Tracking spatial regimes in animal communities: Implications for resilience-based management

Authors

Caleb P. Roberts, University of Arkansas, Fayetteville
Daniel R. Uden, University of Nebraska - LincolnFollow
Craig R. AllenFollow
David G. Angeler, Swedish University of Agricultural Sciences
Larkin A. Powell, University of Nebraska-LincolnFollow
Brady W. Allred, University of Montana
Matthew O. Jones, University of Montana
Jeremy D. Maestas, USDA Natural Resources Conservation Service
Dirac L. Twidwell Jr, University of Nebraska‐LincolnFollow

Date of this Version

1-11-2022

Citation

C.P. Roberts et al. Ecological Indicators 136 (2022) 108567 2. https://doi.org/10.1016/j.ecolind.2022.108567

Comments

open access article

Abstract

Spatial regimes (the spatial extents of ecological states) exhibit strong spatiotemporal order as they expand or contract in response to retreating or encroaching adjacent spatial regimes (e.g., woody plant invasion of grasslands) and human management (e.g., fire treatments). New methods enable tracking spatial regime boundaries via vegetation landcover data, and this approach is being used for strategic management across biomes. A clear advancement would be incorporating animal community data to track spatial regime boundaries alongside vegetation data. In a 41,170-hectare grassland experiencing woody plant encroachment, we test the utility of using animal community data to track spatial regimes via two hypotheses. (H1) Spatial regime boundaries identified via independent vegetation and animal datasets will exhibit spatial synchrony; specifically, grassland:woodland bird community boundaries will synchronize with grass:woody vegetation boundaries. (H2) Negative feedbacks will stabilize spatial regimes identified via animal data; specifically, frequent fire treatments will stabilize grassland bird community boundaries. We used 26 years of bird community and vegetation data alongside 32 years of fire history data. We identified spatial regime boundaries with bird community data via a wombling approach. We identified spatial regime boundaries with vegetation data by calculating spatial covariance between remotely-sensed grass and woody plant cover per pixel. For fire history data, we calculated the cumulative number of fires per pixel. Setting bird boundary strength (wombling R² values) as the response variable, we tested our hypotheses with a hierarchical generalized additive model (HGAM). Both hypotheses were supported: animal boundaries synchronized with vegetation boundaries in space and time, and grassland bird communities stabilized as fire frequency increased (HGAM explained 38% of deviance). We can now track spatial regimes via animal community data pixel-by-pixel and year-by-year. Alongside vegetation boundary tracking, tracking animal community boundaries can inform the scale of management necessary to maintain animal communities endemic to desirable ecological states. Our approach will be especially useful for conserving animal communities requiring large-scale, unfragmented landscapes—like grasslands and steppes.