Linking vegetation patterns to environmental gradients and human impacts in a mediterranean-type island ecosystem

Kyla M. Dahlin, Gregory P. Asner, Christopher B. Field

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

Vegetation patterns at the landscape scale are shaped by myriad processes and historical events, and understanding the relative importance of these processes aids in predicting current and future plant distributions. To quantify the influence of different environmental and anthropogenic patterns on observed vegetation patterns, we used simultaneous autoregressive modeling to analyze data collected by the Carnegie Airborne Observatory over Santa Cruz Island (SCI; California, USA). SCI is a large continental island, and its limited suite of species and well documented land use history allowed us to consider many potential determinants of vegetation patterns, such as topography, substrate, and historical grazing intensity. As a metric of vegetation heterogeneity, we used the normalized difference vegetation index (NDVI) stratified into three vegetation height classes using LiDAR (short, medium, and tall). In the SAR models topography and substrate type were important controls, together explaining 8–15 % of the total variation in NDVI, but historical grazing and spatial autocorrelation were also key components of the models, together explaining 17–21 % of the variation in NDVI. Optimal spatial autocorrelation distances in the short and medium height vegetation models (600–700 m) were similar to the home range sizes of two crucial seed dispersers on the island– the island fox (Urocyon littoralis santacruzae) and the island scrub-jay (Aphelocoma insularis)—suggesting that these animals may be important drivers of the island’s vegetation patterns. This study highlights the importance of dynamic processes like dispersal limitation and disturbance history in determining present-day vegetation patterns.

Original languageEnglish (US)
Pages (from-to)1571-1585
Number of pages15
JournalLandscape Ecology
Volume29
Issue number9
DOIs
StatePublished - Jan 1 2014
Externally publishedYes

Fingerprint

environmental gradient
anthropogenic effect
vegetation
geography
NDVI
history
land use
animal
driver
autocorrelation
determinants
event
grazing
present
topography
substrate
island ecosystem
scrub
home range
range size

Keywords

  • Airborne remote sensing
  • AVIRIS
  • Carnegie airborne observatory
  • Ecosystem assembly
  • Simultaneous autoregressive modeling
  • Spatial autocorrelation

ASJC Scopus subject areas

  • Geography, Planning and Development
  • Ecology
  • Nature and Landscape Conservation

Cite this

Linking vegetation patterns to environmental gradients and human impacts in a mediterranean-type island ecosystem. / Dahlin, Kyla M.; Asner, Gregory P.; Field, Christopher B.

In: Landscape Ecology, Vol. 29, No. 9, 01.01.2014, p. 1571-1585.

Research output: Contribution to journalArticle

@article{e1fd7fd0b4514be0bc8d895463b82a9b,
title = "Linking vegetation patterns to environmental gradients and human impacts in a mediterranean-type island ecosystem",
abstract = "Vegetation patterns at the landscape scale are shaped by myriad processes and historical events, and understanding the relative importance of these processes aids in predicting current and future plant distributions. To quantify the influence of different environmental and anthropogenic patterns on observed vegetation patterns, we used simultaneous autoregressive modeling to analyze data collected by the Carnegie Airborne Observatory over Santa Cruz Island (SCI; California, USA). SCI is a large continental island, and its limited suite of species and well documented land use history allowed us to consider many potential determinants of vegetation patterns, such as topography, substrate, and historical grazing intensity. As a metric of vegetation heterogeneity, we used the normalized difference vegetation index (NDVI) stratified into three vegetation height classes using LiDAR (short, medium, and tall). In the SAR models topography and substrate type were important controls, together explaining 8–15 {\%} of the total variation in NDVI, but historical grazing and spatial autocorrelation were also key components of the models, together explaining 17–21 {\%} of the variation in NDVI. Optimal spatial autocorrelation distances in the short and medium height vegetation models (600–700 m) were similar to the home range sizes of two crucial seed dispersers on the island– the island fox (Urocyon littoralis santacruzae) and the island scrub-jay (Aphelocoma insularis)—suggesting that these animals may be important drivers of the island’s vegetation patterns. This study highlights the importance of dynamic processes like dispersal limitation and disturbance history in determining present-day vegetation patterns.",
keywords = "Airborne remote sensing, AVIRIS, Carnegie airborne observatory, Ecosystem assembly, Simultaneous autoregressive modeling, Spatial autocorrelation",
author = "Dahlin, {Kyla M.} and Asner, {Gregory P.} and Field, {Christopher B.}",
year = "2014",
month = "1",
day = "1",
doi = "10.1007/s10980-014-0076-1",
language = "English (US)",
volume = "29",
pages = "1571--1585",
journal = "Landscape Ecology",
issn = "0921-2973",
publisher = "Springer Netherlands",
number = "9",

}

TY - JOUR

T1 - Linking vegetation patterns to environmental gradients and human impacts in a mediterranean-type island ecosystem

AU - Dahlin, Kyla M.

AU - Asner, Gregory P.

AU - Field, Christopher B.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Vegetation patterns at the landscape scale are shaped by myriad processes and historical events, and understanding the relative importance of these processes aids in predicting current and future plant distributions. To quantify the influence of different environmental and anthropogenic patterns on observed vegetation patterns, we used simultaneous autoregressive modeling to analyze data collected by the Carnegie Airborne Observatory over Santa Cruz Island (SCI; California, USA). SCI is a large continental island, and its limited suite of species and well documented land use history allowed us to consider many potential determinants of vegetation patterns, such as topography, substrate, and historical grazing intensity. As a metric of vegetation heterogeneity, we used the normalized difference vegetation index (NDVI) stratified into three vegetation height classes using LiDAR (short, medium, and tall). In the SAR models topography and substrate type were important controls, together explaining 8–15 % of the total variation in NDVI, but historical grazing and spatial autocorrelation were also key components of the models, together explaining 17–21 % of the variation in NDVI. Optimal spatial autocorrelation distances in the short and medium height vegetation models (600–700 m) were similar to the home range sizes of two crucial seed dispersers on the island– the island fox (Urocyon littoralis santacruzae) and the island scrub-jay (Aphelocoma insularis)—suggesting that these animals may be important drivers of the island’s vegetation patterns. This study highlights the importance of dynamic processes like dispersal limitation and disturbance history in determining present-day vegetation patterns.

AB - Vegetation patterns at the landscape scale are shaped by myriad processes and historical events, and understanding the relative importance of these processes aids in predicting current and future plant distributions. To quantify the influence of different environmental and anthropogenic patterns on observed vegetation patterns, we used simultaneous autoregressive modeling to analyze data collected by the Carnegie Airborne Observatory over Santa Cruz Island (SCI; California, USA). SCI is a large continental island, and its limited suite of species and well documented land use history allowed us to consider many potential determinants of vegetation patterns, such as topography, substrate, and historical grazing intensity. As a metric of vegetation heterogeneity, we used the normalized difference vegetation index (NDVI) stratified into three vegetation height classes using LiDAR (short, medium, and tall). In the SAR models topography and substrate type were important controls, together explaining 8–15 % of the total variation in NDVI, but historical grazing and spatial autocorrelation were also key components of the models, together explaining 17–21 % of the variation in NDVI. Optimal spatial autocorrelation distances in the short and medium height vegetation models (600–700 m) were similar to the home range sizes of two crucial seed dispersers on the island– the island fox (Urocyon littoralis santacruzae) and the island scrub-jay (Aphelocoma insularis)—suggesting that these animals may be important drivers of the island’s vegetation patterns. This study highlights the importance of dynamic processes like dispersal limitation and disturbance history in determining present-day vegetation patterns.

KW - Airborne remote sensing

KW - AVIRIS

KW - Carnegie airborne observatory

KW - Ecosystem assembly

KW - Simultaneous autoregressive modeling

KW - Spatial autocorrelation

UR - http://www.scopus.com/inward/record.url?scp=84939898105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84939898105&partnerID=8YFLogxK

U2 - 10.1007/s10980-014-0076-1

DO - 10.1007/s10980-014-0076-1

M3 - Article

AN - SCOPUS:84939898105

VL - 29

SP - 1571

EP - 1585

JO - Landscape Ecology

JF - Landscape Ecology

SN - 0921-2973

IS - 9

ER -