Effects of sample size on the performance of species distribution models

A wide range of modelling algorithms is used by ecologists, conservation practitioners, and others to predict species ranges from point locality data. Unfortunately, the amount of data available is limited for many taxa and regions, making it essential to quantify the sensitivity of these algorithms to sample size. This is the first study to address this need by rigorously evaluating a broad suite of algorithms with independent presence-absence data from multiple species and regions. We evaluated predictions from 12 algorithms for 46 species (from six different regions of the world) at three sample sizes (100, 30, and 10 records). We used data from natural history collections to run the models, and evaluated the quality of model predictions with area under the receiver operating characteristic curve (AUC). With decreasing sample size, model accuracy decreased and variability increased across species and between models. Novel modelling methods that incorporate both interactions between predictor variables and complex response shapes (i.e. GBM, MARS-INT, BRUTO) performed better than most methods at large sample sizes but not at the smallest sample sizes. Other algorithms were much less sensitive to sample size, including an algorithm based on maximum entropy (MAXENT) that had among the best predictive power across all sample sizes. Relative to other algorithms, a distance metric algorithm (DOMAIN) and a genetic algorithm (OM-GARP) had intermediate performance at the largest sample size and among the best performance at the lowest sample size. No algorithm predicted consistently well with small sample size (n < 30) and this should encourage highly conservative use of predictions based on small sample size and restrict their use to exploratory modelling.

Prediction of soil orders with high spatial resolution: response of different classifiers to sampling density

The objective of this work was to evaluate sampling density on the prediction accuracy of soil orders, with high spatial resolution, in a viticultural zone of Serra Gaúcha, Southern Brazil. A digital elevation model (DEM), a cartographic base, a conventional soil map, and the Idrisi software were used. Seven predictor variables were calculated and read along with soil classes in randomly distributed points, with sampling densities of 0.5, 1, 1.5, 2, and 4 points per hectare. Data were used to train a decision tree (Gini) and three artificial neural networks: adaptive resonance theory, fuzzy ARTMap; self‑organizing map, SOM; and multi‑layer perceptron, MLP. Estimated maps were compared with the conventional soil map to calculate omission and commission errors, overall accuracy, and quantity and allocation disagreement. The decision tree was less sensitive to sampling density and had the highest accuracy and consistence. The SOM was the less sensitive and most consistent network. The MLP had a critical minimum and showed high inconsistency, whereas fuzzy ARTMap was more sensitive and less accurate. Results indicate that sampling densities used in conventional soil surveys can serve as a reference to predict soil orders in Serra Gaúcha.

Overcoming limitations of modelling rare species by using ensembles of small models

1. Species distribution models (SDMs) have become a standard tool in ecology and applied conservation biology. Modelling rare and threatened species is particularly important for conservation purposes. However, modelling rare species is difficult because the combination of few occurrences and many predictor variables easily leads to model overfitting. A new strategy using ensembles of small models was recently developed in an attempt to overcome this limitation of rare species modelling and has been tested successfully for only a single species so far. Here, we aim to test the approach more comprehensively on a large number of species including a transferability assessment. 2. For each species numerous small (here bivariate) models were calibrated, evaluated and averaged to an ensemble weighted by AUC scores. These 'ensembles of small models' (ESMs) were compared to standard Species Distribution Models (SDMs) using three commonly used modelling techniques (GLM, GBM, Maxent) and their ensemble prediction. We tested 107 rare and under-sampled plant species of conservation concern in Switzerland. 3. We show that ESMs performed significantly better than standard SDMs. The rarer the species, the more pronounced the effects were. ESMs were also superior to standard SDMs and their ensemble when they were independently evaluated using a transferability assessment. 4. By averaging simple small models to an ensemble, ESMs avoid overfitting without losing explanatory power through reducing the number of predictor variables. They further improve the reliability of species distribution models, especially for rare species, and thus help to overcome limitations of modelling rare species.

Geospatial approach for defining the Wildland-Urban Interface in the Alpine environment

Problems related to fire hazard and fire management have become in recent decades one of the most relevant issues in the Wildland-Urban Interface (WUI), that is the area where human infrastructures meet or intermingle with natural vegetation. In this paper we develop a robust geospatial method for defining and mapping the WUI in the Alpine environment, where most interactions between infrastructures and wildland vegetation concern the fire ignition through human activities, whereas no significant threats exist for infrastructures due to contact with burning vegetation. We used the three Alpine Swiss cantons of Ticino, Valais and Grisons as the study area. The features representing anthropogenic infrastructures (urban or infrastructural components of the WUI) as well as forest cover related features (wildland component of the WUI) were selected from the Swiss Topographic Landscape Model (TLM3D). Georeferenced forest fire occurrences derived from the WSL Swissfire database were used to define suitable WUI interface distances. The Random Forest algorithm was applied to estimate the importance of predictor variables to fire ignition occurrence. This revealed that buildings and drivable roads are the most relevant anthropogenic components with respect to fire ignition. We consequently defined the combination of drivable roads and easily accessible (i.e. 100 m from the next drivable road) buildings as the WUI-relevant infrastructural component. For the definition of the interface (buffer) distance between WUI infrastructural and wildland components, we computed the empirical cumulative distribution functions (ECDF) of the percentage of ignition points (observed and simulated) arising at increasing distances from the selected infrastructures. The ECDF facilitates the calculation of both the distance at which a given percentage of ignition points occurred and, in turn, the amount of forest area covered at a given distance. Finally, we developed a GIS ModelBuilder routine to map the WUI for the selected buffer distance. The approach was found to be reproducible, robust (based on statistical analyses for evaluating parameters) and flexible (buffer distances depending on the targeted final area covered) so that fire managers may use it to detect WUI according to their specific priorities.

Predictors of smoking craving during virtual reality exposure

Cue exposure treatment (CET) consists of controlled and repeated exposure to drugrelated stimuli in order to reduce cue-reactivity. Virtual reality (VR) has proved to be a promising tool for exposition. However, identifying the variables that can modulate the efficacy of this technique is essential for selecting the most appropriate exposure modality. The aim of this study was to determine the relation between several individual variables and self-reported craving in smokers exposed to VR environments. Fortysix smokers were exposed to seven complex virtual environments that reproduce typical situations in which people smoke. Self-reported craving was selected as the criterion variable and three types of variables were selected as the predictor variables: related to nicotine dependence, related to anxiety and impulsivity, and related to the sense of presence in the virtual environments. Sense of presence was the only predictor of self-reported craving in all the experimental virtual environments. Nicotine dependence variables added predictive power to the model only in the virtual breakfast at home. No relation was found between anxiety or impulsivity and self-reported craving. Virtual reality technology can be very helpful for improving CET for substance use disorders. However, the use of virtual environments would make sense only insofar as the sense of presence was high. Otherwise, the effectiveness of exposure might be affected. © 2012 by the Massachusetts Institute of Technology.

Invasive species and habitat degradation in Iberian streams: an analysis of their role in freshwater fish diversity loss

Mediterranean endemic freshwater fish are among the most threatened biota in the world. Distinguishing the role of different extinction drivers and their potential interactions is crucial for achieving conservation goals. While some authors argue that invasive species are a main driver of native species declines, others see their proliferation as a co-occurring process to biodiversity loss driven by habitat degradation. It is difficult to discern between the two potential causes given that few invaded ecosystems are free from habitat degradation, and that both factors may interact in different ways. Here we analyze the relative importance of habitat degradation and invasive species in the decline of native fish assemblages in the Guadiana River basin (southwestern Iberian Peninsula) using an information theoretic approach to evaluate interaction pathways between invasive species and habitat degradation (structural equation modeling, SEM). We also tested the possible changes in the functional relationships between invasive and native species, measured as the per capita effect of invasive species, using ANCOVA. We found that the abundance of invasive species was the best single predictor of natives’ decline and had the highest Akaike weight among the set of predictor variables examined. Habitat degradation neither played an active role nor influenced the per capita effect of invasive species on natives. Our analyses indicated that downstream reaches and areas close to reservoirs had the most invaded fish assemblages, independently of their habitat degradation status. The proliferation of invasive species poses a strong threat to the persistence of native assemblages in highly fluctuating environments. Therefore, conservation efforts to reduce native freshwater fish diversity loss in Mediterranean rivers should focus on mitigating the effect of invasive species and preventing future invasions

Commitment to military service among Finnish conscripts

This research establishes the primary components, predictors, and consequences of organizational commitment in the military context. Specifically, the research examines commitment to the military service among Finnish conscripts and whether initial affective commitment prior to service predicts later commitment, attitudes, behavior, and performance, and, furthermore, analyzes the changes in commitment and its possible outcomes. The data were collected from records as well as by surveys from 1,387 rank and file soldiers, immediately after they reported for duty, near the end of basic training, and near the end of 6 to 12 months of service. The data covered a wide array of predictor variables, including background items, attitudes toward conscription, mental and physical health, sociability, training quality, and leadership. Moreover, the archival data included such items as rank, criminal record, performance ratings, and the number of medical examines and exemptions. The measures were further refined based on the results of factor analysis and reliability tests. The results indicated that initial commitment significantly corresponded with expected adjustment, intentions to stay in the military, and acceptance of authority. Moreover, initial commitment moderately related to personal growth, perceived performance, and the number of effective service days at the end of service. During basic training, affective commitment was mostly influenced by challenging training, adjustment experiences, regimentation, and unit climate. At the end of service, committed soldiers demonstrated more personal growth and development in service, had higher-level expected performance, and less malingering during their service. Additionally, they had significantly more positive attitudes toward national defense. The results suggest that affective commitment requires adequate personal adjustment, experiences of personal growth and development, and satisfaction with unit dynamics and training. This research contributes to the theoretical discussion on organizational commitment and the will to defend the nation and advances developing models to support and manage conscript training, education, leadership, and personnel policy. This is achieved by determining the main factors and variables, including their relative strength, that affect commitment to the military service. These findings may also facilitate in designing programs aimed at reducing unwanted discharges and inadequate performance. In particular, these results provide tools for improving conscripts’ overall attachment to and identification with the military service.

Factors associated with the onset of hypertension in women of 50 years of age or more in a city in Southeastern Brazil

PURPOSE:To evaluate factors associated with hypertension in Brazilian women of 50 years of age or more.METHODS:A cross-sectional population based study using self-reports. A total of 622 women were included. The association between sociodemographic, clinical and behavioral factors and the woman's age at the onset of hypertension was evaluated. Data were analyzed according to cumulative continuation rates without hypertension, using the life-table method and considering annual intervals. Next, a Cox multiple regression analysis model was adjusted to analyze the occurrence rates of hypertension according to various predictor variables. Significance level was pre-established at 5% (95% confidence level) and the sampling plan (primary sampling unit) was taken into consideration.RESULTS:Median age at onset of hypertension was 64.3 years. Cumulative continuation rate without hypertension at 90 years was 20%. Higher body mass index (BMI) at 20–30 years of age was associated with a higher cumulative occurrence rate of hypertension over time (coefficient=0.078; p<0.001). Being white was associated with a lower cumulative occurrence rate of hypertension over time (coefficient= -0.439; p=0.003), while smoking >15 cigarettes/day was associated with a higher rate over time (coefficient=0.485; p=0.004).CONCLUSION:The results of the present study highlight the importance of weight control in young adulthood and of avoiding smoking in preventing hypertension in women aged ≥50 years.

Biomass and mineralmass estimates in a "cerrado" ecosystem

This work aimed to develop allometric equations for tree biomass estimation, and to determine the site biomass in different "cerrado" ecosystems. Destructive sampling in a "campo cerrado" (open savanna) was carried out at the Biological Reserve of Moji-Guaçu, State of São Paulo, southeastern Brazil. This "campo cerrado" (open savanna) grows under a tropical climate and on acid, low nutrient soils. Sixty wood plants were cut to ground level and measurements of diameter, height and weight of leaves and stems were taken. We selected the best equations among the most commonly used mathematical relations according to R² values, significance, and standard error. Both diameter (D) and height (H) showed good relationship with plant biomass, but the use of these two parameters together (DH and D²H) provided the best predictor variables. The best equations were linear, but power and exponential equations also showed high R² and significance. The applicability of these equations is discussed and biomass estimates are compared with other types of tropical savannas. Mineralmass was also estimated. "Cerrados" proved to have very important carbon reservoirs due to their great extent. In addition, high land-use change that takes place nowadays in the "cerrado" biome may significantly affect the global carbon cycle.

Factors affecting a woman's decision to leave an abusive relationship : the theory of planned behaviour /

The present study tested the appHcabiUty of Ajzen's (1985) theory of planned behaviour (TPB), an extension of Fishbein and Ajzen's (1975) theory of reasoned action (TRA), for the first time, in the context of abused women's decision to leave their abusive relationships. The TPB, as a means of predicting women's decision to leave their abusive partners' was drawn from Strube's (1988, 1991) proposed decision-making model based on the principle that the decision-making process is a rational, deliberative process, and regardless of outcome, was a result of a logical assessment of the available data. As a means of predicting those behaviours not under volitional control, Ajzen's (1985) TPB incorporated a measure of perceived behavioural control. Data were collected in two phases, ranging from 6 months to 1 year apart. It was hypothesized that, to the extent that an abused woman held positive attitudes, subjective norms conducive to leaving, and perceived control over leaving, she would form an intention to leave and thus, increase the likelihood of actually leaving her partner. Furthermore, it was expected that perceptions of control would predict leaving behaviour over and above attitude and subjective norm. In addition, severity and frequency of abuse were assessed, as were demographic variables. The TPB failed to account significantly for variability in either intentions or leaving behaviour. All of the variance was attributed to those variables associated with the theory of reasoned action, with social influence emerging as the strongest predictor of a woman's intentions. The poor performance of this model is attributed to measurement problems with aspects of attitude and perceived control, as well as a lack of power due to the small sample size. The insufficiency of perceived control to predict behaviour also suggests that, on the surface at least, other factors may be at work in this context. Implications of these results, and recommendations such as, the importance of obtaining representative samples, the inclusion of self-esteem and emotions as predictor variables in this model, a reevaluation of the target behaviovu" as nonvolitional, and longitudinal studies spanning a longer time period for future research within the context of decision-making are discussed.

Children's rights and parenting beliefs : a study of attitudes, values and emotions

The conceptualization of childhood has changed over the centuries and appears to be undergoing further change in our post-modern culture. While the United Nations Convention on the Right of the Child is designed to give children everywhere basic human rights while taking into consideration their special needs, no recent research has examined adult attitudes toward those rights. In an attempt to understand the attitudes adults hold regarding autonomy rights and to look for some factors that could predict those attitudes, the current study considers values, parenting style, emotions and the issue of parent status as possible predictor variables. A total of 90 participants took part in the research, which had both written and interview components. Results generally failed to establish a reliable set of predictors. However, some interesting information was obtained regarding the endorsement of children's autonomy rights and some general conclusions were reached about our view of children and their rights at the end of the twentieth century.

The Moderating Effects of Appearance Commentary on the Relationship Between Weight Status and Physical Activity Participation in Female College Students

When looked at individually, overweight status and a higher frequency of negative appearance commentary (AC) are associated with lower physical activity (PA) levels. However, the combined effect has yet to be examined. The purpose of this study was to examine if the frequency of AC moderated the relationship between weight status and PA in college-aged females. No significant differences in PA levels (F(1,99)=2.41, p=.12) were found between the Never Overweight and Previously/Presently Overweight groups. Significant correlations existed for both negative AC (r=-.30, p=.00) and positive AC (r=.20, p=.05) with PA participation. AC did not significantly moderate the relationship between weight status and PA (F(2,95)=.65, p=.52, R2 adjusted=.13) as the interaction term did not account for any additional increase in variance (ΔR2=.01). Overall, AC frequency does not moderate the relationship between weight status and PA; other predictor variables should be explored.

Factors Contributing to Reading Performance: An In-Depth Analysis of the “Boy Crisis”

This research used a quantitative study approach to investigate the “boy crisis” in Canada. Boy crisis advocates suggest that boys are being surpassed by girls on reading assessments and promote strategies to assist male students. A feminist framework was used in this study that allowed for an investigation and discussion of the factors that mediate between gender and success at reading comprehension, interpretation, and response to text without ignoring female students. Reading scores and questionnaire data compiled by the Pan-Canadian Assessment Program were used in this research, specifically the PCAP-13 2007 assessment of approximately 30,000 13-year-old students from all Canadian provinces and Yukon Territory (CMEC, 2008). Approximately 20,000 participants wrote the reading assessment, while 30,000 students completed the questionnaire responses. Predictor variables were tested using parametric tests such as independent samples t-test, one-way ANOVA, chi-square analysis, and Pearson r. Findings from this study indicate that although boys scored lower than girls on the PCAP-13 2007 reading assessment, factors were found to influence the reading scores of both male and female students to varying degrees. Socioeconomic status, perceptions of the reading material used in language arts classrooms, reading preference, reading interest, parental involvement, parental encouragement for reading, and self-efficacy were all found to affect the reading performance of boys and girls. Relationships between variables were also found and are discussed in this research. The analysis presented in this study allows parents, educators, and policy makers to begin to critically examine and re-evaluate boy crisis literature and offers suggestions on how to improve reading performance for all students of all socioeconomic backgrounds.

La thérapie conjugale en milieu naturel: Étude du lien entre attachement amoureux, satisfaction conjugale, mandat thérapeutique et résultat de la consultation

L’objectif général de cette thèse est d’examiner le lien entre l’attachement amoureux des conjoints, la satisfaction conjugale, le mandat thérapeutique et le résultat de la thérapie conjugale telle que conduite en milieu naturel. Afin d’atteindre cet objectif, des couples se présentant en thérapie conjugale ont d’abord complété une batterie de questionnaires comprenant l’Échelle d’ajustement dyadique (Spanier, 1976) et le Questionnaire sur les expériences d’attachement amoureux (Brennan, Clark, & Shaver, 1998). Les thérapeutes ont par la suite indiqué le mandat thérapeutique poursuivi avec chaque couple (réconciliation ou résolution de l’ambivalence), suivant la classification de Poitras-Wright et St-Père (2004). À la fin de la consultation, le jugement du thérapeute a été utilisé pour classer chacun des cas comme ayant abandonné ou complété le traitement. Les couples ayant complété la thérapie ont rempli l’Échelle d’ajustement dyadique au post-traitement. Dans le premier des articles composant cette thèse, le lien entre l’attachement amoureux et la satisfaction conjugale a été examiné auprès d’un échantillon de 172 couples en détresse débutant une thérapie conjugale, de même qu’auprès de 56 couples non en détresse débutant également une thérapie conjugale, pour fins de comparaison. Les résultats ont démontré que l’évitement de la proximité semble être une caractéristique distinctive des couples en détresse et que cette dimension de l’attachement est fortement liée à l’insatisfaction conjugale de ce même groupe. Dans le deuxième article, le mandat thérapeutique, l’attachement amoureux et la satisfaction conjugale ont été examinés en tant que prédicteurs de l’abandon de la thérapie conjugale, auprès de 141 couples. Les résultats ont notamment démontré qu’un mandat de résolution de l’ambivalence augmente les probabilités d’abandon de la thérapie conjugale. De plus, les prédicteurs du résultat de la thérapie ont également été examinés dans ce second article. Les résultats obtenus au moyen d’analyses acteur-partenaire ont démontré que la satisfaction conjugale pré-traitement apparaît comme le meilleur prédicteur de la satisfaction conjugale post-traitement, et ce, malgré l’inclusion de l’attachement amoureux parmi les variables investiguées. Considérés dans leur ensemble, les résultats de cette thèse suggèrent que l’insécurité d’attachement serait fortement associée à l’insatisfaction des couples en détresse, mais ne nuirait pas pour autant à l’obtention d’un résultat positif en thérapie conjugale. En somme, cette thèse contribue à l’avancement des connaissances en se penchant sur l’utilité de la théorie de l’attachement en thérapie conjugale et en soulignant la nécessité de tenir compte des mandats thérapeutiques dans les futures études en thérapie conjugale. Les implications cliniques des résultats et des recommandations pour la recherche clinique sont présentées dans la conclusion de l’ouvrage.

L'effort associé à la reconnaissance de la parole chez les adultes et les personnes aînées

L’objectif principal de cette thèse était de quantifier et comparer l’effort requis pour reconnaître la parole dans le bruit chez les jeunes adultes et les personnes aînées ayant une audition normale et une acuité visuelle normale (avec ou sans lentille de correction de la vue). L’effort associé à la perception de la parole est lié aux ressources attentionnelles et cognitives requises pour comprendre la parole. La première étude (Expérience 1) avait pour but d’évaluer l’effort associé à la reconnaissance auditive de la parole (entendre un locuteur), tandis que la deuxième étude (Expérience 2) avait comme but d’évaluer l’effort associé à la reconnaissance auditivo-visuelle de la parole (entendre et voir le visage d’un locuteur). L’effort fut mesuré de deux façons différentes. D’abord par une approche comportementale faisant appel à un paradigme expérimental nommé double tâche. Il s’agissait d’une tâche de reconnaissance de mot jumelée à une tâche de reconnaissance de patrons vibro-tactiles. De plus, l’effort fut quantifié à l’aide d’un questionnaire demandant aux participants de coter l’effort associé aux tâches comportementales. Les deux mesures d’effort furent utilisées dans deux conditions expérimentales différentes : 1) niveau équivalent – c'est-à-dire lorsque le niveau du bruit masquant la parole était le même pour tous les participants et, 2) performance équivalente – c'est-à-dire lorsque le niveau du bruit fut ajusté afin que les performances à la tâche de reconnaissance de mots soient identiques pour les deux groupes de participant. Les niveaux de performance obtenus pour la tâche vibro-tactile ont révélé que les personnes aînées fournissent plus d’effort que les jeunes adultes pour les deux conditions expérimentales, et ce, quelle que soit la modalité perceptuelle dans laquelle les stimuli de la parole sont présentés (c.-à.-d., auditive seulement ou auditivo-visuelle). Globalement, le ‘coût’ associé aux performances de la tâche vibro-tactile était au plus élevé pour les personnes aînées lorsque la parole était présentée en modalité auditivo-visuelle. Alors que les indices visuels peuvent améliorer la reconnaissance auditivo-visuelle de la parole, nos résultats suggèrent qu’ils peuvent aussi créer une charge additionnelle sur les ressources utilisées pour traiter l’information. Cette charge additionnelle a des conséquences néfastes sur les performances aux tâches de reconnaissance de mots et de patrons vibro-tactiles lorsque celles-ci sont effectuées sous des conditions de double tâche. Conformément aux études antérieures, les coefficients de corrélations effectuées à partir des données de l’Expérience 1 et de l’Expérience 2 soutiennent la notion que les mesures comportementales de double tâche et les réponses aux questionnaires évaluent différentes dimensions de l’effort associé à la reconnaissance de la parole. Comme l’effort associé à la perception de la parole repose sur des facteurs auditifs et cognitifs, une troisième étude fut complétée afin d’explorer si la mémoire auditive de travail contribue à expliquer la variance dans les données portant sur l’effort associé à la perception de la parole. De plus, ces analyses ont permis de comparer les patrons de réponses obtenues pour ces deux facteurs après des jeunes adultes et des personnes aînées. Pour les jeunes adultes, les résultats d’une analyse de régression séquentielle ont démontré qu’une mesure de la capacité auditive (taille de l’empan) était reliée à l’effort, tandis qu’une mesure du traitement auditif (rappel alphabétique) était reliée à la précision avec laquelle les mots étaient reconnus lorsqu’ils étaient présentés sous les conditions de double tâche. Cependant, ces mêmes relations n’étaient pas présentes dans les données obtenues pour le groupe de personnes aînées ni dans les données obtenues lorsque les tâches de reconnaissance de la parole étaient effectuées en modalité auditivo-visuelle. D’autres études sont nécessaires pour identifier les facteurs cognitifs qui sous-tendent l’effort associé à la perception de la parole, et ce, particulièrement chez les personnes aînées.