Detecting the Influence of Best Management Practices on Vegetation Near Ephemeral Streams With Landsat Data

Various best management practices (BMPs) have been implemented on rangelands with the goals of controlling nonpoint source pollution, reducing the impact of livestock in ecologically important riparian areas, and improving grazing distribution. Providing off-stream water sources to livestock in pastures, cross-fencing, and rotational grazing are common rangeland BMPs that have demonstrated success in drawing livestock grazing pressure away from streams. We evaluated the effects of rangeland BMP implementation with six commercial-scale pastures in the northern mixed-grass prairie. Four pastures received a BMP suite consisting of off-stream water, cross-fencing, and deferred-rotation grazing, and two pastures did not receive BMPs. We hypothesized that the BMPs increased the quantity of riparian vegetation cover relative to the conditions in these pastures during the pre-BMP period and to the two pastures that did not receive BMPs. We used a series of 30-m Landsat normalized difference vegetation index (NDVI) images to track the spatial and temporal changes (1984–2010, n = 24) in vegetation cover, to which NDVI has been well correlated. Validation indicated that the remotely sensed signal from in-channel vegetation was representative of ground conditions. The BMP suite was associated with a 15% increase in the in-channel NDVI (0–30 m from stream centerline) and 18% increase in the riparian NDVI (30–180 m from stream center line). Conversely, the in-channel and riparian NDVI of non-BMP pastures declined 30% and 18% over the study period. The majority of change occurred within 2 yr of BMP implementation. The patterns of in-channel NDVI among pastures suggested that BMP implementation likely altered grazing distribution by decreasing the preferential use of riparian and in-channel areas. We demonstrated that satellite imagery time series are useful in retrospectively evaluating the efficacy of conservation practices, providing critical information to guide adaptive management and decision makers.     

Comparison of Ground-Measured and Image-Classified Mesquite (Prosopis glandulosa) Canopy Cover

Remote sensing has long been recognized as a rapid, inexpensive, nondestructive, and synoptic technique to study rangeland vegetation and soils. With respect to the worldwide phenomenon of woody plant invasion on many grasslands and rangelands, there is increasing interest in accurate and cost-effective quantification of woody plant cover and distribution over large land areas. Our objectives were to 1) investigate the relationship between ground-measured and image-classified honey mesquite (Prosopis glandulosa Torr.) canopy cover at three sites in north Texas using high spatial resolution (0.67-m) aerial images, and 2) examine the suitability of aerial images with different spatial resolutions (0.67-m, 1-m, and 2-m) for accurate estimation of mesquite canopy cover. The line intercept method and supervised maximum likelihood classifier were used to measure mesquite cover on the ground and on images, respectively. Images all were taken in September when mesquite foliage was photosynthetically active and most herbaceous vegetation was dormant. The results indicated that there were robust agreements between classified and ground-measured mesquite cover at all three sites with the coefficients of determination (r²) ≥ 0.95. Accuracy of lower spatial resolution images ranged from r² = 0.89–0.93, with the 2-m spatial resolution image on one of the sites at r² = 0.89. For all sites, the overall, producer's, and user's accuracies, and kappa statistics were 92% and 97%, 91% and 99%, 85% and 96%, and 0.82 and 0.95 for 2-m and 0.67-m spatial resolution images, respectively. Results showed that images at all three spatial resolution levels were effective for estimating mesquite cover over large and remote or inaccessible areas.     

Long-term impacts of livestock grazing and browsing in the Succulent Karoo: a 20-year study of vegetation change under different grazing regimes in Namaqualand§

This study used a fence-line contrast approach to investigate the long-term impact of high grazing pressure on the vegetation at a site in Namaqualand, South Africa. Forty pairs of permanently marked plots were surveyed in 1996, 2006 and 2016. The main objective was to investigate changes in the vegetation structure and species composition between the near-continuously grazed communal rangelands and the relatively lightly grazed commercial rangelands over the 20-year period. The results showed a decline in total vegetation cover in both commercial and communal rangelands in 2016 relative to the two earlier sampling periods. This can be attributed to the low rainfall in 2016 and was due largely to a reduction in annual plant cover, especially on the communal rangeland. Perennial shrub species provide a fodder bank that can be utilised by livestock in times of drought and can buffer short-term deficits in forage supply. However, the annuals that dominate the vegetation of the communal rangeland do not form such fodder banks and consequently do not have the same multi-year buffering capacity as perennial shrubs. This provides the mechanism whereby long-term continuous grazing decreases resilience to rainfall fluctuations and increases livestock variability, thereby promoting non-equilibrium-type dynamics in the system.     

Canadian Prairie Rangeland and Seeded Forage Classification Using Multiseason Landsat 8 and Summer RADARSAT-2

Rangeland and seeded forage in Canada’s Prairie provinces represent productive landscapes that provide multiple ecosystem services. Past efforts to map these resources at regional scales have not achieved consistently high accuracies as they are spatially variable in both ecology and management. In particular, Agriculture and Agri-food Canada needs to distinguish these land use classes from each other and from cropland in its annual national agricultural land cover inventory. Given the potential to distinguish these classes based on seasonal phenological differences, this study used multi-season Landsat 8 top-of-atmosphere reflectance data and derived vegetation and phenological indices, as well as mid-summer RADARSAT-2 data in random forest classification of two ecoregions in Alberta and Manitoba. Classification accuracy was compared for single and multi-date Landsat 8 variables, the vegetation index and phenological variable groups, RADARSAT-2 VV and VH backscatter intensity, and combined datasets. Variable importance analysis showed that spring Landsat 8 reflectance generally contributed most to class discrimination, but accuracy improved with the addition of Landsat 8 data from the other seasons. Vegetation indices and phenological variables produced similar accuracies and were deemed to not warrant the additional processing effort to derive them. RADARSAT-2 VH backscatter was the most important variable for the Manitoba study area, which is wetter with more vegetation structure variability than the Alberta study area. Backscatter intensity significantly increased overall accuracy when it was combined with one or two-season Landsat 8 data. The best overall accuracy was achieved using the three seasons of Landsat 8 and mid-summer RADARSAT-2 data, but it was not significantly better than that for two season Landsat 8 + RADARSAT-2. The methods presented in this paper provide a process for accurate and efficient classification of seeded forage, rangeland and cropland that can be applied over large areas in operational agricultural land cover inventory.     

A description of rangeland on commercial and communal land,Peddie district,South Africa

Analysis of a Landsat TM image from a rangeland near Peddie, Eastern Cape, revealed differences in two vegetation indices (normalised difference vegetation index, NDVI, and moving standard deviation index, MSDI) between communal and commercial rangeland. It was suggested that the difference in the MSDI reflected differences in rangeland condition. To assess whether or not any differences could be detected in the field, vegetation parameters were recorded (cover, species composition) along ten, paired 20-m transects. Based on species forage factors for commercial live-stock production, the commercial grassland was in a significantly better condition than the communal land. The commercial farmland had a higher occurrence of palatable species while the communal land was richer in non-palatable species. A TWINSPAN classification and the NDVI and MSDI values confirmed the marked difference between the communal and the commercial land. Both the vegetative field survey data and the satellite imagery showed that the communal land was transformed in comparison to the commercial land, and this difference can be attributed to differences in land-use.     

Assessing the Success of Postfire Reseeding in Semiarid Rangelands Using Terra MODIS

Successful postfire reseeding efforts can aid rangeland ecosystem recovery by rapidly establishing a desired plant community and thereby reducing the likelihood of infestation by invasive plants. Although the success of postfire remediation is critical, few efforts have been made to leverage existing geospatial technologies to develop methodologies to assess reseeding success following a fire. In this study, Terra Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data were used to improve the capacity to assess postfire reseeding rehabilitation efforts, with particular emphasis on the semiarid rangelands of Idaho. Analysis of MODIS data demonstrated a positive effect of reseeding on rangeland ecosystem recovery, as well as differences in vegetation between reseeded areas and burned areas where no reseeding had occurred (P < 0.05). We conclude that MODIS provides useful data to assess the success of postfire reseeding.     

2000-2019年青海地区植被总初级生产力遥感估算及时空变化分析

青海作为三江源所在地,监测其生态系统变化对我国的生态文明建设具有战略意义.植被总初级生产力(GPP)是陆地生态系统碳循环的重要组分.采用MODIS卫星遥感数据和土壤背景校正NIRv模型,结合3个地面站点的通量观测数据,估算了2000-2019年青海地区的GPP,并结合土地利用数据和气象数据分析了其时空分异特征及对气候变...     

Spatially Explicit Representation of State-and-Transition Models

The broad-scale assessment of natural resource conditions (e.g., rangeland health, restoration needs) requires knowledge of their spatial distribution. We argue that creating a database that links state-and-transition models (STMs) to spatial units is a valuable management tool for structuring ground-based observations, management planning for landscapes, and for housing information on the responses of land areas to management actions. To address this need, we introduce a multifactor classification system based on ecological sites and STMs that is directly linked to recent concepts of vegetation dynamics in rangelands. We describe how this classification was used as a basis for creating a spatial database and maps of ecological states. We provide an example of how the classification and mapping has been applied in over 1.2 million ha of public rangelands in southern New Mexico using aerial photo interpretation supplemented with existing inventory data and rapid field assessments. The resulting state map has been used by the Bureau of Land Management: 1) to design landscape-level shrub control efforts, 2) to structure and report district-wide rangeland health assessments, and 3) to evaluate locations for energy development. We conclude by discussing options for the development of state maps and their current limitations, including the use of satellite imagery and concepts for defining states. We argue that cataloging ecological states in a spatial context has clear benefits for rangeland managers because it connects STM concepts to specific land areas. State mapping provides a means to generate and store spatially explicit data resulting from tests of the propositions in STMs and conservation practices.     

Monitoring of Livestock Grazing Effects on Bureau of Land Management Land

Public land management agencies, such as the Bureau of Land Management (BLM), are charged with managing rangelands throughout the western United States for multiple uses, such as livestock grazing and conservation of sensitive species and their habitats. Monitoring of condition and trends of these rangelands, particularly with respect to effects of livestock grazing, provides critical information for effective management of these multiuse landscapes. We therefore investigated the availability of livestock grazing-related quantitative monitoring data and qualitative region-specific Land Health Standards (LHS) data across BLM grazing allotments in the western United States. We then queried university and federal rangeland science experts about how best to prioritize rangeland monitoring activities. We found that the most commonly available monitoring data were permittee-reported livestock numbers and season-of-use data (71% of allotments) followed by repeat photo points (58%), estimates of forage utilization (52%), and, finally, quantitative vegetation measurements (37%). Of the 57% of allotments in which LHS had been evaluated as of 2007, the BLM indicated 15% had failed to meet LHS due to livestock grazing. A full complement of all types of monitoring data, however, existed for only 27% of those 15%. Our data inspections, as well as conversations with rangeland experts, indicated a need for greater emphasis on collection of grazing-related monitoring data, particularly ground cover. Prioritization of where monitoring activities should be focused, along with creation of regional monitoring teams, may help improve monitoring. Overall, increased emphasis on monitoring of BLM rangelands will require commitment at multiple institutional levels.     

基于STARFM的草地地上生物量遥感估测研究——以甘肃省夏河县桑科草原为例

遥感数据具有实时、动态、大范围等特点,在草地资源监测与管理研究中获得了广泛应用。然而,单一的遥感植被指数无法同时满足草地地上生物量观测中时空分辨率的需求。因此,本研究基于时间序列Landsat NDVI和MODIS NDVI数据,结合时空融合算法（spatial and temporal adaptive reflectance fusion model, STARFM）,生成了2000-2016年高时空分辨率的植被指数数据集（NDVI_STARFM,时间分辨率为16 d,空间分辨率为30 m）,并基于2013-2016年地面实测草地地上生物量数据,构建了夏河县桑科草原高寒草地地上生物量遥感反演模型,分析了2000-2016年研究区草地地上生物量生长状况和变化趋势。结果表明：1）基于NDVI_STARFM的最优估测模型为乘幂模型,其R²为0.58,均方根误差（root mean square error, RMSE）为795.62 kg·hm^-2,模型的表现能力次于Landsat NDVI最优估测模型（R²=0.76,RMSE=634.83 kg·hm^-2）,而优于MODIS NDVI最优估测模型（R²=0.24,RMSE=937.79 kg·hm^-2）;2）基于NDVI_STARFM最优估测模型对各样区草地地上生物量总产的估测精度优于MODIS NDVI而次于Landsat NDVI,总体精度达84.05%;3）2000-2016年来,夏河县研究区草地地上生物量总体呈现增加趋势,其中90%左右的区域年增量大于30 kg·hm^-2,草地地上生物量呈现减少趋势的区域仅占2.30%。     

A Process for Assessing Wooded Plant Cover by Remote Sensing

The ability to map the extent of wooded vegetation cover over large areas using remote sensing is important for managing and assessing rangelands. Currently, applied techniques are inadequate because they 1) do not directly measure the amount of land covered by woody plants and rely on low-resolution images, 2) require considerable training-area data to train a classifier, and 3) describe only a limited number of land cover types. This paper presents an innovative methodology for creating a land-cover map that requires little to no traditional, training-area data collection before classification. The procedure combines both high-resolution aerial photography (resampled to 2.5-m pixels) and lower-resolution satellite imagery (30-m pixels) to produce a detailed and easily producible data set. The resulting data set also categorizes regions into a wide variety of land cover types in addition to differing levels of wooded cover. This new methodology was applied to the Upper Guadalupe River watershed in Texas, which is composed of varying amounts of brush cover between herbaceous range and dense cover. Validation by comparison to aerial imagery demonstrated a 74.4% success rate for all land cover classes. Validation was also performed by ground survey for several brush-covered points and showed a 90.0% success rate. As a result of the ground survey, modifications to the methodology were recommended to reduce classification errors and improve the process.     

Extent of Coterminous US Rangelands: Quantifying Implications of Differing Agency Perspectives

Rangeland extent is an important factor for evaluating critical indicators of rangeland sustainability. Rangeland areal extent was determined for the coterminous United States in a geospatial framework by evaluating spatially explicit data from the Landscape Fire and Resource Management Planning Tools (LANDFIRE) project describing historic and current vegetative composition, average height, and average cover through the viewpoints of the Natural Resources Inventory (NRI) administered by the Natural Resources Conservation Service and the Forest Inventory and Analysis (FIA) program administered by the US Forest Service. Three types of rangelands were differentiated using the NRI definition encompassing rangelands, afforested rangelands, and transitory rangelands. Limitations in the FIA definition permitted characterization of only two rangeland types: rangeland and rangeland vegetation with a small patch size. These classes were similar to those from the NRI definition but differed in tree canopy cover threshold requirements. Estimated rangeland area resulting from the NRI- and FIA-LANDFIRE models were 268 and 207 Mha, respectively. In addition, the NRI-LANDFIRE model identified 19 Mha of afforested rangelands due principally to encroachment and increased density by species classified as trees belonging to the genera Quercus, Prosopis, and Juniperus. The biggest discrepancies between acreage estimates derived from NRI- and FIA-LANDFIRE models occurred in oak, pinyon-juniper, and mesquite woodlands. The differences in area estimates between the NRI and FIA perspectives demonstrate the need for development of unified, objective methods for determining rangeland extent that can be applied consistently to all rangelands regardless of ownership or jurisdiction. While the models and geospatial information developed here are useful for national-scale estimates of rangeland extent, they are subject to the limitations of the LANDFIRE data products.     

Understanding Mediterranean Pasture Dynamics: General Tree Cover vs. Specific Effects of Individual Trees

The study investigated the effect of general and homogeneous tree cover on grassland composition on an extensive Mediterranean rangeland with sparse oak trees in central Spain. We analyzed this effect together with other significant factors identified in this type of rangeland: topography and plowing. Data were collected in the 1984 growing season and they form part of a historical database on the characteristics of vegetation and livestock behavior; these data refer to grasslands below and away from the tree crowns of 91 individual trees, located in different topographical positions and in areas that were last plowed at different times. We used multivariate analyses to identify the main compositional trends of variation in pasture communities. The results indicate that the herbaceous community below tree crowns was more similar to that of the lowland areas than to the nearby areas away from the tree. This result supports the idea of tree cover in semiarid rangelands as a factor attenuating the effects on pastures of environmental conditions typical of high and intermediate topographical positions—generally presenting low soil moisture and fertility. Coupled with this, we also found effects of some individual trees related with the way livestock uses them as shelter and resting places. Our results indicate that the role played by dispersed trees in the management of this type of rangeland should be analyzed at two complementary spatial scales: the overall effect of tree cover as a factor acting at landscape scale and the specific effect of some individual trees acting at a more detailed scale.     

Remote Sensing of Spatial and Temporal Vegetation Patterns in Two Grazing Systems

One constraint that range scientists must face in grazing studies is the lack of accurate and repeatable techniques for discriminating grazing effects from both temporal variability and spatial heterogeneity of vegetation. Both forms of variability contribute to inconsistent grazing system effects on vegetation response and forage production in semiarid ecosystems. Remote sensing may be an efficient tool for detecting differences in spatial and temporal patterns of grazing impact on vegetation. The purpose of this study was to evaluate the spectral data derived from satellite images as a tool for comparing grazing system impacts on spatial and temporal vegetation patterns. We evaluated the effect of two grazing systems, “Continuous” (C) and “Two-Paddocks Rest-Rotation” (TPRR), on vegetation cover from 1996 to 2006 in a semiarid ecosystem of Argentina. We compared grazing effects on vegetation cover using two indices derived from the Normalized Difference of Vegetation Index (NDVI) data from Landsat Thematic Mapper images. We observed a slight advantage in NDVI improvement for the TPRR over the C. Even though, in both grazing systems, an upward vegetation trend occurred only in areas located far from the watering points, TPRR showed higher relative vegetation cover near the watering point than C. We consider this methodology an important step for monitoring vegetation changes and making management decisions in livestock systems of semiarid regions because grazing system impacts may be compared for both spatial and temporal vegetation patterns. However, we think that the key next step is to develop procedures that discriminate between forage and nonforage components.     

Fire,Plant Invasions,and Erosion Events on Western Rangelands

Millions of hectares of rangeland in the western United States have been invaded by annual and woody plants that have increased the role of wildland fire. Altered fire regimes pose significant implications for runoff and erosion. In this paper we synthesize what is known about fire impacts on rangeland hydrology and erosion, and how that knowledge advances understanding of hydrologic risks associated with landscape scale plant community transitions and altered fire regimes. The increased role of wildland fire on western rangeland exposes landscapes to amplified runoff and erosion over short- and long-term windows of time and increases the risk of damage to soil and water resources, property, and human lives during extreme events. Amplified runoff and erosion postfire are a function of storm characteristics and fire-induced changes in site conditions (i.e., ground cover, soil water repellency, aggregate stability, and surface roughness) that define site susceptibility. We suggest that overall postfire hydrologic vulnerability be considered in a probabilistic framework that predicts hydrologic response for a range of potential storms and site susceptibilities and that identifies the hydrologic response magnitudes at which damage to values-at-risk are likely to occur. We identify key knowledge gaps that limit advancement of predictive technologies to address the increased role of wildland fire across rangeland landscapes. Our review of literature suggests quantifying interactions of varying rainfall intensity and key measures of site susceptibility, temporal variability in strength/influence of soil water repellency, and spatial scaling of postfire runoff and erosion remain paramount areas for future research to address hydrologic effects associated with the increased role of wildland fire on western rangelands.     

A Multi-Scale Approach to Predict the Fractional Cover of Medusahead (Taeniatherum Caput-Medusae)

Medusahead is an aggressive, winter annual that is of dire concern for the health and sustainability of western rangelands in the United States. Medusahead reduces plant diversity, alters ecosystem function, and reduces carrying capacities for both livestock and wildlife. The species has competitive advantages over cheatgrass and native grasses that causes an increased amount of fine fuels deposited on western rangelands. The Channeled Scablands of eastern Washington in the United States represent a typical example of a region being challenged by the expansion of this weed. The costs of the invasion are high and financial constraints can limit successful management. Managers need the ability to identify medusahead across entire landscapes, so they can work towards effective and efficient management approaches. Remote sensing offers the ability to measure vegetation cover at large spatial scales, which can lead to a better understanding of the invasive characteristics of problematic species like medusahead. For instance, research has been successful in creating large-scale distribution maps of cheatgrass over western rangelands. Many applications rely on the phenological characteristics of a target plant which can present problems in separating two species with similar phenologies (i.e. cheatgrass & medusahead). A medusahead-specific map gives managers the flexibility to prioritize and direct management needs when attempting to control the spread of medusahead into non-invaded areas. This study integrated GPS acquired field locations from three study sites (Sites S, C, & N) and imagery from two remote sensing platforms (1-m aerial imagery & 30-m Landsat), to model and predict fractional cover of medusahead over 37,000+ ha of rangelands in the Channeled Scabland region of eastern Washington. Using a multi-scaled approach, this research showed that regression tree algorithms can model the complex spectral response of senesced medusahead using late summer Landsat scenes. The predictive performances resulted in a R² of 0.80 near the model's training site (Site S) and an average R² of 0.68 away from the training site (Sites C & N). This research provides a non-phenological approach to produce accurate large-scale, distribution maps of medusahead which can aid land managers who are challenged by its invasion.     

Departures of Rangeland Fractional Component Cover and Land Cover from Landsat-Based Ecological Potential in Wyoming,USA

Monitoring rangelands by identifying the departure of contemporary conditions from long-term ecological potential allows for the disentanglement of natural biophysical gradients driving change from changes associated with land uses and other disturbance types. We developed maps of ecological potential (EP) for shrub, sagebrush (Artemisia spp.), perennial herbaceous, litter, and bare ground fractional cover in Wyoming, USA. EP maps correspond to the potential natural vegetation cover expected by environmental conditions in the absence of anthropogenic and natural disturbance as represented by the greenest and least disturbed period of the Landsat archive. EP was predicted using regression tree models with inputs of soil maps and spectral data associated with the 75th percentile of the Normalized Difference Vegetation Index in the Landsat archive. We trained our EP models with 2015 component cover maps on ecologically intact sites with relatively lower bare ground than expected. We generated departure of vegetation cover by comparing the EP and 2015 fractional cover. The departures represent land cover change from potential land cover and/or within-state changes in 2015. Next, we converted EP and 2015 fractional cover maps into thematic land cover and evaluated departure to determine if it was great enough to result in land cover change. The 2015 conditions showed reduced shrub, sagebrush, litter, and perennial herbaceous cover and increased bare ground relative to EP. Known disturbances, such as energy development, fires, and vegetation treatments, are clearly visible on the departure maps, but not on EP component maps. The most frequent departure from EP land cover was shrubland conversion to grassland. Land cover departures can be explained only in small part by known disturbance, and instead are ostensibly related to climate and land management practices. These drivers result in land cover departures that broadened the ecotone between shrubland and grassland relative to EP.     

Spatial Predictions of Cover Attributes of Rangeland Ecosystems Using Regression Kriging and Remote Sensing

Sound rangeland management requires accurate information on rangeland condition over large landscapes. A commonly applied approach to making spatial predictions of attributes related to rangeland condition (e.g., shrub or bare ground cover) from remote sensing is via regression between field and remotely sensed data. This has worked well in some situations but has limited utility when correlations between field and image data are low and it does not take advantage of all information contained in the field data. I compared spatial predictions from generalized least-squares (GLS) regression to a geostatistical interpolator, regression kriging (RK), for three rangeland attributes (percent cover of shrubs, bare ground, and cheatgrass [Bromus tectorum L.]) in a southern Idaho study area. The RK technique combines GLS regression with spatial interpolation of the residuals to improve predictions of rangeland condition attributes over large landscapes. I employed a remote-sensing technique, object-based image analysis (OBIA), to segment Landsat 5 Thematic Mapper images into polygons (i.e., objects) because previous research has shown that OBIA yields higher image-to-field data correlations and can be used to select appropriate scales for analysis. Spatial dependence, the decrease in autocorrelation with increasing distance, was strongest for percent shrub cover (samples autocorrelated up to a distance [i.e., range] of 19 098 m) but present in all three variables (range of 12 646 m and 768 m for bare ground and cheatgrass cover, respectively). As a result, RK produced more accurate results than GLS regression alone for all three attributes when predicted versus observed values of each attribute were measured by leave-one-out cross validation. The results of RK could be used in assessments of rangeland conditions over large landscapes. The ability to create maps quantifying how prediction confidence changes with distance from field samples is a significant benefit of regression kriging and makes this approach suitable for landscape-level management planning.     

Development of a ground truthing method for determination of rangeland biomass using canopy reflectance properties

Veld condition assessment and establishment of grazing capacity norms provide guidelines for the formulation of sustainable practices. However, conventional monitoring methods are becoming inadequate to meet future challenges, where quantification of spatial and temporal variation is required. This study proposes an on-site remote sensing method for monitoring above-ground biomass in rangelands. A preliminary model was formulated, based on simple regression relationships between canopy reflectance properties and aboveground biomass. This model was validated in semi-arid environments (Nama-karoo and Kalahari) within the framework of spatial and temporal experimental trials. Model accuracy was found to be primarily a function of canopy structure and vegetation composition where increases in dwarf shrub presence resulted in greater variations of both NDVI (Normalized Difference Vegetation Index) and LAI (Leaf Area Index) measurements. Temporal variation in model accuracy could also be observed which seemed to be associated with precipitation events. It was concluded that the proposed remote sensing method has potential as a ground truthing technique for determination of rangeland biomass. Indications are that this method is well suited for use in grass dominated veld types. With further refinement, implementation of this technique should also be possible in dwarf shrublands.     

Linking Phenology and Biomass Productivity in South Dakota Mixed-Grass Prairie

Assessing the health of rangeland ecosystems based solely on annual biomass production does not fully describe the condition of the plant community; the phenology of production can provide inferences about species composition, successional stage, and grazing impacts. We evaluated the productivity and phenology of western South Dakota mixed-grass prairie in the period from 2000 to 2008 using the normalized difference vegetation index (NDVI). The NDVI is based on 250-m spatial resolution Moderate Resolution Imaging Spectroradiometer (MODIS) satellite imagery. Growing-season NDVI images were integrated weekly to produce time-integrated NDVI (TIN), a proxy of total annual biomass production, and integrated seasonally to represent annual production by cool- and warm-season species (C₃ and C_4, respectively). Additionally, a variety of phenological indicators including cool-season percentage of TIN were derived from the seasonal profiles of NDVI. Cool-season percentage and TIN were combined to generate vegetation classes, which served as proxies of the conditions of plant communities. TIN decreased with precipitation from east to west across the study area. However, the cool-season percentage increased from east to west, following patterns related to the reliability (interannual coefficient of variation [CV]) and quantity of midsummer precipitation. Cool-season TIN averaged 76.8% of the total TIN. Seasonal accumulation of TIN corresponded closely (R² > 0.90) to that of gross photosynthesis data from a carbon flux tower. Field-collected biomass and community composition data were strongly related to TIN and cool-season percentage. The patterns of vegetation classes were responsive to topographic, edaphic, and land management influences on plant communities. Accurate maps of biomass production, cool- and warm-season composition, and vegetation classes can improve the efficiency of land management by facilitating the adjustment of stocking rates and season of use to maximize rangeland productivity and achieve conservation objectives. Further, our results clarify the spatial and temporal dynamics of phenology and TIN in mixed-grass prairie.