Inferring local competition intensity from patch size distributions: A test using biological soil crusts

Matthew A Bowker, Fernando T. Maestre

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Dryland vegetation is inherently patchy. This patchiness goes on to impact ecology, hydrology, and biogeochemistry. Recently, researchers have proposed that dryland vegetation patch sizes follow a power law which is due to local plant facilitation. It is unknown what patch size distribution prevails when competition predominates over facilitation, or if such a pattern could be used to detect competition. We investigated this question in an alternative vegetation type, mosses and lichens of biological soil crusts, which exhibit a smaller scale patch-interpatch configuration. This micro-vegetation is characterized by competition for space. We proposed that multiplicative effects of genetics, environment and competition should result in a log-normal patch size distribution. When testing the prevalence of log-normal versus power law patch size distributions, we found that the log-normal was the better distribution in 53% of cases and a reasonable fit in 83%. In contrast, the power law was better in 39% of cases, and in 8% of instances both distributions fit equally well. We further hypothesized that the log-normal distribution parameters would be predictably influenced by competition strength. There was qualitative agreement between one of the distribution's parameters (μ) and a novel intransitive (lacking a 'best' competitor) competition index, suggesting that as intransitivity increases, patch sizes decrease. The correlation of μ with other competition indicators based on spatial segregation of species (the C-score) depended on aridity. In less arid sites, μ was negatively correlated with the C-score (suggesting smaller patches under stronger competition), while positive correlations (suggesting larger patches under stronger competition) were observed at more arid sites. We propose that this is due to an increasing prevalence of competition transitivity as aridity increases. These findings broaden the emerging theory surrounding dryland patch size distributions and, with refinement, may help us infer cryptic ecological processes from easily observed spatial patterns in the field.

Original languageEnglish (US)
Pages (from-to)1914-1922
Number of pages9
JournalOikos
Volume121
Issue number11
DOIs
StatePublished - Nov 2012

Fingerprint

soil crusts
soil crust
patch size
arid lands
dry environmental conditions
vegetation
power law
biogeochemistry
testing
facilitation
aridity
vegetation types
hydrology
lichens
mosses and liverworts
researchers
ecology
patchiness
lichen
moss

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics

Cite this

Inferring local competition intensity from patch size distributions : A test using biological soil crusts. / Bowker, Matthew A; Maestre, Fernando T.

In: Oikos, Vol. 121, No. 11, 11.2012, p. 1914-1922.

Research output: Contribution to journalArticle

@article{435566db931b4ac7ba140029df23d6ab,
title = "Inferring local competition intensity from patch size distributions: A test using biological soil crusts",
abstract = "Dryland vegetation is inherently patchy. This patchiness goes on to impact ecology, hydrology, and biogeochemistry. Recently, researchers have proposed that dryland vegetation patch sizes follow a power law which is due to local plant facilitation. It is unknown what patch size distribution prevails when competition predominates over facilitation, or if such a pattern could be used to detect competition. We investigated this question in an alternative vegetation type, mosses and lichens of biological soil crusts, which exhibit a smaller scale patch-interpatch configuration. This micro-vegetation is characterized by competition for space. We proposed that multiplicative effects of genetics, environment and competition should result in a log-normal patch size distribution. When testing the prevalence of log-normal versus power law patch size distributions, we found that the log-normal was the better distribution in 53{\%} of cases and a reasonable fit in 83{\%}. In contrast, the power law was better in 39{\%} of cases, and in 8{\%} of instances both distributions fit equally well. We further hypothesized that the log-normal distribution parameters would be predictably influenced by competition strength. There was qualitative agreement between one of the distribution's parameters (μ) and a novel intransitive (lacking a 'best' competitor) competition index, suggesting that as intransitivity increases, patch sizes decrease. The correlation of μ with other competition indicators based on spatial segregation of species (the C-score) depended on aridity. In less arid sites, μ was negatively correlated with the C-score (suggesting smaller patches under stronger competition), while positive correlations (suggesting larger patches under stronger competition) were observed at more arid sites. We propose that this is due to an increasing prevalence of competition transitivity as aridity increases. These findings broaden the emerging theory surrounding dryland patch size distributions and, with refinement, may help us infer cryptic ecological processes from easily observed spatial patterns in the field.",
author = "Bowker, {Matthew A} and Maestre, {Fernando T.}",
year = "2012",
month = "11",
doi = "10.1111/j.1600-0706.2012.20192.x",
language = "English (US)",
volume = "121",
pages = "1914--1922",
journal = "Oikos",
issn = "0030-1299",
publisher = "Wiley-Blackwell",
number = "11",

}

TY - JOUR

T1 - Inferring local competition intensity from patch size distributions

T2 - A test using biological soil crusts

AU - Bowker, Matthew A

AU - Maestre, Fernando T.

PY - 2012/11

Y1 - 2012/11

N2 - Dryland vegetation is inherently patchy. This patchiness goes on to impact ecology, hydrology, and biogeochemistry. Recently, researchers have proposed that dryland vegetation patch sizes follow a power law which is due to local plant facilitation. It is unknown what patch size distribution prevails when competition predominates over facilitation, or if such a pattern could be used to detect competition. We investigated this question in an alternative vegetation type, mosses and lichens of biological soil crusts, which exhibit a smaller scale patch-interpatch configuration. This micro-vegetation is characterized by competition for space. We proposed that multiplicative effects of genetics, environment and competition should result in a log-normal patch size distribution. When testing the prevalence of log-normal versus power law patch size distributions, we found that the log-normal was the better distribution in 53% of cases and a reasonable fit in 83%. In contrast, the power law was better in 39% of cases, and in 8% of instances both distributions fit equally well. We further hypothesized that the log-normal distribution parameters would be predictably influenced by competition strength. There was qualitative agreement between one of the distribution's parameters (μ) and a novel intransitive (lacking a 'best' competitor) competition index, suggesting that as intransitivity increases, patch sizes decrease. The correlation of μ with other competition indicators based on spatial segregation of species (the C-score) depended on aridity. In less arid sites, μ was negatively correlated with the C-score (suggesting smaller patches under stronger competition), while positive correlations (suggesting larger patches under stronger competition) were observed at more arid sites. We propose that this is due to an increasing prevalence of competition transitivity as aridity increases. These findings broaden the emerging theory surrounding dryland patch size distributions and, with refinement, may help us infer cryptic ecological processes from easily observed spatial patterns in the field.

AB - Dryland vegetation is inherently patchy. This patchiness goes on to impact ecology, hydrology, and biogeochemistry. Recently, researchers have proposed that dryland vegetation patch sizes follow a power law which is due to local plant facilitation. It is unknown what patch size distribution prevails when competition predominates over facilitation, or if such a pattern could be used to detect competition. We investigated this question in an alternative vegetation type, mosses and lichens of biological soil crusts, which exhibit a smaller scale patch-interpatch configuration. This micro-vegetation is characterized by competition for space. We proposed that multiplicative effects of genetics, environment and competition should result in a log-normal patch size distribution. When testing the prevalence of log-normal versus power law patch size distributions, we found that the log-normal was the better distribution in 53% of cases and a reasonable fit in 83%. In contrast, the power law was better in 39% of cases, and in 8% of instances both distributions fit equally well. We further hypothesized that the log-normal distribution parameters would be predictably influenced by competition strength. There was qualitative agreement between one of the distribution's parameters (μ) and a novel intransitive (lacking a 'best' competitor) competition index, suggesting that as intransitivity increases, patch sizes decrease. The correlation of μ with other competition indicators based on spatial segregation of species (the C-score) depended on aridity. In less arid sites, μ was negatively correlated with the C-score (suggesting smaller patches under stronger competition), while positive correlations (suggesting larger patches under stronger competition) were observed at more arid sites. We propose that this is due to an increasing prevalence of competition transitivity as aridity increases. These findings broaden the emerging theory surrounding dryland patch size distributions and, with refinement, may help us infer cryptic ecological processes from easily observed spatial patterns in the field.

UR - http://www.scopus.com/inward/record.url?scp=84867742745&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867742745&partnerID=8YFLogxK

U2 - 10.1111/j.1600-0706.2012.20192.x

DO - 10.1111/j.1600-0706.2012.20192.x

M3 - Article

AN - SCOPUS:84867742745

VL - 121

SP - 1914

EP - 1922

JO - Oikos

JF - Oikos

SN - 0030-1299

IS - 11

ER -