Other Titles


With over a million plant and animal species facing the threat of extinction and countless others experiencing steep population declines, ecosystem restoration is being emphasized as a key strategy to stem biodiversity declines. Tropical forests, which both house 80% of global biodiversity and account for 90% of deforestation from 2000-2018, are a particular focus of international initiatives like the Bonn Challenge and the UN Decade on Ecosystem Restoration (2021-2030). The success of restoration projects for conserving biodiversity – as well as evaluation of success – can be limited by a poor understanding of the interactions of restorations with natural ecological variation and co-occurring stressors like drought and fire that are exacerbated by climate change. For example, although a common goal in forest restoration is to recover biodiversity to levels typically encountered in reference mature forests, mature forests often vary strongly across space in vegetation structure and species composition. Most evaluations of biodiversity outcomes of restoration do not explicitly consider how this natural heterogeneity should inform the identification of biodiversity baselines or targets. Ultimately, restoration initiatives may fall short of biodiversity targets unless an understanding of the local ecology and of risks from co-occurring environmental stressors informs design and management of restoration projects, and monitoring and evaluation of biodiversity outcomes. My doctoral research used a model taxon – birds – to address three key challenges related to restoration project design, implementation, and monitoring that are fundamentally grounded in ecology: (1) selecting ecologically relevant targets, baselines, and metrics; (2) managing risk in the context of climate change; and (3) monitoring biodiversity outcomes in seasonal environments. Working within a mosaic landscape of regenerating cattle pastures and mature lowland tropical forest in Guatemala’s Maya Biosphere Reserve (MBR), I investigated how metrics of biodiversity recovery in restored lands may be influenced by the selection of mature forest baseline(s) in regions with high beta diversity, and how the value of regenerating lands for biodiversity is mediated by fire and rainfall seasonality. Although restoration can incorporate a range of strategies (e.g., active tree planting versus assisted or passive natural regeneration), my research focused on natural regeneration which is generally the most cost-effective and scalable approach in tropical regions that face resource constraints. Birds were surveyed at 317 sites situated within mature forests (n = 144) or naturally regenerating pastures (1-11 years old; n = 173). At each survey location, we counted all birds seen or heard during both wet (May - August 2019) and dry (February - April 2019) seasons and quantified local and landscape habitat attributes. Birds were surveyed at an additional 111 sites in mature forest >30 km from the regenerating pasture landscape between February - September in 2018 or 2019. Selecting ecologically relevant targets: Consistent with local knowledge and natural history observations, bird communities differed between two distinct forest types within the MBR: bosque bajo (short, low-lying, seasonally flooded forest) and bosque alto (tall, open-understory forest found at slightly higher elevations). My research builds upon this knowledge by showing that species associated with bajo (short-stature) forest used regenerating pastures more readily than species associated with alto (tall-stature) forest. Indeed, we found that fine-scale habitat associations explained the propensity of species to use regenerating habitat better than did commonly used species traits like degree of forest dependence or foraging guild. Furthermore, results show for the first time that trajectories of biodiversity recovery in restored lands can differ between distinct forest baselines that reflect natural heterogeneity within the reference ecosystem. As canopy height increased within regenerating pastures, avian communities approached a bajo forest baseline more quickly than an alto forest baseline. Overall, setting meaningful baselines for restoration efforts is best achieved by understanding species’ responses to the natural heterogeneity of the local ecosystem, which is often overlooked in the context of restoration projects and ecological studies of restoration alike. Managing climate-associated risk: Despite the emergence of natural and human-set fires in recent decades as a major phenomenon in the humid tropics due to agricultural expansion and regional drying, most restoration efforts do not account for potential impacts of fires on the value of tropical regenerating forests for biodiversity. In the second chapter of my dissertation, I investigated how occupancy rates of forest-dependent and threatened bird species differed between repeatedly burned and unburned pasture. To better understand the effect of fire on pasture use, I also examined how species’ opportunistic use of regenerating pastures is a function of vegetation succession at local scales. Repeated burning throughout the first decade of restoration reduced or eliminated pasture use for 56% of forest-dependent and 53% of threatened species. Local vegetation succession explained the likelihood of colonizing pastures for nearly half of forest species that used regenerating pastures. Species closely associated with closed-canopy vegetation were most likely to be partially or fully excluded from burned pasture, suggesting that fire excluded certain species by impeding forest regrowth. This study is the first to reveal that burning is broadly detrimental for forest species using early-successional restored lands, and worse for more highly forest-dependent species, highlighting the critical importance of preventing human-caused fires in restored tropical landscapes. Monitoring biodiversity outcomes in seasonal environments: Although seasonal variation in environmental conditions and habitat use by species means that regenerating lands may not hold the same biodiversity value year-round, most restoration projects that collect data on species recovery do so at one time of year or collapse surveys from multiple seasons into a snapshot measure of diversity. I investigated how use of regenerating pastures and mature forest by tropical resident birds changed within a year, with a focus on the transition from the dry to the wet season. Rainfall seasonality is a defining attribute of many tropical ecosystems that can potentially affect the use of restored habitats by species, especially given that early-successional lands are usually drier and less vegetatively complex than mature forests. Roughly half of resident birds within our Guatemalan lowland forest landscape showed evidence of seasonal habitat use, and over 20% of forest-associated species – including 13 species of high conservation concern – used pastures inconsistently over time, either sporadically or during the wet season only. The underlying dynamic rates of local colonization and extinction that drove occupancy patterns varied temporally for over half of species, and notably, even widely disparate temporal patterns of colonization and extinction produced equivalent static occupancy patterns. My work suggests that the concurrent examination of both occurrence and underlying dynamic rates will best elucidate future changes in habitat use. Overall, this research demonstrates the need to integrate ecological concepts into restoration practice to set baselines and biodiversity targets, manage risk in the context of climate change, and monitor biodiversity outcomes in the context of natural environmental variability. The overarching recommendations for the design, management, and monitoring of restoration projects for biodiversity are distilled in my final chapter. Ultimately, this dissertation highlights that more nuanced understandings of the spatial and temporal variability that characterizes tropical forests will improve the ability of restoration projects to deliver meaningful and enduring benefits for biodiversity, particularly within the context of anthropogenic climate change.

Journal / Series

Volume & Issue


164 pages


Date Issued




biodiversity monitoring; birds; fire; rainfall seasonality; restoration; tropical forest


Effective Date

Expiration Date




Union Local


Number of Workers

Committee Chair

Rodewald, Amanda

Committee Co-Chair

Committee Member

Poveda, Katja
Dhondt, Andre

Degree Discipline

Natural Resources

Degree Name

Ph. D., Natural Resources

Degree Level

Doctor of Philosophy

Related Version

Related DOI

Related To

Related Part

Based on Related Item

Has Other Format(s)

Part of Related Item

Related To

Related Publication(s)

Link(s) to Related Publication(s)


Link(s) to Reference(s)

Previously Published As

Government Document




Other Identifiers


Attribution-NonCommercial 4.0 International


dissertation or thesis

Accessibility Feature

Accessibility Hazard

Accessibility Summary

Link(s) to Catalog Record