Black-and-white aerial photography at a scale of 1:24,000 was the principal source of information used to assess distribution and abundance of SAV in Chesapeake Bay, its tributaries, and Chincoteague Bay in 1994. There were 1,509 photographs from 139 flight lines which were carefully examined to identify all SAV beds visible on the photography. Outlines of SAV beds were subsequently drawn onto USGS 7.5 minute quadrangles and then digitized, which provided a digital database for analysis of bed areas and locations. Ground-survey information collected in 1994 was tabulated, placed onto the same 7.5 minute quadrangles, and entered into the SAV digital database.
The 1994 SAV aerial photography was obtained by Air Photographics (Martinsburg, West Virginia) using a Wild RC-20 camera with a 153 mm (6 inch) focal length Aviogon lens and Agfa Pan 200 film. The camera was mounted in the bottom fuselage of Air Photographics' Piper Aztec, a twin engine reconnaissance aircraft. Photography was acquired at an altitude of approximately 12,000 feet, which yielded 1:24,000 scale photographics.
There were 139 flight lines covering 1,771 miles of shoreline and yielding 1,509 exposures. Flight lines included land features that were necessary to establish control points for accurate mapping (Figure 6). Flight lines to obtain the photography were predetermined by Air Photographics to include all areas known to have SAV, as well as most areas which could potentially have SAV in the middle and upper zones [i.e., all areas where water depths were less than 2 meters at mean low water (mlw)]. In the lower zone, sections of the upper Rappahannock and upper York rivers, and most of the James River were not photographed for analysis because of the continued absence of SAV in these areas as evidenced by ground truth.
Flight lines were prioritized by sections. Flights were timed to occur during the peak growing season of species known to occur in the sections. In addition, specific areas with significant SAV coverage were given priority. Dates of photography for each quadrangle are noted on each map in Appendix B.
General guidelines followed during acquisition of aerial photography (Table 1) address tidal stage, plant growth, sun elevation, water and atmospheric transparency, turbidity, wind, sensor operation, and plotting. Adherence to these guidelines assured acquisition of photography under nearly optimal conditions for detection of SAV, thus insuring accurate photointerpretation. Deviation from any of these guidelines required prior approval by VIMS staff. Quality assurance and calibration procedures were consistently followed. The altimeter was calibrated annually by the Federal Aviation Administration. Camera settings were selected by automatic exposure control. Sun angle was measured with a sensor on the plane. Flight lines were plotted on 1:250,000 scale maps to allow for overlap of photography. To minimize image degradation due to sun glint, the camera was equipped with a computer controlled intervalometer which established 60% line overlap and 20% sidelap. An automatic bubble level held the camera to within one degree tilt. The scale/altitude/film/focal length combination was coordinated so that SAV patches of one square meter could be resolved. Ground-level wind speed was monitored hourly. Under normal operating conditions, flights were usually conducted under wind speeds less than 10 mph. Above this speed, wind-generated waves stir bottom sediments which can easily obscure SAV beds in less than one hour. The pilot used experiential knowledge to determine what acceptable level of turbidity would allow complete delineation of SAV beds. During optimum flight conditions the pilot was able to distinguish bottom features such as SAV or algae at low tide. Excessively turbid conditions precluded photography. Determination of optimum cloud cover level was based on pilot experience. Records of this parameter were kept in a flight notebook. Every attempt was made to acquire photographs when there was no cloud cover below 12,000 feet. Cloud cover did not exceed 5% of the area covered by the camera frame. A thin haze layer above 12,000 feet was generally acceptable. Experience with the Chesapeake Bay has shown that optimal atmospheric conditions generally occur two to three days following passage of a cold front, when winds have shifted from north-northwest to south and have moderated to less than 10 mph. Within the guidelines given for prioritizing and executing the photography, the flights were planned to coincide with these atmospheric conditions where possible. All film was processed by Air Photographics. A 9 inch x 9 inch black-and-white contact print was produced for each exposed frame. Each photograph was labeled with the date of acquisition as well as flight line number. Film and photographs were stored under appropriate environmental conditions to prevent degradation.
For this analysis, USGS 7.5 minute quadrangle maps were utilized for mapping SAV beds from aerial photography, for digitizing the SAV beds, for mapping ground-truth data, and for compiling SAV bed area measurements. Figure 7 gives locations of 181 quadrangles in the study area which includes all regions with potential for SAV growth. Most quadrangles are sequentially numbered for efficient access to data. The name corresponding to each quadrangle in Figure 7 is listed in Table 2. Identification and delineation of SAV beds by photointerpretation utilized all available information including: knowledge of aquatic grass signatures on film, distribution of SAV in 1994 from aerial photography, 1994 ground-truth information, and aerial site surveys. USGS 7.5 minute quadrangle maps (1:24,000 scale), printed by the Mid-Continent Mapping Center of the National Cartographic Information Center on stable transparent mylar, were used as base maps from which to make copies. Distortion-free, identical copies of these base maps were made at the same scale on stable transparent mylar using a contact print process.
SAV beds from the 1994 aerial photographs were then mapped onto these mylar copies of USGS 7.5 minute quadrangles. Delineation of each SAV bed was facilitated by superimposing the photographic print with the appropriate mylar quadrangle on a light table. SAV bed boundaries were then traced directly onto the mylar quadrangle with a pencil. Where minor scale differences were evident between a photograph and a quadrangle, or where significant shoreline erosion or accretion had occurred since USGS publication of a map, either a best fit was obtained or shoreline changes were noted on the quadrangle. All photointerpretation of 1994 aerial photography for SAV beds was done by one scientist who also photointerpreted the 1971 to 1993 aerial photographs.
In addition to delineating SAV bed boundaries, an estimate of SAV density within each bed was made by visually comparing each bed to an enlarged Crown Density Scale (Figure 8.) similar to those developed for estimating forest tree crown cover from aerial photography (Paine, 1981). Bed density was categorized into one of four classes based on a subjective comparison with the density scale. These were: 1, very sparse (<10% coverage); 2, sparse (10-40%); 3, moderate (40-70%); or 4, dense (70-100%). Either the entire bed or subsections within the bed were assigned a bed density number (1 to 4) corresponding to the above density classes. Some beds were subsectioned to delineate where variations in SAV density occurred. Additionally, each distinct SAV unit (bed or bed subsection) was assigned an identifying two letter designation unique to its map. Subsections were further identified as contiguous beds by the addition of two letters unique to that sequence. These contiguous bed identifications aid the tracking and analysis of single natural bed units that were subsectioned due to variation in SAV density. Coupled with the appropriate SAV map number and year of photography, these two letter designations uniquely identify each SAV bed in the database.
The perimeters of all SAV beds mapped from the aerial photography onto the mylar copies were digitized in ARC/INFO, using an Altek Model 41 tablet, with a resolution of .001 inches (.00254 cm) and an accuracy of .005 inches (.0127 cm). The beds for each quadrangle were digitized twice in two separate ARC/INFO coverages. Each coverage was plotted at an exact scale of 1:24,000 on translucent plotter paper and overlaid on the original mylar for visual checking. In instances where the digitized SAV bed boundaries did not correspond to within 0.5 mm of the original, the bed was re-digitized. Once the SAV outlines on both coverages passed visual inspection, a bed-by-bed comparison of the areas (sq. meters) was made as an additional quality assurance check. Individual beds were rejected and redigitized if they were larger than 0.1 hectare and there was a difference of greater than 5% area between the two coverages, or larger than 1 hectare and there was a difference of greater than 1% area between the two coverages. The bed-by-bed comparison was useful in identifying instances where SAV beds were incorrectly labelled, thus eliminating coding errors.
Prior to each digitization session, the Altek instrument was checked manually against a digitizing standard. This was accomplished by first securing a mylar quadrangle with SAV polygons to the digitizing tablet. The mylar standard was then secured to the same quadrangle and digitized. The digitized area of each standard was compared to the known area of the standard. If a variation between the known and the mean of the observed areas exceeded 1.0%, the maps were redigitized. In addition, the digitized standard was plotted and checked visually against its location on the map to verify positional accuracy.
After all quadrangles were digitized, the resulting digital data was combined to form a single data set for the entire Bay. The quadrangle edges were then scanned to ensure that the SAV polygons were consistent on both sides of the border (edgematching). Inconsistencies were resolved by checking the mylar maps and re-interpreting the photography if necessary.
Maximum accuracy was maintained by exclusively using mylar quadrangles and standards rather than paper ones, which can change scale as a result of changes in air temperature and humidity in the digitizer room.
Standard operating procedures (SOPs) were developed to facilitate orderly and efficient processing of the 1994 SAV maps and the SAV computer files produced from them, and to comply with the need for consistency, quality assurance, and quality control. SOPs developed include: a detailed procedure for digitization of SAV maps; a digitizer log in which all operations were recorded and dated, which was used to guide and record editing operations; and a flow chart used to track progress of all operations.
The SAV coverages in Universal Transverse Mercator (UTM), ARC/INFO, Zone 18 format were used to calculate area in square meters for all SAV beds. These areas are reported as USGS 7.5 minute quadrangle, segment, and zone totals in the tables in the Results section. Segment and zone totals were calculated by using an overlay operation of the segment and zone regions on the SAV beds in ARC/INFO. The definition of the segments used in this analysis are provided in Table 3. The 1991-1993 data were also edgematched as above and area totals recomputed.
SAV distribution data are presented and discussed based on different segmentation and zonation schemes from those used in the previous SAV distribution and abundance reports. The segmentation scheme used in this report was that adopted by the Chesapeake Bay Program (Flemer et al., 1983) (Tables 3 and 4; Figure 9). The Upper, Middle, and Lower zonation scheme used in the previous reports, as established by Orth and Moore (1982) and modified by Orth et al., (1989) was adapted to the new segmentation scheme. It was followed as closely as possible but, necessarily, had to be modified to accommodate the new segment boundaries (Figure 9). Data are presented for the years 1991-1994, based on these new segmentation and zonation schemes, in order to follow the trends report (Orth et al., 1995) which covers the years 1971-1991 using the Chesapeake Bay Program segmentation scheme.
The area between the Chesapeake Bay Bridge and the Susquehanna Flats is referred to as the Upper Bay zone. The salinity within each zone roughly coincides with the major salinity zones of estuaries: polyhaline (18-25 ppt), Lower zone; mesohaline (5-18 ppt), Middle zone; oligohaline (0.5-5 ppt), Upper zone. Although the major rivers and smaller tributaries of Chesapeake Bay have their own salinity regimes, the distribution of SAV in each river is discussed within the zone where it connects to the Bay. SAV distribution in Chincoteague Bay is presented and discussed separately from Chesapeake Bay.
Ground surveys were accomplished by cooperative efforts from a number of agencies and individuals. Although not all areas of the Bay were surveyed, the data did provide valuable supplemental information. The surveys confirmed the existence of some SAV beds mapped from the 1994 aerial photography, as well as SAV beds not visible from the photography because they were too small at 1:24,000 scale. The surveys also provided species data for many of the SAV beds. Ground-survey information supplied to VIMS researchers was included on the SAV distribution and abundance maps reproduced in Appendix B. Each survey was designated by a unique symbol to identify the different methods of sampling. In most cases the symbols on the SAV maps (Appendix B) were enlarged and offset from the actual sampling point to avoid confusion with the mapped SAV bed. Where species information was available, it was included on the map. Because of space limitations on the maps reproduced in Appendix B, occasionally one or more survey points were combined where the information was duplicated. All ground-survey data supplied to VIMS are tabulated in Appendix D.
In Maryland, ground-survey data were obtained in 1994 by VIMS, Stan Kollar of Harford Community College, the USGS National Center, the Maryland DNR, Patuxent River Park staff, and by the Citizens' volunteer survey. The USGS National Center provided ground-survey data for the Potomac River. Patuxent River ground-survey data were obtained by the Maryland-National Capital Parks and Planning Commission Patuxent River Park staff and the Maryland Department of Natural Resources (Naylor and Kazyak, 1995). The Citizens' volunteer survey, including the Ocean Pines Boat Club of Berlin, Maryland, and the Essex Community College SAV group of Baltimore County, Maryland, under the guidance of the USFWS and the Chesapeake Bay Foundation (CBF), identified SAV locations and SAV species when possible throughout various areas of the Chesapeake and Chincoteague bays. Volunteers, who were recruited through press releases, newsletters, and personal letters, were provided with a SAV identification guide, reduced 1992 SAV maps to aid in the location of SAV beds, and data sheets for reporting visits to numerous sites around the bays. USFWS staff mapped the data on copies of 1992 SAV distribution maps (USGS 7.5 minute quads with 1992 SAV beds). These maps were supplied to VIMS SAV researchers and transferred to the 1994 SAV distribution maps reproduced in Appendix B. Data from the Patuxent River Park staff, and the Citizens' surveys were compiled and tabulated by USFWS. This table became the basis of the much expanded table published in Appendix D.
One 1994 SAV research project being conducted on the Susquehanna Flats by Stan Kollar of Harford Community College, Maryland, also provided data in the form of species presence by estimated percent cover, although these percentages are not reported here.
For those areas in Virginia waters where aerial photographic evidence of SAV beds was inconclusive, photoverification was accomplished by ground-truth surveys. Observations were principally made from small boats and by divers snorkeling over areas indicated from the photographs. In the York, Piankatank, and Rappahannock rivers, where VIMS researchers transplanted SAV (principally eelgrass), transplant sites were also examined carefully by divers for any extant SAV. VIMS scientists also surveyed a number of sites in the Chesapeake Bay as part of an intensive quantitative SAV study (VIMS, unpublished data). Data for Virginia waters were also collected by the Citizens' volunteer survey (compiled by the USFWS). In addition, a great deal of ground-survey information could be extrapolated from earlier studies (Orth et al., 1979; Orth and Moore, 1982). SAV beds in the lower Bay contained primarily one or two species and most areas underwent wide fluctuations in distribution and abundance since the first bay-wide survey in 1978.
Ground-survey data from all sources reported here are presented in Appendix D.