Léon Fairmaire

One of the best ways to share and disseminate biodiversity information is through the digitization of data and making it available via online databases. The rapid growth of publicly available biodiversity data is not without problems which may decrease the utility of online databases. In this study we analyze taxonomic, geographic and temporal data gaps, and bias related to existing data on selected marine invertebrate occurrences along the coastline of two African countries, Mozambique and São Tomé and Príncipe. The final marine invertebrate dataset comprises of 19.910 occurrences, but 32% of the original dataset occurrences were excluded due to data gaps. Most marine invertebrates in Mozambique were collected in seagrasses, whereas in São Tomé and Príncipe they were mostly collected offshore. The dataset has a temporal coverage from 1816 to 2019, with most occurrences collected in the last two decades. This study provides baseline information relevant to a better understanding of marine invertebrate biodiversity data gaps and bias in these habitats along the coasts of these countries. The information can be further applied to complete marine invertebrate data gaps contributing to design informed sampling strategies and advancing refined datasets that can be used in management and conservation plans in both countries.

The amount of biological data available in online repositories is increasing at an exponential rate. However, data on marine invertebrate biodiversity resources from Mozambique and São Tomé and Príncipe are still sparse and scattered. Online repositories are useful instruments for biodiversity resea…

随机森林(Random forest)模型在2001年发表后得到广泛的关注。由于随机森林可以进行回归和判别等多种统计分析,而且不受正态性、方差齐性和自变量独立性等参数检验的前提条件的制约,其应用日益普遍,有被看作万能模型的趋势。实际上,随机森林是一种特点鲜明的模型,应用局部优化拟合观察值,在分析有偏效应关系的数据时,其结果往往不准确。本文以蝉科(Cicadidea)物种的分布数据为例,比较了随机森林在回归分析时与多元线性回归、广义可加模型和人工神经网络模型的差别,在判别分析时与线性判别分析的差别,强调了随机森林预测时的碎片化特点。结果显示随机森林在处理有多元共线性和交互作用的数据时,以及在判别…

Although Madagascar is one of the world’s most important biodiversity hotspots, the knowledge of its faunistic diversity is still incomplete, notwithstanding many field campaigns were organized since the 17th century until nowadays, leading to a huge number of vertebrate and invertebrate records. In…

In multiple regression Y ~ β0 + β1X1 + β2X2 + β3X1 X2 + ɛ., the interaction term is quantified as the product of X1 and X2. We developed fractional-power interaction regression (FPIR), using βX1M X2N as the interaction term. The rationale of FPIR is that the slopes of Y-X1 regression along the X2 gr…

Madagascar is one of the most biodiverse countries in Africa, due to its level of endemism and species diversity. However, the pressure of human activities threatens the last patches of natural vegetation in the country and conservation decisions are undertaken with limited data availability. In thi…

Determining the mechanisms that underlie species distributions and assemblages is necessary to effectively preserve biodiversity. This cannot be accomplished by examining a single taxonomic group, as communities comprise a plethora of interactions across species and trophic levels. Here, we examine …