Explanatory Parameter Research Articles

Use of a flashiness index to predict phosphorus losses from subsurface drains on a Swedish farm with clay soils

Risk assessment for elevated leaching losses of phosphorus (P) from agricultural land is commonly based on indices, since such losses are highly episodic and difficult to predict. Here a flashiness index (FI) representing changes in daily water flow from drainage systems was estimated from measured discharge (agrohydrological years 2004–2013) after reconstruction of subsurface drainage systems in 16 fields on a former swine farm. The fields were analysed for ammonium lactate-extractable soil P (P-AL), clay, carbon and other soil parameters in 2004. Transport of total P (TP), dissolved reactive P (DRP) and unreactive P (UP) was estimated from concentrations in composite water samples taken flow-proportionally up to 20 times per year. On average, 2.20kgTPha−1yr−1 was leached, with 27% in DRP form, from the entire farm. FI was significantly negatively correlated (Pearson correlation coefficient p<0.05) to mean yearly discharge from each field. Stepwise regression demonstrated that FI index was the most important single explanatory parameter for flow-proportional yearly mean concentration of unreactive P losses (UP) from each field, with a coefficient of determination (r2) of 0.67. The corresponding concentration of dissolved reactive P (DRP) was significantly positively correlated (p=0.015) to soil P-AL and FI. A regression model for TP leaching losses based on FI, P-AL and yearly discharge (Q) from 11 of the fields over nine years (r2=0.67, p=0.002) was validated against TP leaching from the remaining five fields (32% of farm area). Root mean square error (RMSE) was 0.43, which represented 20% of measured leaching (mean 2.14kgTPha−1yr−1). For individual years, RMSE for different fields was 37–80% of measured TP leaching (0.8–3.7kgTPha−1yr−1). The FI index could be used together with soil P test to predict P leaching from individual fields of a drained farm.

Building vulnerability to hydro-geomorphic hazards: Estimating damage probability from qualitative vulnerability assessment using logistic regression

Summary The focus of this study is an analysis of building vulnerability through investigating impacts from the 8 February 2013 flash flood event along the Avenida Venezuela channel in the city of Arequipa, Peru. On this day, 124.5 mm of rain fell within 3 h (monthly mean: 29.3 mm) triggering a flash flood that inundated at least 0.4 km 2 of urban settlements along the channel, affecting more than 280 buildings, 23 of a total of 53 bridges (pedestrian, vehicle and railway), and leading to the partial collapse of sections of the main road, paralyzing central parts of the city for more than one week. This study assesses the aspects of building design and site specific environmental characteristics that render a building vulnerable by considering the example of a flash flood event in February 2013. A statistical methodology is developed that enables estimation of damage probability for buildings. The applied method uses observed inundation height as a hazard proxy in areas where more detailed hydrodynamic modeling data is not available. Building design and site-specific environmental conditions determine the physical vulnerability. The mathematical approach considers both physical vulnerability and hazard related parameters and helps to reduce uncertainty in the determination of descriptive parameters, parameter interdependency and respective contributions to damage. This study aims to (1) enable the estimation of damage probability for a certain hazard intensity, and (2) obtain data to visualize variations in damage susceptibility for buildings in flood prone areas. Data collection is based on a post-flood event field survey and the analysis of high (sub-metric) spatial resolution images (Pleiades 2012, 2013). An inventory of 30 city blocks was collated in a GIS database in order to estimate the physical vulnerability of buildings. As many as 1103 buildings were surveyed along the affected drainage and 898 buildings were included in the statistical analysis. Univariate and bivariate analyses were applied to better characterize each vulnerability parameter. Multiple corresponding analyses revealed strong relationships between the “Distance to channel or bridges”, “Structural building type”, “Building footprint” and the observed damage. Logistic regression enabled quantification of the contribution of each explanatory parameter to potential damage, and determination of the significant parameters that express the damage susceptibility of a building. The model was applied 200 times on different calibration and validation data sets in order to examine performance. Results show that 90% of these tests have a success rate of more than 67%. Probabilities (at building scale) of experiencing different damage levels during a future event similar to the 8 February 2013 flash flood are the major outcomes of this study.

Phosphorus release from anaerobic peat soils during convective discharge — Effect of soil Fe:P molar ratio and preferential flow

Phosphorus (P) accumulation in drained agricultural lowlands causes a risk for P pollution to the aquatic environment following wetland restoration. While extensive knowledge is available on P sorption and desorption from anaerobic soils, very limited information is available on the interacting influence of soil geochemistry and local scale active flow volume. Combining batch incubation experiments and continuous column discharge experiments, we investigated iron (Fe) reduction and P release from 10 anoxic Fe-dominated (oxalate extractable Fe (Feox) from ~5500 to 50,000mgkg−1) lowland peat soils (TOC from 5 to 39%) with a gradient in Fe:P molar ratio (molar ratio between bicarbonate dithionite extractable Fe and P (FeBD:PBD) from 3 to 112) and degree of non-equilibrium (preferential) flow. Short-term batch incubation experiments (21days) indicated that concurrent Fe and P release was controlled by reductive Fe(III) dissolution, and was well predicted from the soil FeBD:PBD molar ratio. Continuous convective column discharge with oxygen-free deionised water at 1mmh−1 for 10 effluent pore volumes resulted in highly variable in situ redox potential (Eh from −200 to 300mV), effluent Fe(II) concentrations (23 to 2000μM) and effluent dissolved reactive phosphorus (DRP) concentrations (<6.5 to 316μM). Effluent P forms changed from dominantly particular P/dissolved organic P (PP/DOP) to dominantly DRP as Eh decreased in all soils. Total phosphorus (TP) release rates during convective discharge (3–66μmolkg−1day−1) were negatively non-linearly correlated with the soil FeBD:PBD molar ratio, which explained 71–73% of the variability, and with FeBD:PBD of 10 as a critical threshold ratio. Fe and P release rates from batch experiments were poorly correlated with convective discharge Fe and P release rates, indicating the overall influence of soil structure. Diffusion was found to be the rate-limiting step for P release after prolonged leaching. Although the soil FeBD:PBD molar ratio turned out as a key explanatory parameter in predicting P release rates following rewetting of these peat soils, the results did indicate the influence of preferential flow in decreasing P release rates. Although, sufficient available P is present for leaching in the initial phase after rewetting P enriched lowland soils, we do expect that soils with pronounced preferential flow will become more rapidly exhausted in available P, and hence limit P release after prolonged discharge.

Comparative Analysis of Car-Following Models for Emissions Estimation

Recent studies have indicated that the accuracy of the emissions estimation in a traffic simulation model can be little improved by using the traditional model calibration approaches. Instead, the model's internal mechanism in depicting the second-by-second vehicle activities needs to be investigated. Since the car-following model is the core component of a traffic simulation model, this paper attempts to conduct a comparative study of car-following models concerning their effects on the explanatory parameter of vehicle emissions, namely, the vehicle specific power (VSP) distribution. The car-following models selected for the analysis are the optimal velocity model (OVM), generalized force model (GFM), full velocity difference model (FVDM), Wiedemann model, and the Fritzsche model. Massive field car-following trajectories are collected, and a numerical simulation method is designed for each car-following model to generate its vehicle trajectories and the speed-specific VSP distributions. By a comparison of VSP distributions collected from the field and generated by car-following models, it was found that OVMs and GFMs generate unrealistic VSP distributions, which will lead to significant emissions estimation errors. By adding the variable of positive velocity difference, the FVDM can effectively improve the accuracy of the VSP distribution and emissions estimation. The VSP distribution of the Wiedemann model differs largely from the field data, which overestimate the peak VSP fraction and the fractions in aggressive driving modes. The Fritzsche model produces VSP distributions consistent with the field distributions. It is also found that the speed-specific VSP distribution is highly correlated with the acceleration distribution. Therefore, improving the accuracy of the speed-specific acceleration distribution is an effective measure to improve the accuracy of the VSP distribution and thus the emissions estimation of the car-following models.

Influence diagnostics in generalized symmetric linear models

The aim of this paper is to introduce generalized symmetric linear models (GSLMs) in the same sense of generalized linear models (GLMs), in which a link function is defined to establish a relationship between the mean values of symmetric distributions and linear predictors. The class of symmetric distributions contains various distributions with lighter and heavier tails than normal and hence offers a more flexible basis for analyzing symmetric data. An iteratively reweighed least squares (IRLS) algorithm is derived to obtain maximum likelihood estimates. The local influence methodology is applied to study the sensitivity of the maximum likelihood estimates under some usual perturbation schemes, such as case-weight, response variable, continuous explanatory variable and scale parameter perturbations. We also discuss generalized leverage and residual analysis. Finally, an illustration is given in which the methodology developed in this paper is applied to a real data set.

Phenylpropanoids from maize pericarp: resistance factors to kernel infection and fumonisin accumulation by Fusarium verticillioides

The aim of this work was to investigate the role of pericarp phenylpropanoids as resistance factors to F. verticillioides in eleven maize genotypes. Disease severity and kernel fumonisin accumulation were measured after inoculation with F. verticillioides and related to contents of pericarp phenylpropanoids in field trials conducted during 2 years. Grain fumonisin concentrations were highly dependant on disease severity of the genotypes (r = 0.88). A detailed analysis of pericarp phenylpropanoids indicated the presence of trans-ferulic acid (tFA), cis-ferulic acid (cFA), p-coumaric acid (pCA), and five diferulates (DFAs). The most prominent diferulates were 8,5′-diferulic acid benzofuram form (8,5′-DFAbz), followed by 8,5′-DFA and 8,8′-DFA. Except for cFA, the most resistant genotypes exhibited high levels of phenylpropanoids which were related to low levels of disease severity and grain fumonisin concentration (−0.61 > r > −0.90). A stepwise regression analysis revealed that total diferulates was the best explanatory parameter for variability of disease severity (r2 = 0.71). Grain fumonisin concentration was well depicted by contents of total diferulates, 8,5′DFAbz and pCA (r2 = 0.82). Our findings suggest that high level of phenylpropanoids in the kernel pericarp is a trait associated to less disease severity and fumonisin accumulation caused by F. verticillioides. Further research is in progress to map quantitative trait loci for these cell wall components in bi-parental populations derived by crossing resistant and susceptible genotypes included in this study.

Applicability of Traffic Microsimulation Models in Vehicle Emissions Estimates

Efforts in estimating emissions by an integration of traffic simulation models and emissions models have become a fast-evolving research area. However, because of the lack of effective methods and indicators to characterize traffic behaviors, the accuracy of emissions by such an approach has not been effectively verified or evaluated. The current study is intended to examine the applicability of traffic microsimulation models in vehicle emissions estimates on the basis of the explanatory parameter of vehicle emissions—the vehicle-specific power (VSP) distribution. Analyzing massive real-world and simulated vehicle activity data showed that the results from traffic simulation could not represent real-world driving behaviors for emissions estimates. The simulated vehicle- specific power distribution led to errors that were as high as 82.8%, 53.6%, and 29.6% for nitrogen oxides, hydrocarbons, and carbon monoxide emissions, respectively. Then a sensitivity analysis of 16 adjustments on eight parameters of simulation models was conducted to determine their effects on simulated VSP distributions: systematic errors existed in the use of traffic simulation models to represent second-by-second driving behaviors. The errors could not be reduced by parameter calibration on the simulation model. This study concluded that the traditional approach of integrating traffic simulation models with emissions models was not applicable for vehicle emissions estimates. The primary reasons for the errors need to be investigated further from the internal mechanism of submodels of microsimulation. On the basis of these findings, several recommendations are proposed for future studies.

Social risk assessment and social capital: A significant parameter for the formation of climate change policies

Public policy issues arising from climate change projections are becoming increasingly important in terms of the range and the scope of their effects. In order to effectively confront them it is important to address social, environmental and economic aspects as fully as possible in decision-making processes. In this context, social risk assessment techniques have begun to be applied to explore citizens’ risk perceptions of climate change projections. The present study aims to contribute to this area by examining the influence of four social capital parameters (social trust, institutional trust, social norms and social networks) on public risk perceptions of climate change. An empirical study was conducted for this purpose in Greece's fourth largest city, Heraklion. Results demonstrate that social capital is a significant explanatory parameter for citizens’ risk perceptions. Specifically, individuals with lower levels of social capital tend to perceive higher risks from climate change impacts. The most important parameter explaining this result is the low level of institutional trust, revealing that Heraklion's citizens do not believe that existing public institutions will be able to effectively manage projected climate change impacts.

Willingness to pay for drinking water quality improvement and the influence of social capital

Several factors have been identified as determinants of willingness to pay (WTP), including socio-economic and environmental elements. The present article aims to investigate the influence of individual social capital on willingness to pay for environmental goods. In particular, through an empirical study, a multi-dimensional measurement of social capital is conducted and its influence on WTP for water quality improvement is explored. According to the results of the survey it is observed that social capital is a significant explanatory parameter of WTP.

Conceptualising aerospace outsourcing: Airbus UK and the lean supply approach

Lean production and supply concepts are amongst the most influential management theories. Yet, they have scarcely been applied to aerospace, an industry sector of great importance. This paper identifies generic trends in aerospace outsourcing and explores to what extent the lean supply approach is fit to conceptualise them. It investigates to what degree the outsourcing strategies of Airbus UK and the wider Airbus operations are explicable with the lean approach. The paper finds that lean is capable of conceptualising most generic trends in aerospace outsourcing and most of the Airbus policies. Lean neglects the relevance of knowledge to outsourcing, however, and fails to recognise the diversity of the Airbus supply approach. Knowledge management in this process is now crucial. Technological capabilities are central to the Airbus strategies, illustrating the need to acknowledge knowledge as a key explanatory parameter of outsourcing theory. In conclusion, we find the higher the knowledge content, the more partnering occurs, the lower the content, the more hierarchical the relationship are.

Efficacy of recombinant activated factor VII vs. activated prothrombin complex concentrate for patients suffering from haemophilia complicated with inhibitors: a Bayesian meta‐regression

The optimal on-demand treatment of joint bleeds in haemophilia patients with inhibitors is a source of debate, with studies reporting various efficacy levels for different drugs and dosage regimens. To analyse, in a unified Bayesian meta-regression model, the published efficacy of recombinant activated factor VII (rFVIIa) and/or activated prothrombin complex concentrate (aPCC) as on-demand treatments for joint bleeds in haemophilia patients with inhibitors. A systematic search was carried out to identify studies reporting on dosage and efficacy of rFVIIa and aPCC in the treatment of joint bleeds in the target patient population. Data were abstracted and included in the model and adjusted for potential sources of heterogeneity. Pooled efficacy levels for typical rFVIIa and aPCC regimens were estimated. Seventeen studies, collectively reporting on >2000 joint bleeds, were included. Medication type combined with dosage was the only significant explanatory parameter. The model predicts that a typical regimen of 90 microg kg(-1) rFVII repeated every 3 h if needed results in cumulative joint bleed resolution of 66%, 88% and 95% after 12, 24 and 36 h, respectively. In comparison, a typical regimen of 75 IU kg(-1) aPCC repeated every 12 h if needed results in cumulative joint bleed resolution of 39%, 62% and 76%, respectively. These differences were statistically significant and were also robust in sensitivity analyses. This analysis suggests that a typical rFVIIa regimen will resolve joint bleeds more effectively than a typical aPCC regimen after 12, 24 and 36 h.

Elevated erythrocyte carbonic anhydrase activity is a novel clinical marker in hyperventilation syndrome

The aim of this study was to evaluate the erythrocyte carbonic anhydrase (CA) activity in patients with hyperventilation syndrome. A total of 71 patients with hyperventilation syndrome and 60 controls were recruited in this study for measurement of the erythrocyte CA activity. CA activity was analyzed from erythrocyte using CA esterase activity analysis. The erythrocyte CA activity was significantly elevated in the patients with hyperventilation syndrome compared to controls (31.07+/-1.37 U/g Hb vs. 24.67+/-0.99 U/g Hb, p=0.003). The standardized beta value and significant R(2) value of total CA activity for prediction of hyperventilation was 0.155 (p=0.016) and 0.024, respectively. Moreover, a significant negative correlation was found between total CA activity and partial pressure of carbon dioxide (PaCO(2)) (r=-0.185, p=0.03, n=131). Furthermore, the adjusted odds ratios for patients with hyperventilation were 11.6 (95% CI: 4.8-27.9) and 51.0 (95% CI: 5.8-445.2) for individuals with either PaCO(2)<or=28.1 mm Hg or CA activity >or=29.71 U/g hemoglobin (Hb), and for individuals with both PaCO(2)<or=28.1 mm Hg and CA activity >or=29.71 U/g Hb, respectively, compared to individuals with both PaCO(2)>28.1 mm Hg and CA activity <29.71 U/g Hb. Erythrocyte CA activity was significantly elevated in patients with hyperventilation. Detection of erythrocyte CA activity may provide a potential explanatory parameter for the prediction of hyperventilation syndrome.

Serum IL-18 levels in patients with type 1 diabetes: Relations to metabolic control and microvascular complications

The study was designed to examine serum IL-18 level and its relation to metabolic control parameters and microvascular complications in type 1 diabetes mellitus (DM). Sixty two patients with type 1 DM and 30 healthy individuals were enrolled in the study. Serum IL-18 levels of patients with type 1 DM were significantly increased compared to controls (293.4±133.4 vs 211.2±63.9pg/ml, P=0.003). Patients with poor glycemic control had higher levels of IL-18 than patients with well glycemic control (329.9±141.0 vs 226.3±89.6pg/ml, P=0.02). There was no significant difference between the serum IL-18 levels of patients with microvascular complications and those of patients without microvascular complications (307.6±127.6 vs 293.2±145.6pg/ml, P>0.05). IL-18 correlated positively with HbA1c (r=0.32, P=0.01) and postprandial blood glucose (PPBG) (r=0.26, P=0.02); and negatively with HDL-cholesterol (HDL-C) (r=−0.38, P=0.007). By linear regression analysis, PPBG was determined as the most explanatory parameter for the alterations in serum IL-18 levels (P=0.02). High levels of IL-18 in patients with type 1 DM is related to short and long term glycemic control and HDL-C levels but not to microvascular complications.

Model for optimizing energy efficiency through controlling speed and gear ratio

Specific fuel consumption of vehicles has been improved through technical changes in the last three decades. There has been little study of the impact of behavioral aspects on the specific fuel consumption of vehicles. Analytical tools developed recently have included options for quantifying the specific fuel consumption in certain predefined driving cycles. Empirical studies have shown that real driving cycles are different from predefined ones. It has also been observed that optimal coordination of both speed of vehicle and gear ratio are detrimental to the improved performance of a vehicle. Previous studies have only considered the vehicle speed as the explanatory parameter, and gear ratio has rarely been discussed. Therefore, an optimal model of vehicle fuel consumption has been developed on the basis of microeconomic theories. The application of this model helps to identify the optimal driving strategy of a vehicle in a certain real driving cycle. The model has then been applied to estimate the optimal fuel consumption of a vehicle in a given real case, and the results of model estimate have been compared with measured data. It has been found that the implementation of optimal driving strategy based on the model estimates may help to reduce specific fuel consumption by 3.2 l/100 km on average (37% of actual specific consumption). It is also observed that the potential for energy saving prevails when the traffic flow is dense and slow. Hence, the implementation of optimal driving strategy based on the coordination of speed and gear ratio would lead to the realization of energy saving potential that is considerable.

Influence of Transportation Access and Market Dynamics on Property Values

This paper presents housing price models by using multilevel modeling techniques. The key motivation of using the multilevel modeling technique is that it clearly identifies and differentiates between-cluster heterogeneity (i.e., intrinsic differences across aggregated units) and heterogeneity between units of analysis that are nested within aggregated clusters. Two different specifications are tested: two-level spatial and mixed two-level spatiotemporal random effects models. Whereas the first specification assumes that dwelling units are nested within spatial clusters (i.e., neighborhoods), the second specification hypothesizes that dwelling units are nested within spatiotemporal clusters (neighborhoods in a given time period). The unique contribution of this paper is that it accounts for temporal heterogeneity simultaneously with spatial heterogeneity in the housing price models. The study uses an extensive sample of more than 250,000 housing property transactions in 1987–1995 in the Greater Toronto Area of Canada. The paper examines the functional form of the hedonic price model and chooses a semilogarithmic model for subsequent multilevel housing price modeling. The results suggest that the spatiotemporal model performs better in terms of explanatory power and parameter estimates.

Development of Shopping Mode Choice Model with Bayesian Estimation Method

In this paper, we developed shopping mode choice multinomial probit models basedon the shopping trip database in a suburban newtown with maximum likelihood estimation and Bayesian estimation using MCMC methods. According to our study, both estimation results were similar, while Bayesian estimation method could take account of standard deviation as well as mean of each coefficient. In addition, we found that the standard deviation for each explanatory parameter was smaller forthe multi-purpose shopping than the shopping-only case.

Evaluating the Behavioural Performance of Alternative Logit Models: An Application to Corporate Takeovers Research

Abstract: Econometric models involving a discrete outcome dependent variable abound in the finance and accounting literatures. However, much of the literature to date utilises a basic or standard logit model. Capitalising on recent developments in the discrete choice literature, we examine three advanced (or non‐IID) logit models, namely: nested logit, mixed logit and latent class MNL. Using an illustration from corporate takeovers research, we compare the explanatory and predictive performance of each class of advanced model relative to the standard model. We find that in all cases the more advanced logit model structures, which correct for the highly restrictive IID and IIA conditions, provide significantly greater explanatory power than standard logit. Mixed logit and latent class MNL models exhibited the highest overall predictive accuracy on a holdout sample, while the standard logit model performed the worst. Moreover, the analysis of marginal effects of all models indicates that use of advanced models can lead to more insightful and behaviourally meaningful interpretations of the role and influence of explanatory variables and parameter estimates in model estimation. The results of this paper have implications for the use of more optimal logit structures in future research and practice.

Genetic Background Determines the Size and Structure of the Endocrine Pancreas

Key parameters of the endocrine pancreas, such as islet number, islet mass, beta-cell mass, and alpha-cell mass, were studied in different strains of inbred mice to investigate the impact of genetic background on the size and structure of the endocrine pancreas. Six mice from each of seven different strains of inbred mice were included in the study. For all parameters investigated, there was a pronounced interstrain variation. ANCOVA showed that only mouse strain was statistically significant as an explanatory parameter for the number of islets. Mouse strain, body weight, and pancreas mass reached statistical significance as explanatory parameters for the islet mass, with mouse strain as the most significant predictor. These data show that genetic background is the most important predictor of both the number of islets and total islet volume. We also conclude that inbred mice could be a valuable resource to identify the genes responsible for the size and structure of the endocrine pancreas.

Key site variables governing the functional characteristics of Dissolved Natural Organic Matter (DNOM) in Nordic forested catchments

The objective of this study was to extract and evaluate key catchment characteristics explaining the main variation in properties of dissolved natural organic matter (DNOM) in surface waters draining each catchment. A number of classical and sophisticated DNOM characterisation methodologies were used for this purpose. The explanatory capability of site characteristics on the variation in descriptors of DNOM, including elemental analysis (EA), DOC fractionation, SEC, CZE, CGE, UV/Vis- and FTIR spectra, FES, TLS, ESR, ESI/MS, 13C-CPMAS-NMR, potentiometric pH titration as well as functional characteristics such as biodegradability, and the response of macrophyte and PAH partitioning coefficients were evaluated. The sampling sites were all natural, coniferous, heathery-forested catchments with bogs, forming a climatic, anthropogenic deposition and proportional biotope cover gradient. Correlations to the main principal components, as well as a correlation matrix, singled out total S-deposition as the key explanatory site parameter, accounting for most of the variation in DNOM descriptors. Other significant factors were H+ concentration, length of growing season, biotope coverage (i.e., peatland/ or lake/watershed ratio), site elevation and TOC concentration. Total sulphur (S) deposition was strongly negatively correlated to the molecular weight, aromaticity, carboxylic acidity and sorption capacity of PAH. A greater proportion of lake and Histosol biotopes of the sampling sites were reflected in a greater aliphatic character of the DNOM. PCA clustering of samples from the same site in spring and fall showed a basic site-specific DNOM quality and a weak season effect. The Hasse diagram technique (a method of partial order theory) was used to show, for instance, how the relative Histosol coverage may determine NOM quality and, in turn, how this coincides with high macrophyte response.

An ordination study to view vegetation structure dynamics in disturbed and undisturbed mangrove forests in Kenya and Sri Lanka

The mangrove vegetation of a disturbed and undisturbed site in both Kenya and Sri Lanka was investigated in the field for three vegetation layers: adult trees, young trees, and juvenile trees. A minimum of 25 sample points, in which the vegetation was described and environmental factors (salinity, light intensity, land/water ratio, abundance of herbivorous crabs and snail abundance) were measured or estimated, were taken on each site. Detrended correspondence analysis (DCA) and canonical correspondence analysis (CCA) were used to summarize the data bulk, to investigate the vegetation dynamics (e.g., comparability of species distribution in the three vegetation layers), and/or to link the vegetation data to the environmental factors. Results showed that species clusters were relatively easy to delineate, whether mangrove zonation was present or not. Among the environmental factors, the abundance of propagule predators (mostly sesarmid crabs) contributed significantly to the variation in vegetation and could be an explanatory parameter for the observed vegetation data in a majority of sites. In the site where it was not, the most important factor in the ordination was the land/water ratio, which is important at the ecological level as well (link between water level and vegetation dynamics). However, none of the environmental factors could successfully explain the total variability in the vegetation data suggesting that other, more determining factors exist. Our results further provide information on the dynamic or non-dynamic nature of a forest and on its ability to rejuvenate, and may contribute to appropriate forestry management guidelines in the future.