Lidar provides novel insights into the effect of pixel size and grazing intensity on measures of spatial heterogeneity in a native bunchgrass ecosystem
There is a strong link between vegetation heterogeneity and biodiversity in grassland ecosystems. However, quantifying spatial patterns of key metrics, such as aboveground biomass, at landscape scales remains a challenge. This stems from difficulties in accurately estimating grassland biomass at fin...
Saved in:
| Published in: | Remote sensing of environment Vol. 235; p. 111432 |
|---|---|
| Main Authors: | , , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
New York
Elsevier Inc
15-12-2019
Elsevier BV |
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | There is a strong link between vegetation heterogeneity and biodiversity in grassland ecosystems. However, quantifying spatial patterns of key metrics, such as aboveground biomass, at landscape scales remains a challenge. This stems from difficulties in accurately estimating grassland biomass at fine scales over large areas and determining what spatial scale is most appropriate to monitor how grassland impacts (e.g., livestock grazing) affect spatial patterns of biomass (i.e., spatial heterogeneity). Here, we use lidar metrics (volume, max height, and intensity) in Random Forest models to quantify fine-resolution (pixel size 1.0668 m (3.5 ft)) aboveground biomass estimates (pseudo R2 = 0.59; RMSD = 139.4 g m-2) across a bunchgrass prairie grassland system. To determine both the effects of grazing on the spatial heterogeneity of aboveground biomass and which pixel size is most sensitive to the effects of livestock grazing on grassland heterogeneity, we aggregated fine-resolution biomass maps to coarser pixel resolutions (3 m, 5 m, 8 m, 20 m, 30 m) across 23 pastures with varying levels of grazing intensity. Following aggregation to coarser pixel resolutions, we observed that semivariogram models produced statistically different (α = 0.05) measures of biomass heterogeneity. The range statistic was the only pasture-level semivariogram metric sensitive to grazing, and this relationship was only significant when using the finer-resolution datasets (~1 m to 8 m pixels). Our results demonstrate 1) the applicability of lidar data for quantifying biomass in short-statured grasslands, 2) that grazing in pacific northwest bunchgrass prairie can decrease spatial heterogeneity of aboveground biomass and 3) that fine-resolution satellite data (<10 m pixel sizes) are necessary to effectively monitor the effects of grazing on the spatial heterogeneity of vegetation biomass, an indirect metric of biodiversity at management scales (pasture sizes ranged from 40 to 745 ha) in this grassland ecosystem.
•Canopy and intensity lidar metrics were used to predict grassland biomass.•The range statistic was sensitive to grazing at 1 and 8 m resolution data.•The range statistic was not sensitive to grazing at 20 m and 30 m resolutions.•Grazing decreased spatial heterogeneity in this grassland habitat. |
|---|---|
| ISSN: | 0034-4257 1879-0704 |
| DOI: | 10.1016/j.rse.2019.111432 |