Estimating Extinction Risk with Metapopulation Models of Large‐Scale Fragmentation

Habitat loss is the principal threat to species. How much habitat remains—and how quickly it is shrinking—are implicitly included in the way the International Union for Conservation of Nature determines a species’ risk of extinction. Many endangered species have habitats that are also fragmented to different extents. Thus, ideally, fragmentation should be quantified in a standard way in risk assessments. Although mapping fragmentation from satellite imagery is easy, efficient techniques for relating maps of remaining habitat to extinction risk are few. Purely spatial metrics from landscape ecology are hard to interpret and do not address extinction directly. Spatially explicit metapopulation models link fragmentation to extinction risk, but standard models work only at small scales. Counterintuitively, these models predict that a species in a large, contiguous habitat will fare worse than one in 2 tiny patches. This occurs because although the species in the large, contiguous habitat has a low probability of extinction, recolonization cannot occur if there are no other patches to provide colonists for a rescue effect. For 4 ecologically comparable bird species of the North Central American highland forests, we devised metapopulation models with area‐weighted self‐colonization terms; this reflected repopulation of a patch from a remnant of individuals that survived an adverse event. Use of this term gives extra weight to a patch in its own rescue effect. Species assigned least risk status were comparable in long‐term extinction risk with those ranked as threatened. This finding suggests that fragmentation has had a substantial negative effect on them that is not accounted for in their Red List category. Estimación del Riesgo de Extinción Mediante Modelos Metapoblacionales de Fragmentación a Gran Escala     

Reorienting Systematic Conservation Assessment for Effective Conservation Planning

Abstract: Systematic conservation assessment (an information‐gathering and prioritization process used to select the spatial foci of conservation initiatives) is often considered vital to conservation‐planning efforts, yet published assessments have rarely resulted in conservation action. Conservation assessments may lead more directly to effective conservation action if they are reoriented to inform conservation decisions. Toward this goal, we evaluated the relative priority for conservation of 7 sites proposed for the first forest reserves in the Union of the Comoros, an area with high levels of endemism and rapidly changing land uses in the western Indian Ocean. Through the analysis of 30 indicator variables measured at forest sites and nearby villages, we assessed 3 prioritization criteria at each site: conservation value, threat to loss of biological diversity from human activity, and feasibility of reserve establishment. Our results indicated 2 sites, Yiméré and Hassera‐Ndrengé, were priorities for conservation action. Our approach also informed the development of an implementation strategy and enabled an evaluation of previously unexplored relations among prioritization criteria. Our experience suggests that steps taken to ensure the closer involvement of practitioners, include a broader range of social data, encourage stakeholder participation, and consider the feasibility of conservation action can improve the relevance of assessments for conservation planning, strengthen the scientific basis for conservation decisions, and result in a more realistic evaluation of conservation alternatives.     

Toward quantification of the impact of 21st‐century deforestation on the extinction risk of terrestrial vertebrates

Conservation actions need to be prioritized, often taking into account species’ extinction risk. The International Union for Conservation of Nature (IUCN) Red List provides an accepted, objective framework for the assessment of extinction risk. Assessments based on data collected in the field are the best option, but the field data to base these on are often limited. Information collected through remote sensing can be used in place of field data to inform assessments. Forests are perhaps the best‐studied land‐cover type for use of remote‐sensing data. Using an open‐access 30‐m resolution map of tree cover and its change between 2000 and 2012, we assessed the extent of forest cover and loss within the distributions of 11,186 forest‐dependent amphibians, birds, and mammals worldwide. For 16 species, forest loss resulted in an elevated extinction risk under red‐list criterion A, owing to inferred rapid population declines. This number increased to 23 when data‐deficient species (i.e., those with insufficient information for evaluation) were included. Under red‐list criterion B2, 484 species (855 when data‐deficient species were included) were considered at elevated extinction risk, owing to restricted areas of occupancy resulting from little forest cover remaining within their ranges. The proportion of species of conservation concern would increase by 32.8% for amphibians, 15.1% for birds, and 24.7% for mammals if our suggested uplistings are accepted. Central America, the Northern Andes, Madagascar, the Eastern Arc forests in Africa, and the islands of Southeast Asia are hotspots for these species. Our results illustrate the utility of satellite imagery for global extinction‐risk assessment and measurement of progress toward international environmental agreement targets.     

Hierarchical Bayesian Spatial Models for Multispecies Conservation Planning and Monitoring

Abstract: Biologists who develop and apply habitat models are often familiar with the statistical challenges posed by their data's spatial structure but are unsure of whether the use of complex spatial models will increase the utility of model results in planning. We compared the relative performance of nonspatial and hierarchical Bayesian spatial models for three vertebrate and invertebrate taxa of conservation concern (Church's sideband snails [Monadenia churchi], red tree voles [Arborimus longicaudus], and Pacific fishers [Martes pennanti pacifica]) that provide examples of a range of distributional extents and dispersal abilities. We used presence–absence data derived from regional monitoring programs to develop models with both landscape and site‐level environmental covariates. We used Markov chain Monte Carlo algorithms and a conditional autoregressive or intrinsic conditional autoregressive model framework to fit spatial models. The fit of Bayesian spatial models was between 35 and 55% better than the fit of nonspatial analogue models. Bayesian spatial models outperformed analogous models developed with maximum entropy (Maxent) methods. Although the best spatial and nonspatial models included similar environmental variables, spatial models provided estimates of residual spatial effects that suggested how ecological processes might structure distribution patterns. Spatial models built from presence–absence data improved fit most for localized endemic species with ranges constrained by poorly known biogeographic factors and for widely distributed species suspected to be strongly affected by unmeasured environmental variables or population processes. By treating spatial effects as a variable of interest rather than a nuisance, hierarchical Bayesian spatial models, especially when they are based on a common broad‐scale spatial lattice (here the national Forest Inventory and Analysis grid of 24 km² hexagons), can increase the relevance of habitat models to multispecies conservation planning.     

Effects of Dam‐Induced Landscape Fragmentation on Amazonian Ant–Plant Mutualistic Networks

Mutualistic networks are critical to biological diversity maintenance; however, their structures and functionality may be threatened by a swiftly changing world. In the Amazon, the increasing number of dams poses a large threat to biological diversity because they greatly alter and fragment the surrounding landscape. Tight coevolutionary interactions typical of tropical forests, such as the ant–myrmecophyte mutualism, where the myrmecophyte plants provide domatia nesting space to their symbiotic ants, may be jeopardized by the landscape changes caused by dams. We analyzed 31 ant–myrmecophyte mutualistic networks in undisturbed and disturbed sites surrounding Balbina, the largest Central Amazonian dam. We tested how ant–myrmecophyte networks differ among dam‐induced islands, lake edges, and undisturbed forests in terms of species richness, composition, structure, and robustness (number of species remaining in the network after partner extinctions). We also tested how landscape configuration in terms of area, isolation, shape, and neighborhood alters the structure of the ant–myrmecophyte networks on islands. Ant–myrmecophytic networks were highly compartmentalized in undisturbed forests, and the compartments had few strongly connected mutualistic partners. In contrast, networks at lake edges and on islands were not compartmentalized and were negatively affected by island area and isolation in terms of species richness, density, and composition. Habitat loss and fragmentation led to coextinction cascades that contributed to the elimination of entire ant–plant compartments. Furthermore, many myrmecophytic plants in disturbed sites lost their mutualistic ant partners or were colonized by opportunistic, nonspecialized ants. Robustness of ant–myrmecophyte networks on islands was lower than robustness near lake edges and in undisturbed forest and was particularly susceptible to the extinction of plants. Beyond the immediate habitat loss caused by the building of large dams in Amazonia, persistent edge effects and habitat fragmentation associated with dams had large negative effects on animal–plant mutualistic networks. Efectos de la Fragmentación del Paisaje Inducida por Presas sobre Redes Mutualistas Hormiga‐Planta Amazónicas     

Characterizing Spatial Uncertainty when Integrating Social Data in Conservation Planning

Recent conservation planning studies have presented approaches for integrating spatially referenced social (SRS) data with a view to improving the feasibility of conservation action. We reviewed the growing conservation literature on SRS data, focusing on elicited or stated preferences derived through social survey methods such as choice experiments and public participation geographic information systems. Elicited SRS data includes the spatial distribution of willingness to sell, willingness to pay, willingness to act, and assessments of social and cultural values. We developed a typology for assessing elicited SRS data uncertainty which describes how social survey uncertainty propagates when projected spatially and the importance of accounting for spatial uncertainty such as scale effects and data quality. These uncertainties will propagate when elicited SRS data is integrated with biophysical data for conservation planning and may have important consequences for assessing the feasibility of conservation actions. To explore this issue further, we conducted a systematic review of the elicited SRS data literature. We found that social survey uncertainty was commonly tested for, but that these uncertainties were ignored when projected spatially. Based on these results we developed a framework which will help researchers and practitioners estimate social survey uncertainty and use these quantitative estimates to systematically address uncertainty within an analysis. This is important when using SRS data in conservation applications because decisions need to be made irrespective of data quality and well characterized uncertainty can be incorporated into decision theoretic approaches.     

Proposed Conservation Landscape for Giant Pandas in the Minshan Mountains,China

Abstract: The giant panda (Ailuropoda melanoleuca), is one of the world's most endangered species. Habitat loss and fragmentation have reduced its numbers, shrunk its distribution, and separated the population into isolated subpopulations. Such isolated, small populations are in danger of extinction due to random demographic factors and inbreeding. We used least‐cost modeling as a systematic approach to incorporate satellite imagery and data on ecological and behavioral parameters of the giant panda collected during more than 10 years of field research to design a conservation landscape for giant pandas in the Minshan Mountains. We identified 8 core habitats and 4 potential linkages that would link core habitats CH3, CH4, and CH5 with core habitats CH6, CH7, and CH8. Establishing and integrating the identified habitats with existing reserves would create an efficient reserve network for giant panda conservation. The core habitats had an average density of 4.9 pandas/100 km² and contained approximately 76.6% of the giant panda population. About 45% of the core habitat (3245.4 km²) existed outside the current nature reserves network. Total estimated core habitat decreased between 30.4 and 44.5% with the addition of residential areas and road networks factored into the model. A conservation area for giant panda in the Minshan Mountains should aim to ensure habitat retention and connectivity, improve dispersal potential of corridors, and maintain the evolutionary potential of giant pandas in the face of future environmental changes.     

Extirpation of Macroalgae (Sargassum spp.) on the Subtropical East Australian Coast

Abstract: Populations of large brown algae of the Laminariales and Fucales (Phaeophyta) have declined or been extirpated from many locations on temperate coasts worldwide. We conducted field surveys and a literature review, and examined herbarium specimens, through which we discovered previously unreported extirpations of large brown algal species from a tropical and subtropical coastline. Sargassum amaliae, S. aquifolium, S. carpophyllum, S. polycystum, and S. spinifex were common habitat‐forming macroalgae that supported diverse assemblages of invertebrates and smaller algae before urbanization began in 1970 along the 45‐km length of Sunshine Coast in Queensland, Australia. Causes of these extirpations are not known, but are consistent with losses of other large brown algal species from coastal areas undergoing urbanization or eutrophication. Sargassum spp. do not have the characteristics thought to protect marine species from extinction (large geographical ranges, occurrence on many different substrata, long‐distance dispersal). Some local Sargassum spp. are endemic to eastern Australia. Abundance of Sargassum is limited by suitable substrata on the sandy southern Queensland coast (370 km). These substrata are 12 rocky headlands separated by long (5–105 km) sandy beaches. Most multicellular propagules (the only motile stage in Sargassum) settle within 1–3 m of parental thalli, which restricts long‐distance dispersal needed to maintain connectivity among populations and to recolonize areas of the headlands from which populations have been extirpated. Local Sargassum spp. could be categorized as data deficient by the International Union for Conservation of Nature (IUCN), but the IUCN vulnerable category is more accurate given extirpations, limited habitat, and the lack of connectivity among populations.     

Ants as a Measure of Effectiveness of Habitat Conservation Planning in Southern California

Abstract: In the United States multispecies habitat conservation plans were meant to be the solution to conflicts between economic development and protection of biological diversity. Although now widely applied, questions exist concerning the scientific credibility of the conservation planning process and effectiveness of the plans. We used ants to assess performance of one of the first regional conservation plans developed in the United States, the Orange County Central‐Coastal Natural Community Conservation Plan (NCCP), in meeting its broader conservation objectives of biodiversity and ecosystem‐level protection. We collected pitfall data on ants for over 3 years on 172 sites established across a network of conservation lands in coastal southern California. Although recovered native ant diversity for the study area was high, site‐occupancy models indicated the invasive and ecologically disruptive Argentine ant (Linepithema humile) was present at 29% of sites, and sites located within 200 m of urban and agricultural areas were more likely to have been invaded. Within invaded sites, native ants were largely displaced, and their median species richness declined by more than 60% compared with uninvaded sites. At the time of planning, 24% of the 15,133‐ha reserve system established by Orange County NCCP fell within 200 m of an urban or agricultural edge. With complete build out of lands surrounding the reserve, the proportion of the reserve system vulnerable to invasion will grow to 44%. Our data indicate that simply protecting designated areas from development is not enough. If habitat conservation plans are to fulfill their conservation promise of ecosystem‐level protection, a more‐integrated and systematic approach to the process of habitat conservation planning is needed.     

Evaluating the Success of Conservation Actions in Safeguarding Tropical Forest Biodiversity

Abstract: We reviewed the evidence on the extent and efficacy of conservation of tropical forest biodiversity for each of the classes of conservation action defined by the new International Union for Conservation of Nature (IUCN) classification. Protected areas are the most tested conservation approach, and a number of studies show they are generally effective in slowing deforestation. There is some documentation of the extent of sustainable timber management in tropical forest, but little information on other landscape‐conservation tactics. The extent and effectiveness of ex situ species conservation is quite well known. Forty‐one tropical‐forest species now survive only in captivity. Other single‐species conservation actions are not as well documented. The potential of policy mechanisms, such as international conventions and provision of funds, to slow extinctions in tropical forests is considerable, but the effects of policy are difficult to measure. Finally, interventions to promote tropical conservation by supporting education and livelihoods, providing incentives, and furthering capacity building are all thought to be important, but their extent and effectiveness remain poorly known. For birds, the best studied taxon, the sum of such conservation actions has averted one‐fifth of the extinctions that would otherwise have occurred over the last century. Clearly, tropical forest conservation works, but more is needed, as is critical assessment of what works in what circumstances, if mass extinction is to be averted.     

Mapping Human and Social Dimensions of Conservation Opportunity for the Scheduling of Conservation Action on Private Land

Abstract Spatial prioritization techniques are applied in conservation‐planning initiatives to allocate conservation resources. Although typically they are based on ecological data (e.g., species, habitats, ecological processes), increasingly they also include nonecological data, mostly on the vulnerability of valued features and economic costs of implementation. Nevertheless, the effectiveness of conservation actions implemented through conservation‐planning initiatives is a function of the human and social dimensions of social‐ecological systems, such as stakeholders’ willingness and capacity to participate. We assessed human and social factors hypothesized to define opportunities for implementing effective conservation action by individual land managers (those responsible for making day‐to‐day decisions on land use) and mapped these to schedule implementation of a private land conservation program. We surveyed 48 land managers who owned 301 land parcels in the Makana Municipality of the Eastern Cape province in South Africa. Psychometric statistical and cluster analyses were applied to the interview data so as to map human and social factors of conservation opportunity across a landscape of regional conservation importance. Four groups of landowners were identified, in rank order, for a phased implementation process. Furthermore, using psychometric statistical techniques, we reduced the number of interview questions from 165 to 45, which is a preliminary step toward developing surrogates for human and social factors that can be developed rapidly and complemented with measures of conservation value, vulnerability, and economic cost to more‐effectively schedule conservation actions. This work provides conservation and land management professionals direction on where and how implementation of local‐scale conservation should be undertaken to ensure it is feasible.     

Conserving tigers in working landscapes

Tiger (Panthera tigris) conservation efforts in Asia are focused on protected areas embedded in human‐dominated landscapes. A system of protected areas is an effective conservation strategy for many endangered species if the network is large enough to support stable metapopulations. The long‐term conservation of tigers requires that the species be able to meet some of its life‐history needs beyond the boundaries of small protected areas and within the working landscape, including multiple‐use forests with logging and high human use. However, understanding of factors that promote or limit the occurrence of tigers in working landscapes is incomplete. We assessed the relative influence of protection status, prey occurrence, extent of grasslands, intensity of human use, and patch connectivity on tiger occurrence in the 5400 km² Central Terai Landscape of India, adjacent to Nepal. Two observer teams independently surveyed 1009 km of forest trails and water courses distributed across 60 166‐km² cells. In each cell, the teams recorded detection of tiger signs along evenly spaced trail segments. We used occupancy models that permitted multiscale analysis of spatially correlated data to estimate cell‐scale occupancy and segment‐scale habitat use by tigers as a function of management and environmental covariates. Prey availability and habitat quality, rather than protected‐area designation, influenced tiger occupancy. Tiger occupancy was low in some protected areas in India that were connected to extensive areas of tiger habitat in Nepal, which brings into question the efficacy of current protection and management strategies in both India and Nepal. At a finer spatial scale, tiger habitat use was high in trail segments associated with abundant prey and large grasslands, but it declined as human and livestock use increased. We speculate that riparian grasslands may provide tigers with critical refugia from human activity in the daytime and thereby promote tiger occurrence in some multiple‐use forests. Restrictions on human‐use in high‐quality tiger habitat in multiple‐use forests may complement existing protected areas and collectively promote the persistence of tiger populations in working landscapes.     

Conservation Planning with Multiple Organizations and Objectives

Abstract: There has been a dramatic increase in the number of conservation organizations worldwide. It is now common for multiple organizations to operate in the same landscape in pursuit of different conservation goals. New objectives, such as maintenance of ecosystem services, will attract additional funding and new organizations to conservation. Systematic conservation planning helps in the design of spatially explicit management actions that optimally conserve multiple landscape features (e.g., species, ecosystems, or ecosystem services). But the methods used in its application implicitly assume that a single actor implements the optimal plan. We investigated how organizational behavior and conservation outcomes are affected by the presence of autonomous implementing organizations with different objectives. We used simulation models and game theory to explore how alternative behaviors (e.g., organizations acting independently or explicitly cooperating) affected an organization's ability to protect their feature of interest, and investigated how the distribution of features in the landscape influenced organizations’ attitudes toward cooperation. Features with highly correlated spatial distributions, although typically considered an opportunity for mutually beneficial conservation planning, can lead to organizational interactions that result in lower levels of protection. These detrimental outcomes can be avoided by organizations that cooperate when acquiring land. Nevertheless, for cooperative purchases to benefit both organizations’ objectives, each must forgo the protection of land parcels that they would consider to be of high conservation value. Transaction costs incurred during cooperation and the sources of conservation funding could facilitate or hinder cooperative behavior.     

Using citizen science data to identify the sensitivity of species to human land use

Conservation practitioners must contend with an increasing array of threats that affect biodiversity. Citizen scientists can provide timely and expansive information for addressing these threats across large scales, but their data may contain sampling biases. We used randomization procedures to account for possible sampling biases in opportunistically reported citizen science data to identify species’ sensitivities to human land use. We analyzed 21,044 records of 143 native reptile and amphibian species reported to the Carolina Herp Atlas from North Carolina and South Carolina between 1 January 1990 and 12 July 2014. Sensitive species significantly associated with natural landscapes were 3.4 times more likely to be legally protected or treated as of conservation concern by state resource agencies than less sensitive species significantly associated with human‐dominated landscapes. Many of the species significantly associated with natural landscapes occurred primarily in habitats that had been nearly eradicated or otherwise altered in the Carolinas, including isolated wetlands, longleaf pine savannas, and Appalachian forests. Rare species with few reports were more likely to be associated with natural landscapes and 3.2 times more likely to be legally protected or treated as of conservation concern than species with at least 20 reported occurrences. Our results suggest that opportunistically reported citizen science data can be used to identify sensitive species and that species currently restricted primarily to natural landscapes are likely at greatest risk of decline from future losses of natural habitat. Our approach demonstrates the usefulness of citizen science data in prioritizing conservation and in helping practitioners address species declines and extinctions at large extents.     

Some Guiding Concepts for Conservation Biology

Abstract: The search for generalities in ecology has often been thwarted by contingency and ecological complexity that limit the development of predictive rules. We present a set of concepts that we believe succinctly expresses some of the fundamental ideas in conservation biology. (1) Successful conservation management requires explicit goals and objectives. (2) The overall goal of biodiversity management will usually be to maintain or restore biodiversity, not to maximize species richness. (3) A holistic approach is needed to solve conservation problems. (4) Diverse approaches to management can provide diverse environmental conditions and mitigate risk. (5) Using nature's template is important for guiding conservation management, but it is not a panacea. (6) Focusing on causes not symptoms enhances efficacy and efficiency of conservation actions. (7) Every species and ecosystem is unique, to some degree. (8) Threshold responses are important but not ubiquitous. (9) Multiple stressors often exert critical effects on species and ecosystems. (10) Human values are variable and dynamic and significantly shape conservation efforts. We believe most conservation biologists will broadly agree these concepts are important. That said, an important part of the maturation of conservation biology as a discipline is constructive debate about additional or alternative concepts to those we have proposed here. Therefore, we have established a web‐based, online process for further discussion of the concepts outlined in this paper and developing additional ones.     

Wood Density as a Conservation Tool: Quantification of Disturbance and Identification of Conservation‐Priority Areas in Tropical Forests

Abstract: Inventories of tree species are often conducted to guide conservation efforts in tropical forests. Such surveys are time consuming, demanding of expertise, and expensive to perform and interpret. Approaches to make survey efforts simpler or more effective would be valuable. In particular, it would be good to be able to easily identify areas of old‐growth forest. The average density of the wood of a tree species is closely linked to its successional status. We used tree inventory data from eastern Borneo to determine whether wood density can be used to quantify forest disturbance and conservation importance. The average density of wood in a plot was significantly and negatively related to disturbance levels, with plots with higher wood densities occurring almost exclusively in old‐growth forests. Average wood density was unimodally related to the diversity of tree species, indicating that the average wood density in a plot might be a better indicator of old‐growth forest than species diversity. In addition, Borneo endemics had significantly heavier wood than species that are common throughout the Malesian region, and they were more common in plots with higher average wood density. We concluded that wood density at the plot level could be a powerful tool for identifying areas of conservation priority in the tropical rain forests of Southeast Asia.     

Relation between extinction and assisted colonization of plants in the arctic‐alpine and boreal regions

Assisted colonization of vascular plants is considered by many ecologists an important tool to preserve biodiversity threatened by climate change. I argue that assisted colonization may have negative consequences in arctic‐alpine and boreal regions. The observed slow movement of plants toward the north has been an argument for assisted colonization. However, these range shifts may be slow because for many plants microclimatic warming (ignored by advocates of assisted colonization) has been smaller than macroclimatic warming. Arctic‐alpine and boreal plants may have limited possibilities to disperse farther north or to higher elevations. I suggest that arctic‐alpine species are more likely to be driven to extinction because of competitive exclusion by southern species than by increasing temperatures. If so, the future existence of arctic‐alpine and boreal flora may depend on delaying or preventing the migration of plants toward the north to allow northern species to evolve to survive in a warmer climate. In the arctic‐alpine region, preventing the dispersal of trees and shrubs may be the most important method to mitigate the negative effects of climate change. The purported conservation benefits of assisted colonization should not be used to promote the migration of invasive species by forestry.     

A robust nonparametric method for quantifying undetected extinctions

How many species have gone extinct in modern times before being described by science? To answer this question, and thereby get a full assessment of humanity's impact on biodiversity, statistical methods that quantify undetected extinctions are required. Such methods have been developed recently, but they are limited by their reliance on parametric assumptions; specifically, they assume the pools of extant and undetected species decay exponentially, whereas real detection rates vary temporally with survey effort and real extinction rates vary with the waxing and waning of threatening processes. We devised a new, nonparametric method for estimating undetected extinctions. As inputs, the method requires only the first and last date at which each species in an ensemble was recorded. As outputs, the method provides estimates of the proportion of species that have gone extinct, detected, or undetected and, in the special case where the number of undetected extant species in the present day is assumed close to zero, of the absolute number of undetected extinct species. The main assumption of the method is that the per‐species extinction rate is independent of whether a species has been detected or not. We applied the method to the resident native bird fauna of Singapore. Of 195 recorded species, 58 (29.7%) have gone extinct in the last 200 years. Our method projected that an additional 9.6 species (95% CI 3.4, 19.8) have gone extinct without first being recorded, implying a true extinction rate of 33.0% (95% CI 31.0%, 36.2%). We provide R code for implementing our method. Because our method does not depend on strong assumptions, we expect it to be broadly useful for quantifying undetected extinctions.     

The Potential for Species Conservation in Tropical Secondary Forests

Abstract: In the wake of widespread loss of old‐growth forests throughout the tropics, secondary forests will likely play a growing role in the conservation of forest biodiversity. We considered a complex hierarchy of factors that interact in space and time to determine the conservation potential of tropical secondary forests. Beyond the characteristics of local forest patches, spatial and temporal landscape dynamics influence the establishment, species composition, and persistence of secondary forests. Prospects for conservation of old‐growth species in secondary forests are maximized in regions where the ratio of secondary to old‐growth forest area is relatively low, older secondary forests have persisted, anthropogenic disturbance after abandonment is relatively low, seed‐dispersing fauna are present, and old‐growth forests are close to abandoned sites. The conservation value of a secondary forest is expected to increase over time, as species arriving from remaining old‐growth forest patches accumulate. Many studies are poorly replicated, which limits robust assessments of the number and abundance of old‐growth species present in secondary forests. Older secondary forests are not often studied and few long‐term studies are conducted in secondary forests. Available data indicate that both old‐growth and second‐growth forests are important to the persistence of forest species in tropical, human‐modified landscapes.