg) under or close to

67

equilibrium conditions in marine sediments. 2) The second hypothesis proposes that the activity of

68

benthic anaerobic bacteria, including those with the capacity to methylate HgII, is controlled by the

2


Page 3 of 19


69

availability of metabolic electron donors.10,23,25-26 These electron donors are largely constituted by

70

short-chain fatty acids originating predominantly from autochthonous NOM deposited to the

71

sediment. This hypothesis has been used to explain observed increases in HgII methylation rate or

72

MeHg concentration with amount25 or the fraction of autochthonous algal NOM (manifested by a low

73

C/N ratio10,26). These two hypotheses are not necessarily alternatives, as both processes may take

74

place in parallel, yet different studies have suggested opposing net effects on MeHg formation from

75

increases in pelagic biomass production and subsequent organic carbon content or C/N ratio in

76

sediments.

77 78

It is a challenge to experimentally investigate the combined effects of nutrient loading and chemical

79

speciation of HgII on methylation rates. Mesocosm experiments, in which natural environmental

80

conditions can be simulated and factors affecting MeHg formation controlled, have proven powerful

81

in addressing these issues. Using mesocosms, Jonsson et al. [Jonsson et al. in prep] demonstrated

82

that the net methylation of HgII bonded to thiol groups in NOM, or as metacinnabar (β-HgS), in an

83

estuarine sediment was enhanced by 10‒40% in response to a twofold increase in pelagic biomass

84

production caused by enhanced nutrient loading. Here, using a similar system, we further investigate

85

how rates of methylation and demethylation of different Hg geochemical pools are affected by the

86

nutrient loading (N and P) to the pelagic zone. We hypothesized that Hg pools with a low availability

87

to methylating bacteria will show less of a response to nutrient loading in comparison to Hg pools

88

with a higher availability. Sediment-brackish water mesocosm model ecosystems (n=6, 5 m high,

89

0.75 m ⌀ cylinders) were constructed and rates of HgII methylation were compared for two

90

principally different types of enriched Hg isotope tracers. One set of tracers were added in situ to the

91

mesocoms, either as defined pre-equilibrated solid/adsorbed chemical forms where HgII and MeHg

92

was bonded to sulfide and/or thiols, or as dissolved complexes which were equilibrated in the

93

mesocoms’ water phase prior to deposition to the sediment. A second set of tracers were added ex

94

situ to sediment sampled from the mesocosm sediments as non-equilibrated labile dissolved

95

complexes. Nutrients were added to the water column at low to moderate concentrations during

96

weeks 1‒2, and at high concentrations during weeks 2‒4 of the experiment to study the short-term

97

effects of increased nutrient loading on MeHg formation/degradation for different chemical forms of

98

Hg.

99 100

EXPERIMENTAL SECTION

101

Sampling of sediment and water. The study site, Öre river estuary (located in the Bothnian Sea at

102

the Sweden east coast), has been described in detail in a previous study.16 Intact sediment cores 3



Page 4 of 19

103

were manually sampled from the estuary by divers at site 63° 33.905′ N, 19° 50.898′ E, at a water

104

depth of 5‒7 m using custom-made sampling devices. The sampling cylinders (~0.2 m × 0.63 m ⌀)

105

were made of high density polyethylene (HDPE) with a detachable bottom and lid. The cylinders

106

were immersed into undisturbed sediment, and surrounding sediment was cleared to insert the

107

bottom plate and lid of the core sampler. The intact sediment cores were stored (in the dark, at 15

108

°C) in barrels filled with brackish water to avoid leakage and to avoid direct contact with air for up to

109

2 months. The mesocosm cylinders were filled with 2000 L of unfiltered brackish water (salinity of 4

110

Practical Salinity Units) 3 days before the start of the experiment. Water was collected using a

111

peristaltic pump system with two pipes located 800 m from land and at a water depth of 2 and 8 m (a

112

50:50 mixture of water from these two depths were used). To assure that all mesocosms had similar

113

distribution of planktonic organisms, water was added to all tanks in parallel.

114

Mesocosm system and setup. The experiment was conducted at a mesocosm facility located at the

115

Umeå Marine Sciences Center. This facility includes 12 double-mantled HDPE tubes (5 m × 0.75 m ⌀)

116

regularly used for studies of the pelagic food web ecology. For this experiment, 6 of the systems

117

were utilized. Three of them included collected sediment cores and a 5 m high water column. The

118

remaining three were used to study the deposition of organic matter from the pelagic zone, using

119

sedimentation traps placed at the bottom of the mesocosm tanks, and included a 5 m water column.

120

Temperature of the water column was maintained at 15 oC in the upper 3.5 m and 10 oC in the lower

121

part (below 3.5 m) during the course of the experiment, via an outer glycol layer. The convection of

122

the upper part of the water column was obtained by purging with ~20 mL s-1 air at a depth of 3.2 m

123

from the water surface. 150 W metal halogen lamps (Master colour CDM-T 150w/942 G12 1CT) were

124

used as light sources with 12/12 hour on/off cycles. When turned on, the lamps resulted in average

125

photon flux through the water column of 410, 57 and 2.7 µmol s-1 m-2 at ~0.05 m, 1 m and 4 m depth,

126

respectively. Further technical specifications of mesocosm system compartments are given in Table

127

S1. The total experiment time was 30 days from 9th February (referred to as day 1) to 10th March (day

128

30). Nutrients (nitrate (NO3-), phosphate (PO43-) and ammonium (NH4+) prepared from salts of NaNO3,

129

NaH2PO4×H2O and NH4Cl) were added at moderate levels (corresponding to 19.5 µg L-1 NO3-, 3.49 µg

130

L-1 NH4+ and 3.33 µg L-1 H2PO4- in the 2000 L mesocosm water phase) equivalent to 3-10% of the

131

concentrations typical for winter conditions in the Bothnian Sea (600 µg L-1 NO3-, 30 µg L-1 NH4+ and

132

70 µg L-1 H2PO4-)27 during the first two weeks of the experiments (additions were done at day 1, 8 and

133

11). The nutrient concentrations were increased by a factor of five (corresponding to 97.5 µg L-1 NO3-,

134

17.5 µg L-1 NH4+ and 16.7 µg L-1 H2PO4- in the 2000 L mesocosm water phase) during the last two

135

weeks of the experiments to induce a plankton bloom effect (additions were done at day 15 and 22).

4


Page 5 of 19


136

The in situ net methylation and demethylation of Hg were measured using five HgII or MeHg tracers

137

added to the water column (denoted

138

The tracers were synthesized from HgII enriched

139

(96.41%),

140

Laboratory (TN, USA). β‒201HgSsed and isotopically enriched MeHg tracers were synthesized as

141

described by Jonsson et al.13 and Snell et al.,28 respectively. The 200HgII‒NOMsed and Me198Hg‒NOMsed

142

tracers were prepared by adding

143

previously characterized by Skyllberg and Drott.29 The Hg/RSH molar ratio was kept in the range

144

giving 1:2 HgII:RSH and 1:1 MeHg:RSH complex stoichiometry.30 A slurry mixture of β‒201HgSsed,

145

200

146

and injected at 1 cm depth into the ~0.2 m deep × 0.63 m ⌀ intact sediment cores one day prior to

147

the start of the experiment using an electronic 12 channel pipette (VWR, for 10‒200 µL). The

148

injections were made with help of a custom made grid system which was placed above the sediment

149

surface. Totally 3384 injections were done per sediment core, resulting in 113 injections per dm2

150

(Table S2) covering 95% of the sediment surface. Sediment sections not covered by addition of tracer

151

were in the outer part (close to the wall) and no sediment sub-samples were taken from these areas

152

during the experiment. The intact sediment cores were then placed into the water filled mesocosm

153

with the lid retained during the placement to protect the sediment surface. This is denoted day 1 of

154

the experiment. The average ambient HgII and MeHg concentrations in the sediment were 170 pmol

155

g-1 and 2.9 pmol g-1 (dry weight), respectively. The added concentration of HgIIsed (sum of β‒201HgSsed

156

and

157

HgII and MeHg, respectively. Aqueous stock solutions of Me199Hgwt and

158

extraction from toluene28 and dissolution of the oxide salt, respectively in 0.1 M HCl. A 20 mL mixture

159

of Me199Hgwt and

160

immediately added to the water column at day 1 and 15 with one aliquot of 13.3 mL at 1.5 m depth

161

and 6.7 mL at 4 m depth from the water surface (Table S3). The two additions were made to sustain

162

the concentration of Hg in the water phase, and were based on a previously observed average

163

residence time of 10-15 days for Hg in the mesocosm water column.16

164

Sampling and analysis. The sampling and analytical measurement procedures applied here have

165

been previously described in detail by Jonsson et al.16 Briefly, water samples were collected during

166

the experiment at specific depths using teflon tubing and peristaltic pumps. These samples were

167

then used to measure photosynthetic primary and bacterial production rate (by incorporation

168

experiments of NaH14CO3 and [3H-methyl]-thymidine, respectively), HgII methylation rate constants

169

(km), MeHg demethylation rate constants (kd) and concentrations of chlorophyll a (Chl α), DOC,

201

Hg (98.11%) and

204

wt

tracers) or injected into the sediment (denoted sed tracers). 196

Hg (50%),

198

Hg (92.78%),

199

Hg (91.95%),

200

Hg

Hg (98.11%) (as HgO or HgCl2) purchased from Oak Ridge National

200

HgII(aq) and Me198Hg(aq) to a homogenized humic soil extract

HgII‒NOMsed and Me198Hg‒NOMsed (2.59 µM, 3.89 µM and 0.25 µM, respectively) was prepared

200

HgII‒NOMsed) and Me198Hg‒NOMsed tracers (Table S2) constituted 50% and 130% of ambient

204

204

HgIIwt were prepared by

HgIIwt (0.22 µM and 2.24 µM, respectively) was diluted with MQ water and

5



Page 6 of 19

170

nutrients and MeHg. Pressure, turbidity and O2 saturation were measured in situ in the water column

171

using a Seaguard CTD with attached microelectrode sensors (Unisense), and light penetration was

172

measured using a Spherical Quantum Sensor (LI-COR, 193-SA and LI-COR 1400 unit). Sediment sub-

173

cores (diameter of 4.2 cm) were sampled during the experiment using a custom made sampler

174

designed for the mesocosm system.16 Sediment was sampled for determination of km and kd, dry

175

weight, concentrations of total C, total N, HgT and MeHg. The concentrations of MeHg in water and

176

of HgT and MeHg in sediment were determined using

177

isotope dilution analysis by inductively coupled plasma mass spectrometry (ICPMS) for HgT and Gas

178

Chromatography‒ICPMS analysis for MeHg.31 Concentrations of HgT and MeHg for ambient Hg and

179

tracers were then calculated from mass bias corrected signals using signal deconvolution.32

180

Determination of km and kd (ex situ experiments). The

181

the water column in situ were also added to sediment slurries incubated ex situ to determine the km

182

and kd, respectively). The sediment slurries were prepared from the top 2 cm of sediment sub cores

183

collected from the mesocosm during the experiment. The concentration of 204HgII and Me199Hg added

184

ex situ were on average 7.0 and 0.60 pmol g-1 (d.w.), respectively, and exceeded the concentration of

185

204

186

(d.w.) for

187

simultaneously, t0 and t48. The ex situ incubations were terminated by freezing directly after the

188

addition of ex situ tracer for t0 samples and after incubating for 48 hours under a N2(g) atmosphere

189

(using a glove box) and dark condition for t48 samples. The amount of MeHg from the 204HgIIwt tracers

190

added in situ (which deposited from the water column and methylated in the sediments) in t0

191

samples was subtracted from the amount formed from the tracer added ex situ in t48 samples. The kd

192

and km were calculated using equations 1 and 2.

193

km = ([Me204Hg]t48 - [Me204Hg]t0) / ([204HgII]added × time of incubation)

(d-1)

(1)

194

kd = -1 × (ln[Me199Hg]t48 - ln[Me199Hg]t0) / time of incubation

(d-1)

(2)

195

where subscripts “t48” and “t0” refer to measured MeHg concentrations at 48 h and 0 h of

196

incubation, respectively, and subscript “added” refers to the concentration of HgII added to sediment

197

in the incubation assay.

196

204

HgII or Me196Hg as internal standard for

Hg(aq) and Me199Hg(aq) tracers added to

HgII and Me199Hg already present in the sediment from in situ added tracers (3.4 and 0.25 pmol g-1 204

HgII and Me199Hg, respectively). Two sets of sediment slurries were prepared

198 199

RESULTS AND DISCUSSION

200

Pelagic productivity and sedimentation of organic matter to the benthic zone. Selected key

201

ecological and biogeochemical factors describing the model ecosystems are given in Table S4. At the

6


Page 7 of 19


202

start of the experiment, rates of photosynthetic primary biomass production and bacterial biomass

203

production and the concentration of Chl α were 0.7 and 4.4 µmol C dm-2 h-1 and 1.1 µg dm-3,

204

respectively. These conditions corresponded well with those typically observed during mid-winter

205

with ice cover in the Bothnian Sea.33 Both primary production and the concentration of Chl α in the

206

mesocosm water phase (Figure 1a) responded quickly to the combined treatment of light exposure

207

and nutrient addition and reflected growth of phytoplankton. At day 22, the primary production and

208

the concentration of Chl α reached a maximum and corresponded to typical spring bloom conditions

209

in the Bothnian Sea (primary production of ~8 µmol C dm-2 h-1)27. After day 22 both variables

210

decreased despite a maintained nutrient loading, which may be explained by the growth rate of

211

predatory zooplankton which commonly increases subsequent to phytoplankton growth, generating

212

a predator-prey dynamic.34 The pelagic bacteria biomass production rate remained fairly constant at

213

4.1 ± 1.1 (given as average ± CI throughout the text unless otherwise stated, p=0.05, n=12) µmol C

214

dm-2 h-1 throughout the experiment. The structure of the food web was thus shifted during the

215

experiment, from a bacteria dominated (heterotrophic) to a photosynthetic primary production

216

dominated (autotrophic) web, in response to the mesocosm light and nutrient addition treatment.

217

Following the increases in primary production rate and Chl α concentration, the C/N molar ratio in

218

the material collected in sedimentation traps (placed at the bottom of triplicate mesocosms without

219

sediment) decreased from 10.7 ± 0.3 during weeks 0–2 to 8.6 ± 0.9 (p=0.05, n=6) during weeks 2–4 of

220

the experiment. This decrease in C/N molar ratio suggested an increased fraction of autochthonous

221

over allochthonous NOM in the deposited material during the second half of the experiment.35 As a

222

comparison, the C/N molar ratio in the top 0‒5 cm of the sediment was 11.4 ± 0.3 (p=0.05, n=12).

223

The total amount of deposited organic carbon during the 28 days of the experiment constituted 0.1%

224

of the total organic carbon in the top 0-2 cm of the sediment. It is however expected that this

225

fraction is important for controlling the activity of saprotrophic microbial organisms,17-18 including

226

bacteria with the capability of methylating HgII. The redox conditions of the bulk sediment, below an

227

oxidized surface layer of 1‒2 mm (yellowish-brownish color by visual inspection), was in the

228

ferruginous to low sulfidic range with a concentration of dissolved sulfide in the pore water of < 0.05

229

to 1.0 µM. This is in good agreement with previous studies on sediment from the same area.13,16,25

230 231

Effect of nutrient loading on rates of methylation and demethylation of labile Hg tracers added ex

232

situ. Benthic anaerobic bacteria primarily utilize metabolic electron donors originating from

233

autochthonous rather than allochthonous NOM.17-18 In line with this, high MeHg formation rates

234

have been related to low C/N molar ratios in estuarine sediment,10,26 suggesting that the activity of

235

mercury methylating bacteria could be driven by autochthonous NOM. As described above, the 7



Page 8 of 19

236

primary production and Chl α concentration increased during the second half of the experiment, and

237

the C/N molar ratio was indicative of an increased proportion of planktonic material in the material

238

deposited to the benthic zone in the last two weeks of the experiment. The HgII methylation rate

239

constant (km) for the tracer (dissolved labile Hg(OH)2(aq) complex) added ex situ to sediment

240

subsamples increased in parallel with increasing primary production and Chl α concentration (Figure

241

1b). The constant increased by a factor of four from day 9 to day 16 followed by a decrease after day

242

23. A significant correlation (R2=0.96) was found between the Chl α concentration and km. The MeHg

243

demethylation rate constant (kd) in the sediment increased slightly towards the end of the

244

experiment, when primary production rates and Chl α concentrations decreased after day 23. Overall

245

kd was only moderately affected by the treatments and varied less than 40% during the 21 days’ time

246

period covered by measurements (day 9‒30), while km increased by a factor of 3.9. When HgII and

247

MeHg tracers are added as labile dissolved complexes to determine km and kd in short term

248

incubation experiments, they typically have a higher availability for transformation reactions as

249

compared to ambient Hg or the type of solid/adsorbed tracers added in situ to the mesocosms in this

250

study.13,16 The results in Figure 1 therefore suggest that the methylation of HgII chemical forms having

251

a high availability for methylation is rapid and largely responds to changes in pelagic primary

252

production rate and Chl α biomass following altered nutrient loading rate. These results also highlight

253

that although the amount of freshly deposited autochthonous organic carbon constituted only a

254

small fraction (0.1%) of the total organic carbon in the surface sediment, it significantly enhanced

255

MeHg formation.

256 257

Effect of nutrient loading on net methylation of Hg tracers added in situ. Net methylation of tracers

258

added in situ (injected in the sediment at a depth of 1 cm; β‒201HgSsed,

259

NOMsed, and added to the water column; 204HgIIwt and Me199Hgwt) was determined as the MeHg/HgII

260

molar ratio16. The average net methylation was significantly different for the different HgII tracers

261

(ANOVA, p

Effects of Nutrient Loading and Mercury Chemical Speciation on the

Recommend Documents