Vegetation composition differentiation and species-environment relationships in the northern part of Isfahan Province,Iran

A considerable proportion of Iran's territory is covered with arid and semi-arid rangelands and mismanagement and overexploitation of those rangelands have resulted in serious ecological degradation. Thus, the need is pressing to examine the present species composition and the relationships with environmental factors for providing the needed scientific references to species conservation and ecological rehabilitation efforts. The aims of this study were to examine the species composition and to delineate the most important factors influencing the distributions of plant species and groups in the northern rangelands of Isfahan Province(Iran) using two-way indicator species analysis(TWINSPAN), detrended correspondence analysis(DCA), principal component analysis(PCA), and canonical correspondence analysis(CCA). Field investigations were conducted in the growing season of 2014 using stratified random method in 22 homogeneous sampling units. In total, 75 plant species belonging to 52 genera and 19 families were identified. The most important families were Asteraceae and Papilionaceae, the most important genera were Astragalus, Cousinia, and Acanthophyllum, and the most important species were Artemisia aucheri and Artemisia sieberi. Plant species were classified into 10 groups using TWINSPAN. DCA was used to estimate the magnitude of changes in species composition along the first two ordination axes to provide gradient length estimations for PCA and CCA ordinations. The first three PCA axes and the first three CCA axes demonstrated similar cumulative percentage of variance, indicating that the environmental factors(selected by PCA) used in CCA ordination were acceptable for explaining the species composition and the distributions. CCA ordination showed that the first axis was closely related to elevation, slope, surface bare soil cover, surface litter cover, gravel proportion, organic matter, total nitrogen, Ca CO3 content, and grazing intensity and that the second axis was closely related to sand proportion, silt proportion, clay proportion, and saturation percentage. Among these factors, elevation was the most effective factor to separate the plant groups and grazing was the major cause of rangeland degradation.     

基于MaxEnt模型的菜豆象和蚕豆象在中国的适生区预测

为减少外来入侵物种菜豆象Acanthoscelides obtectus和蚕豆象Bruchus rufimanus对中国造成的潜在威胁,收集这2种豆象的全球地理分布数据,采用Pearson相关性分析和主成分分析分别从19个环境变量中筛选关键环境变量,采用MaxEnt模型对历史气候条件下和未来气候情景下这2种豆象在中国的适生区进行预测,并对预测结果进行分析。结果显示,经Pearson相关性分析共筛选出4个关键环境变量用于菜豆象适生性区的模型构建,分别为最暖季度平均温度、最干月份降水量、年气温变化范围及最湿季度降水量,其对MaxEnt模型的累积贡献率分别为31.6%、28.4%、26.3%和13.7%;经Pearson相关性分析共筛选出4个主要关键环境变量用于蚕豆象适生性区的模型构建,分别为最冷季度平均温度、最干月份降水量、最热月份最高温度和最湿月份降水量,其对MaxEnt模型的累积贡献率分别为48.5%、39.5%、7.8%和4.2%。MaxEnt模型重复运行10次后,菜豆象训练数据的平均AUC值为0.938,蚕豆象训练数据的平均AUC值为0.963,均显著高于随机模型的AUC值,表明基于MaxEnt模型的菜豆象和蚕豆象在中国适生区的预测结果准确。未来气候情景下,这2种豆象在中国的适生区均呈现向北扩张的趋势,需加强对这2种豆象的检疫与防治,严防发生区域进一步扩大。     

苋属4种外来有害杂草在中国的适生区预测

［目的］ 明确反枝苋(Amaranthus retroflexus)、凹头苋(A. lividus)、刺苋(A. spinosus)、皱果苋(A. viridis)4种有害杂草在中国的适生区,为有效制定检疫措施和防治决策提供依据。［方法］基于4种苋属杂草已有的分布点数据,使用GARP和Maxent两个生态位模型对其在中国的适生区进行预测。［结果］ 对4种苋属植物适生区影响最大的环境因子主要为雨日频率、海拔、极端低温、水汽压、坡度。模型评价表明,Maxent和GARP两模型对4种杂草的分布均能较好地进行预测,Maxent的结果稍好于GARP。以Maxent为主,GARP作参考,得出苋属4个种在中国的适生区主要集中在华东地区、华北的部分地区、西北和东北的少数地区、除西藏和四川西部以外的西南地区以及中南的大部分地区。［结论］ 建议相关部门对4种杂草适生区域及其周边做好相应的预警和治理工作。     

Potential distribution of two Ambrosia species in China under projected climate change

The invasion of Ambrosia artemisiifolia and Ambrosia trifida from their native range to occupy large areas in China has raised considerable concern. Using the maximum entropy (Maxent) method, we developed models for each Ambrosia species, based on occurrence records from both native ranges (North America) and their invaded ranges (e.g. northern and south‐western Europe) to predict the availability and distribution of suitable habitats for these two species in China. For each species, we also assessed potential shifts in habitat suitability for the year 2050, using three general circulation models (GCMs) and two emission scenarios. Elevation and average mean precipitation in October contributed most to model development for both species. Potential distribution projections under future climatic change scenarios suggested an averaged percentage of suitable area (2.21%) and habitat gain (1.49%) in A. artemisiifolia distribution, with further expansion to environmentally favourable locations in south‐east coastal regions, northern Taiwan and the Beijing–Tianjin–Tangshan area in northern China. Future predicted percentage of suitable area for A. trifida was 0.03% with a very limited suitable habitat gain of <1% although this species had the potential to continue to spread in northern China. Our findings suggest that management priorities should be focused on A. artemisiifolia, whilst effective control strategies for A. trifida may be optimised by concentrating efforts on those relatively fewer regions of China where the species is currently abundant.     

Mapping the current and future distributions of Onosma species endemic to Iran

Climate change may cause shifts in the natural range of species especially for those that are geographically restricted and/or endemic species. In this study, the spatial distribution of five endemic and threatened species belonging to the genus Onosma (including O. asperrima, O. bisotunensis, O. kotschyi, O. platyphylla, and O. straussii) was investigated under present and future climate change scenarios: RCP2.6 (RCP, representative concentration pathway; optimistic scenario) and RCP8.5 (pessimistic scenario) for the years 2050 and 2080 in Iran. Analysis was conducted using the maximum entropy (MaxEnt) model to provide a basis for the protection and conservation of these species. Seven environmental variables including aspect, depth of soil, silt content, slope, annual precipitation, minimum temperature of the coldest month, and annual temperature range were used as main predictors in this study. The model output for the potential habitat suitability of the studied species showed acceptable performance for all species (i.e., the area under the curve (AUC)>0.800). According to the models generated by MaxEnt, the potential current patterns of the species were consistent with the observed areas of distributions. The projected climate maps under optimistic and pessimistic scenarios (RCP2.6 and RCP8.5, respectively) of 2050 and 2080 resulted in reductions and expansions as well as positive range changes for all species in comparison to their current predicted distributions. Among all species, O. bisotunensis showed the most significant and highest increase under the pessimistic scenario of 2050 and 2080. Finally, the results of this study revealed that the studied plant species have shown an acute adaptability to environmental changes. The results can provide useful information to managers to apply appropriate strategies for the management and conservation of these valuable Iranian medicinal and threatened plant species in the future.     

Impacts of livestock grazing on a savanna grassland in Kenya

The dynamics of most rangelands in Kenya remain to be poorly understood.This paper provides baseline information on the response of a semiarid rangeland under different livestock grazing regimes on land inhabited by the Massai people in the east side of Amboseli National Park in Kenya.The data were collected from grasslands designated into four types:(1) grassland from previous Massai settlements that had been abandoned for over twenty years;(2) grassland excluded from livestock grazing for eight years;(3) a dry season grazing area;and (4) a continuous grazing area where grazing occurred throughout all seasons.Collected data included grass species composition,grass height,inter-tuft distance,standing grass biomass and soil characteristics.The results indicated that continuous grazing area in semiarid rangelands exhibited loss of vegetation with negative,long-term effects on grass functional qualities and forage production,whereas grassland that used traditional Maasai grazing methods showed efficiency and desirable effects on the rangelands.The results also showed that abandoned homestead sites,though degraded,were important nutrient reservoirs.     

Modelling the impact of climate change on rangeland forage production using a generalized regression neural network: a case study in Isfahan Province,Central Iran

Monitoring of rangeland forage production at specified spatial and temporal scales is necessary for grazing management and also for implementation of rehabilitation projects in rangelands. This study focused on the capability of a generalized regression neural network(GRNN) model combined with GIS techniques to explore the impact of climate change on rangeland forage production. Specifically, a dataset of 115 monitored records of forage production were collected from 16 rangeland sites during the period 1998–2007 in Isfahan Province, Central Iran. Neural network models were designed using the monitored forage production values and available environmental data(including climate and topography data), and the performance of each network model was assessed using the mean estimation error(MEE), model efficiency factor(MEF), and correlation coefficient(r). The best neural network model was then selected and further applied to predict the forage production of rangelands in the future(in 2030 and 2080) under A1 B climate change scenario using Hadley Centre coupled model. The present and future forage production maps were also produced. Rangeland forage production exhibited strong correlations with environmental factors, such as slope, elevation, aspect and annual temperature. The present forage production in the study area varied from 25.6 to 574.1 kg/hm~2. Under climate change scenario, the annual temperature was predicted to increase and the annual precipitation was predicted to decrease. The prediction maps of forage production in the future indicated that the area with low level of forage production(0–100 kg/hm~2) will increase while the areas with moderate, moderately high and high levels of forage production(≥100 kg/hm~2) will decrease both in 2030 and in 2080, which may be attributable to the increasing annual temperature and decreasing annual precipitation. It was predicted that forage production of rangelands will decrease in the next couple of decades, especially in the western and southern parts of Isfahan Province. These changes are more pronounced in elevations between 2200 and 2900 m. Therefore, rangeland managers have to cope with these changes by holistic management approaches through mitigation and human adaptations.     

基于MaxEnt和ArcGIS的黑腹尼虎天牛潜在地理分布预测

为明确我国检疫性有害生物黑腹尼虎天牛Neoclytus acuminatus的潜在地理分布范围，基于MaxEnt模型、ArcGIS软件及全球分布数据预测当前气候和未来气候（2个情景）条件下黑腹尼虎天牛在全球和中国的潜在分布区域，并分析影响黑腹尼虎天牛分布的关键环境变量。结果显示，MaxEnt模型的曲线下面积（area under curve，AUC）为0.962，表明模型预测结果可靠；影响黑腹尼虎天牛潜在地理分布的5个关键环境变量分别是5月平均降雨量、11月平均最高温度、温度变化方差、7月平均降雨量和最湿季度平均气温，贡献率分别为40.5%、33.2%、23.9%、2.2%和0.1%。在当前气候条件下，黑腹尼虎天牛在全球的适生区较广泛，总面积约为3 928.63×10⁴ km²，且在我国湖北、安徽及浙江等省存在高适生区和中适生区；在未来气候条件下，黑腹尼虎天牛在全球范围内的适生区总面积会进一步增加，并且在我国的高适生区面积也会进一步扩大。     

Rangelands of Central Asia: challenges and opportunities

Rangelands of Central Asia(referring to Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan in this study), the largest contiguous area of grazed land in the world, serve as an important source of livelihood for pastoral and agro-pastoral communities in this region. They also play an important role in absorbing CO2 as a global carbon sink. However, unsustainable management of rangelands has led to their degradation hugely by downgrading their potential agro-ecological, environmental and socio-economical roles. This paper reviewed the rangeland degradation in Central Asia, a topic which so far has received only scant coverage in the international scientific literature. It also provided examples of successful experiences and outlined possible options that land managers can adopt to enhance the sustainable management of these vast degraded rangelands. The experiences and lessons described in this paper may also be relevant for other degraded rangeland areas, especially in the developing countries. The causes of rangeland degradation within the Central Asian region are numerous, complex and inter-related. Therefore, while addressing the factors associated with improper rangeland management may shed some light on the causes of rangeland degradation, the scope of this paper would not be all-encompassing for the major causes of degradation. There is a need to develop and widely apply the viable and locally accepted and adapted packages of technical, institutional and policy options for sustainable rangeland management. Incentivizing the collective action of small-scale pastoralists who group together to facilitate access to remote pastures can reduce the degree of overgrazing within community pastures, such as those near the settlements. We also found that migratory grazing through pooling of resources among small-scale pastoralists can increase household income. After their independence, most Central Asian countries adopted various rangeland tenure arrangements. However, the building of enhanced capacities of pasture management and effective local rangeland governance structures can increase the likelihood, which will be sustainable and equitable. Finally, this paper presented several promising technical options, aiming at reversing the trend of rangeland degradation in Central Asia.     

Estimation of spatial and temporal changes in net primary production based on Carnegie Ames Stanford Approach (CASA) model in semi-arid rangelands of SemiromCounty,Iran

Net primary production (NPP) is an indicator of rangeland ecosystem function. This research assessed the potential of the Carnegie Ames Stanford Approach (CASA) model for estimating NPP and its spatial and temporal changes in semi-arid rangelands of Semirom County, Iran. Using CASA model, we estimated the NPP values based on monthly climate data and the normalized difference vegetation index (NDVI) obtained from the MODIS sensor. Regression analysis was then applied to compare the estimated production data with observed production data. The spatial and temporal changes in NPP and light utilization efficiency (LUE) were investigated in different rangeland vegetation types. The standardized precipitation index (SPI) was also calculated at different time scales and the correlation of SPI with NPP changes was determined. The results indicated that the estimated NPP values varied from 0.00 to 74.48 g C/(m²?a). The observed and estimated NPP values had different correlations, depending on rangeland conditions and vegetation types. The highest and lowest correlations were respectively observed in Astragalus spp.-Agropyronspp. rangeland (R²=0.75) with good condition and Gundeliaspp.-Cousiniaspp. rangeland (R²=0.36) with poor and very poor conditions. The maximum and minimum LUE values were found in Astragalus spp.-Agropyronspp. rangeland (0.117 g C/MJ) with good condition and annual grasses-annual forbs rangeland (0.010 g C/MJ), respectively. According to the correlations between SPI and NPP changes, the effects of drought periods on NPP depended on vegetation types and rangeland conditions. Annual plants had the highest drought sensitivity while shrubs exhibited the lowest drought sensitivity. The positive effects of wet periods on NPP were less evident in degraded areas where the destructive effects of drought were more prominent. Therefore, determining vegetation types and rangeland conditions is essential in NPP estimation. The findings of this study confirmed the potential of the CASA for estimating rangeland production. Therefore, the model output maps can be used to evaluate, monitor and optimize rangeland management in semi-arid rangelands of Iran where MODIS NPP products are not available.     

基于物种分布模型对未来气候变化下云南松毛虫在四川省适生区的预测

为明确云南松毛虫Dendrolimus houi于未来气候变化下在四川省的分布情况,运用最大熵（maximum entropy,MaxEnt）模型中的刀切法和Pearson相关系数分析法对未来气候数据（2050年和2070年）、林地因子及人为因子进行权重划分,筛选出影响云南松毛虫潜在分布的重要且相关系数较低的环境因子,并结合openModeller中的人工神经网络（artificial neural network,ANN）模型、生物气候（BioClim）模型、气候空间（climate space,CS）模型、气候信封（envelope score,ES）模型、基于统计概率和规则集的遗传算法（genetic alorithm for rule-set production,GARP）模型、MaxEnt模型和支持向量机（support vector machine,SVM）模型对云南松毛虫未来的适生区进行预测,利用AUC（area under curve）评价模型的精确度。结果发现,最暖季降水量、人类足迹指数、最冷季降水量和海拔对云南松毛虫在未来气候条件下的潜在分布有较强影响,在2050年贡献率分别为27.2%、16.0%、2.0%和4.9%,在2070年贡献率分别为20.6%、16.8%、9.7%和4.9%。对比7种模型的预测结果,发现SVM模型在2050年和2070年对云南松毛虫适生区预测的AUC为0.93,预测精确度最高,具有较高的可信度;该模型预测结果显示,从2050年至2070年云南松毛虫在四川省的总适生面积增加了4 269.8 km²,其中,中、低适生区面积共减少了17 185.8 km²,高适生区面积增加了21 455.6 km²。     

草地植物入侵的预防和控制

外来植物入侵属于生物入侵的范畴,对生态、环境和生物多样性具有破坏性影响。植物入侵的途径很多,但有意引进是最主要途径。本文阐述了国外一些地方由于植物入侵所造成的损失和教训。重点介绍了我国植物入侵的形势及主要有害外来植物种类、来源、入侵途径、分布范围及危害程度,并根据国际、国内目前应对植物入侵的作法和未来趋势,提出了预防和控制植物入侵的办法和措施。认为预防和控制植物入侵是草地保护的一项重要任务。     

基于生态位模型的番荔枝实蝇潜在适生性分布预测(英文)

番荔枝实蝇Ceratitis anonae(Graham)是一种重要的外来入侵性检疫害虫。在广东口岸,其幼虫连续从入境旅客所携带的水果中被检出。目前关于番荔枝实蝇潜在适生性分布的研究进行得很少,但对于我国的生物生态安全却有重要意义。本研究中,我们使用3种生态位模型(ENFA模型,马氏典型性模型和Maxent模型)对番荔枝实蝇在中国以及全球范围内的潜在适生性分布区域进行了预测分析。结果显示:Maxent模型拥有最好的预测精确度,马氏典型性模型次之,而ENFA模型的预测精确度最差;Maxent模型和马氏典型性模型的预测精确度无显著性差异;根据Maxent模型的预测结果,番荔枝实蝇在中国的潜在适生区主要是广西、广东、海南以及云南的少部分地区。分析结果显示,番荔枝实蝇从境外传入中国南部地区并最终在上述地区定殖的风险可能性存在,但风险较小。另外,折刀法(Jackknife)分析显示,6种环境因子,例如地面霜冻频率、年平均降雨量、十月降雨量、四月降雨量、年最低温度以及蒸气压,对于番荔枝实蝇在全球和局部地区的分布模式有显著的影响。     

Assessing the potential distribution of insect pests: case studies on large pine weevil (Hylobius abietis L) and horse‐chestnut leaf miner (Cameraria ohridella) under present and future climate conditions in European forests

Forest insect pests represent a serious threat to European forests and their negative effects could be exacerbated by climate change. This paper illustrates how species distribution modelling integrated with host tree species distribution data can be used to assess forest vulnerability to this threat. Two case studies are used: large pine weevil (Hylobius abietis L) and horse‐chestnut leaf miner (Cameraria ohridella Deschka & Dimi?) both at pan‐European level. The proposed approach integrates information from different sources. Occurrence data of insect pests were collected from the Global Biodiversity Information Facility (GBIF), climatic variables for present climate and future scenarios were sourced, respectively, from WorldClim and from the Research Program on Climate Change, Agriculture and Food Security (CCAFS), and distributional data of host tree species were obtained from the European Forest Data Centre (EFDAC), within the Forest Information System for Europe (FISE). The potential habitat of the target pests was calculated using the machine learning algorithm of Maxent model. On the one hand, the results highlight the potential of species distribution modelling as a valuable tool for decision makers. On the other hand, they stress how this approach can be limited by poor pest data availability, emphasizing the need to establish a harmonised open European database of geo‐referenced insect pest distribution data.     

Discriminating and monitoring rangeland condition classes with MODIS NDVI and EVI indices in Iranian arid and semi-arid lands

Monitoring is essential for appropriate rangeland management. The present study aimed to examine the potential of moderate resolution imaging spectroradiometer (MODIS) satellite imagery in rangeland condition assessment and monitoring within and across vegetation types in the arid and semi-arid rangelands of central Iran. The normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) were calculated from MODIS Aqua/Terra Level 1B data (related to 2003–2013). The obtained values were compared with vegetation cover measurements and rangeland condition classes at 110 sampling sites using linear regression, one-way analysis of variance (ANOVA), independent-samples t-tests, and Tukey’s pairwise comparisons. The results showed that two indices made stronger predictions of vegetation cover within a vegetation type (R²?>?0.87, P?R²?>?0.51, P?P?≤?0.001) in predicting rangeland condition classes (poor, fair, and good), but their performance varied between vegetation types. The NDVI classified about 73, 19, and 7.5% of the rangelands in poor, fair, and good condition classes, respectively. The good performance of MODIS NDVI index at different landscapes indicates that this index has high potential in detecting vegetation cover and discriminating different condition classes, therefore, it can be used to aid field- based techniques in rangeland condition assessment and monitoring.     

基于GIS的三江源地区物种多样性保护优先性分析

根据我们的多次科考调查和青海省对野生动物的调查结果等有关资料图件,对三江源地区的物种多样性和人类干扰因素等进行了分析和评估。选择9种鸟类、15种哺乳动物、20种植物作为本研究的指示物种,利用GIS制作了主要的珍稀濒危野生动植物分布图。选择放牧、人口、居民点、公路等因素作为生物多样性的主要干扰因素,制作出三江源地区干扰强度空间分布图。然后根据物种的丰富度和特有性、脆弱性和干扰程度等指标分析了三江源地区物种多样性保护优先性地区,做出了物种多样性保护优先地区空间分布图。     

Herbaceous vegetation response to controlled fire and postfire defoliation in a semiarid rangeland of Argentina

Semiarid rangelands of central Argentina are shaped by fire and grazing. The aim of this study was to quantify the effect of controlled fire on aboveground biomass production of (1) herbaceous species as a whole group and (2) herbaceous species subclassified as desirable or undesirable perennial grasses and as herbaceous forbs at the end of two growing seasons after an experimental fire in a semiarid rangeland of central Argentina. The effect of the length of the resting period after fire on aboveground biomass production of desirable perennial grass species was also assessed; treatments consisted in defoliations at the end of the first and second postfire growing seasons, at the beginning and at the end of the second postfire growing season and only at the end of the second postfire growing season. Total biomass production of the burned treatments was higher than that of the control treatment, the difference being attributed to an increase in desirable perennial grass species biomass. This in turn was negatively affected by defoliation at the end of the first postfire growing season, suggesting that the beneficial effect of fire on biomass production might have been overridden by the stress imposed by an early defoliation. Overall, the results of the present study suggest that grazing should be held back for at least a whole year after the fire to achieve the pursued aim of increasing forage production by the use of controlled fire.     

The Struggle for a Uniform Monitoring System

Development of rangeland inventory and monitoring was initiated as early as in 1908. The early efforts were simple methods to establish rangeland stocking rates. With advancing scientific methodologies and the recognition for moitoring needs, the evaluation of range conditions and management planning increase well after WWII by a number of US federal agencies. By 1970s the environmental protection measures were first emphasized in the USA and were applied also to rangeland utilization. Since then the approaches and methodologies for evaluation of rangelands have changed periodically by accommodating the tenets of federal administrations. The methodologies applied have been geared either to support the environmental preservation or to support economic exploitation. There is still no system, however, for national or regional reporting on range conditions and trends. Local monitoring and management systems often lack scientific and objective basis. Local surveys, however, are important for evaluating existing conditions and should not be based on national reporting requirements. The ecological site concept has been adopted as the toxonomic unit for rangeland classification. It is important that monitoring methods are not biased, based on sufficient sample sizes, and not based on "estimations". The forage allocation based on one-time inventories are seriously flawed and the methods for evaluating stocking rates must be reevaluated. The same applies to the evaluating the rangeland "health." A unified national reporting system must be developed on scientific and objective basis.     

Predicting weed distribution at the regional scale in the native range: environmental determinants and biocontrol implications of Phyla nodiflora (Verbenaceae)

Alien plants produce severe environmental and economic losses in the territories they invade. Modelling the spatial distribution of alien species as a function of the environment in the native range has therefore become an essential first step in the struggle against invasions. Phyla nodiflora var. minor is a fast‐growing perennial herb native to South America that has spread through three continents, where it poses a major threat to biodiversity and significantly impacts on conservation and grazing systems, mainly in riparian areas. To assess the distribution of native Argentine populations of P. nodiflora as a function of the environment, we conducted long‐term roadside surveys and associated the occurrence of the plant with climatic, geographical, demographical and vegetation cover variables in a generalised linear mixed model. The plant was recorded in 230 of 431 sites, mostly east of 66°W and north of 39°S. According to the best model, which predicted the data 58% better than random assignment, its occurrence was associated with temperature variables (mean annual values and daytime range) and relative humidity. Based on these associations, we generated a presence probability map for the occurrence of P. nodiflora in southern South America. Understanding the environmental determinants of the distribution of weeds in their native range provides valuable baseline data to further manage the spread of alien species.     

What are the best predictors for invasive potential of weeds? Transferability evaluations of model predictions based on diverse environmental data sets for Flaveria bidentis

Ecological niche models are widely used in the study of weed invasions, yet best approaches for selecting ecologically relevant environmental predictors for weeds remain unresolved. Here, we evaluate niche model transferability based on diverse environmental data sets for an invasive herb, Flaveria bidentis. This species is native to South America, but has established populations in China that pose a threat to agriculture and animal husbandry. Relevant environmental data sets were selected via five statistical approaches: permutation importance (PI) and jackknife test (JK) in Maxent, variable importance identified by boosting regression trees (BRT), ecological niche factor analysis (ENFA) and a newly released algorithm based on a fluctuation index (FI). Climate spaces occupied by native South American and introduced Chinese populations were compared based on these environmental data sets. Native niche model predictions in China were compared across environmental data sets and model settings (i.e. default versus fine‐tuned Maxent settings). Results suggest that native and introduced populations occupy two distinct climate spaces, but that this divergence likely results from background effects. Niche models based on fine‐tuned Maxent settings generally showed better discrimination ability than those based on default settings. The best model discrimination in China was attained in the FI model using fine‐tuned settings, followed by the BRT model on default settings. The best models suggest that highly suitable areas at risk of invasion in China are to the west and north‐east of present distributional areas. Results presented here provide predictions for F. bidentis in particular, but also shed light on procedures for selecting ecologically relevant predictors for invasive species distributional predictions more generally.