Date of Completion


Embargo Period



Mutualism, climate change, species distribution, range shift, species-area relationship, seed dispersal, spatial mismatch, knock-on species loss, secondary extinction, California flora

Major Advisor

Chris S. Elphick

Co-Major Advisor

Morgan W. Tingley

Associate Advisor

John A. Silander, Jr.

Associate Advisor

Cynthia S. Jones

Associate Advisor

Elizabeth L. Jockusch

Field of Study

Ecology and Evolutionary Biology


Doctor of Philosophy

Open Access

Open Access


Climate change has caused species range shifts, with more predicted in coming decades. Range shifts could result in secondary threats such as spatial mismatch with mutualist partners and movement out of protected areas. We found that due to range shifts, shrub species richness will be lost at higher elevations, and species turnover will peak at middle elevations. Areas of bird species turnover will only partially overlap with areas of shrub species turnover, which could result in broken interactions between partners. Our projections suggest that climate change will result in clear winners and losers with some species gaining and others losing extent within a large protected area. Our findings add to growing evidence that currently protected species lose protection as they shift their ranges with changing climate.

The species-area relationship is a foundational idea in ecology and conservation biology which has been used to predict the number of species lost with habitat destruction and climate change. We extend it to biotic interactions. We present theory for how interactions scale with space and provide mathematical relationships with the species-area curve. Our interactions-area curve accounts for connectance, a measure of interactions per species within a network. We find that the interactions-area curve from an empirical seed dispersal network fits our theoretical equation.

Habitat loss can result in species loss from a mutualist network, causing additional species to become secondarily disconnected from the network. The number of disconnected species from habitat loss is poorly known. We simulated a null model network with random species loss according to the species-area relationship, and enumerated the disconnected species, varying both the number of species within the network and connectance. Our network simulations show more species detached at lower connectivity and with greater disparity in the species richness of the two mutualist groups. Our empirical example also displayed a wide range of outcomes: 0-5 species/10 km2 of simulated habitat loss. As available habitat is lost to land use conversion and climate change, community level repercussions are greater than predicted by simple species loss, but there is uncertainty in how severe these repercussions will be, from minimal to catastrophic.