Big Data Resource Planning for Food Safety: a Preliminary Exploration of the “Environment,Food and Health” Information Chain

Environmental pollution, food safety and health are closely linked. A key challenge in addressing the problem of food safety and protecting public health is building an integrated knowledge base to inform policy and strengthen governance. This requires breaking down the trans-departmental information barrier across the environment, food and health domains to ensure the effective flow of data and the efficient utilization of resources, and facilitate the collaborative governance of food safety. Achieving this will be crucial for the development of health and medical care in China in the era of big data. Currently, the information resources commanded by various departments are incomplete and fragmented. Data resources are also organized in vertical silos and there is a lack of data sharing within and across policy streams. To provide the basis for more effective integrated collection and analysis of data in future, this study summarizes the information resources of various departments whose work relates to interactions between environment, food and health, and presents measures to strengthen top-down design, and establish unified data standards and a big data sharing platform. It also points to the need for increased training of data analysts with interdisciplinary expertise.     

Ecosystem Carbon Allocation of a Temperate Mixed Forest and a Subtropical Evergreen Forest in China

Ecosystem carbon allocation can indicate ecosystem carbon cycling visually through its quantification within different carbon pools and carbon exchange. Using the ecological inventory and eddy covariance measurement applied to both a mature temperate mixed forest in Changbai Mountain (CBM) and a mature subtropical evergreen forest in Dinghu Mountain (DHM), we partitioned the ecosystem carbon pool and carbon exchange into different components, determined the allocation and analyzed relationships within those components. Generally, the total carbon stock of CBM was slightly higher than that of DHM due to a higher carbon stock in the arbor layer at CBM. It was interesting that the proportions of carbon stock in vegetation, soil and litter were similar for the two mature forests. The ratio of vegetation carbon pool to soil carbon stock was 1.5 at CBM and 1.3 at DHM. However, more carbon was allocated to the trunk and root from the vegetation carbon pool at CBM, while more carbon was allocated to foliage and branches at DHM. Moreover, 77% of soil carbon storage was limited to the surface soil layer (0-20 cm), while there was still plentiful carbon stored in the deeper soil layers at DHM. The root/shoot ratios were 0.30 and 0.25 for CBM and DHM, respectively. The rates of net ecosystem productivity (NPP) to gross ecosystem productivity (GPP) were 0.76 and 0.58, and the ratios of ecosystem respiration (Re) to GPP were 0.98 and 0.87 for CBM and DHM, respectively. The net ecosystem carbon exchange/productivity (NEP) was 0.24 t C ha^-1 yr^-1 for CBM and 3.38 t C ha^-1 yr^-1 for DHM. Due to the common seasonal and inter-annual variations of ecosystem carbon exchange resulting from the influence of environmental factors, it was necessary to use the long record dataset to evaluate the ecosystem sink capacity.     

基于倾向值匹配法的城市建成环境对居民生理健康的影响

伴随着城市化和机动化进程,肥胖及其引致的慢性疾病已成为中国严峻的社会问题。本文利用2010年中国社会综合调查、空间兴趣点POI、道路网等数据,探究城市建成环境对居民生理健康状况的影响,并通过倾向值匹配法控制自选择机制的干扰。研究发现：① 高密度的土地利用对居民总体的身体健康状况具有负向影响,而功能混合、支路网通达的城市肌理、以及充足的健康设施在降低身体质量指数BMI、抑制超重和减少慢性病方面发挥着积极作用。② 各类建成环境特征对中高社会阶层的影响集中在主观的健康感知,对中低阶层人群的影响则主要作用于客观的健康指标。③ 小尺度范围内的建成环境与中低阶层群体的身体健康水平存在更为密切的关联,但这一规律在中高社会阶层并不明显。本研究证明了主动式空间干预手段在促进居民生理健康过程中的有效性,进而就建成环境的优化策略提出了初步建议。     

广西森林转型与森林扰动遥感监测研究

近20年来由于封山育林政策的实施和人工乔木林的大面积种植,广西森林面积大幅增加,正在经历深刻的森林转型,森林转型过程中农林交错带和森林内部的结构重组都给生态系统带来了不容忽视的扰动。本文基于2000—2016年MODIS-EVI影像数据、1∶100万地貌图和4期土地利用数据,采用变化矢量法和Sen+Mann-Kendall对广西进行森林转型背景下的森林转型路径和森林内部扰动特征分析。研究表明:自2000年以来广西森林的EVI变化强度以无变化和低变化类型为主,总体EVI上升趋势略大于下降趋势,EVI显著下降主要分布于十万大山、大瑶山、海洋山和越城岭等山地区域和桂西北的百色水利枢纽和龙滩水利枢纽周围,显著上升部分主要集中于桂中南的左江-邕江-郁江流域平原丘陵区、桂西南喀斯特区和桂东北山地区地势低平的河谷地带。广西森林面积增加的土地来源于耕地和草地,森林转型路径表现为"经济增长型"和"森林短缺型"两种路径并存,或者存在更为复杂的复合型路径。广西森林内部扰动的减少型和扰动型分布于桂西北喀斯特和桂东北山地起伏度较大的陡坡区域,增加型分布于桂中南的平原丘陵地区。由于山地陡坡区域和喀斯特地区的生态敏感性,应更多关注扰动型和减少型在桂西北和桂东北地区的集中分布问题。     

The Role of Coffee in the Development of Southwest Ethiopia's Forests: Farmers' Strategies,Investor Speculation,and Certification Projects

This article examines the role of coffee in the agriculture of southwest Ethiopian highlands, a context characterized by recent huge private coffee estate developments, coffee certification projects and smallholdings. The analysis shows that smallholder farmers prioritize staple crops, and allocate land and labor to coffee only once food production objectives have been reached. It also underscores the diversity of the local peasantry, often treated as a homogeneous group in development projects designed to boost farmers' income through coffee production. Based on this analysis, the article evaluates the effects of recent coffee-oriented development policies: state forest concessions for coffee investment, and two coffee certification projects (UTZ-Certified and Fair-Trade/Organic). Because they omit the differentiated farmers' strategies, these certification initiatives fail to improve local populations' livelihoods and accelerate incipient socioeconomic inequalities. Moreover, when applied to big private plantations, they participate in farmers' dispossession from critical forest resources essential to households' food self-sufficiency.     

Spatial patterns of subsistence extraction of forest products – An indirect approach for estimation of forest degradation in dry forest

Successful implementation of a forest based climate change mitigation mechanism such as REDD + depends on robust and available methods for measurement and estimation of forest degradation. Currently available methods are for application in single-hit degradation incidents in high density humid forests. However, it has been suggested that gradual degradation, especially in dry forests, is more widespread and that methods are needed for measuring and estimating associated emissions. We assess the applicability of an indirect remote sensing approach for monitoring forest degradation: infrastructure and other indicators of human activities are mapped and used for spatial prediction of degradation activities. For proxy variables we tested distance to forest edge, distance to roads, and population pressure calculated as the sum of inhabitants per pixel in the Landscan 2010 population raster dataset multiplied by an inverse power distance decay function. Wood extraction incidents were counted in 160 plots in two dry forests in Tanzania with infrastructural entry from one side only. We analyzed the spatial pattern of forest degradation as a function of the chosen proxy variables using zero inflated count models which allows for an excess of zero counts. A jack-knife bootstrap using 10,000 runs was applied to optimize the population distance decay function. We found that the impact of forest degradation is highest near high population concentration, above 1000 individuals. Furthermore, distance to nearest forest edge or road was a significant proxy for estimation of the number of wood extraction incidents (p < 0.001), where degradation incidents decreased with increasing distance to forest edge or road. At 3000 m from the forest edge towards the forest core the probability of wood extraction is 20% and dropping. The population distance decay function was found to have a steep decline indicating a relative small impact on forest degradation. Further, and perhaps larger, studies are needed to be able to recommend a distance decay function for general application in Tanzania. However, the results are useful for understanding spatial patterns of wood harvesting as a function of distance to nearest forest edge or road in dry Miombo woodland areas with average population pressure at 1685 ± 101 persons within a radius of 4000 m from the wood extraction sites.     

The Geography of Malnutrition

The purpose of this article is to provide an overview of an emerging theme within the subfield of nutritional geography we call the geography of malnutrition. Work relating to malnutrition is a high-priority research topic, with growing relevance to geographical concepts, but there is no overview of geographical approaches to this theme. Using keyword searches in Google Scholar and Web of Science to obtain relevant publications, we identified the major foci of work within this theme: undernutrition, diseases that cause malnutrition, the nutrition transition, and critical and feminist approaches to malnutrition. We review these foci, provide examples of prominent work, and identify areas of research concerning malnutrition that are highly spatial but have yet to be effectively studied using geographic techniques.     

中国天山西部山区森林积雪消融过程观测分析——以巩乃斯河谷为例

通过对2013年春季中国科学院天山积雪与雪崩研究站区内阳坡无林地和阴坡不同开阔度森林内积雪深度、融雪速率以及常规气象的观测,分析了融雪期不同开阔度森林积雪的消融过程以及积雪表面能量平衡特征。结果表明:不同开阔度林冠下积雪深度具有相同的变化趋势,森林的林冠开阔度越大,林下积雪深度越大,林下积雪开始消融和完全消融的时间越晚,消融期也越长。森林积雪融雪开始和结束时间比阳坡无林地区晚20~30 d左右。融雪前期林冠开阔度越大,其林下融雪速率越小。融雪后期则森林开阔度越大,森林积雪的融雪速率越大。不同时期由于不同开阔度林冠下雪面能量收支以及雪层深度等物理特性的差异,从而使不同开阔度林冠下森林积雪融雪速率的相对大小,融雪速率最大值出现时间和日变化特征均不相同。晴天森林积雪的消融速率和日变化特征取决于净短波辐射和长波辐射变化特征。降水期间,其融雪速率的变化则主要受降水形式、降水量以及积雪深度等雪层特性的影响。     

基于GWR模型的道路网络对森林碳密度干扰的地理变异

以闽江上游地区为例,在分析三明市2007—2016年森林碳密度时空动态的基础上,采用常规的以及改进后的道路网络测度指标,应用缓冲区分析方法和地理加权回归（Geographically Weighted Regression,GWR）模型,从线上和面上分别探讨道路网络对森林碳密度干扰的地理变异规律。结果表明：① 碳密度受到道路网络的较大影响,路网影响域内外碳密度的大小排序为：路网影响域内＜整个研究区＜路网影响域外;多条道路影响域重叠区的碳密度(26.330 Mg/hm²)明显低于单条道路影响域的碳密度(37.406 Mg/hm²);不同等级道路影响域的碳密度由大到小依次为县道＞高速＞省道＞其它道路＞国道＞乡道。道路网络对2007—2016年碳密度的降低也有明显影响。② GWR模型的分析结果表明,路网对碳密度的影响程度随着样点的变化而变化,具有“空间非平稳性”。碳密度随着路网密度的增加而降低,而随着离道路距离的增加而增加。③ 研究区西北部和中部,GWR的回归系数及相关系数均较大,表明这2个区域道路对碳密度影响大且解释力皆较强。     

Latitudinal variation of leaf morphological traits from species to communities along a forest transect in eastern China

Comprehensive information on geographic patterns of leaf morphological traits in Chinese forests is still scarce. To explore the spatial patterns of leaf traits, we investigated leaf area (LA), leaf thickness (LT), specific leaf area (SLA), and leaf dry matter content (LDMC) across 847 species from nine typical forests along the North-South Transect of Eastern China (NSTEC) between July and August 2013, and also calculated the community weighted means (CWM) of leaf traits by determining the relative dominance of each species. Our results showed that, for all species, the means (± SE) of LA, LT, SLA, and LDMC were 2860.01 ± 135.37 mm², 0.17 ± 0.003 mm, 20.15 ± 0.43 m² kg^–1, and 316.73 ± 3.81 mg g^–1, respectively. Furthermore, latitudinal variation in leaf traits differed at the species and community levels. Generally, at the species level, SLA increased and LDMC decreased as latitude increased, whereas no clear latitudinal trends among LA or LT were found, which could be the result of shifts in plant functional types. When scaling up to the community level, more significant spatial patterns of leaf traits were observed (R ² = 0.46–0.71), driven by climate and soil N content. These results provided synthetic data compilation and analyses to better parameterize complex ecological models in the future, and emphasized the importance of scaling-up when studying the biogeographic patterns of plant traits.