Search USGS⌕ Search
Home/Archive/USGS

USGS · 70040645

Choosing a DIVA: a comparison of emerging digital imagery vegetation analysis techniques

Christopher F. Jorgensen·Ryan J. Stutzman·Lars C. Anderson·Suzanne E. Decker·Larkin A. Powell·Walter H. Schacht·Joseph J. Fontaine
View original source ↗Export metadata ↓

Abstract

Question: What is the precision of five methods of measuring vegetation structure using ground-based digital imagery and processing techniques? Location: Lincoln, Nebraska, USA Methods: Vertical herbaceous cover was recorded using digital imagery techniques at two distinct locations in a mixed-grass prairie. The precision of five ground-based digital imagery vegetation analysis (DIVA) methods for measuring vegetation structure was tested using a split-split plot analysis of covariance. Variability within each DIVA technique was estimated using coefficient of variation of mean percentage cover. Results: Vertical herbaceous cover estimates differed among DIVA techniques. Additionally, environmental conditions affected the vertical vegetation obstruction estimates for certain digital imagery methods, while other techniques were more adept at handling various conditions. Overall, percentage vegetation cover values differed among techniques, but the precision of four of the five techniques was consistently high. Conclusions: DIVA procedures are sufficient for measuring various heights and densities of standing herbaceous cover. Moreover, digital imagery techniques can reduce measurement error associated with multiple observers' standing herbaceous cover estimates, allowing greater opportunity to detect patterns associated with vegetation structure.

Explore related subjects

Applied Vegetation ScienceArticle
90° N90° S · 180° W ← longitude → 180° E
Source-reported bounding extent: 40.7113° to 40.8899° latitude; -96.8041° to -96.5773° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Christopher F. Jorgensen, Ryan J. Stutzman, Lars C. Anderson, Suzanne E. Decker, Larkin A. Powell, Walter H. Schacht, Joseph J. Fontaine. 2013-04-13. Choosing a DIVA: a comparison of emerging digital imagery vegetation analysis techniques. https://doi.org/10.1111/avsc.12037

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2025

Daily fluctuating flows affect riparian plant species distributions from local to regional scales

Aims The number of hydropower dams has grown globally over recent decades, with significant impacts on downstream riparian plant communities. Many of these dams generate daily fluctuating flows known as hydropeaking to meet sub-daily variation in energy demands. Hydropeaking can significantly impact riparian plant communities, with obligate riparian species tending to experience the greatest negative effects on habitat suitability. Whether this pattern holds in arid biomes where daily soil moisture enhancements could benefit some plants is an open question. Location Colorado River, Grand Canyon, Western USA. Methods We used occurrence records to model species responses to variation in daily flow fluctuations across 32 689 river segments in the Western United States. We then applied estimates of hydropeaking responses derived from those models to understanding the abundance and fine scale hydrologic niches of riparian plant species in the Colorado River ecosystem downstream of Glen Canyon Dam, which has experienced vegetation expansion attributed to river regulation, including hydropeaking that began in 1964. Results At the regional scale, species with greater wetland dependence exhibited increasingly negative responses to hydropeaking across 1 496 species, consistent with previous studies at smaller scales. At the local scale of the Colorado River, we found that species inhabiting near-channel habitat characterized by daily inundation and exposure had positive modeled responses to hydropeaking, consistent with a long history of selection for species tolerant of hydropeaking. In contrast, species inhabiting the zone immediately above peak daily river stage had negative modeled responses to hydropeaking, suggesting that they are being excluded from otherwise suitable habitat nearer the channel. Conclusions These results demonstrate that hydropeaking can impact species distributions from local to regional scales by excluding obligate wetland species and reducing habitat suitability for some facultative wetland species. These results from an arid river system are consistent with those reported from other biomes.

Arizona↗
USGS2023

Satellite-derived prefire vegetation predicts variation in field-based invasive annual grass cover after fire

Aims Invasion by annual grasses (IAGs) and concomitant increases in wildfire are impacting many drylands globally, and an understanding of factors that contribute to or detract from community resistance to IAGs is needed to inform postfire restoration interventions. Prefire vegetation condition is often unknown in rangelands but it likely affects variation in postfire invasion resistance across large burned scars. Whether satellite-derived products like the Rangeland Analysis Platform (RAP) can fulfill prefire information needs and be used to parametrize models of fire recovery to inform postfire management of IAGs is a key question. Methods We used random forests to ask how IAG abundances in 669 field plots measured in the 2-3 years following megafires in sagebrush steppe rangelands of western USA responded to RAP estimates of annual:perennial prefire vegetation cover, the effects of elevation, heat load, postfire treatments, soil moisture–temperature regimes, and land-agency ratings of ecosystem resistance to invasion and resilience to disturbance. Results Postfire IAG cover measured in the field was 22¯% and RAP-estimated prefire annual herbaceous cover was 15.7¯%. The random forest model had an R 2 of 0.36 and a root-mean-squared error (RMSE) of 4.41. Elevation, postfire herbicide treatment, and prefire estimates from RAP for the ratio of annual:perennial and shrub cover were the most important predictors of postfire IAG cover. Threshold-like relationships between postfire IAG cover and the predictors indicate that maintaining annual:perennial cover below 0.4 and shrub cover below <10% prior to wildfire would decrease invasion, at low elevations below 1400 m above sea level. Conclusion Despite known differences between RAP and field-based estimates of vegetation cover, RAP was still a useful predictor of variation in IAG abundances after fire. IAG management is oftentimes reactive, but our findings indicate impactful roles for more inclusively addressing the exotic annual community, and focusing on prefire maintenance of annual:perennial herbaceous and shrub cover at low elevations.

Applied Vegetation Science↗
USGS2023

Do seeding and seedling planting result in similar restored plant communities?

Aims Restoration practitioners often face a tradeoff between low cost but risky seeding vs expensive but more reliable seedling planting to meet revegetation goals. Knowing under what environmental and management conditions direct seeding vs seedling planting benefit different species could improve restoration practice. Methods We compared seed emergence to planted-seedling survival among perennial herbaceous species commonly used in restoration across eight experimental restoration sites on the Colorado Plateau, USA. We used linear models to assess relationships between emergence and survival among species, and to assess the effects of site climate and seeding pre-treatments on those relationships. Results We found that among species, emergence was positively correlated with survival in the cooler sites, meaning that species with high emergence also had high survival and vice versa, but had no relationship in the hottest sites. Furthermore, pre-treatments to enhance soil moisture in seeded plots, specifically microtopography (pits) and mulch, also resulted in positive relationships between emergence and survival among species, while seeding without additional soil pre-treatments did not. Seedling planting cost 14 times as much as direct seeding alone, dropping to nine times as much when pre-treatments were combined with seeding. Conclusions Taken together, these results suggest that investments in seedling planting at hotter dryland sites, or in creating microtopography or mulching prior to seeding across sites, are likely to promote establishment success compared to simple seeding methods in degraded dryland ecosystems. These findings also identify opportunities for hybrid seeding and planting approaches that balance tradeoffs between risk and cost, respectively.

Applied Vegetation Science↗