Grantee Research Project Results
Final Report: Economic Effects of HABs on Coastal Communities and Shellfish Culture in Florida
EPA Grant Number: R831707Title: Economic Effects of HABs on Coastal Communities and Shellfish Culture in Florida
Investigators: Larkin, Sherry L. , Degner, Robert L. , Adams, Charles M.
Institution: University of Florida
EPA Project Officer: Packard, Benjamin H
Project Period: November 1, 2004 through October 31, 2006 (Extended to October 31, 2007)
Project Amount: $91,959
RFA: Ecology and Oceanography of Harmful Algal Blooms (2004) RFA Text | Recipients Lists
Research Category: Environmental Statistics , Aquatic Ecosystems , Water
Objective:
(1) To estimate the change in gross revenues to various business sectors of coastal communities affected by HAB (e.g., red tide) events;
(2) To calculate the costs incurred by coastal communities to address the effects of HAB events including planning efforts, contingency plans, beach patrols and cleanup, etc.; and
(3) To quantify the effects of HABs and HAB-related harvest regulations on commercial molluscan shellfish operations.
Empirical application will be restricted to Florida for manageability and effectiveness.
Summary/Accomplishments (Outputs/Outcomes):
Objective (1):
One component of this study employed the use of proprietary data to estimate sales reductions (economic losses) due to red tide events (Morgan, Larkin, and Adams, in submission (a); Morgan, 2007). Aside from the quantitative results, implications for public managers, funding agencies, scientists, and private business owners abound. For example, recent scientific advances have resulted in improved forecasting models and a suite of potential prevention, control and mitigation strategies for red tides. The cost of these activities will need to be compared with potential benefits, which could be proxied by preventable losses to affected businesses, such as the restaurants measured in this research. The estimated benefits resulting from this study provide support to the hypothesis that the restaurant sector is affected during red tides, as are the traditional commercial fisheries or marine-related recreation and tourism sectors which seem to have received the most media attention.
The characteristics of red tides (i.e., as an exogenous environmental event with localized impacts) could support the development of private business disruption insurance such as that available for floods or hurricanes. While such an industry would need fairly accurate prediction models capable of forecasts and associated probability of bloom estimates, the state of the science is progressing and could support relatively low-cost annual premiums. This is because red tide conditions can affect a relatively large geographic region but also vary across short distances as shown in this study. Such a mitigation approach may be the most appropriate in addressing the impacts of red tides as opposed to other forms of public assistance, including control and mitigation activities.
If another red tide should occur, the restaurant sector could demonstrate its eligibility for economic recovery assistance, such as loans offered by the Small Business Administration, using the empirical approach developed in this study. Statistically significant daily sales losses due to red tide conditions are close to the Small Business Administration (SBA) loan values provided to Florida restaurants. For example, the SBA provided loans ranging from $4,832 to $81,912 per event to individual restaurants from 1996-2002. By comparison, this study found sales reductions ranging from $868 to $22,404 per red tide event (for six consecutive days of recorded red tide conditions). This suggests that cumulative sales losses resulting from the persistent presence of a red tide over a period of several days may have driven historical SBA loan values to the maximum end of the range.
Lastly, the average cell count measures that corresponded to the days when red tide conditions were noted by the restaurant managers greatly exceed the 5,000 cells per liter threshold level that is used to close commercial shellfish harvesting areas. Thus, cell count measurements may need to be much higher to impact beachfront restaurant patrons when compared to the shellfish closure threshold levels, particularly in light of corresponding wind speed and wind direction. If these thresholds are supported in other studies, they could be used to estimate red tide effect thresholds for any beachfront business and, thereby, provide support for the future use of subjective red tide data in empirical analyses. Firm-level analysis is, however, essential since red tides were not found to affect all restaurants in this study despite the close proximity of the restaurants to one another.
Overall, the methodology and results of this study may provide a means to extrapolate findings to a regional level for the restaurant sector using the cell count thresholds, number of days, and number and type of beachfront restaurants. Such losses can be compared to estimated costs of any proposed red tide prevention, control and mitigation strategy for other businesses dependent on the economy’s tourism sector. This methodology could be applied to other businesses sectors (commercial fishing, lodging, beach attendance, etc.) to get a better estimate of regional costs and benefits associated with proposed red tide control, mitigation and management.
Another approach to providing quantitative statistical evidence of red tide impacts on coastal business is to use routinely collected data from the Florida Department of Revenue (FDOR). At this time, we have summarized the monthly gross sales and number of reporting businesses for 108 zip codes that are nearest to the Gulf of Mexico. The data within the zip codes represent only those zip codes with at least four firms reporting, which is an improvement over previous studies that have been restricted to using only data if at least 10 firms reported. This is an important distinction for the zip codes that cover a relatively small geographical area. The data cover two sectors (i.e., restaurants and lodging or kind codes 08 and 39, respectively) that are located across nine coastal counties (i.e., Okaloosa, Franklin, Gulf, Pinellas, Sarasota, Manatee, Charlotte, Lee, and Collier) from February 1998 through January 2003. The data set was restricted to the two largest kind codes to reduce the size of the task at the request of FDOR. In addition, the archive tapes containing data prior to February 1998 are "unreadable" so could not be incorporated into the analysis (Dale Snyder, FDOR, personal communication). Data for the most recent period (i.e., February 2003 - December 2005), which is needed to include a period with severe red tides, has not been obtained despite repeated assurances that delivery was imminent (see Appendix A for additional details on this data). Thus, analysis of the existing data set containing the 1998-2003 data has commenced.
To evaluate the potential impact of a red tide on the lodging sector a fax survey of hotels was initiated by the Lee County Chamber of Commerce and resulted in 47 responses out of a total 142 hotels with valid addresses (33.1%). The complete survey protocol involved four total contacts, two by fax and two by mail. The hotel manager was asked to indicate the effect that red tide had on gross room rental sales in 2005 using an 11-point scale (-5 = extremely negative, 0=no effect, +5=extremely positive). While 35.6% of responding hotels (N=45 to this question) indicated that red tide had no effect on revenue from room rentals in 2005 (a year with generally bad red tide conditions), the remaining 64.4% did report a negative effect (the majority, 45.6% of the total, selected ratings from -3 to -5). The losses associated with hotels providing negative effects on room sales averaged $26,764. A similar analysis was done for gross food service and all other service sales. Only 29 of the 45 (64.4%) responding to queries about the room rates responded, suggesting that not all hotels offer such services or that it is a more resilient sector. Of the 29 respondents, 48.3% indicated no effect on sales from food service or other services, which means that 51.7% were negatively affected. For comparison, 31.0% provided ratings of -3 to -5. Losses of sales for food and other services averaged $10,440.
Since red tide impacts are most commonly associated with interrupting marine activities, and beach going is the number one marine activity in Southwest Florida (and area that experiences frequent red tide events), another study sought to examine secondary data on park attendance. In particular, this study tested whether attendance at Lover’s Key State Park was affected by red tide events and, if so, determine the absolute and relative size of that affect. (King, Larkin and Adams, 2008). It was hypothesized that several seasonal and environmental factors would also affect attendance, as would any media attention to such events.
The analysis indicated that on days when red tide was present, the informational data taken from local newspapers is a better proxy for explaining change in attendance than the actual red tide cell count measurements themselves. This information also supports previous work on human responses to red tide events, which showed that 70% of coastal residents stated newspapers were the most important single source of information on red tide events. The decrease of park attendance averaged of 398 visitors (21% of average daily visitors). A decrease in daily park attendance by 398 visitors, coupled with the average person spending $42.60 per day at the park, the analysis predicts a potential daily loss to Lee County of $16,954.80 in revenue attributable to declining visitation at Lovers Key State Park when an article is published about red tide in a local newspaper. In the three-year study period, there were 263 days where the keyword “red tide” was published in a local newspaper article. The 263 published articles in the three-year period created a decrease in attendance by 104,674 visitors. With the average visitor spending $42.60 per day, the decrease in attendance resulted in a $4,459,112 loss in revenue in the three-year study period. The loss in revenue results regardless of the context in which the red tide reference was used. This creates a notion that any press is bad press in regards to “red tide.” After analyzing these results, the media must understand that red tide is an issue to which the public reacts negatively and must be careful in the context where “red tide” is used especially when explaining issues such as mitigation or control techniques.
The study has shown that informational data are relevant to understanding the impact of red tide to park attendance and should be available to every park. With complete park attendance, weather, and red tide data, this analysis may be repeated at other coastal parks in Florida in order to show an assessment of attendance losses due to environmental events. With recreation and tourism numbers continuing to grow, it is beneficial to understand the extent with which environmental and news related factors affect park attendance. An especially promising mitigation strategy is the real-time reporting of beach conditions on the Internet and available by telephone (Kirkpatrick et al., 2007). This system employs lifeguard as consistent and trained sentiments to disseminate timely information to direct the public to the best recreational opportunities at the time. At the same time, such a system employed in Lover’s Key State Park could prevent loss of recreational dollars to the system as a whole by redirecting recreationists to other areas instead of losing them altogether, which is what the majority of residents were found to do when confronted with red tide conditions at the beach.
Prior to this study, the project investigators conducted a survey of residents in Southwest Florida to better understand their knowledge level concerning red tides and how they react to such events. This study was conducted after a period of relatively high red tide activity (i.e., throughout 2000). Red tide events were also generally widespread in this region in 2005, and more recently in the northeast part of the state (nearing Georgia in Atlantic waters) in late 2007. At this time, the state has initiated several research efforts to help mitigate red tide related losses under the Red Tide Control and Mitigation Program. As such, an examination of the 2001 survey data serves to measure previous impacts and preview the potential for amplification of negative economic impacts in cases where effective public education efforts are not forthcoming (Morgan, Larkin, and Adams, in submission (c); Morgan, 2007). This study is important because armful algal blooms occur worldwide and can cause severe economic losses to waterfront communities. Blooms of Karenia brevis (red tide), a species native to the Gulf of Mexico, can kill fish and marine mammals and irritate respiratory systems of humans. Knowing how and why participation in marine-based activities (beach going, fishing, and coastal restaurant patronage) is affected during a red tide is fundamental to estimating the change in use values that occur during blooms. Using a sample of residents in Southwest Florida, we find that factors influenced by extension activities have a larger impact than socioeconomic factors commonly hypothesized to affect individual response behavior.
This study found that he share of residents surveyed that were affected by red tide events ranged from a low of 37% for restaurant patronage to a high of 70% for beach going; this is not surprising as beach going entails more direct interaction with affected waters and aerolized toxins. Using a model to explain a respondents overall reaction (or lack thereof) to a red tide event in terms of their participation in any one of four marine-based activities, revealed that the probability of a reaction (i.e., that an individual would react negatively by cutting short, delaying or relocating) increases with the number of activities they participate in and their knowledge level regarding red tides (measured either one of two ways). None of the demographic variables were found to have a statistically significant affect on this overall probability. When estimated by activity, the probability of a reaction was directly related to the level of participation for the beach going, pier fishing and restaurant patronage activities only; boat fishing was the exception.
The probability of a reaction for boat fishing was only affected by how many red tide effects the individual could describe: the more effects listed (such as fish kills and offensive odor) the higher the probability an individual will react (i.e., cut short, delay, or relocate). For beach or pier fishing, on the other hand, higher reaction probabilities were associated with more years of residence, lower income levels, and higher participation levels. Higher probabilities of losses to restaurant patronage were found for year-round residents, those that frequented restaurants more often, those that could describe more effects, and for those that had more information sources. Lastly, for the beach going activity, the residents who have lived in Florida the longest, spent more days at the beach on average, were able to recall accurate red tide effects, and were more knowledgeable of red tide effects were more likely to react.
In terms of potential strategies to mitigate losses from red tide events, the knowledge-related variable results offer some clues. The probabilities of an effect on beach-going and the probability of boat fishing being affected increase from about 50% to over 90% when individuals can list more than five red tide effects. The effect on probability for restaurant patronage reveals the same magnitude of increase, but at a lower overall probability, which is to be expected given that restaurant patronage is lower overall.
These results can be contrasted with those of the variable that best describes what the individuals actually know as ascertained from their ability to correctly answer 17 true and false questions about the nature and health effects of a red tide bloom. Although the overall probability of a reaction is less sensitive, the probability of beach-going being affected during a red tide (by individuals choosing to cut short, delay, or relocate) was only approximately 30% with one correct answer, however, this probability increased to over 60% with six correct answers and exceeded 90% with 15 correct answers. Continued comparison of the overall and beach-going models revealed similar sensitivity with respect to a resident’s mention of red tide effects, where individuals that described one to five red tide effects resulted in higher probabilities of reacting of 50% to 90%, respectively. Respondents engaged in at least one activity had a 50% probability of reacting to a red tide that increased to over 90% with participation in all four activities, and those that spent a single day on average at the beach had a 70% probability of reacting to a red tide.
The final empirical analyses in this paper estimated the probabilities of specific reactions, namely, whether an individual would cut short, delay or relocate their participation in response to a red tide. Notably, some additional demographic variables were statistically significant at this finer level of resolution; males had a higher absolute probability of relocating their beach-going activities and a higher probability of relocation away from a waterfront restaurant. Individuals that were over age 55 had a slightly lower probability of relocating to another restaurant. Those individuals that had attended some college were more likely to cut short fishing from a pier. Full time residents were, perhaps the least impacted; the probabilities of cutting short a boat fishing trip or of relocating patronage of a restaurant were lower. Figure 4 also shows the impact of red tide information sources and how much residents think they know; the former increased the probabilities of delaying a boat fishing trip or relocating restaurant patronage, the latter increased the probability of cutting short a beach trip.
Collectively these results should spur private and public sector beach-related concerns to develop appropriate educational materials and make them available in both time and location-sensitive manner in an effort to minimize both short and long run probable losses. Given the general finding that each additional piece of red tide information and source resulted in increased likelihood that a resident would alter his or her participation in a marine-related activity (and, thus, incur lost revenues), extension materials could focus on a few direct educational messages provided at a single venue to targeted recreationists, rather than an abundance of literature and messages scattered across many sources and aimed at all residents.
Extrapolation of study results to all residents within the study counties generates absolute numbers of affected residents based on participation in each activity. For example, survey results revealed that 76% of residents engaged in beach-going, which suggests that 503,989 people visit Manatee and Sarasota beaches at least once per year. Of these, 70%, or 352,793 people have indicated that red tide affected their beach participation. Fifty-eight percent, or 204,620 residents, delayed their day at the beach, while 84,670 cut short a beach day and 63,503 relocated away from red-tide affected beaches. These changes in participation levels revealed in this study can be combined with estimated values of a day at the beach, or on a boat fishing trip, or beach and pier fishing, or restaurant’s average daily sales to estimate potential losses to local businesses and communities reliant on marine-related activities. Future resident surveys could focus on their willingness-to-pay for red tide prevention, mitigation, and control measures to avoid disruption in recreational activities that are affected by red tide events, particularly in the case of beach users.
These findings offer strong support for provision of extension materials targeted towards residents engaged in marine-related activities. Public managers may affect awareness and, therefore, participation decisions by providing information where conditions are optimal for beach-going as opposed to information on where conditions may be unpleasant or even harmful. Such reports are analogous to daily alpine ski reports, although providing them for red tides is more complicated in that bloom effects are not limited to a single, isolated beach location. That said, coastal communities offer a variety of nature-related activities that could be included such reports. These types of education activities may foster an atmosphere where coastal populations are protected from the negative aspects of a red tide bloom, while simultaneously encouraging residents to frequent alternative, unaffected, local marine-related businesses and therefore retain consumer spending in the area.
Objective (2):
This objective sought to determine the financial and managerial impacts of red tide events on nine county and eighteen city governments charged with management of public beaches along the Gulf coast in Florida (Morgan, Larkin, and Adams, in submission (b); Morgan, 2007). In these areas the majority of funds for red tide-related cleanups were generated by tourism tax dollars, with only two counties relying strictly on their regular county dollars, perhaps due to the lack of public beaches in these areas (e.g., none were reported in Franklin and only one in Charlotte). In all, four counties and two cities were able to provide actual dollar amounts specific to red tide events that occurred on their public beaches. These six locations provided red tide-specific costs totaling $653,890 over the 2004-07 time period, with total expenditures per event (including labor, equipment, supplies and vendor fees) ranging from $11,114 to $250,000. Only two cities, Longboat Key and Naples, have placed red tide cleaning costs as a line-item in the annual budget, in the amounts of $100,000 and $50,000, respectively.
Although Sarasota County provided the only official written protocol outlining specific policies and procedures in the case of a red tide event, each of the other counties and cities appeared to follow a similar pattern of activity. Initially, a complaint of odor from a red tide-related fish kill was received by the agency, either from a member of the public or from beach or park personnel. An agency member, or private consultant, with some level of resource management experience, was sent to the area to investigate the claim and establish a cleaning protocol that would meet any human welfare, environmental and access restrictions (e.g., human health hazard, turtle nesting site, protected dunes, etc.). At this point, cleaning personnel were assigned from existing staff, outside labor agencies, or prison trustees, while machinery was also either diverted from usual uses or rented from local suppliers. Once the debris was collected in either trucks or garbage bags, it was hauled to local waste disposal sites following prescribed regulatory procedures (e.g., dead fish might be bagged, buried, or incinerated in designated locations).
In addition to data concerning red tide fiscal costs, respondents provided insight into the difficulties associated with cleaning public beaches in the event of a fish kill. For example, many of the Gulf County beaches harbor protected nesting areas for turtles and seabirds, as well as native flora that have low tolerance levels for invasive mechanized equipment. Several beaches have strict environment protocols in place to limit or prevent removal of washed up marine materials for a set period of time in an effort to preserve the natural state of coastal ecosystems. Such policies include criteria such as “no-rake” areas, cleaning only when there are “significant numbers” of dead fish, or they require “one large fish per foot of shoreline” or “substantial portion of the beach be covered by fish for 24-48 hours, or to a depth of six inches” before cleaning can occur; adherence to environmental policies must be enforced by public officials on private businesses, and in some cases exceptions have been granted for resorts that have established cleaning policies. On at least one occasion, the state health department stepped in and required a county to clean private homeowners’ beaches as the fish kill was deemed a human health hazard.
Five of the counties, and only one city, mentioned public notification of an ongoing red tide event, typically by placing warning signs on the beach and sending alerts to tourism-related businesses. However, a few counties and cities mentioned financial support of the grassroots organization START, or Solutions To Avoid Red Tide, which has active membership in most of the responding regions and works to educate the public and businesses about red tide. Manatee and Sarasota counties have equipped their lifeguards with Blackberries, which are used to send twice-daily messages concerning red tide and other beach conditions.
An important finding is the estimated costs of a red tide event per linear foot of beach. Sarasota spent an average of $4.87 per linear foot of beach to provide the labor and equipment necessary to remove the dead fish resulting from a single red tide event that occurred in October 2006 through February 2007. In Pinellas, seven cities were reimbursed an average of $14.27 per linear foot of beach for red tide-related cleaning required throughout 2005; however, incidence and duration of the events were not mentioned, and city expenditures may have exceeded county reimbursements due to in-kind labor and equipment reallocations. This information may provide a useful baseline for estimation of red tide-related budget needs for other cities and counties that are responsible for public beach management. However, it should be noted that public government protocols associated with red tide events are strongly dependent on all of the following factors: the timing, duration and severity of an event; size of budget and labor force; overall importance of tourism (evidenced by tourism tax collections); quantity and accessibility of public beaches; and the environmental regulations that are specific to each locality.
Objective (3):
Another component of the project was to quantify the effects of HABs and HAB-related harvest regulations (i.e., closures at 5,000 K. brevis cells per liter and higher) on commercial molluscan shellfish operations. To that end, a survey of hard clam and oyster dealers was conducted in early 2006 (Larkin, Adams, and Degner, 2007). While the oyster industry is not primarily affected by red tide, the dealers require the same license and they are subject to supply disruptions due to environmental events and are, thus, included for comparison. A total of 91 dealers with valid ISSC certificates, and that purchased from Florida harvesters in 2005, were surveyed by mail in June 2006. The census of dealers produced 41 responses but only 33 completed surveys (36.3%). The respondents were found to be concentrated in their activities; of those reporting majority dealings of clams, oysters, or other marine products, the share of those species that comprised total landings were 98%, 88%, and 86%, respectively. Gross revenues from marine species averaged $5.1 million in 2005. Using an 11-point scale from 0 (not important) to 10 (extremely important), red tide was found to have a more important affect on sales than other factors presented. When asked to rate how detrimental closures of increasing length would be to gross and net sales, on average, the rating increased at a decreasing rate indicating that initial losses are relatively high but begin to dissipate as dealers are able to compensate. Ratings by dealers specializing in hard clams were higher, on average, than rating by dealers specializing in sales of oysters. For hard clam dealers, regaining markets and buyers was the most important concern following a closure, followed by the need to arrange for substitute products and the need to regain employees. On average, dealers lost $21,330 in gross revenue ($8,378 in net revenue) as a result of a one-week unanticipated supply disruption due, for example, to red tide. Clam dealers or firms with more employees would experience larger losses. Retail suppliers, dealers with more dependence on out-of-state supplies, and those with higher sales would experience smaller losses. All results were quantified and statistically estimated and tested using regression analysis.
Journal Articles:
No journal articles submitted with this report: View all 24 publications for this projectSupplemental Keywords:
EPA region 4, FL, Florida, harmful algal blooms, social science,, RFA, Scientific Discipline, Water, Ecosystem Protection/Environmental Exposure & Risk, Health Risk Assessment, Oceanography, Economics, algal blooms, Environmental Monitoring, Ecological Risk Assessment, Ecology and Ecosystems, marine ecosystem, shellfish transport, bloom dynamics, economic assessments, water quality, Florida, human health, algal bloom detectionProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.