Premium
Using trophic hierarchy to understand food web structure
Author(s) -
Scotti Marco,
Bondavalli Cristina,
Bodini Antonio,
Allesina Stefano
Publication year - 2009
Publication title -
oikos
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.672
H-Index - 179
eISSN - 1600-0706
pISSN - 0030-1299
DOI - 10.1111/j.1600-0706.2009.17073.x
Subject(s) - trophic level , food web , hierarchy , topology (electrical circuits) , binary number , link (geometry) , position (finance) , ecological network , constant (computer programming) , ecology , mathematics , computer science , environmental science , biological system , biology , ecosystem , combinatorics , arithmetic , finance , economics , market economy , programming language
Link arrangement in food webs is determined by the species’ feeding habits. This work investigates whether food web topology is organized in a gradient of trophic positions from producers to consumers. To this end, we analyzed 26 food webs for which the consumption rate of each species was specified. We computed the trophic positions and the link densities of all species in the food webs. Link density measures how much each species contributes to the distribution of energy in the system. It is expressed as the number of links species establish with other nodes, weighted by their magnitude. We computed these two metrics using various formulations developed in the ecological network analysis framework. Results show a positive correlation between trophic position and link density across all the systems, regardless the specific formulas used to measure the two quantities. We performed the same analysis on the corresponding binary matrices (i.e. removing information about rates). In addition, we investigated the relation between trophic position and link density in: a) simulated binary webs with same connectance as the original ones; b) weighted webs with constant topology but randomized link strengths and c) weighted webs with constant connectance where both topology and link strengths are randomized. The correlation between the two indices attenuates, vanishes or becomes negative in the case of binary food webs and simulated data (weighted and unweighted). According to our analysis, link density in food webs decreases with trophic position so that it is greatly reduced toward the top of the trophic hierarchy. This outcome, that seems to challenge previous conclusions based on null models, strongly depends on link quantification. Including interaction strengths may improve substantially our understanding of food web organization, and possibly contradict results based on the analysis of binary webs.