Figure 1 - uploaded by Samuel E. Cox
Content may be subject to copyright.
A, Technician acquiring nadir digital image along a stretched transect tape for later analysis with SamplePoint (SP). Good images will be nadir, shadow-free, in good focus, and taken when plant phenology allows easiest plant type or species detection. The use of two short sticks laid perpendicular along the tape at 82-cm intervals allows users of different heights to capture equal-sized fields of view without a camera stand by using the sticks to frame the scene. The inlay shows the common mistake of ‘‘short-arming’’ the shot and taking the image off- nadir. B, Screenshot of a project image being analyzed in SP. Note that the classification area is a single pixel in the image, and is the center pixel of a nine-pixel array outlined by the crosshair. When the point in question is magnified 3 23, the pixel is unquestionably representing rabbitbrush.
Source publication
There is global recognition that sustainable land use requires monitoring that will detect change on a scale that protects the resource. That fundamental necessity is threatened where labor-intensive methods and high labor costs cause sampling deficiencies and increased Type-II error rates (false negatives). Ground-based imaging is a monitoring met...
Citations
... Along each transect 0.5 m 2 images were taken at 5 m intervals, resulting in 9 images per transect or 18 images per each well pad and paired reference location. All images were taken by hand with a 50 mega-pixel camera on a Samsung Galaxy A14 taking images perpendicular to the ground (freehand, nadir technique) [49]. The ground sample distance in each photo was ~0.2 mm, which is similar to previously published research using image analysis to identify vegetation on reclamation sites and in rangeland areas [13,50,51]. ...
Natural gas extraction is a critical driver of the economy in western North America. Ecological
reclamation is important to ensure surface disturbance impacts associated with natural gas
development are not permanent and to assist native biota. Previous studies in semi-arid natural gas
fields within Sublette County, Wyoming, USA have shown insects respond favorably to 1–3-yearold
well pads undergoing reclamation compared to older successional reference vegetation communities
dominated by Wyoming big sagebrush (Artemisia tridentata spp. Wyomingensis). Here, we
examined well pads which were initially seed 5, 8, 10, 11, and 12 years prior to our study. We used
a free, image-based software called SamplePointv. 1.60 to quantify vegetation on these well pads
and adjacent reference areas from cell phone camera photographs. Insects were collected with a
sweep net and identified to the family and morphospecies level. Statistical analyses were conducted
to compare both vegetation and insect communities between reclamation sites and their paired reference
area. We found little statistical difference between vegetation communities across our study
but found significantly more insect abundance on reclaimed well pads than reference areas in 3 of
5 years and significantly higher family and morphospecies richness on reclaimed well pads in 4 of
5 years. A total of 2036 individual insects representing 270 species from 71 families across 11 orders
were identified across this study. A total of 1557 individuals (76.5%) were found on reclamation
sites, whereas 479 (23.5%) were found in reference areas across the entire study. A total of 233 species
(86.3% of total) were found on reclamation sites, whereas 121 species (44.8% of total) were found
in reference areas across the entire study. A total of 67 families (94.4% of total) were found on reclamation
sites, whereas 45 families (63.4% of total) were found in reference areas across the entire
study. All 11 orders found in the study were found on reclamation sites, whereas 9 orders were
found in reference areas across the entire study. Our results suggest reclamation of natural gas well
pads within an old successional stand of sagebrush continues to support higher levels of insect biodiversity
and abundance for at least 12 years. As insects are the most diverse group of animals on
Earth and because they provide a wide array of ecosystem services, our findings suggest ecological
reclamation plays an important role in returning biodiversity and ecosystem functionality to a semiarid
and old successional sagebrush–steppe ecosystem.
... At each plot we captured nadir (ground-facing) images using a Casio camera (EX-H2OG IS, Shibuya-ku, Tokyo, Japan). We positioned our nadir images from directly overhead to capture a 1-m × 1-m L frame of 1.27-cm diameter polyvinyl chloride pipe placed on the ground so that the edges of the frame aligned with the edges of the photograph (Cagney et al. 2011; Figure 2). We then used SamplePoint version 1.56 software (Booth et al. 2006) to distribute a grid of 49 (7 × 7) points over each digital image and assigned a ground cover category of bare ground, grass, forb, rock, litter, woody, bovid scat, cactus, or unknown to each point. ...
Overwintering is a key demographic stage for migratory birds but remains poorly understood, especially among multiple declining grassland bird species. The non‐breeding ranges all 4 species of longspur (i.e., chestnut‐collared [Calcarius ornatus], Smith's [C. pictus], Lapland [C. lapponicus], thick‐billed [Rhynchophanes mccownii]) overlap in Oklahoma and the Texas Panhandle, USA, making this region ideal to study their wintering ecology. We evaluated the relationship between wintering longspur occurrence and fine‐scale habitat characteristics using a combination of standardized bird surveys and vegetation plot sampling. Our study encompassed large, representative tracts of 3 prairie ecosystems (i.e., shortgrass, mixed‐grass, and tallgrass prairies) that intersect within the Southern Great Plains, during winters of 2018–2019 and 2019–2020. Using randomization tests and classification trees, we characterized longspur habitats and compared these associations across the 3 prairie ecosystems. Fine‐scale winter habitats (horizontal structure, vertical structure, and species compositions) varied among all 4 longspur species, varied at very fine scales, and differed between grassland types. Our findings can be applied to the management of grasslands such as decreasing vegetation height in mixed‐grass prairies for chestnut‐collared longspurs or removing woody vegetation in shortgrass prairies for thick‐billed longspurs to help develop full‐life cycle conservation for longspurs, which have experienced population declines. We described wintering habitat associations of the 4 species of longspur overwintering in the Southern Great Plains across 3 different prairie ecotypes. We found that longspurs each use different habitats, and that key habitat associations change for various species depending on prairie ecotype.
... An advantage of the freehand method is that a monopod and remote shutter release cable are not required. However, a significant disadvantage is a higher likelihood of obtaining off-nadir, blurred, and/or inconsistent-FOV imagery (Cagney et al. 2011). ...
... SamplePoint produces foliar cover measurements with potential accuracy exceeding 90% (Booth et al. 2006). SamplePoint field measurements have shown equal precision to line point intercept but do not always return the same cover measurement (Cagney et al. 2011), though one method is not assumed to be more accurate than the other. SamplePoint measures foliar cover but not basal cover; inherently, this method is incapable of collecting multiple canopy cover data. ...
Vegetation monitoring is integral to maintaining healthy and productive public lands. Virtually all activities on public lands have the potential to degrade native vegetation, with cascading effects of soil erosion, loss of resources, and diminishing ecosystem services. Monitoring vegetation allows land managers to recognize problems and implement management solutions in a timely manner to preserve resources and ecosystem benefits. This technical note describes ground-based image collection for vegetation monitoring, including best practices and equipment details. This technical note also describes image analysis using SamplePoint software, which produces foliar cover measurements with potential accuracy exceeding 90% (Booth et al. 2006). By acquiring and analyzing ground-based images, land managers can monitor more area during a growing season, analyze imagery during the off-season, improve statistical power through larger sample sizes, and maintain permanent records of resources. These benefits allow land managers to make more informed and defensible management decisions.
... Nadir (ground-facing) and orthogonal images were obtained with a PowerShot ELPH 115 IS (Canon, Melville, New York) camera. For nadir images, we placed a 0.82 · 0.61-m (0.5-m 2 ) rectangular frame on the ground 2 m to the right (and perpendicular to) the transect and photographed from directly above so that the edges of the frame aligned with the edges of the photograph (Cagney et al., 2011). We then placed a grid of 100 points over each digital image using SamplePoint v1.56 software (Booth et al., 2006). ...
The Sprague's pipit (Anthus spragueii) is a passerine of North American prairies, breeding in the northern United States and southern Canada, and overwintering in the grasslands and deserts of the southern United States and northern Mexico. In recent decades, pipit populations have declined in both breeding and wintering ranges because of degradation and loss of native prairie. Previous research on this species has focused primarily on its breeding habitat and provided limited information on its wintering habitat. To help fill in these data gaps, we assessed wintering density along the Texas Gulf Coast Plain, as well as characteristics of the associated habitat. We used a distance-based line-transect sampling technique to estimate pipit densities during the winters of 2013–2014 and 2014–2015 at seven conservation sites and found Sprague's pipits at all sites. The highest number of detections and the highest density estimates were at the Attwater Prairie Chicken National Wildlife Refuge. Habitat associated with pipits included open grasslands with mixed forbs and low vertical cover. Although wildlife biologists consider pipits to be native grassland specialists in their breeding range, our analyses did not identify proportion of native grasses as an important variable. Instead, the physical structure of the vegetation (short and sparse grass cover) was most important to pipit presence. Our results suggest that habitat management to benefit wintering populations of this species should focus on vegetation structure, though native grassland cover might be important at a larger scale.
... Increasing the amount of data that can be sampled within a given area in a short amount of time should improve statistical reliability of the data compared to more timeconsuming traditional techniques. Images are permanent records that can be analyzed by multiple observers after field data are collected 6 . Additionally, many cameras are equipped with global positioning systems (GPS), so images can be geotagged with a collection location 18 , 20 . ...
... Because travel costs are the most expensive part of large-scale vegetation monitoring programs 4 , the ability to rapidly collect monitoring data is critical. The permanence of an image allows for analyses to be conducted long after it is collected6 , which suggests that the methods proposed here could allow for robust amounts of data to be collected in short periods of time with the ability to analyze field data at a later date and potentially by multiple individuals or interest groups. ...
Rangeland ecosystems cover 3.6 billion hectares globally with 239 million hectares located in the United States. These ecosystems are critical for maintaining global ecosystem services. Monitoring vegetation in these ecosystems is required to assess rangeland health, to gauge habitat suitability for wildlife and domestic livestock, to combat invasive weeds, and to elucidate temporal environmental changes. Although rangeland ecosystems cover vast areas, traditional monitoring techniques are often time-consuming and cost-inefficient, subject to high observer bias, and often lack adequate spatial information. Image-based vegetation monitoring is faster, produces
permanent records (i.e., images), may result in reduced observer bias, and inherently includes adequate spatial information. Spatially balanced sampling designs are beneficial in monitoring natural resources. A protocol is presented for implementing a spatially balanced sampling design known as balanced acceptance sampling (BAS), with imagery acquired from ground-level cameras and unmanned aerial systems (UAS). A route optimization algorithm is used in addition to solve the ‘travelling salesperson problem’ (TSP) to increase time and cost efficiency. While UAS images can be acquired 2–3x faster than handheld images, both types of images are similar
to each other in terms of accuracy and precision. Lastly, the pros and cons of each method are discussed and examples of potential applications for these methods in other ecosystems are provided.
... A nadir photo was taken of each quadrat and all species present in the quadrat were listed. Photos were analyzed in SamplePoint 1.59 using a grid of 225 pixels ( Booth et al. 2006 ;Cagney et al. 2011 ). Each pixel was classified as graminoid, forb, litter, soil (including rocks), or unknown (usually deep shadows). ...
Solarization (covering soil and vegetation with plastic) has long been used in agriculture to control undesirable plants, but solarization of invasive plants in rangelands has shown mixed and species-specific results. Yellow bluestem (Bothriochloa ischaemum (L.) Keng var. songarica (Rupr. ex Fisch & C.A. Mey) Celarier & Harlan), an invasive perennial C4 grass, is common throughout the southern Great Plains and is not controlled by winter prescribed fire. We tested whether solarization (tarping) with black plastic, combined with winter prescribed fire, could control yellow bluestem. We applied three treatments (with four replicates): solarization (August to November 2017) + fire (January 2018), trimming + fire, and fire only. Results after two growing seasons show that total yellow bluestem cover in solarized + fire plots was reduced to 54% ± 10% (mean ± standard error), lower than trimmed + fire (82% ± 5%, p < 0.01) and fire only plots (78% ± 6%, p = 0.01). Forb cover in solarized + fire plots (15% ± 4%) was much higher than trimmed + fire (4% ± 1%, p < 0.01) and fire only plots (3% ± 1%, p < 0.01). Native forb richness was only slightly higher in solarized + fire plots (16 ± 2 species) compared to fire only (10 ± 2 species, p = 0.08) and trimmed + fire plots (10 ± 1 species, p = 0.08). Interestingly, native forb richness in all plots increased compared to pre-treatment values (2 ± 1 species for all treatments, p < 0.01). Solarization + winter fire can slightly decrease yellow bluestem cover and greatly increase native forb cover, creating islands of diversity in otherwise low-diversity grasslands. However, repeated treatments or alternative techniques will be needed for full control of yellow bluestem.
... To achieve more accuracy of drone imagery, several authors also mention the importance of the choice of software and the type of aerial vehicle. In addition, flight altitude, and fluctuations in altitude, photographic equipment, topography, calibrations, and even varying acuity of vision among participating researchers or technicians can all influence results (Booth et al. 2006;Booth & Cox 2008;Cagney et al. 2011;Laliberte et al. 2011). We note that an additional cause to the mentioned errors in the countings from NDVI images in this study was the low resolution of each monochrome sensor (1.2 MP) for detect desert species with very low growth and absent or tiny leaves. ...
Field data and unmanned aerial vehicles, also known as drones, were used to analyze a long-term experiment in a severely degraded area of the "Monte," one of the most arid and degraded biomes in Argentina. Field methods and drone orthoimages obtained from flights of a quadcopter at 20 and 40 m were used to compare plantation plots versus natural recovery on scarified sites, and the field performance of six putative framework species under four different treatments was evaluated. After 5 years of recovery in soils with almost no organic matter and mean annual rainfall averaging 145 mm, plantation plots had high survival rates (>65%) for all species as determined through both field survey and drone imagery. Three species were planted with a greater number of seedlings due to their availability in nurseries. For these species experimental treatments were studied statistically. Independently of treatment applied (control; 1/2 L hydrogel; 1 L hydrogel; 1 L of hydrogel + 1/2 kg of organic compost), highest survival rates were found for P. flexuosa var. depressa (88% ± 14.8) and A. lampa (84% ± 14.8). Drone images and field data both showed that these two putative framework species had the highest plant cover. In sharp contrast, mechanical scarification without planting or other treatments induced very low recovery (<2%). We demonstrate that drone imagery provides a new and very valuable tool for evaluating and monitoring restorative interventions in drylands.
... Second, LPI methods are time and labor intensive, making them expensive for large projects (West 1999). Due to LPI time requirements, phenological changes in some species may result in misidentification or underestimation (Cagney et al. 2011). Third, LPI may overestimate standing cover in rangeland ecosystems (Cook & Stubendieck 1986). ...
... Alternative methods, such as image-based sampling, have been shown to reduce monitoring costs and time while increasing statistical power and reducing observer bias (Cagney et al. 2011;Duniway et al. 2011). Additionally, since nearby units associated with environmental surveys tend to be spatially auto-correlated, a spatially balanced sample is likely to improve monitoring design (Stevens & Olsen 2004). ...
... Additionally, since nearby units associated with environmental surveys tend to be spatially auto-correlated, a spatially balanced sample is likely to improve monitoring design (Stevens & Olsen 2004). In addition to reducing costs compared to traditional LPI methods, Cagney et al. (2011) highlighted that images provide a permanent record, which can be reevaluated if data are questionable, if software improves, or if management objectives change. Here, we add a spatially-balanced sampling design, called balanced acceptance sampling (BAS) (Robertson et al. 2013;Robertson et al. 2017), along with geo-tagged imagery, to sample reclaimed natural gas well pads and reference sites. ...
Land reclamation associated with natural gas development has become increasingly important to mitigate land surface disturbance in western North America. Since well pads occur on sites with multiple land use and ownership, the progress and outcomes of these efforts are of interest to multiple stakeholders including industry, practitioners and consultants, regulatory agents, private landowners and the scientific community. Reclamation success criteria often vary within, and among, government agencies and across land ownership type. Typically, reclamation success of a well pad is judged by comparing vegetation cover from a single transect on the pad to a single transect in an adjacent reference site and data are collected by a large number of technicians with various field monitoring skills. We utilized ‘SamplePoint’ image analysis software and a spatially balanced sampling design, called balanced acceptance sampling, to demonstrate how spatially‐explicit quantitative data can be used to determine if sites are meeting various reclamation success criteria and use chi‐square tests to show how sites in vegetation percent cover differ from a statistical standpoint. This method collects field data faster than traditional methods. We demonstrate how quantitative and spatially‐explicit data can be utilized by multiple stakeholders, how it can improve upon current reference site selection, how it can satisfy reclamation monitoring requirements for multiple regulatory agencies, how it may help improve future seed mix selection, and discuss how it may reduce costs for operations responsible for reclamation and how it may reduce observer bias.
This article is protected by copyright. All rights reserved.
... Archived field plot imagery can therefore be reviewed by many observers at later times using potentially improved or multiple techniques to record land cover. High resolution imagery, less than 1 cm, is being used by a number of researchers (e.g., Breckenridge et al. 2011;Cagney et al. 2011;Karl et al. 2012;Mirik and Ansley 2012). Using high resolution imagery, Pilliod and Arkle (2013) found that photography-based grid point intercept in Great Basin plant communities was strongly correlated to line point intercept but it was 20-25 times more efficient, identified 23% more plant species, and was more precise in determining percent cover. ...
Natural rangelands are one of the significant pillars of support for the Libyan national economy. The total area of rangelands in Libya is c.13.3 million hectares. This resource plays an important role providing part of the food needs of the large numbers of grazing animals, in turn providing food for human consumption. In the eastern Libyan rangelands, vegetation cover has changed both qualitatively and quantitatively due to natural factors and human activity. This raises concerns about the sustainability of these resources. Observation methods at ground- based sites are widely used in studies assessing rangeland degradation in Libya. However, observations across the periods of time between the studies are often not integrated nor repeatable, making it difficult for rangeland managers to detect degradation consistently. The cost of such studies can be high in comparison to their accuracy and reliability, in terms of the time and resources required. These costs are not expected to encourage the local administrators of rangelands to make repeated or continuous observations in order to determine possible changes in managed areas. This has led to a lack of time-series data, and a lack of regularly updated information. The sustainability of rangelands requires effective management, which in turn is dependent upon accurate and timely monitoring data to support the assessment of rangeland deterioration.
The aim of this research has been to develop a framework for monitoring and evaluating rangeland condition in the east of Libya with a prediction of the future condition based on a historical assessment. This approach was achieved through the utilisation of medium resolution satellite imagery to classify vegetation cover using vegetation indices. A number of vegetation indices applied in arid and semi- arid rangelands similar to the study area were assessed using ground-based colour vertical photography (GBVP) methods to identify the most appropriate index for classifying percentage vegetation cover. The vegetation cover data were integrated with climate data, topography and soil erosion assessment using the RUSLE system to form a Rangeland Assessment Management Information System (RAMIS). These data were used to assess the historical and predicted future rangeland condition.
The MSAVI2 vegetation index was identified as the most appropriate index to map vegetation cover as this had good correlation with the ground data (R2 = 0.874). The RUSLE prediction identified that over 1,300,000 hectares were affected by soil loss over the time period from 1986 to 2010 representing nearly 97% of the study area. The RAMIS output indicated that most of the study area in 1986 was affected by a high risk of rangeland degradation, with less than 10 % of the area having a moderate and low-risk of degradation. The rangeland condition up to 2010 indicated a slight improvement in degradation distribution, with a slight decrease from 90% to 85% in the high risk of degradation area and the area having a low risk of degradation increasing from 2% to 8% in 2010. The result of the predictions made showed that the area of low cover class, which in 2017 reached about 1,280,000 hectares continues to increase through 2030 to 2050, to some 1,400,000 hectares, with a consequent increase in areas of high risk of rangeland degradation.
The result of implementing the RAMIS framework over the historical period illustrates changes in the rangeland condition, reflecting the fluctuation in the effectiveness of rangeland management development projects linked to the financial resources available in the 1980s, with increasing numbers of grazing animals exceeding the rangeland capacity and the expansion of rangeland cultivation. Libyan rangeland managers need to focus more on expanding the fenced area, conducting soil survey, and implementing soil erosion studies that can be used in erosion model calibration at a large scale to better inform rangeland management planning. Otherwise, the future projections of change up to 2050 indicate a continuance of the deterioration of rangeland condition, increasing the areas of low vegetation. However, this projection is based only on the vegetation data as the lack of available climate data did not permit its incorporation into the prediction.
... Digital photo analyses have been used for measurements of plant cover (e.g. Chen et al. 2010) or for ground truthing of high-resolution remote sensing of vegetation cover (Cagney et al. 2011). Using digital photos for rapid assessment of ES has the advantage of maximizing field sampling of an area in a short period of time (Chen et al. 2010;Getzin et al. 2012), thus reducing the fieldwork costs, and at the same time, they can be reanalyzed if needed. ...
Decision makers and stakeholders need high-quality data to manage ecosystem services (ES) efficiently. Landscape-level data on ES that are of sufficient quality to identify spatial tradeoffs, co-occurrence and hotspots of ES are costly to collect, and it is therefore important to increase the efficiency of sampling of primary data. We demonstrate how ES could be assessed more efficiently through image-based point intercept method and determine the tradeoff between the number of sample points (pins) used per image and the robustness of the measurements. We performed a permutation study to assess the reliability implications of reducing the number of pins per image. We present a flexible approach to optimize landscape-level assessments of ES that maximizes the information obtained from 1 m² digital images. Our results show that 30 pins are sufficient to measure ecosystem service indicators with a crown cover higher than 5% for landscape scale assessments. Reducing the number of pins from 100 to 30 reduces the processing time up to a 50% allowing to increase the number of sampled plots, resulting in more management-relevant ecosystem service maps. The three criteria presented here provide a flexible approach for optimal design of landscape-level assessments of ES.
