Preliminary Survey

One participant (7096) operated a larger vessel (10 m) and primarily fished offshore in U.S. federal waters (4.8–322.0 km), whereas the remaining five individuals operated smaller center-console vessels (<7.6 m) and offered inshore and nearshore (0.0–4.8 km) charters. Participants in the study reported that they took, on average, between 40 and 200 for-hire fishing trips per year. Full-time professional fishing experience ranged from 5 to 16 years. While all six participants were comfortable with mobile phones and used voice services anywhere from 10 to 50 times/d, reported use of other applications on mobile phones was minimal. Four of the six participants indicated that they had sent at least one text message before the study began, while only three individuals used text messaging on a daily or weekly basis. The two participants who had never used text messaging before were, however, aware of the technology.

Effort and Catch

All fishing activity took place in estuarine and marine waters of southeastern North Carolina. Text message fishing reports were submitted between March 15 and July 31, 2008. The number of fishing reports submitted per participant ranged from 5 to 40. Over the course of the study, six individuals submitted 128 text message fishing reports on behalf of 326 fishing clients (anglers). One individual (9950) indicated that only 20% to 40% of trips were reported, and another individual (4338) dropped out of the study after April; the remaining four participants indicated that all fishing trips and finfish encounters were reported during the study period. The mean number of anglers taken per trip varied by participant, as did fishing effort (Table 1).

A total of 1,957 finfish representing 28 of the 57 species or species groupings included in the RECTEXT syntax were reported by participants. The most frequently reported species were Spanish mackerel Scomberomorus maculatus (n = 881; 45%), spotted seatrout Cynoscion nebulosus (n = 204; 10.4%), bluefish Pomatomus saltatrix (n = 183; 9.4%), red drum (n = 163; 8.3%), and kingfishes Menticirrhus spp. (n = 150; 7.7%). Collectively, these taxa accounted for 80.8% of the fish reported.

In addition to species identification and number caught, participants were asked to designate the disposition of the catch (kept or released) and were given the opportunity to report total length measurements in inches for any single observation. Clients kept 1,123 fish and released 834, and the percentages varied by species and participating captain. Of the finfish that were retained by participants, Spanish mackerel (n = 611), kingfishes (n = 150), spotted seatrout (n = 131), and bluefish (n = 72) made up 49% of the harvest. While popular inshore species like red drum and southern flounder Paralichthys lethostigma had high release rates (90% and 58%, respectively), several species or species groupings were not retained by any for-hire operation. Three of the six participants utilized the reporting system's ability to include total length measurements for a total of 13 fish.

System Performance

No detectable errors were observed in either the operation of the prepaid mobile phones or Twitter, the text message aggregating service used in this study. Because we were able to activate the phones, load minutes, and register the phones with the aggregator prior to delivering the phones to participants, pitfalls related to system registration were avoided.

Although we received a small percentage of fishing reports with detectable errors, all information submitted by participants using the RECTEXT syntax was usable for analysis. Because reports were first posted in raw code form on the aggregator website, we were able to review them for typing errors, assuming that all transmissions submitted were received. Of the 128 reports received, only five (3.9%) had detectable errors (i.e., data fields not separated by spaces, number of interactions placed before species codes, etc.). Considering that the 128 reports contained a total of 548 separate data fields (sections of code separated by spaces), 99.1% of the data submitted was usable when it was received without further action on our part. Four of the five errors observed were submitted by the two participants who had never before sent or received a text message.

In addition to the data entered manually by participants, time of submission was automatically recorded as part of each fishing report. As individuals were asked only to submit the report as soon as possible after the completion of their trip, we received reports at various times of day throughout the study (Figure 3). While most participants indicated that reports were submitted directly after trips or at latest, several hours after the trips, participants 7105 and 4338 for the most part did not submit individual trip reports. These individuals usually submitted reports in batches after several consecutive trips. When time of report submission was plotted against date for the entire study period and participants 7105 and 4338 were excluded, it was apparent that the majority of reports (76%) were submitted between 1200 and 2400 hours. As the majority of fishing trips were initiated in the morning hours, we did not expect to receive a large number of reports before noon.

Postdata Collection Survey

A poststudy survey completed by participants indicated that the fisheries reporting system was easy or relatively easy to use. All participants agreed that we provided sufficient training and instruction materials on system use. Prior to this study, all six participants indicated that they kept a personal paper logbook of their catch history for their own records; therefore, it was not difficult for them to compile the information necessary for each text message report. Individuals were also asked which method of reporting should be developed if mandatory reporting for the for-hire sector is required in the future. Of the four participants who responded to the question, three indicated that some form of electronic data submission (i.e., mobile phone, website data entry, etc.) would be preferred over traditional methods, such as paper-based logbooks or trip-tickets.

System Evaluation

The results of this study indicate that the text message based fisheries reporting system described here is capable of handling basic effort and catch information submitted directly from recreational anglers in the field and displaying that information in near real-time in a dedicated database. Although we chose to focus on marine recreational fisheries because of the relative lack of information available for many U.S. marine stocks (NOAA Fisheries 2008a), additional syntaxes could easily be developed for freshwater fisheries applications as well as for many other volunteer environmental monitoring programs (USEPA 1998; Savan et al. 2003). We were able accomplish our objectives in part by developing the compact syntax RECTEXT such that we could work within the limitations of the SMS infrastructure. Twitter provided not only a free text message aggregating service for this trial but also a portal by which we could review participants' fishing reports for errors in coding prior to transcription. Comments we received from participants indicate that this method of data submission and collection has potential and should be further evaluated.

Errors made by participants using the RECTEXT language were fewer than anticipated, considering that 34 reports (27%) were submitted by individuals who had never used text messaging prior to this study. As we were unable to find information on data entry error rates associated with other self-reporting programs, we can assume that errors might occur during transcription if paper forms are used (Gouveia et al. 2004) or during data entry through electronic methods. Using demographic information obtained from the preliminary survey, we determined that four of the five detectable errors in our study were unknowingly made by the two oldest participants (age range = 45–55+), who also happened to be the ones that had never before used text messaging. This is not surprising as other researchers have observed that middle-aged and older users were not as adept at text messaging as younger users (Kurniawan 2008; Soriano et al. 2006). Having anticipated this, we felt it was imperative to provide brief training sessions with individuals (if needed) prior to the start of data collection. Positive feedback from users indicated to us that the pocket reference guide and the time spent initially instructing participants on system use were beneficial and appreciated. While it is possible that additional errors, such as those related to species identification, also occurred during this study, we felt this was unlikely considering that professional captains submitted all the reports for this study.

Although reports submitted by participants were received by the data aggregator moments after submission, there was often considerable delay (up to 6 h) in the time it took to route some of these reports to our website for assimilation and viewing purposes. This delay probably occurred because at the time of this study, Twitter only had limited capacity to perform noncritical functions, such as RSS feeds, in addition to those required to perform basic services. As real-time display of fishing reports on our website was not required for our evaluation, we recommend that to truly take advantage of this reporting technique, a fee-based text message aggregating service should be used to provide actual real-time assimilation of self-reported data (Brown et al. 2007).

For this study, participants were neither informed of the data aggregation website nor given access to the website used to display the fishing reports. As a consequence, participants were not able to determine whether text messages successfully submitted from their phones were received by the system. As text messages are routed through mobile communication networks by using a store and forward mechanism (Brown et al. 2007), transmission of text messages to the intended receiver cannot be guaranteed. Although lost or failed transmissions rarely occur, the benefit of a text message based reporting system is that electronic “receipts” can be retrieved from originating phones, providing evidence of text message submission attempts in case validation is needed. For this study, participants were instructed to resubmit text message reports only if the message failed to send, and to our knowledge such failures did not occur. As we did not ask participants to keep an additional record of reports for validation purposes, we assumed that all text messages submitted were received by the data aggregator and displayed on our website. In future surveys, we suggest that participating anglers be able to browse their personal text message reports for the dual purpose of error identification and verification of any possible text message submission failure.

Text Messaging Compared with Existing Electronic Self-reporting Mechanisms

Paper logbooks, catch cards, and angler diaries have been the most popular types of recording device for angler-reported data in past surveys as these are inexpensive to manufacture, are easily customized, and can be used for small groups (Gartside et al. 1999; Campbell et al. 2003) or even statewide surveys (Bray and Schramm 2001). Despite their benefits, paper-based methods require significant time and resources to manage effectively (Cooke et al. 2000). Repeated handling of the data also reduces the time that administrators can spend interacting with and responding to the participants. In a review of 47 angler diary and voluntary logbook programs in Canada, Cooke et al. (2000) indicated that of the 70% of programs that failed, many did not establish adequate feedback mechanisms to respond to angler inquiries or improve the survey once initiated. While real-time data acquisition may not be necessary for all applications, the use of paper-based self-reporting methods all but eliminates this opportunity.

Use of a web-based portal allows registered anglers to log in to a website and fill out customizable forms detailing trip-level fishing activity, resulting in timely assimilation of data that are usable by fisheries managers. Established agency-based programs along the Atlantic coast, such as the Maryland Volunteer Angler Survey (MDDNR 2009), have been able to provide year-to-year comparisons between angler-reported data and MRFSS data for a variety of species, including recreationally important striped bass Morone saxatilis and summer flounder Paralichthys dentatus. The program has inspired other state fisheries agencies in Maine, New Hampshire, and New Jersey to adopt similar web-based reporting programs for anglers. While this reporting approach can reduce the cost of survey administration and improve angler confidence in MRFSS data, the drawbacks of this technique are that (1) anglers must have access to the internet and (2) information is usually submitted after the fishing event, subjecting the data to potential recall and prestige biases (Pollock et al. 1994; Connelly et al. 2000). The web-based reporting method also makes it problematic to validate report contents unless the database can be configured to receive reports submitted electronically from the field by using a network or internet-enabled device.

Dedicated electronic logbooks provide yet another method to submit and assimilate self-reported data. These applications allow for detailed data entry, on-site storage, and report submission at-sea to managers primarily through periodic internet connections with satellite or mobile communication networks. Example programs, such as Catchlog ( www.Catchlog.com) and Olfish (www.olfish.com), both used primarily in commercial fishing operations, are operated from laptop computers on the vessel. Third-party applications like these often bundle additional features to benefit the fisherman in improving operations and making informed, timely marketing decisions. In cooperation with interagency project partners (Alaska Department of Fish and Game and International Pacific Halibut Commission), NOAA Fisheries has developed its own Interagency Electronic Reporting System for select commercial fisheries in the Pacific Northwest (NOAA Fisheries 2009). The obvious disadvantages of electronic logbooks include the requirement for a computer and communication terminal onboard the vessel as well as any required costs for the application itself (if third party) and associated subscription or annual maintenance fees. If the primary objective is to collect basic, single-trip-level effort and catch information from recreational anglers at the completion of trips, a mobile phone-based reporting method would be much less costly and would be accessible to more users than applications requiring internet- or satellite-based communications (CTIA–The Wireless Association 2008; Horrigan 2008). In addition, a text message based reporting program as described in this study is highly scalable and customizable due to the small data “footprint” of SMS compared with voice calls (Brown et al. 2007) and the use of open-source applications and off-the-shelf components.

Despite the advantages of using mobile phones for self-reporting, there are currently drawbacks to this electronic approach. For example, SMS is limited to a maximum of 160 characters and does not support more data-rich file types, such as photos and videos (Brown et al. 2007). Therefore, the technical limitations of SMS might make it programmatically difficult to report extremely detailed information from anglers. In comparison with using a full-sized keyboard to submit data via a website application, usability of SMS is arguably dependent on the type or style of phone used (Balackrishnan et al. 2005). The prepaid, “candy bar” phones used in this study were small and required that participants toggle through keys to select the appropriate character. Use of a compact syntax like that developed for this study on mobile phones with full or slide-out keyboards or phones with predictive text capabilities (MacKensie and Soukeroff 2002) would probably minimize reporting errors and decrease time required to compose reports. Implementation of a mobile phone reporting system would also require that communication networks exist in a survey area. Because the number of subscribers and geographic reach of mobile communication networks have expanded at a rapid rate over the last several years (CTIA–The Wireless Association 2008), this is likely to be an insignificant issue in the years to come.

The real-time nature, on-site submission capability, and growing acceptance of this universal, low-cost communication method are arguably the most compelling aspects of this approach when compared with existing electronic self-reporting methods. As text message reports are received momentarily after submission by an angler, it is possible for administrators to simultaneously view this data on a website from either an internet-accessible mobile phone (i.e., smartphone) or a portable computer while on-site. Conveniently, this method would only require administrators to have smartphones or computers for viewing reports, whereas individual anglers would only need standard mobile phones to submit reports.

Consideration should also be given to the fact that the reliance on mobile communication devices and the resulting technology are growing at a rapid rate in this country. Mobile phone subscriptions in the United States have dramatically increased in recent years from 194.4 million in June 2005 to 262.7 million in June 2008, reaching a penetration rate (percent of total population with a mobile phone subscription) of 84.7% (CITA–The Wireless Association 2008). During that same time period, the number of text messages sent increased from 57.2 × 10⁹ to 600.5 × 10⁹. In a recent survey conducted by the Pew Internet and American Life Project (Horrigan 2008), over 2,000 individuals were asked how hard it would be to give up a specific technology; respondents indicated that the mobile phone would be the hardest to do without, followed by the internet, television, and landline telephone. The concept that a cost-effective, easy-to-use, portable, real-time reporting system does not exist for the individual angler is no longer an issue. The mobile phone and its ability to communicate with data networks clearly indicate that it will remain a staple of American households for the foreseeable future.

Suggestions for Future Evaluation and Implementation

Any self-reporting method is likely to have a higher acceptance rate by all end users if the contents of reports can be easily validated (Pollock et al. 1994). As the primary objective of our study was to develop and evaluate the actual data collection process, future studies should consider incorporating on-site intercept surveys to validate report contents. The method described here is capable of assembling small amounts of information from a large number of anglers during a short period of time. We suggest that further evaluation be focused on fishing practices or special events that should be monitored but that for various reasons have been difficult to characterize with traditional survey methods. In addition to the for-hire reporting method described here as well as potential for personalized electronic diaries for individual anglers, the text message approach may provide a mechanism to collect verifiable effort and catch data from anglers participating in organized clubs, associations, and fishing tournaments.

Despite the large number of anglers that participate in marine fishing tournaments throughout the country (Falk et al. 1989; Oh et al. 2007), characterization of the effort and catch data associated with these events is usually avoided by fisheries managers simply because of the inherent biases and logistical difficulties associated with sampling these types of events (Essig and Holliday 1991). In a typical tournament setting, a large number of anglers expend high amounts of fishing effort over a very short time period, often targeting larger fish than would be targeted during normal fishing activities (Loomis and Ditton 1987). Despite the difficulty in quantifying tournament fishing impacts on fisheries resources, the number of tournaments appears to be increasing nationally and the effort and catch associated with these events could be significant in relation to the total recreational harvest for many species. Where traditional intercept methods applied in a tournament setting would likely lead only to data that are of limited use for fisheries management purposes (Pollock et al. 1994), simultaneous, real-time reporting by all registered anglers may enable managers to track general effort and catch statistics for these events.

A tournament setting would allow the opportunity for a sampling protocol to be established to validate angler's reports. As it may be logistically difficult to verify all anglers' text message reports for content, a design could be developed such that a percentage of tournament anglers is requested at random to complete follow-up exit surveys regardless of fishing success (Figure 4). An automated text message notification sent to anglers immediately after report submission (via the aggregating service) would expedite the postfishing intercept process in often hectic tournament sampling situations. By requesting that anglers submit reports as soon as possible after completion of fishing activity but before the tournament deadline, administrators would have time to conduct staggered exit interviews with anglers. At exit interviews, self-reported information, such as species identification, length, and weight estimates, could be validated on site by survey administrators. In the unlikely event that the internet database does not receive the text message report, each angler would have an electronic receipt available on the mobile phone to provide proof of attempted transmission during the designated reporting period. Such a design would allow survey administrators to collect and verify content submitted by successful anglers (those submitting fish to the tournament) and by anglers that voluntarily exit the tournament after fishing for any variety of reasons.

This format of “required” self-reporting, albeit with paper catch cards and shoreside surveyors, has been used successfully in bass tournaments (Quertermus 1991) and in a national park where fishing licenses were surrendered in exchange for mandatory trip reports (Larson et al. 1986). Data collected over several years could allow fisheries managers the opportunity to characterize the fishing practices of tournament anglers. If successful, data collected may provide an additional scientific index to facilitate the development of future management actions. One distinct advantage of such a customizable data collection program is that anglers would have an opportunity to contribute effort and catch data directly to management agencies in a format that would ultimately be usable by anglers, fishery managers, and tournament organizers.