Center for Ecological Research

Since 1994, we have been studying many aspects of the breeding and wintering biology of Florida Grasshopper Sparrows. For a more complete listing of the peer-reviewed articles that we have published on this endangered species, please refer to our publications: by subject page.

The following are highlights of some of our recent, major findings.

Source - Sink Dynamics for Florida Grasshopper Sparrows

Fragmentation and edge effects adversely affect passerines in North America, primarily by reducing territory density, reproduction and survival. As natural landscapes become increasingly altered and fragmented by human development, it is important to understand the demographic parameters of remaining sub-populations. We wanted to determine if remaining dry prairie fragments in central Florida were acting as population sources or sinks for the federally endangered Florida Grasshopper Sparrow (Ammodramus savannarum floridanus).  We obtained sparrow survival estimates from 2 populations in central Florida, and combined these with productivity estimates from 3 sites from 1996-1998 to determine if a sub-population was acting as a source or sink.

We determined that there were specific sub-units within a site that consistently functioned as sources.  For Florida Grasshopper Sparrows, we found that interior core areas, >400m from edge, were consistently sources.  We think that the only way that Florida Grasshopper Sparrows will be able to persist at these sites is if the core source areas produce enough surplus young to compensate for the sink habitat along the wide borders of these prairie fragments.   We think that large prairie fragments, possibly >4000 ha, are necessary for maintaining source habitat for Florida Grasshopper Sparrows and, possibly, other grassland bird species.

Population Viability Analysis for Florida Grasshopper Sparrows

 We developed a stage-structured, spatially explicit population viability (PVA) model for Florida Grasshopper Sparrows (Ammodramus savannarum floridanus) and used these results to develop reserve designs for this taxon. The baseline model that represents the present status predicted an 18% chance of falling below the extinction threshold of 60 individuals within 50 years. However, we caution that the true value of PVAs are not to predict absolute values of viability or number of individuals, but rather to use sensitivity analyses to evaluate the relative impact of different management options.

We conducted a sensitivity analysis on the PVA to determine which parameters had the greatest affect on population viability. We found that increased environmental variation (stochasticity) was very important and had a 32% affect on the final metapopulation. The model was also very sensitive to survival and fecundity for adults and juveniles.  If adult and juvenile survival and fecundity were increased or reduced by 5%, this had a 24-35% affect on the final metapopulation abundance. Other parameters were less sensitive.

In an effort to find ways to increase long-term metapopulation viability for Florida Grasshopper Sparrows, we tried to determine what would happen if we created several management options that would improve some of the parameters in our initial model. Positive options included: re-introducing sparrows to new prairie sites (either large or small), increasing prairie size by removing exotic pines at Avon Park, restoring improved pasture near Three Lakes WMA and Kissimmee Prairie State Preserve (hereafter Kissimmee Prairie), improving survivorship by reducing predation, and increasing dispersal.

We also modeled several “negative” possibilities, such as habitat loss, loss of private land that presently supports sparrows, or decreasing the frequency of prescribed fires on protected prairies.  Infrequent prescribed burning increased the probability of the entire metapopulation would fall below 60 adult males to 33%. Loss of dry prairie located on adjacent private land at Three Lakes and Kissimmee Prairie reduced the amount of core habitat (>400 meters from an edge) at each site. If all prairie habitats on private land were lost, the probability of falling below the extinction threshold increased to 29%.

We found that increasing the amount of prairie habitat, especially core habitat, had a strong positive affect on viability. This was especially true at Kissimmee Prairie. A concentrated effort to introduce large numbers of juvenile sparrows to a new, large, previously unoccupied prairie also benefited metapopulation viability.  We modeled the USFWS plan for Florida Grasshopper Sparrow recovery (10 sub-populations with at least 50 males per site) and estimated that the metapopulation had a 90% probability of remaining above the metapopulation extinction threshold. Other management options tended to have more local benefits.  We found that population viability decreased substantially should any of the negative possibilities occur. 

Finally, we used the findings from the PVA work to develop five reserve designs that could be used to improve long-term sparrow viability.  To do this, we modeled what would happen if two or more management options were employed simultaneously.  Our first design incorporated three options that were likely to have the greatest positive affect on sparrow viability. These options included: habitat restoration on private land adjoining Three Lakes and Kissimmee Prairie, and introducing 15% of the juveniles from these two sites to a new large site.  This design increased long-term metapopulation viability to 97% and predicted an average total abundance after 50 years of  1,156 individuals.

 Our design for habitat improvement at Avon Park included creating connectivity between Delta/OQ and Bravo by removing exotic pines, exotic pine removal elsewhere on the base, and predator control to improve survivorship.  We found that this design did not have a strong affect on metapopulation viability but it did increase the average number of males present after 50 years at Delta/OQ from 1.4 (under the baseline model) to 82.2.  This design also predicted an increase in the average number of years sparrows were present at Delta/OQ from 4.7 to 26.2 years (out of 50 years). 

 Our efficient design employed three relatively inexpensive options: introduction of 15% of the juveniles from Three Lakes and Kissimmee Prairie to a new large prairie fragment, pine removal at Echo, Bravo, and Delta/OQ ranges, and improved connectivity between Delta/OQ and Bravo.  This deign reduced the probability of falling below the metapopulation threshold to 7% and predicted final abundance of 738 individuals.

 In our negative design, we theorized that: 1) all native prairies on private land that adjoined protected state or federal land was lost, 2) protected sites were burned less frequently.  This design had a 45% probability of falling below the metapopulation threshold.

 In conclusion, these analyses demonstrate that:

• Without further action, Florida Grasshopper Sparrows are in jeopardy of falling below the metapopulation threshold within the next 50 years.
• Active management, especially use of prescribed fire, of existing populations is critical for sparrow persistence.
• There are management options that are likely to help increase the sparrow metapopulation, which will reduce the probability of extinction.
• Employing two or more options in a reserve design is likely to minimize the likelihood that the sparrow metapopulation will fall below the extinction threshold.

Response to winter prescribed burning by breeding Florida Grasshopper Sparrows

  Populations of Florida grasshopper sparrow are small and declining. Prescribed burning is the primary management tool used to maintain their grassland habitats, but the effects of this management practice on the breeding density and reproductive success of these populations are poorly understood. We conducted a three-year spot-mapping study of three winter burn classes (0.5 yr, 1.5 yr, and 2.5 yr post fire) in native dry prairie on two sites in central Florida to determine the effects of fire management on habitat selection and reproductive success of these two sparrows.

Florida grasshopper sparrow densities were greater on recently burned plots (both 0.5 and 1.5 yr post-burn) than on plots that had not been burned in 2.5 yrs. Grasshopper sparrow reproductive success was also higher in the recently burned (0.5 yr post-burn) plots than in 2.5 yr burn plots. Our results indicate that a fire rotation of < 3 years is necessary to maintain suitable breeding habitat for Florida grasshopper sparrows.

Florida Grasshopper Sparrow Survivorship during the Non-breeding (Winter) Season

We tried to determine survival rates and identify specific causes of mortality using radio-marked Florida Grasshopper Sparrows during the non-breeding seasons (August-March) from August 1996 to March 1998.  We established two study areas for this project: Avon Park Bombing Range and Three Lakes Wildlife Management Area. We captured Florida Grasshopper Sparrows by flushing them into mist nets and attached 0.9-1.2 g radio transmitters to sparrows in adult plumage. We found that predation was the primary source of mortality for Florida Grasshopper Sparrows, and that avian predators were implicated in 10 of 12 deaths. Hawk density increased substantially during the non-breeding season with the arrival of migrant raptors that overwinter on dry prairie habitats. 

In the first field season (1996-1997), the survival rate for Florida grasshopper sparrows during the 7-month non-breeding season was 34.2% using Kaplan-Meier analysis and 44.1% using the Mayfield method. During the second field season (1997-1998), the season-long survival rate was 31.0% using Kaplan-Meier analysis and 38.3% using the Mayfield method.  When we used the Mayfield method without partitioning probabilities by month, season-long survival rates for Florida Grasshopper Sparrows were 51.8% and 51.4% for the first and second field seasons, respectively.  These survival rates were somewhat lower than published estimates of annual adult male Florida Grasshopper Sparrow survivorship (mean = 0.598, Delany et al. 1993; 0.483 and 0.533, Perkins and Vickery 2001). Though survival probability during the breeding season has not been estimated, mortality risk during the non-breeding season appears to be much greater than during the breeding season.     

Ancestral polymorphisms in genetic markers obscure detection of evolutionarily distinct populations in the endangered Florida grasshopper sparrow

Research conducted by: NatalieL. Bulgin, H. Lisle Gibbs, Peter D. Vickery, and Allan J. Baker

Genetic analyses of bird subspecies designated as conservation units can address whether they represent units with independent evolutionary histories and provide insights into the evolutionary processes that determine the degree to which they are genetically distinct. Here we use mitochondrial DNA control region sequence and six microsatellite DNA loci to examine phylogeographical structure and genetic differentiation among five North American grasshopper sparrow (Ammodramus savannarum) populations representing three subspecies, including a population of the endangered Florida subspecies ( A. s. floridanus). This federally listed taxon is of particular interest because it differs phenotypically from other subspecies in plumage and behavior and has also undergone a drastic decline in population size over the past century. Despite this designation, we observed no phylogeographical structure among populations in either marker: mtDNA haplotypes and microsatellite genotypes from floridanus samples did not form clades that were phylogenetically distinct from variants found in other subspecies. However, there was low but significant differentiation between Florida and all other populations combined in both mtDNA (F_ST = 0.069) and in one measure of microsatellite differentiation ( = 0.016), while the non-Florida populations were not different from each other. Based on analyses of mtDNA variation using a coalescent-based model, the effective sizes of these populations are large (80 000 females) and they have only recently diverged from each other (< 26 000 ybp). These populations are probably far from genetic equilibrium and therefore the lack of phylogenetic distinctiveness of the floridanus subspecies and minimal genetic differentiation is due most probably to retained ancestral polymorphism. Finally, levels of variation in Florida were similar to other populations supporting the idea that the drastic reduction in population size which has occurred within the last 100 years has not yet had an impact on levels of variation in floridanus. We argue that despite the lack of phylogenetic distinctiveness of floridanus genotypes the observed genetic differentiation and previously documented phenotypic differences justify continued designation of this subspecies as a protected population segment.

Molecular Ecology 12:831-844.