The area in and around JNDDNWR Complex is hit periodically by tropical storms and by minor and major hurricanes (Category 3-5). The landscape has repeatedly been sculpted by wind and waves from tropical cyclones. Prior to Hurricane Charley, three major hurricanes occurred in the area since 1900. The Great Miami Hurricane of 1926 first devastated Miami as a Category 4 storm then passed over San Carlos Bay and Captiva Island as a Category 3 storm. In 1944, an unnamed Category 3 storm passed west of the area making landfall near the Sarasota County line. The last major hurricane to impact the area was in 1960 when Hurricane Donna took a storm track similar to Charley. Donna made landfall as a Category 4 Hurricane near Naples and cut a path north to Ft. Myers and across the peninsula to re-enter the Atlantic Ocean near Daytona Beach. The storm track of the eye of Donna was east of the Complex, but the size of the storm was immense, and the Charlotte harbor area was subjected to hurricane force and greater winds for over four hours (Dunion and others, 2003).

The relative inactivity of hurricane-status storms since Donna has been due to a larger global climate pattern. Hurricane activity in the western Atlantic is known to occur in multi-decade cycles that alternate between active and quiet phases of 25-40 years each (Gray 1990, Landsea 1993). In 1995, we entered a new cycle of increased activity that is expected to continue for the next 20 to 40 years (Landsea and others, 1996). Global warming is not expected to alter this multi-decade cycle, but a recent analysis concludes that it will likely result in hurricanes of higher intensity and precipitation rates than experienced in the past with a greater risk of occurrence of highly destructive Category 5 storms (Knutson and Tuleya, 2004).

Habitat damage and effects on bird populations

Hurricanes have direct and indirect effects on birds (Wiley and Wunderle, 1993). The storm's direct effect kills birds (Hooper, 1990, Meyers and others, 1993, Semple, 1936) or displaces them 1,000's of kilometers from their original home ranges (Chapin, 1934). Bird mortality in the short-term, caused by indirect effects from storms, can be precipitated by a loss of food supplies, foraging substrates, nesting and roosting sites, or from increases in predation, conflicts with humans, and changes in microclimate (Cely, 1991, Hooper and others, 1990, Waide, 1991a ,1991b, Wiley and Wunderle, 1993, Wunderle ,1995). Seed, fruit, and nectar feeding birds may suffer the most, immediately after the hurricane (Askins and Ewert, 1991, Waide, 1991a, 1991b, Wunderle and others, 1992, Wiley and Wunderle, 1993). More than 87% of 1,765 Red-cockaded Woodpecker (Picoides borealis) cavity trees were lost in South Carolina during and after Hurricane Hugo (Hooper, 1990). Positive impacts on bird populations also occur. Many insectivorous and omnivorous birds increase after hurricanes and may be adapted to storms by the plasticity of their diets and habitat requirements (Waide, 1991b), but some investigators suggest that birds move away temporarily (Askins and Ewert, 1991). Lost of foraging habitat could cause birds to move elsewhere, especially in winter. Wunderle and others (1992) found large declines of wintering Black-and-white Warblers (Minotilta varia) in Jamaica after Hurricane Gilbert, which was probably related to large losses of tree trunks for foraging. We expect similar effects on birds at JNDDNWR Complex from Hurricane Charley.

Birds may also shift and adapt to storm damage, e.g., by feeding in different habitat (canopy birds feeding in understory, Wunderle, 1995) or by building nests on fallen debris (e.g., Brown Pelicans [Pelecanus occidentalis], Pierce, 1990) or at new locations nearby (Cely, 1991). We expect Brown Pelicans at JNDDNWR Complex to build nests on damaged habitat and successfully rear their young given that fish populations in the area have not been impacted negatively and that spring high tides do not destroy nests. Erosion of pelican nesting islands during Hurricane Hugo and loss of nesting substrate on three islands in South Carolina caused failure of most of the 3,000 nesting pelicans a year after the storm because of lower nesting substrate and a high tide in late May (Marsh and Wilkerson, 1991). Increased numbers of nesting pelicans, however, arrived shortly after that destructive high tide event at a nesting area 45 km to the SW, which was unaffected by the Hurricane Hugo storm surge and spring high tides. Pierce (1990) found that Brown Pelicans learned to build nests on damaged habitat or fallen debris and actually increased the number of fledglings the year following Hurricane Hugo. We expect pelicans at JNDDNWR Complex to also successfully maintain their breeding colonies. Site fidelity is strong for this species based on years of experience with construction related habitat changes and continual increases of nesting pairs (2,12, to >300 pairs) in Mobile Bay, Alabama (J. M. Meyers, personal observation and unpublished data).

Only one Bald Eagle nest was reported lost at JNDDNWR Complex (K. Pednault-Willett, personal communication). Bald Eagles rebuilt 21 of 24 destroyed nests at their original nesting territory after Hurricane Hugo (Cely, 1991). We would expect eagles nesting at the complex, which have lost nests, to rebuild within the same area, although the nest may be built in shorter trees or man-made objects.

Bird population declines from storms over the long-term may be caused by loss of habitat, but habitat usually recovers with plant succession. Forest interior birds may be absent for 1-2 years in former forest habitat after a storm (e.g., 17 mo after Hurricane Joan in Nicaragua, Wiley and Wunderle 1993). The bird community changes in this situation, to one associated with edge and second growth (shrub-scrub) habitat until forest structure develops. An exception to this type of recovery would be severe to catastrophic hurricane damage to old growth forests where recovery may take from decades to a century (Hooper 1990, Wiley and Wunderle, 1993). Some bird populations may decline in old growth mangrove and oak hammock habitats of JNDDNWR Complex, especially those habitats with severe to catastrophic impacts from the storm, e.g., major losses of tree canopy from tree snaps and tips. Monitoring species richness and abundance will provide valuable information on recovery from the hurricane as well as benefiting the refuge's mission and management goals. Methods that account for differences in species detection probabilities will provide better estimates of population and community changes (Buckland and others, 1993, Boulinier and others, 1998, Hines and others, 1999, Nichols and Conroy 1996; Nichols and others, 1998a, 1998b). Residents, such as the Mangrove Cuckoo, and migrants, such the Black-whiskered Vireo or other high priority Nearctic- Neotropical migrants (see Rich and others, 2004) should be the focus of monitoring. Monitoring for these species can be done by distance sampling on transects or by species presence-absence from point counts.

Waterbird breeding colonies in slightly damaged nesting habitat from Hurricane Hugo, showed quite different results in breeding after the storm passed over Pumpkinseed Island, South Carolina (Sheppard and others, 1991). White Ibis (Eudocimus albus) breeding pairs in one colony declined from 10,000 pairs to zero after the storm. Although ibises were in the area, they failed to initiate breeding and may have moved south. There was, however, evidence that freshwater wetlands used by feeding ibises were modified by saltwater intrusion caused by a high storm surge. Nesting White Ibis in the JNDDNWR Complex will probably continue nesting there if freshwater feeding areas were not inundated by saltwater. We expect that a minimum of damage occurred to freshwater feeding areas because of a relatively low storm surge from Hurricane Charley.

Nesting Great Egrets (Ardea alba) and Tricolored Herons (Egretta tricolor) declined at Pumpkinseed Island colonies affected by Hurricane Hugo. Nesting Great Egrets declined 44% because of nesting habitat loss (marsh elder, Iva frutescens). This species also did not change its average nesting height after the storm. On Pumpkinseed Island, Great Egrets showed site fidelity by nesting on damaged habitat, but some may have moved 7 km south. This will probably also occur at JNDDNWR Complex in 2005 with some waterbird colonies nesting on damaged habitat, at lower heights, or even on the ground while other colonies may shift to new nesting areas, if available nearby. Nesting on the ground will probably increase mortality from high spring tidal overwash, which caused an 86% abandonment rate of Snowy Egret (Egretta thula) nests after Hurricane Hugo in South Carolina (Shepherd and others, 1991). About 80-90% of all Snowy Egrets nested on the ground a year after Hurricane Hugo with few nests surviving overwashes that year. Pierce (1990, 1991) reported loss of breeding for one year at Red-footed Booby (Sula sula) colonies in the Virgin Islands after Hurricane Hugo; however, Sooty Tern (Sterna fuscata) colonies increased in breeding pairs in the same area. Saliva (1989) also reported loss of waterbird nesting habitat (loss of rocks or vegetation) and nesting waterbirds on the Culebra Islands, Puerto Rico, but also noted that Hurricane Hugo created more tern nesting habitat (sandy areas).

Wunderle and others (1992) found increases in Mangrove Cuckoos four months after Hurricane Gilbert, but a limited sample size (n = 0 pre-storm, n = 3 post-storm) and increases in detection rates makes this increase doubtful. Overall, Wunderle and others (1992) found that total bird abundance increased in mangrove forest four months after Hurricane Gilbert, but this also could be a result of increased detection rates for birds after the storm. Mangrove Cuckoos are common in Florida and Caribbean coastal mangroves and oak hammocks (e.g., 4.7 Mangrove Cuckoos/km on transects in Puerto Rico, Kepler and Kepler, 1978) and would be an ideal species to monitor post-Charley at JNDDNWR, especially if pre-storm data exist. The species also has lost 60% of its habitat in south Florida to land cleared for residential and agricultural uses (Hughes, 1997). Cuckoos forage on lepidopterans, Anolis (become compressed or more dense in lower vegetation after hurricanes) and orthopterans, which should all increase post-Charley based on previous hurricane research (Hughes, 1997, Reagan, 1991, Torres, 1992, Waide, 1991a). Mangrove Cuckoos, however, may be absent from forest fragments of <12.8 ha (Bancroft and others, 1995).

Another songbird, the Black-whiskered Vireo, may also be an excellent indicator species of mangrove forest habitat conditions at JNDDNWR. This vireo inhabits coastal mangroves and hardwood forests in southern Florida, and specializes in mangroves in Florida (Chace and others, 2002). Black-whiskered Vireos are foliage gleaners of insects or fruit found high in trees (Cruz, 1980b, 1987), which makes them ideal for assessing recovery of old growth mangrove forests. The species is also highly susceptible to parasitism (>50%, Chace and others, 2002) from a recent invader, the Shiny Cowbird (Molothrus bonariensis), a species that will probably increase after the storm and should also be monitored.

Some bird populations increase with changes in habitat caused by storms (e.g., species of shrubs, grasslands, and wetlands; Wiley and Wunderle, 1993). Arengo and Baldessarre (1999) believed that hurricanes may benefit Greater Flamingos (Phoenicopterus ruber) in wetland habitats of the Yucatán Peninsula, Mexico, by changing salinity, food quality, density, and availability, i.e., water depth. By monitoring species presence and absence from sites, managers could track changes in bird communities, bird species extinctions and additions, and turnover rates on the refuge (Boulinier and others, 1998, Nichols and Conroy, 1996, Nichols and others, 1998a, 1998b).

Seagrass beds: manatee habitat

Research suggests that intact seagrass beds are generally resistant to the physical forces of extreme storms (Tilmant and others, 1994, Whitfield and others, 2002). Furthermore storms are thought to improve seagrass meadows by removing detritus and necrotic tissue, resulting in healthy growth after the storm (Whitfield and others, 2002). Significant localized damage, however, can occur from sediment deposition after hurricane-induced erosion, as occurred at the Captiva breach and in Mississippi Sound after Hurricane Camille (Eleuterius and Miller, 1976); or from wave action on patchy, fragmented beds destabilized by propeller scars, motor vessel groundings, or natural blowouts. Past research has shown that scarring can be enlarged and recovery of scars slowed or reversed under hurricane forces (Whitfield and others, 2002). The USGS assessment team could not evaluate the hurricane's impact to seagrass in the entire refuge complex. Given the degree of propeller scars reported for the region in 1995 (Sargent and others, 1995) and 2004 (Madley 2004), it is possible more localized damage occurred in those areas with scarring. The new scars by tree drags may have further destabilized the area near the breach making it vulnerable to new storms.

The USGS assessment team recommends that refuge staff examine the post-hurricane aerial photographs for any large injuries to the seagrass beds within the refuge complex boundaries. If an on-site visit to these sites shows major instability, it may warrant management action. Whitfield and others (2002) and Kenworthy and others (2002) offer management approaches to stabilize degraded areas. Such action should speed recovery and prevent further injury. Continual monitoring for vessel groundings and major scarring, with options to remedy such injuries when they occur, should be considered for inclusion in the Refuge Management Plan. But as every manager knows, prevention is always preferred over remediation.

With seagrass beds extending beyond refuge boundaries, the group also recommends a regional approach to damage assessment and future monitoring and research. The refuge complex has an excellent partnership with the Sanibel-Captiva Conservation Foundation and its Marine Laboratory. Two studies on seagrass integrity and ecological function are part of their core research program (Bortone and others, 2004) and include monitoring within the JNDDNWR Complex. A larger regional collaboration with the South Florida Water Management District and the Florida Department of Environmental Protection is part of these studies and offers possibilities to develop a larger partnership. The refuge, foundation, and its partners may want to consider expanding monitoring to sites beyond Sanibel and Captiva islands and into Pine Island and Matlacha Pass NWRs, perhaps in conjunction with monitoring on some of the mangrove island bird rookeries.

Manatee population

The Florida manatee is listed as endangered under both federal and state law. Concerns regarding Hurricane Charley's impact on manatees focus on three issues: (1) direct mortality, (2) permanent emigration of individuals out of the area as a result of the storm, and (3) the loss or degradation of seagrass beds, an important food source for manatees.

Mortality and emigration concerns are based on analysis of extreme storms that struck the Florida panhandle and north Gulf Coast in the 1980s and 1990s. Mark-recapture statistical analysis of manatee photo-identification data (Langtimm and Beck, 2003) identified lower adult survival rates for manatees of the region during three years with storms rated Category 3 or higher on the Saffir-Simpson hurricane scale: 1985 with Hurricanes Elena and Kate, 1993 with the March “Storm of the Century,” and 1995 with Hurricane Opal. Data suggest that extreme storms have a significant effect on adult survival rates; however, the apparent drop in survival also might be explained, at least in part, by emigration from the region.

Loss or degradation of seagrasses can result in lower manatee reproduction rates as well as a lower number of individuals that the habitat can support. Seagrass beds degraded by propeller are of special concern because they are more vulnerable to hurricane forces (Whitfield and others, 2002) and slower to recover than intact seagrass beds (Kenworthy and others, 2002).

Little is known about what cues manatees use to discern the approach of a major storm or what strategies they use to protect themselves. Given the manatee is a tropical species, one would expect that they have developed behaviors to deal with tropical storms. Several strategies could be employed: move along the coast line away from the storm, rest on the bottom of a deep channel and rise only when needed to breathe, and/or hunker down in a protected inlet or cove. The effectiveness of these behaviors and the vulnerability of manatees to storm forces should depend on (1) the destructiveness of a hurricane, which varies by physical factors such as wind intensity, speed and duration of the storm, storm surge, occurrence of battering waves; (2) physical features of the coastline that can offer protection (barrier islands, protected coves), and (3) coincidental factors such as the density of manatees in the strike area, the number of storms within a season, and occurrence with other health or mortality risks.

Hurricane Charley produced hurricane force winds throughout the JNDDNWR Complex, with the most intense winds crossing North Captiva and northern Pine Island. However, with the exception of the area north of the Matlacha Bridge, relatively few high-use manatee areas experienced the strongest winds. Although Charley was a Category 4 storm when it hit, its potential destruction was reduced by the compact size, short duration, and the lack of a significant storm surge. The outer barrier islands of Sanibel, Captiva, and North Captiva and the inner barrier Pine and Little Pine Islands should have provided some physical protection along with the deeper channels in the pass and sound (Fig. 24). The relatively fast speed of the storm eliminated the development of two additional forces that also could be dangerous to manatees - storm surge and battering waves. All of these factors suggest that the manatee population in the refuge area did not experience a major direct impact from Hurricane Charley. Given the seagrass in the area did not sustain heavy damage it is likely the magnitude of impact to the southwest manatee subpopulation was minimal.

Several studies are already underway to monitor and assess the impact to manatees. These should not require resources from the refuge. USGS Sirenia Project, FWRI, and Mote Marine Laboratory are collaborating under the direction of Dr. Catherine Langtimm to incorporate analysis of effects of the 2004 hurricanes into their long-term study to estimate manatee population parameters using photo-identification data. Estimates of adult survival will require at least two years of data from the winter aggregation sites, but an assessment for impact signatures in the data will be completed after the end of one year. Annual monitoring by Mote Marine Laboratory at Matlacha Isles and the Ft. Myers Power Plant will continue and should provide information on resightings of known individuals and whether a percentage of animals are missing from their usual winter refuge or have been documented at another refuge. Mote Marine Laboratory under the direction of Dr. John Reynolds continues to conduct aerial surveys in Lee County and will be comparing post-hurricane distributions to those from previous years. FWRI continues to monitor manatee carcass recovery, and under the direction of Dr. Holly Edwards will again conduct a winter synoptic survey of manatees at the aggregation sites in the region. All of these data will be analyzed by researchers at their respective institution and reviewed by the Manatee Population Status Working Group under the auspices of the Florida Manatee Recovery and Implementation Team as indicators of possible effects to the population. USGS will be happy to forward information and reports as they become available, put Refuge personnel in touch with key researchers, and provide continued technical support regarding manatee issues.

Other trust species

The refuge staff has observed eastern indigo snakes at drift fence surveys from 2003-2005 (K. Pednault-Willet, personal communication). If a population exists on Sanibel Island, we expect that little or no mortality occurred because of Hurricane Charley. The opening of forest canopy by the storm should provide more suitable habitat and potentially more prey for this threatened snake (J. M. Meyers, personal observation from ongoing telemetry study in Georgia).

Vegetation

We recommend that permanent plots for monitoring vegetation recovery or mortality be established in the refuge's mangrove forests. We believe that monitoring vegetation in a rigorous, quantitative manner with appropriate quality assurance and control may benefit wildlife habitat management. In this regard, the refuge is fortunate to have a partner in the Sanibel-Captiva Conservation Foundation, and its marine laboratory. Professional staff at the lab have established a network of 24 vegetation plots in the mangroves of the JNDDNWR Complex, or nearby sites. Three plots were established at each of eight locations, with four locations representing impounded mangroves and four locations being non-impounded mangrove forest. Data derived from these plots are already proving useful. We recommend some improvements for the plots for long-term monitoring. At present the plots are not permanent. A center stake is needed and individual stems should be tagged, and mapped (Fig. 43). Stem mapping (Fig 44) and the use of permanent, individually numbered, aluminum tree tags will greatly reduce the possibility of error during subsequent surveys of the plots, allow for measuring growth, recruitment, and delayed mortality. Mapping is valuable if another storm occurs; you only need to find a single, tagged stem to reconstruct the entire plot based on geometry. Finally, we recommend that these plots be established also in waterbird islands and the Shell Mound Trail mangrove forests.

- Figure 43. Sanibel-Captiva Conservaton Foundation mangrove plot 7 located west of Dixie-Beach Boulevard, Sanibel Island, Florida. The yellow arrows point to the pink paint used to mark the border of the plot. [larger image] - Figure 44. Plot layout (in red) used by USGS for mangrove vegetation studies in coastal Everglades, Florida. From the permanent center stake, the distance and bearing () to each stem is recorded as well as the diameter breast height (DBH) and species. [larger image]

New establishments of vegetation plots could be two additional plots to each waterbird island, where possible, especially Bird Key, Hemp Island, and Givney Key. An alternate method may be required on the smaller fringe and overwash islands. New plots in non-mangrove vegetation, such as the tropical hammock at Shell Mound Trail and Legion Curve areas would also be beneficial. Because of the more complex nature of the vegetation structure in these areas, particularly higher species richness, larger plot sizes will be needed. Once established, the plots need to be sampled annually for approximately five years and then they can be split into two groups with each group sampled every other year.

The survey team visited only a small portion of the JNDDNWR Complex, yet we observed a variety of invasive exotics. Conducting surveys for this group of plants requires different methods than for monitoring recovery. The only effective method is repeated, 100% coverage, followed by treatment to eradicate the pest plants. Seedlings of some exotic plants existed in the mangrove forests prior to Hurricane Charley, but were suppressed by the intact forest canopy. Loss of the mangrove canopy may allow these seedlings to begin vigorous growth. Furthermore, tip-up mounds will provide a different type of habitat for both native and exotic plants to colonize.

We provide, here, some recommendations for determining sediment elevation in mangrove forests in relation to recovery of the forests. In may not seem to be logical, or feasible, but sampling sediment elevation is necessary. We know that large scale disturbance can result in sediment collapse in mangroves and retard, or even, prevent recovery (Cahoon and others, 2003). We recommend a study to look at sediment elevation changes in the heavily damaged mangroves. At least two sites should be examined, one on a large waterbird island (Lumpkin, Hemp, Givney) and the other in the Shell Mound Trail - Power Line Road area. The idea would be to test potential management actions that could be taken to prevent, slow, or ameliorate elevation loss. Such actions might include planting mangrove propagules, especially the red mangrove, with and without fertilization. An additional factor to examine is wetland water elevation. The experiment would be designed to give an answer as to which course of action (above) is feasible for refuge managers.

Ecosystem function and refuge management

Disturbances, especially large infrequent disturbances (LIDs), affect ecosystem function in south Florida and the Caribbean basin. Ecosystems there have developed and are maintained by the periodic severe hurricanes (LIDs) on approximately 20-25 year intervals or more. Although ecologists have recognized the importance of LIDs in maintaining ecosystems (Turner and Dale, 1998), there is still much to learn (Tanner and others, 1991). In the Caribbean basin, hurricanes are the major LID that organizes natural systems (Walker and others, 1991).

Major ecosystem structural changes occur with hurricanes. One of the most important factors that result from hurricanes is control of species composition (Walker and others, 1991). Plant growth and recolonization occurs rapidly after storms, but predicting effects of storms is difficult because of the heterogeneity of storm damage within relatively small areas and differences between storms (Brokaw and Grear, 1991, Brokaw and Walker, 1991, Walker, 1991). Hurricanes also affect some aspects of ecosystem dynamics, such as processes caused from increased nutrients from large amounts of litter fall (Tanner and others, 1991). All of these hurricane effects will take place on JNDDNWR Complex to some extent, depending on the location and severity of the storm in that area.

Storm effects on animals, after the initial direct mortality (which is usually low), are highly correlated to changes in habitat structure and function. Increase in solar radiation in damaged forests stimulates plant growth, depending on the seed bank, which in turn creates habitat and food for animals (J. Meyers, personal observations). JNDDNWR Complex will experience these changes, which are similar to what has occurred after other storms in the Caribbean basin (Brokaw and Grear, 1991, Brokaw and Walker, 1991, Doyle and others, 1995, Pimm and others, 1994, Scatena and others, 1996, Smith and others, 1994, Tanner and others, 1991, Waide, 1991a, 1991b, Walker, 1991, see also Biotropica Special Issue 28[4a]). Most vegetation at JNDDNWR Complex will recover rapidly. Exceptions may be old growth mangrove forests, which will recover slowly or not at all (see mangrove section). Insects, probably least disturbed by storms of all animal classes, will increase in species and numbers (outbreaks, Waide, 1991b). Butterflies (associated with new growth, see Torres, 1992), frogs (Woolbright, 1991, 1996), lizards (Reagan, 1991), and freshwater shrimp (Covich and others, 1996) increased after Hurricane Hugo. These increases, if they occur on JNDDNWR Complex, may affect populations of trust species, such as the threatened indigo snake and migratory birds. A modified/improved design and analysis for sampling trust species post-storm, e.g., presence/absence especially for birds, would be beneficial for long-term management and may also provide information on hurricane effects for management on other refuges in areas affected/unaffected by the hurricanes of 2004 (Boulinier and others, 1998, Nichols and Conroy, 1996, Nichols and others, 1998a, 1998b). A regional approach using this type of design and analysis for all refuges (damaged and undamaged by hurricanes in coastal Florida and Alabama in 2004) may provide valuable information on the effects of the storms and potential management strategies for many priority, endangered, and trust species.