process safety and environmental protectiondownload.xuebalib.com/xuebalib.com.31962.pdf · process...

9
Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530 Contents lists available at ScienceDirect Process Safety and Environmental Protection journal h om ep age: www.elsevier.com/locate/ps ep Seasonal population changes in the ammonia-oxidizing bacteria community structure of Songhua Lake, China Xinyu Zhao a,b,c , Yue Zhao a , Beidou Xi b , Zimin Wei a,, Junqiu Wu a,b , Taozhi Zhao a a Life Science College, Northeast Agricultural University, Harbin, China b State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing, China c College of Water Sciences, Beijing Normal University, Beijing, China a r t i c l e i n f o Article history: Received 20 December 2015 Received in revised form 13 April 2016 Accepted 25 April 2016 Available online 18 May 2016 Keywords: Ammonia-oxidizing bacteria (AOB) Community structure Seasonal variation Redundancy analysis (RDA) DGGE Songhua Lake a b s t r a c t Ammonia-oxidizing bacteria (AOB) and the effects of environmental characteristics on the AOB community distribution were investigated in the Songhua Lake in May, August, October and December of 2011. Profiles of the AOB communities were generated using denaturing gradient gel electrophoresis (DGGE) to assess the expression of 16S rRNA genes followed by DNA sequence analysis. The dominant AOB groups were affiliated with Nitrosomonas and Nitrosospira spp. The effects of environmental characteristics on the AOB commu- nity distribution were analyzed by the ordination technique of redundancy analysis. The environmental characteristics significantly influencing the AOB community structure were different in the four seasons. The suspended solids (SS), ammonia nitrogen (NH 4 + -N) and nitrate nitrogen (NO 3 -N) differed in May, the pH, NH 4 + -N, NO 3 -N and total nitrogen (TN) differed in August, and the NH 4 + -N, DO and TN differed in December. None of the assessed environmental variables displayed heterogeneity in October. This study suggested that sea- sonal variation and nutrient differences might be responsible for the differences in the key environmental characteristics used to determine the distribution of the AOB communities in all seasons. © 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved. 1. Introduction Nitrification is a critical pathway in the nitrogen cycle and plays a key role in the biogeochemical cycling of nitrogen (Casciotti et al., 2003). It is an important process required to remove excess ammonia to nitrate via nitrite (Camargo and Alonso, 2006). Excess ammonia can stimulate the growth of ammonia-oxidizing bacteria (AOB). This process catalyzes the oxidization of ammonia (NH 4 + -N) to nitrite (NO 2 -N), which is the first and rate-limiting step of nitrification Corresponding author at: Life Science College, Northeast Agricultural University, Harbin 150030, China. Tel.: +86 045155190413; fax: +86 018745724658. E-mail address: [email protected] (Z. Wei). (Arp and Stein, 2003). AOB have traditionally been taxono- mically classified into two monophyletic groups: beta- and gamma-proteobacteria. Gamma-proteobacteria are the minor member of the ammonia-oxidizing community (Ward, 2005). They have been observed in tight clusters within the beta- proteobacteria based on analyses of 16S rRNA sequences with Nitrosomonas and Nitrosospira members (Dang et al., 2010; Phillips et al., 2000). A number of studies have suggested that the varia- tion in the AOB community composition is a response to http://dx.doi.org/10.1016/j.psep.2016.04.020 0957-5820/© 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Upload: others

Post on 03-Feb-2021

1 views

Category:

Documents


0 download

TRANSCRIPT

  • Sao

    XTa

    b

    Ec

    a

    A

    R

    R

    2

    A

    A

    K

    A

    C

    S

    R

    D

    S

    1

    Np(taotw

    f

    h0

    Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530

    Contents lists available at ScienceDirect

    Process Safety and Environmental Protection

    journa l h om ep age: www.elsev ier .com/ locate /ps ep

    easonal population changes in themmonia-oxidizing bacteria community structuref Songhua Lake, China

    inyu Zhaoa,b,c, Yue Zhaoa, Beidou Xib, Zimin Weia,∗, Junqiu Wua,b,aozhi Zhaoa

    Life Science College, Northeast Agricultural University, Harbin, ChinaState Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy ofnvironmental Sciences, Beijing, ChinaCollege of Water Sciences, Beijing Normal University, Beijing, China

    r t i c l e i n f o

    rticle history:

    eceived 20 December 2015

    eceived in revised form 13 April

    016

    ccepted 25 April 2016

    vailable online 18 May 2016

    eywords:

    mmonia-oxidizing bacteria (AOB)

    ommunity structure

    easonal variation

    edundancy analysis (RDA)

    GGE

    a b s t r a c t

    Ammonia-oxidizing bacteria (AOB) and the effects of environmental characteristics on the

    AOB community distribution were investigated in the Songhua Lake in May, August, October

    and December of 2011. Profiles of the AOB communities were generated using denaturing

    gradient gel electrophoresis (DGGE) to assess the expression of 16S rRNA genes followed

    by DNA sequence analysis. The dominant AOB groups were affiliated with Nitrosomonas

    and Nitrosospira spp. The effects of environmental characteristics on the AOB commu-

    nity distribution were analyzed by the ordination technique of redundancy analysis. The

    environmental characteristics significantly influencing the AOB community structure were

    different in the four seasons. The suspended solids (SS), ammonia nitrogen (NH4+-N) and

    nitrate nitrogen (NO3−-N) differed in May, the pH, NH4+-N, NO3−-N and total nitrogen (TN)

    differed in August, and the NH4+-N, DO and TN differed in December. None of the assessed

    environmental variables displayed heterogeneity in October. This study suggested that sea-

    sonal variation and nutrient differences might be responsible for the differences in the key

    environmental characteristics used to determine the distribution of the AOB communities

    onghua Lakein all seasons.

    © 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

    A number of studies have suggested that the varia-

    . Introduction

    itrification is a critical pathway in the nitrogen cycle andlays a key role in the biogeochemical cycling of nitrogen

    Casciotti et al., 2003). It is an important process requiredo remove excess ammonia to nitrate via nitrite (Camargond Alonso, 2006). Excess ammonia can stimulate the growthf ammonia-oxidizing bacteria (AOB). This process catalyzeshe oxidization of ammonia (NH4+-N) to nitrite (NO2−-N),

    hich is the first and rate-limiting step of nitrification

    ∗ Corresponding author at: Life Science College, Northeast Agricultural ax: +86 018745724658.

    E-mail address: [email protected] (Z. Wei).ttp://dx.doi.org/10.1016/j.psep.2016.04.020957-5820/© 2016 Institution of Chemical Engineers. Published by Elsev

    (Arp and Stein, 2003). AOB have traditionally been taxono-mically classified into two monophyletic groups: beta- andgamma-proteobacteria. Gamma-proteobacteria are the minormember of the ammonia-oxidizing community (Ward, 2005).They have been observed in tight clusters within the beta-proteobacteria based on analyses of 16S rRNA sequences withNitrosomonas and Nitrosospira members (Dang et al., 2010;Phillips et al., 2000).

    University, Harbin 150030, China. Tel.: +86 045155190413;

    tion in the AOB community composition is a response to

    ier B.V. All rights reserved.

    http://www.sciencedirect.com/science/journal/09575820www.elsevier.com/locate/psephttp://crossmark.crossref.org/dialog/?doi=10.1016/j.psep.2016.04.020&domain=pdfmailto:[email protected]/10.1016/j.psep.2016.04.020

  • 524 Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530

    ake

    Fig. 1 – Map of Songhua L

    physiological and ecological differences, such as pH, salinityand ammonia concentration in estuarial coastal and deep-seaecosystems (Cilia et al., 1996; de Bie et al., 2001). However,the partial information from these different ecological nichescould not explain the distribution in the AOB communitystructure that responded to environmental changes in com-plex environments, such as anthropogenic activity-impactedlakes. Furthermore, information on the relationships betweenthe environmental characteristics and the AOB communitystructure is not available.

    Songhua Lake is the third largest artificial lake in China.Eutrophication has become one of the most severe problemsin Songhua Lake since 2008 (Shi et al., 2008). Due to the notice-able temperate monsoon climate changes that occur over thefour seasons, seasonally alternations in the climatic condi-tions control the distribution and biodiversity of microbes. Thediversity and composition of the AOB and environmental char-acteristics must also be affected by the seasonal changes (Simset al., 2012). Therefore, it is necessary to determine the sea-sonal population changes in the AOB community compositionand their relationship with the water quality.

    Recent studies on the seasonal variation of ammonifyingand denitrifying bacteria in Songhua Lake have provided apartial understanding of N biogeochemical cycling along agradient of the surface layers in Songhua Lake (Zhao et al.,2015). However, although partial knowledge is emerging, thecorrelation between environmental characteristics and AOBcommunity structure, abundance and distribution across thefour seasons is not well understood in a comprehensive way.Thus, research was carried out to determine to what extentthe environmental characteristics affect the AOB communitycomposition with seasonal population changes.

    In order to achieve a deep understanding of the AOB pro-cess related to water quality and climatic variations, this studyaimed to determine whether changes in the AOB communitycomposition could be related to the water quality and to differ-ences in key environmental characteristics caused by seasonalchanges in the hydrodynamic conditions.

    2. Materials and methods

    2.1. Sample collection and environmentalcharacteristic measurements

    Seven sample sites were established in Songhua Lake, whichis a component of a large drainage system with a catchment

    area that comprises 22.7% of Jilin. This body of water pos-sesses an average water-carrying capacity of 133.2 × 106 m3

    with sampling locations.

    and a depth of 35 m. The sample sites were designated fromupstream to downstream in Songhua Lake based on the distri-bution of the water system (Fig. 1). In 2011, freshwater sampleswere collected in triplicate from the main affluent, whichwas divided into three sections: an upstream (S1, S2), middlestream (S3, S4, S5) and downstream (S6, S7) section in spring(May), summer (August), fall (October), and winter (December),respectively. The deepest point of Songhua Lake is 70 m. Allsamples were collected from the middle level (at a depth of16 m) of the lake, as the surface water (0 m) was vulnerableto various natural factors, including the weather. The bottomlevel of the water contains little oxygen, and the AOB have dif-ficulty surviving in low-oxygen conditions. Approximately 5 Lof water was collected for chemical and molecular analyses.The samples were then rapidly transferred to the laboratory onice. Water samples (200 mL) were filtered through filters with apore size of 0.22 �m and a diameter of 45 mm (Durapore). Thefilters were stored at 20 ◦C until DNA extraction for molecularbiological analysis.

    The temperature, pH, DO concentration, BOD5 and sus-pended solids (SS) were constantly monitored in situ using amultiparameter water quality probe (6600EDS, YSI, USA). Thetotal nitrogen (TN), ammonia nitrogen (NH4+-N), nitrite nitro-gen (NO2−-N), and nitrate nitrogen (NO3−-N) were measuredaccording to a published procedure (Gilliam et al., 2001).

    2.2. DNA extraction, PCR amplification and DGGEanalysis

    DNA extraction was performed using method C with ultra-sonic pretreatment according to a published procedure(Zhang et al., 2011). Briefly, the DNA extraction procedure wasperformed in accordance with a method described by Zhouet al. (1996). High-energy sonication (HES, with a constantfrequency of 20 kHz and 40 W power input) was applied tothe sample for 2 min (Magic-Knezev and Van Der Kooij, 2004).Nest amplification of the PCR products was achieved usingthe AOB 16S rRNA gene fragments as templates. The firstround of amplification was conducted using primers pA (5′-AGAGTTTGATCCTGGCTCAG-3′) and pH (5′-AAGGAGGTGATC-CAGCCGCA-3′) (Rowan et al., 2003; Salles et al., 2004) followedby amplification with the AOB specific primers CTO189f(5′-CCGCCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGAGRA-AAGYAGGGGATCG-3′) and CTO654r (5′-CTAGCYTTGTAGTTT-CAAACGC-3′). These primers were designed to amplify partial16S rDNA sequences (465 bp) from the beta-proteobacteria

    AOB (Kowalchuk et al., 1997). PCR was conducted using anOmn-E programmable thermal cycler (Hybaid Ltd., Middlesex,

  • Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530 525

    Table 1 – Environment characteristics in May, August, October and December, respectively.

    Samples Temperature ( ◦C) pH SS DO BOD5 TN NO3−-N NO2−-N NH4+-N

    MayS1 14.1 7.29 120 10.65 3.25 2.34 2.03 0.0056 0.21S2 14.2 7.46 125 10.80 3.23 2.40 2.03 0.0081 0.25S3 14.2 7.25 105 10.96 2.46 2.40 1.91 0.0087 0.10S4 14.1 7.48 100 10.73 2.56 2.28 1.92 0.0132 0.11S5 13.9 7.26 87 10.68 2.97 2.03 1.83 0.0329 0.18S6 14.0 7.43 83 10.73 2.25 1.94 1.68 0.0459 0.19S7 14.1 7.30 98 10.55 2.66 1.92 1.66 0.0391 0.08

    AugustS1 24.6 9.05 170 8.00 1.93 2.13 2.10 0.0000 0.22S2 24.8 8.85 140 7.93 1.58 2.60 2.55 0.0000 0.23S3 24.9 7.26 142 7.44 1.44 1.87 1.83 0.0001 0.00S4 24.8 7.22 115 5.66 1.56 2.11 2.08 0.0001 0.01S5 25.2 7.48 142 5.94 1.26 2.09 2.07 0.0001 0.00S6 25.3 7.30 127 7.45 1.73 2.62 2.58 0.0002 0.01S7 24.5 7.34 45 5.01 1.12 2.03 2.00 0.0000 0.01

    OctoberS1 9.2 6.380 205 8.70 1.32 2.39 2.21 0.0003 0.03S2 10.2 6.520 130 7.93 1.60 2.35 2.19 0.0006 0.03S3 9.8 6.350 134 5.92 1.61 2.34 2.22 0.0004 0.03S4 10.2 6.330 136 8.98 1.21 2.22 2.21 0.0000 0.01S5 9.8 6.320 130 8.06 1.34 2.41 2.22 0.0003 0.03S6 10.2 6.360 156 7.90 1.27 2.15 2.14 0.0004 0.02S7 10.8 6.410 171 5.20 1.34 2.42 2.01 0.0003 0.06

    DecemberS1 4.0 6.120 210 12.34 1.89 3.02 2.52 0.0001 0.10S2 3.0 6.040 135 12.67 1.96 3.06 2.54 0.0001 0.09S3 4.1 6.890 136 12.96 1.46 3.04 2.45 0.0002 0.11S4 4.1 6.340 150 12.67 1.57 2.92 2.47 0.0001 0.12S5 3.9 6.120 165 12.19 2.08 2.99 2.40 0.0002 0.10S6 2.6 6.330 160 12.97 2.17 2.94 2.37 0.0000 0.11S7 4.0 6.250 170 11.98 1.90 2.92 2.35 0.0001 0.12

    UteUppKiiddaTfsDgac

    2

    TtycmpbU

    K). Different cycling parameters were used depending onhe primers employed (Table 1). The final PCR products werexamined by agarose gel electrophoresis (1.5% agarose, 0.5TBE), and stained with ethidium bromide to confirm theroduct size for visualization via UV illumination. DGGE waserformed following the protocol described and modified byowalchuk et al. (1998). Polyacrylamide gels (8% polyacrylam-

    de; thickness, 1.5 mm; dimensions, 16 cm × 16 cm) were runn 1× TAE buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8). Theenaturant gradient ranged from 30% to 60%, in which a 100%enaturant was used, comprised of 7 M urea and 40% form-mide in 1× TAE buffer as described by Muyzer et al. (1993).he gels were run at 60 ◦C on DcodeTM (Bio-Rad Laboratories)

    or 4 h at 150 V. DNA was stained for 30 min in the stainingolution and visualized by UV transillumination. The centralGGE bands were excised for DNA reamplification. Fourteenel fragments were placed overnight at 4 ◦C. DGGE bands weremplified without the GC clamp CTO primer according to theonditions described.

    .3. Analysis of the DGGE banding pattern

    he DGGE patterns were visually analyzed, and the intensity ofhe bands was estimated through imaging and software anal-sis (Quantity One, Bio-Rad). Hierarchical cluster analysis wasonducted using the unweighted pair-group method of arith-etic averages (UPGMA). Community similarities in the band

    atterns were calculated using the Dice coefficients. Similar

    anding patterns were graphically observed in the form ofPGMA, which resulted in the construction of dendrograms.

    2.4. Multivariate statistical analysis

    The relationship between ammonia-oxidizing bacteriaand the physicochemical properties was performed usingCANOCO for Windows 4.5 (Biometris, Netherlands). Detrendedcorrespondence analysis (DCA) was carried out first in orderto select a linear or unimodal response for the microbialdata. The lengths of the first DCA ordination axis for theammonia-oxidizing bacteria based on the DGGE profiles were0.135, 0.077, 0.097, and 0.094, respectively. Therefore, theredundancy analysis (RDA) was performed to ordinate thecompositions of the ammonia-oxidizing bacteria communityin order to identify species-environment correlations.

    An automated forward selection was used to analyzeinter-sample distances for the DGGE profiles. First, the vari-ance inflation factor for the environmental variables wascalculated. Variables displaying a value greater than 20 wereexcluded from CCA analyses, based on an assumed co-linearity of the respective variable with the other variablesincluded in the examined dataset.

    The analysis was performed without transformation ofthe data or focus scaling of the inter-sample distances. Themanual selection of environmental variables through a par-tial Monte Carlo permutation test (499 permutations) with anunrestricted permutation was performed to investigate thestatistical significance (Salles et al., 2004). The marginal effectsof the environmental variables were selected according totheir significance level (P < 0.05) prior to permutation. Ordina-

    tion biplots including the environmental variables and DGGEsamples (lanes) were used to explain our data.

  • 526 Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530

    omm

    Fig. 2 – DGGE profiles showing the AOB c

    2.5. Nucleotide sequence accession numbers

    All sequences from this study have been deposited in theGenBank database under accession numbers KF373769 toKF373779.

    3. Results

    3.1. Environmental characterization

    Table 1 shows the environmental characterization of SonghuaLake in May, August, October and December. The tempera-ture changed considerably between the four years, especiallyin August and December. NO3−-N was the main nitrogen nutri-ent in this water body, and high NO3−-N concentrations wereobserved in Songhua Lake in all seasons. The accumulationof NO3−-N indicated that the nitrogen removal through deni-trification did not play a major role in the N-cycle process(Herbert, 1999). The level of NH4+-N in Songhua Lake was lowerin August and October than in May and December. In addition,the upstream water at S1 exhibited an extremely high concen-tration of NH4+-N in May, with the water at S2 exhibiting thesecond highest concentration of NH4+-N. DO is a major param-eter of the ecological health of a water system. This parametershowed visible spatial and seasonal variations across the fourseasons studied. Low levels of DO were detected in August,which was due to high temperatures and considerable micro-bial activity. The highest BOD5 values were observed in May,especially in the upstream sites S1 and S2, indicating that thewaterway is accumulating organic matter is being added tothe waterway through its course. SS is an indicator used toevaluate the pollution status of a water body. SS was foundto decrease along the river stream in May and August, whilehigher SS concentrations were observed in December than inAugust and autumn because SS was diluted during the rainyseason but not during the winter months. In addition, thepH values in S1 and S2 in August were alkaline, as shown inTable 1. High concentrations of NH4+-N were also detected inS1 and S2, indicating high nutrient input levels in the waterdue to a large runoff during the summer, which may stimulatealgal growth and increase the pH.

    3.2. Analysis of the DGGE profiles and phylogeneticanalysis of the DGGE bands

    DGGE profiles of the seasonal variation in the AOB commu-nity structure in the Songhua Lake are shown in Fig. 2. The

    unities of samples in the four seasons.

    DGGE patterns of the AOB 16S rRNA products reveal that thedominant bands obtained from seven sampling sites for fourseasons were quite similar. The average numbers of recog-nized DGGE bands from samples collected in May, August,October and December were 14, 12, 12, and 9, respectively.Increased intensities and bands (B5, B6) were only observedat S1 and S2 in May, and these were decreased in the lowerpart of the lake. Cluster analysis of the DGGE profiles based onthe presence or absence of bands is shown in Fig. 3. In May,the samples collected at S1 and S2 formed one cluster sepa-rated from samples at S3/S4/S5/S6/S7 (Fig. 3a). In August, S7formed one cluster, which was separated from the other sam-ples with 60% similarity (Fig. 3b). In October, most sampleswith 84% similarity were grouped into one cluster, which wasseparated from samples S1 and S3 (Fig. 3c). In December, S1formed one cluster that was separate from the other samples,with a similarity of 75% (Fig. 3d).

    A total of 14 bands were sequenced, and all sequences weremost related to the beta subclass ammonia oxidizers (Fig. 4).Five out of the fourteen bands (B3, B4, B8, B9 and B11), whichwere common in all samples, belonged to the Nitrosomonasspp. group. The sequences of B2, B5, B6 and B7 were closelyrelated to the Nitrosospira spp. group, which is the most ubiq-uitous group present in forest soils (Hastings et al., 2000) anddominates in grasslands and agricultural ecosystems (Brunset al., 1999; Stephen et al., 1998). According to the nomen-clature of Francis et al. (2003), the B1, B10, B12, B13 and B14cluster has been called “Nitrosospira-like” in sequence, andthese bands are typical AOB populations that exist in soilfertilized with amended water (Oved et al., 2001). However,Nitrosomonas-like sequences are generally found in effluentsof treated urban sewage (Cébron et al., 2004).

    3.3. Relationship between the AOB communitycomposition and environmental characteristics

    In order to determine to what extent the environmentalcharacteristics affected the AOB community composition, therelationship of the AOB community composition was ana-lyzed by RDA relative to the environmental characteristicsacross the four seasons. The results are shown in Table 2. Thefirst canonical axes for the AOB fingerprints explained 45.4%,55.0%, 36.5% and 39.4% of the variation in May, August, Octoberand December, respectively. Approximately more than 1/2 ofthe variation in the AOB species data (60.2%, 75.3%, 67.8%

    and 50.6%) could be explained by all four significant canonicalaxes in May, August, October and December. The calculated

  • Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530 527

    Fig. 3 – Cluster analysis of the DGGE banding pattern using the unweighted pair group method, based on the presence orabsence of each band. (a–d denote samples collected in May, August, October and December, respectively.).

    Fig. 4 – Neighbor-joining tree based on partial 16S rRNA sequences specific for AOB.

  • 528 Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530

    Table 2 – Eigenvalues and variance decomposition for RDA in samples of May, August, October and December,respectively.

    Axes Eigenvalues Species-environmentcorrelations

    Cumulative percentagevariance of species data (%)

    Cumulative variance ofspecies-environment relation (%)

    MayAxis 1 0.519 0.997 45.4 55.6Axis 2 0.200 0.877 59.2 63.9Axis 3 0.118 0.940 59.8 81.6Axis 4 0.105 0.999 60.2 92.0

    AugustAxis 1 0.568 0.962 55.0 56.8Axis 2 0.239 0.942 73.2 80.7Axis 3 0.160 0.833 73.8 96.7Axis 4 0.023 0.896 75.3 99.1

    OctoberAxis 1 0.372 0.965 36.5 37.2Axis 2 0.330 0.912 43.5 70.3Axis 3 0.187 0.896 43.7 89.0Axis 4 0.066 0.973 50.6 96.5

    DecemberAxis 1 0.473 0.955 39.4 47.3Axis 2 0.253 0.893 50.9 72.6Axis 3 0.126 0.915 64.4 85.2

    Axis 4 0.059 0.844

    species-environment correlations in the samples from thefour seasons for axes 1 were high (>90%).

    The aim of this research was to identify which of theenvironmental characteristics drive the observed changes inthe AOB community composition in May, August, Octoberand December. Forward selection was performed to deter-mine the most influential gradients and extract the variationin the AOB community explained by each of the significantparameters (RDA, Borcard). Table 3 shows the percentages ofvariation explained by each of the significant environmentalcharacteristics. In May, SS alone explained 33.5% (P = 0.03) ofthe variation in the AOB profiles, NH4+-N 24.3% (P = 0.04) andNO3−-N 29.7% (P = 0.05). In August, the variation of the AOBcommunity was best related to the pH, NH4+-N, TN and NO3−-N among the samples. The pH only explained 46.7% (P = 0.044)of the variation in the AOB community, while the NH4+-N,

    TN and NO3−-N explained 45.2% (P = 0.01), 41.9% (P = 0.03) and

    Table 3 – Eigen values, F values and P values testing theinfluence of the significant parameters on the AOBcommunity composition in May, August, October andDecember, respectively.

    Environmentalcharacteristicsincluded in the model

    % variationexplains

    solely

    F value P value

    MaySS 33.5 2.356 0.03NH4+-N 24.3 2.481 0.04NO3−-N 29.7 2.033 0.05

    AugustpH 46.7 4.009 0.04NH4+-N 45.2 3.278 0.01TN 41.9 3.204 0.03NO3−-N 41.5 2.211 0.05

    DecemberNH4+-N 28.0 3.265 0.02DO 25.3 2.361 0.03TN 23.7 2.154 0.04

    67.8 91.1

    41.5% (P = 0.05), respectively. In October, none of the variableshad a significant influence on the AOB community at the 5%level. In December, significant correlation between the AOBcommunity and the environmental characteristics that NH4+-N, DO and TN provided 28% (P = 0.02), 25.3% (P = 0.028) and23.7% (P = 0.037), respectively, were found in the total RDAexplanatory power.

    4. Discussion

    The AOB community in the estuarial and hypoxia area hasbeen widely studied and described in recent years. However,little is known regarding which environmental characteris-tics have greater effects on the variation and diversity ofthe AOB community composition. The relative abundanceof AOB obtained by PCR-DGGE was used to assess the rela-tionships between the AOB community and environmentalcharacteristics in this research. Confirming how environmen-tal characteristics affect the AOB community is an importantstep in improving our understanding of the oxidation ofammonia in Songhua Lake.

    In this study, the environmental characteristics signifi-cantly influencing the AOB community composition weredistinct in the four seasons. NH4+-N was significantly relatedto the AOB community composition in May, August andDecember (Fig. 5). In May, the AOB community compositionwas instead most substantially influenced by SS, NH4+-N andNO3−-N, and SS alone explained up to 33.5% of the variationin the DGGE, while NH4+-N explained 24.3% (Table 3). Thisresult indicated that the decrease of SS along the lake couldbe caused by variations in the AOB community composition.In addition, the variations in the NH4+-N concentrations weresimilar to those in SS. These results might be explained bythe fact that the water contained higher NH4+-N concentra-tions in SS in spring in contrast to the rainy seasons, summerand autumn. As spring is the cultivation time along the river,surplus nitrogen enters the waterway in May, which pro-

    motes the multiplication of AOB. The presence of particularbands affiliated with Nitrosospira (B5 and B6) occur at higher

  • Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530 529

    Fig. 5 – Redundancy analysis (RDA) ordination diagram of AOB communities associated with environmental vaiables a–drepresented the correlation in May, August, October and December, respectively. Environmental variables were indicated asarrows. DGGE samples at S1–S7 were indicated as (©) samples. Environmental vairables indicated by solid lines weresignificant (P < 0.05) .

    cftsctamtNANTinbnmiwnosia(l2Dwl

    oncentrations of SS and NH4+-N at S1 and S2 in May. Samplesrom S1 and S2 with higher SS and NH4+-N grouped togetherhrough clustering analysis (Table 1 and Fig. 3), which alsouggested that SS and NH4+-N greatly influenced the AOBommunity. These sequences might come from upstream inhe Songhua Lake and were probably caused by soil leachingnd agricultural pollution. Under water flow, self-purificationechanisms lighten the pollution in water, and the concentra-

    ion of SS and NH4+-N were thus diluted at downstream sites.O3−-N was another significant parameter that influenced theOB community composition. Higher concentration of NO3−-

    was observed in samples S1 and S2 of Songhua Lake in May.his might be caused by high DO values in May (Table 1), and

    t was the key factor for nitrification. The nitrification processeeds not only active nitrite oxidizer and enough nitrogenut also oxygen in the water body. Hence, although DO wasot significant to have a direct influence on the AOB com-unity composition, it was determined to have an indirect

    nfluence in this study. In August, the pH, NH4+-N and TNere all found to be significantly related to the AOB commu-ity composition (P < 0.05) (Table 3). Because the wet seasonf Songhua Lake has more precipitation than the other sea-ons, significant runoff would bring high levels of nutrientsnto the water, thus stimulating algal growth and leading to

    higher pH in August than in the other months (Table 1)Hong et al., 2010; Van der Steen et al., 2000). However, alka-ine water was harmful for the survival of bacteria (Hong et al.,010). In addition, the decreasing concentration of NH4+-N andO might not be adapted to the survival of AOB. However,

    ith the increase in NO3−-N values (Table 1), it was specu-

    ated that nitrite oxidizers were also active in Songhua Lake

    (Maeda et al., 2010). In October, no environmental charac-teristics were found to be significantly related with the AOBcommunity. This result was consistent with the work of Liuet al. (2012), which indicated that in addition to physical andchemical variables, phytoplankton composition, grazing andviral infection play important roles in shaping the bacterialcommunity structure. In December, the NH4+-N, DO and TNlevels significantly influenced the AOB community composi-tion. The increasing concentrations of NH4+-N, DO and TNare shown in Table 1. However, even with low water temper-atures (Table 1), a high nutrient and DO condition were stillfavorable for AOB survival (Sims et al., 2012). The environ-mental heterogeneity in season in the Songhua Lake, such asseasonal hydrodynamic conditions and nutrient input (Chenet al., 1999) might be the main factors to determine the AOBcommunity composition across the four seasons (Liu et al.,2012).

    Acknowledgments

    This study was partially supported by the Scientific ResearchFoundation for Major Science and Technology Program forWater Pollution Control and Treatment and National Sci-ence (No. 2009ZX07106-001-001) and Technology Program (No.2012BAJ21B02).

    References

    Arp, D.J., Stein, L.Y., 2003. Metabolism of inorganic N compounds

    by ammonia-oxidizing bacteria. Crit. Rev. Biochem. Mol. Biol.38, 471–495.

    http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0005

  • 530 Process Safety and Environmental Protection 1 0 4 ( 2 0 1 6 ) 523–530

    Zhou, J., Bruns, M.A., Tiedje, J.M., 1996. DNA recovery from soils of

    Bruns, M.A., Stephen, J.R., Kowalchuk, G.A., Prosser, J.I., Paul, E.A.,1999. Comparative diversity of ammonia oxidizer 16S rRNAgene sequences in native, tilled, and successional soils. Appl.Environ. Microbiol. 65, 2994–3000.

    Cébron, A., Coci, M., Garnier, J., Laanbroek, H.J., 2004. Denaturinggradient gel electrophoretic analysis of ammonia-oxidizingbacterial community structure in the lower Seine River:impact of Paris wastewater effluents. Appl. Environ. Microbiol.70, 6726–6737.

    Camargo, J.A., Alonso, Á., 2006. Ecological and toxicologicaleffects of inorganic nitrogen pollution in aquatic ecosystems:a global assessment. Environ. Int. 32, 831–849.

    Casciotti, K.L., Sigman, D.M., Ward, B.B., 2003. Linking diversityand stable isotope fractionation in ammonia-oxidizingbacteria. Geomicrobiol. J. 20, 335–353.

    Chen, J., Li, D., Chen, B., Hu, F., Zhu, H., Liu, C., 1999. Theprocesses of dynamic sedimentation in the ChangjiangEstuary. J. Sea Res. 41, 129–140.

    Cilia, V., Lafay, B., Christen, R., 1996. Sequence heterogeneitiesamong 16S ribosomal RNA sequences, and their effect onphylogenetic analyses at the species level. Mol. Biol. Evol. 13,451–461.

    Dang, H., Li, J., Chen, R., Wang, L., Guo, L., Zhang, Z., Klotz, M.G.,2010. Diversity, abundance, and spatial distribution ofsediment ammonia-oxidizing betaproteobacteria in responseto environmental gradients and coastal eutrophication inJiaozhou Bay, China. Appl. Environ. Microbiol. 76, 4691–4702.

    de Bie, M.J., Speksnijder, A.G., Kowalchuk, G.A., Schuurman, T.,Zwart, G., Stephen, J.R., Diekmann, O.E., Laanbroek, H.J., 2001.Shifts in the dominant populations of ammonia-oxidizing �subclass proteobacteria along the eutrophic Schelde estuary.Aquat. Microb. Ecol. 23, 225–236.

    Francis, C.A., O’Mullan, G.D., Ward, B.B., 2003. Diversity ofammonia monooxygenase (amoA) genes acrossenvironmental gradients in Chesapeake Bay sediments.Geobiology 1, 129–140.

    Gilliam, F.S., Somerville, F.C., Lyttle, F.N., Adams, F., 2001. Factorsinfluencing spatial variability in nitrogen processing innitrogen-saturated soils. Sci. World J. 1, 505–513.

    Hastings, R., Butler, C., Singleton, I., Saunders, J., McCarthy, A.,2000. Analysis of ammonia-oxidizing bacteria populations inacid forest soil during conditions of moisture limitation. Lett.Appl. Microbiol. 30, 14–18.

    Herbert, R., 1999. Nitrogen cycling in coastal marine ecosystems.FEMS Microbiol. Rev. 23, 563–590.

    Hong, H., Qiu, J., Liang, Y., 2010. Environmental factorsinfluencing the distribution of total and fecal coliformbacteria in six water storage reservoirs in the Pearl River DeltaRegion, China. J. Environ. Sci. 22, 663–668.

    Kowalchuk, G.A., Bodelier, P.L., Heilig, G.H.J., Stephen, J.R.,Laanbroek, H.J., 1998. Community analysis ofammonia-oxidising bacteria, in relation to oxygen availabilityin soils and root-oxygenated sediments, using PCR, DGGE andoligonucleotide probe hybridisation. FEMS Microbiol. Ecol. 27,339–350.

    Kowalchuk, G.A., Stephen, J.R., De Boer, W., Prosser, J.I., Embley,T.M., Woldendorp, J.W., 1997. Analysis of ammonia-oxidizingbacteria of the beta subdivision of the class Proteobacteria incoastal sand dunes by denaturing gradient gel electrophoresisand sequencing of PCR-amplified 16S ribosomal DNAfragments. Appl. Environ. Microbiol. 63, 1489–1497.

    Liu, M., Xiao, T., Wu, Y., Zhou, F., Huang, H., Bao, S., Zhang, W.,2012. Temporal distribution of bacterial community structure

    in the Changjiang Estuary hypoxia area and the adjacent EastChina Sea. Environ. Res. Lett. 7, 025001.

    Maeda, K., Toyoda, S., Shimojima, R., Osada, T., Hanajima, D.,Morioka, R., Yoshida, N., 2010. Source of nitrous oxideemissions during the cow manure composting process asrevealed by isotopomer analysis of and amoA abundance inbetaproteobacterial ammonia-oxidizing bacteria. Appl.Environ. Microbiol. 76, 1555–1562.

    Magic-Knezev, A., Van Der Kooij, D., 2004. Optimisation andsignificance of ATP analysis for measuring active biomass ingranular activated carbon filters used in water treatment.Water Res. 38, 3971–3979.

    Muyzer, G., De Waal, E.C., Uitterlinden, A.G., 1993. Profiling ofcomplex microbial populations by denaturing gradient gelelectrophoresis analysis of polymerase chainreaction-amplified genes coding for 16S rRNA. Appl. Environ.Microbiol. 59, 695–700.

    Oved, T., Shaviv, A., Goldrath, T., Mandelbaum, R.T., Minz, D.,2001. Influence of effluent irrigation on communitycomposition and function of ammonia-oxidizing bacteria insoil. Appl. Environ. Microbiol. 67, 3426–3433.

    Phillips, C.J., Harris, D., Dollhopf, S.L., Gross, K.L., Prosser, J.I.,Paul, E.A., 2000. Effects of agronomic treatments on structureand function of ammonia-oxidizing communities. Appl.Environ. Microbiol. 66, 5410–5418.

    Rowan, A.K., Snape, J.R., Fearnside, D., Barer, M.R., Curtis, T.P.,Head, I.M., 2003. Composition and diversity ofammonia-oxidising bacterial communities in wastewatertreatment reactors of different design treating identicalwastewater. FEMS Microbiol. Ecol. 43, 195–206.

    Salles, J.F., Van Veen, J.A., Van Elsas, J.D., 2004. Multivariateanalyses of Burkholderia species in soil: effect of crop and landuse history. Appl. Environ. Microbiol. 70, 4012–4020.

    Shi, Y., Liu, J., Li, Y., Shao, W., 2008. Comprehensive assessmentand measurement of eutrophication of Jilin Songhua lake.China Brew. 14, 036.

    Sims, A., Gajaraj, S., Hu, Z., 2012. Seasonal population changes ofammonia-oxidizing organisms and their relationship to waterquality in a constructed wetland. Ecol. Eng. 40, 100–107.

    Stephen, J.R., Kowalchuk, G.A., Bruns, M.-A.V., McCaig, A.E.,Phillips, C.J., Embley, T.M., Prosser, J.I., 1998. Analysis of�-subgroup proteobacterial ammonia oxidizer populations insoil by denaturing gradient gel electrophoresis analysis andhierarchical phylogenetic probing. Appl. Environ. Microbiol.64, 2958–2965.

    Van der Steen, P., Brenner, A., Shabtai, Y., Oron, G., 2000.Improved fecal coliform decay in integrated duckweed andalgal ponds. Water Sci. Technol. 42, 363–370.

    Ward, B.B., 2005. Molecular approaches to marine microbialecology and the marine nitrogen cycle. Annu. Rev. EarthPlanet. Sci. 33, 301–333.

    Zhang, D., Li, W., Zhang, S., Liu, M., Gong, H., 2011. Evaluation ofthe impact of DNA extraction methods on BAC bacterialcommunity composition measured by denaturing gradient gelelectrophoresis. Lett. Appl. Microbiol. 53, 44–49.

    Zhao, X., Wei, Z., Zhao, Y., Xi, B., Wang, X., Zhao, T., Zhang, X.,Wei, Y., 2015. Environmental factors influencing thedistribution of ammonifying and denitrifying bacteria andwater qualities in 10 lakes and reservoirs of the Northeast,China. Microb. Biotechnol. 8, 541–548.

    diverse composition. Appl. Environ. Microbiol. 62, 316–322.

    http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0010http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0015http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0020http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0025http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0030http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0035http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0040http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0045http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0050http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0055http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0060http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0065http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0070http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0075http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0080http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0085http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0090http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0095http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0100http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0105http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0110http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0115http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0120http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0125http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0130http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0135http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140http://refhub.elsevier.com/S0957-5820(16)30037-4/sbref0140htt