February 2004 7 Plant Cover Estimation: A Contiguous Daubenmire Frame By Charles D. Bonham, Daryl E. Mergen, and Sam Montoya lant cover is one of the most measured char- acteristics of vegetation. Cover is used to de- scribe plant communities by individual species, to describe plant-environmental interac- tions within these communities, to monitor effects of changes on plant species within these communi- ties, and to assess restoration or reclamation success of disturbed plant communities. Cover values have been widely used to calculate synthetic variables such as diversity indices and plant community coef ficients. Individual plant species cover has been used to develop models for prediction of individual species biomass History of Cover Classes The use of cover to describe vegetation communi- ties originated in Europe during the first third of the 20" Century. During this period, plant ecologists used rating scales for plant specics abundance in small areas of larger vegetation communities Terms such as dominant, abundant, frequent, and rare were qualitative assessments of species oceur- rences, These terms were considered sufficient to de- scribe vegetation. Each term had to be accompanied by a definition supplied by the ecologist and as a re- sult, comparisons of vegetation composition from place to place were difficult because terms were not consistent among observers. A widely used cover-abundance scale developed by Braun-Blanquet during the 1920s was consid- ered to be absolute because the scale was related to a fixed size of area as determined by the ecologist. Five scales were assigned numbers from | to 5 where | = less than 5% cover, 2 cover, 25-50% cover, 4 10-75% cover, and 5 = greater than 75% covér. The lower scale of 1 was often given an additional symbol to indicate level of abundance The noted USA plant ecologist, R. F. Daubenmire, modified the foregoing cover-abundance scale by division of the > 75% into two scales of 5 = 75-95% and 6 = 95~100% (Daubenmire 1959). He also used a 20 cm x 50 em plot to estimate cover according to these scales. It is important to note that ail historical cover measurements, including the method of Daubenmire, were originally obtained to describe plant communities, not to compare these communities, Cover Definition In spite of historical documentation stating other- wise, Daubenmire has been widely cited for his de- velopment of the cover class method. However he stated that plant cover is an approximation of the area over which a plant exerts its influence on other parts of the plant community. In contrast to those before him, he insisted that cover is not an estimate of the shaded area on the ground and that plants may be present, but only the plant’s area of influence may be measured, This de- finition led to the approximating polygons of plant canopies that occurred within his plot rather than on larger areas used by European ecologists. Because of the difficulty in measuring a plant’s area of influ- ence, other investigators soon interpreted the mea- surement Daubenmire made, with a given plot size and shape, to be synonymous with an estimation of cover rather than a plant’s area of influence. In turn, several definitions of plant cover fol- lowed, but most can be summarized as the percent- age of ground covered by live vegetation material within a given reference area such as a plot. Further Modifications As noted above, Daubenmire derived the mark- ings for his famous plot from the cover classes de-18 RANGELANDS 26 (1) veloped and used by European ecologists to de- scribe areas of vegetation. They used cover scales for individual plant species occurring within an area called a releve’ which was the bounded area of in- terest. Daubenmire insisted that he did not develop a “plot,” but rather he developed “a canopy cover- age method” using a marked plot. His desire was that students of plant ecology be informed of the difference between the plot and the method. Perhaps, he also recognized that the de- scription of his method in the famous 1959 publica- tion had become so corrupted through “modifica- tions,” among vegetation ecologists, that his efforts might be hopeless to insist that we either use his method as it was published or not use his name. So, he personally expressed to the senior author of this paper that at least students of vegetation ecology should be taught that his “canopy coverage method” was just that, a method not a partitioned plot. Particular definitions for plant cover have been and continue to be used according to the interests and needs of the ecologist who plans and conducts the vegetation study. The specific Daubenmire method itself continues to be ignored by all but a few field investigators. Taking only the cover scale part of the Daubenmire method has resulted in a host of modi- fications to both the method and the frame. Changes have included range in cover class intervals, size of frames, shape of frames, and frame placements along transect lines. Cover classes range from the use of equal classes, such as those used by some of the federal land management agencies, to the use by field researchers of a small 0.1% increment class at lower cover values. Still others have used unequal classes at both the lower and upper cover values and retain equal cover class ranges in between. Some of these modifications essentially returned classes to the original cover-abundance scale of Braun-Blanquet. The modern use of class intervals have complicat- ed modifications made by Daubenmire to the origi- nal methods of the European ecologists. Changes in markings have been applied to various plot sizes and shapes. Circles, rectangles, and squares with varying area are used to obtain plant cover mea- surements, Still, these plot descriptions in the literature are referred to as a “modified Daubenmire plot.” It would be more appropriate to refer to one’s modifi- cation as “another modification of the Braun- Blanquet method using the Daubenmire, 20 cm x 50 cm, frame” if the exact frame size and shape is used. It is not appropriate to refer to use of different size and/or shape changes to the original “Daubenmire plot” as a “modified Daubenmire plot or method.” Changes in markings of a 20 em x 50 cm plot should be clearly stated as a modification of Daubenmire’s markings. Visual vs Direct Measures There is much disagreement in the literature over estimates” versus “measurements.” Recording cover obtained by visually guessing is referred to as ‘estimates” by some plant ecologists while others nsist that a “measurement” of cover is obtained from these “estimates” as plot size approaches such a small area that it becomes a point. In any case, nu- merical values representing cover contain errors which in turn lead to inaccurate (statistically) mea- surements. To reduce possible confusion, we sug- gest that all numerical values of vegetation cover be referred to as “measurements” in keeping with the definition of a measure. Observers may be deceived into believing that vi- sual measurements are not as “accurate” as those obtained from pins, singly or in groups, that are placed along a line. These pin-methods are stated by some ecologists to be a “direct” or “accurate” method to obtain an estimate of plant cover. Yet, er- rors occur because of pin diameter size, length of pin, area of a pin point, and so forth, Additionally, decisions must be made by the ob- server as to whether or not the point of the pin makes contact with any portion of a plant as the pin is lowered at a pre-selected angle through veg tion from a pre-selected position above the vegeta- tion. The use of a laser beam to detect a contact does not eliminate other sources of variation, such as effets of air currents, plant water content, shad- ing, and position of the equipment. Some field studies comparing various pieces of equipment (points vs. plots) have been contradicted by results of other studies using the same type of, equipment in a different vegetation type. Then a general problem seems to arise as to what is being measured and what is the subsequent use of this measure for plant cover, All types of equipmentFebruary 2004 19 have in common the measurement of plant cover of an area or volume of space occupied by plants Whether cover is obtained using plots, lines, point frames or single points, equipment affects the re- sulting cover measurement. Accuracy vs Precision It is difficult to ensure that no bias will enter, un- detected, into measurement of plant cover obtained from any equipment, method of measurement, or from any observer. Because this is true, then it is proper to refer to the precision of a measurement rather than its aecuraey Cochran (1977, p. 16) states: “Accuracy refers to the size of deviations from the true mean (mu), whereas precision refers to the size of deviations from the mean (m) obtained by repeated application of the sampling procedure.” Then accuracy is never known in field studies, Measurements can be very precise and not at all accurate because the process of measuring Currently, vegetation on the reclaimed area con- sists of grass or grass/shrub plant communities. Requirements of the New Mexico Energy, Minerals, and Natural Resources Department, Mining and Minerals Division state that amounts of total live plant cover and total cover ( includes plant cover, litter, and surface rocks) are sufficient for phase II consideration of partial bond release. Phase III re~ quires additional demonstrations that cover criteria must be met on that reclaimed area. We refer to our arrangement of individual Daubenmire method plots as the Extended Daubenmire Method which seemed appropriate be- cause it incorporates the marked plot used by Daubenmire. Five of these 20 em by 50 cm plots were placed in a contiguous series within a larger frame (50 em x 101.2 em) (Figure 1). The 1.2 em added to a meter length frame account for the cross bars used in the extended frame. This single large frame was constructed to secure the boundary defi- cover can be repeated by various ob- servers using the same cover class inter- Extended Danhenmira frame vals. Therefore, repeatable results are not m tem - equivalent to accuracy when obtaining cover measurements because the true av- erage remains unknown. 20% f+ 20 94) +20 %| + 20%] + 20%) = 100% Still many articles report results of cover sampling exercises wherein various equip- | 50 exa| ment, measurement methods, and sam- pling designs were compared in terms of precision and accuracy when in fact accu- a racy was not determined. On the other | —™ hand, reporting precision or relative preci- KH sion (relative to the average) may be use- A Area ofeach ful in comparative studies of cover meth- sub-plot }20 em ods. 5% A Field Trial North 3% ‘A frame was constructed to contain five 85cm 20% 50cm contiguous plots that Daubenmire de- 5% seribed for use in his method. The goal was to reduce most sources of error in vi- e 3% sual estimates of plant cover obtained Random fey from plots. A trial evaluation of the frame Point One sub:plot was conducted on reclaimed land of The Pittsburg &.Midway Coal Mining Company's York Canyon Mine Complex figure 1.5 near Raton, New Mexico. Schematic diagram of the extended Dawbenmire Method frame.RANGELANDS 26 (1) Table 1. Example of cover data by species and categories using the extended Daubenmire method. cosy ST 7 me on ma Fal oa u : o iS es a ee n 5 i as 5 : 7 : a ra = 5 ; i ; 2 Boat 5 0 : : = : : 5 5 2 rm feuipegee [o ; 3 i ; 7 nitions for each of the 5 contiguous, smaller plots. The dimensions and markings of the extended frame are in accordance with the S- Daubenmire method plots, each of which retains its individuality in data recording. Each 0.1 m* plot area, as originally marked by Daubenmire, is reduced by a factor of S and repre- sents 20% of the area of the extended frame (0.5 m). The 3% area of the Daubenmire method plot becomes 1% while the original 25% area of the Daubenmire method plot represents 5% of the ¢ tended Daubenmire method frame. Then each indi- vidual plot within the extended frame represents only 20% of the total measurement area (50 om x 100 cm), This arrangement of individual plots re- sults in the original 5% area becoming a 1% area within each original plot (Fig.1). This 1% area al- lows measurements to be more precisely made in 1% increments of an individual species’ cover, total live plant cover, and litter plus rock cover. The summation of cover measurements for plant species over all 5 plots at each sample point pro- les a cover estimate at each sample point. An ex- ample of a placement of one extended frame is given in Figure 2. An abbreviated example of data recording is presented in Table 1. Sample locations were obtained by randomly selecting 30 points from two-dimensional grid lines placed on a map of the reclaimed area. The extended Daubenmire method frame was placed approximately 1.18 m north of each grid lo- cation with the longest axis facing an East-West di- rection. Two steel pins were used to hold the ex- tended frame in place. Percentage cover was visual- ly measured within each Daubenmire plot to the nearest 1% for each plant species, litter + rock, and total live plant cover. Percentage bare ground was also independently measured for each plot and used as a check for measurements of total ground cover (100 - % total ground cover = % bare ground), Data were summarized to obtain averages for the 30 extended frames, Again, the extension of Daubenmire’s method was to obtain more precise visual measurements of cover by increasing the size of a 1% area relative to the original plot size used by Daubenmire. Evaluation The use of the extended Daubenmire method only included the measurement of ground cover as de- fined by the majority of field ecologists. We suggest that the users of the extended frame consider the following: * The extended method should lead to increased re- peatability for one or more observers and provide a more reliable measure of cover to the nearest 1% as required by some state and federal agencies involved in reclamation of surface mined-land, It was tempting to use fractions of less than 1% for cover measurements because of the large 1% area displayed within the individual plots of the extended frame. We recommend that all cover data be recorded to the nearest 1% to preserve the reason for using the extended Daubenmire method frame in the first place. Recall that a major con- cern in using the Daubenmire method plot to mea- sure plant cover is the lack of precision resulting from a single observer or among several observers working in the same vegetation type. Use of theFebruary 2004 a Figure 2. In the field illustration of the Daubenmire method plot arranged within the painted frame that designates area percentages. Each erass bar is removable for easy placement of the frame. nearest 1% measure provided by the extended frame, rather than fractions thereof, should in- crease precision. * Recognize that the data collected from the extend- ed frame should not be used to compare results with other plot sizes and shapes or with other cat- egories for cover. For example, to multiply the first plot in the extended frame by a factor of 5 to convert all plot data into original Daubenmire Method plot data for comparisons may result in biased estimates of cover measurements. Acknowledgments ‘We extend appreciation to The Pittsburg and Midway Coal Mining Company, York Canyon Mine Complex, Raton, New Mexico for financial support of data collection, analyses, and manuscript development. We are also indebted to Dr. C Wayne Cook for his critical review and to Dave Clark. Reclamation Specialist, New Mexico Mining and Minerals Division, for suggested revisions. We express our gratitude to the two anonymous reviewers for their helpful suggestions which improved readability of the paper. Lastly. thanks to Carin Corley and C'isa van Amburg for correcting and placing the photo and the figure of the extended Daubenmire frame into the proper formats for publication, References and Additional Reading Bailey, A. W. and C. E. Poulton, 1968. Plant communities ‘and environmental relationships in a portion of the Tillmook bum. Northwest Oregon, Ecology 49: 1-13 Bonham, C. D. 1989. Measurements for Terrestrial Vegetation, John Wiley and Sons, New York, NY. 338 p. Bureau of Land Management. 1996. Sampling Vegetation Attributes: Interagency Technical Reference. BLM National Applied Resource Sciences Center. BLM/RS/ST- 96/002+1730. Supersedes BLM Technical Reference 4400- 4, Trend Studies, May 1995. 163 p. Cain, S. A. and G. M. Castro, 1959, Manual of Vegetation Analysis. Harper & Brothers, New York, NY. 325p. Cochran, W, G, 1977. Sampling Techniques. John Wiley, New York, NY. 428p. Greig-Smith, P. 1983. Quantitative Plant Ecology. Univ. of Calif. Press. Berkeley. 359 p, Daubenmire, R.F. 1959. A canopy-cover method of vegeta- tional analysis. Northwest Science, 33:43-46. Elzinga, C. L. , D. W. Salzer, and J. W. Willoughby. 1998, Measuring and monitoring plant populations. BLM Technical Reference 1730-1, BLM/RS/ST-98/005+1730, Bureau of Land Management, Denver, CO. 477 p.Hanley, T. A. 1978 A comparison of line-interception and quadrat estimation methods of determining shrub canopy coverage. J. Range Mange. 31:60-62. Kiphart, Paul. 2001. Resource management demonstration at Russian Ridge Preserve. Grasslands 9: 1, 8-11, California Native Grass Assoc. Khan, Ch. M. Anwar. 1974. New adjustable, decimal, col- lapsible quadrat vs three old quadrats and evaluation. J Range Manage, 27:71-75. Jones, D. S. 2002. Vegetation measurement methods and ‘comparison htp:Wwww.cernml colostateedwimethods'methods him Mueller-Dombois, D. and H. Ellenberg. 1974, Aims and Methods of Vegetation Ecology. John Wiley and Sons, New York. 547 p. Peet, R. K., T. R. Wentworth, and Peter S. White. 1998. Method for recording vegetation composition and structure. Castanea 63:2 RANGELANDS 26 (1) Pollak, Oren and Tamara Kan, 1998. The use of prescribed fire to control invasive exotic weeds at Jepson Prairie Preserve. Pp 241-249, In: C, W. Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr., and R. Orduff (eds) Ecology, Conservation, and Management of Vernal Pool Ecosystems. 1996 Proc. California Native Plant Soe... Conifer. CA. Shimwell, D. W. 1971. The Deseription and Classification of ‘Vegetation. Univ. of Washington Press, Seattle, 322 p. Stohigren, T. J., Bull, K. A. and Otsuki, Y. 1998. Comparison of rangeland vegetation sampling techniques. J Range Manage. $1: 164-172, Sorrells, L. and S. Glenn. 1991, Review of sampling tech- niques used in studies of grassland plant communities. Proc. Okla. Acad. Sci, 71:43-45