Worcestershire Record No. 18 April 2005 pp. 32-43

Terry Knight

Most County Floras have distribution maps of plant species. These generally show a grid of symbols showing where the species has been recorded, blanks indicating the absence of records. For the many common species this does not provide much useful information as the map will be more or less covered with the same symbol. The map for one common species will also be much the same as for any other.

During the sixteen years of recent field recording in the south-east of the county an attempt was made to gather more information on common species. The simple method described below was tried to see if it would give any useful results.

The recording was done by 1x1 kilometre squares of the National Grid (monads) and by "parish". The results from the monads were later aggregated into 2x2 kilometre squares (tetrads) for analysis. Trips were made between mid-March and mid-November each year into each monad and "parish". These were spaced out as evenly as possible by week and location to get a representative coverage. For each trip, a list of species was made in the order in which they were seen.

The area selected was governed by the requirements of the ongoing Worcestershire Flora Project so that the records could be used in the production of the soon to be published new County Flora (Reid 2003). All present Worcestershire plus the bits of the core area of the county lost to surrounding administrations since the end of the nineteenth century are included and this has been christened "Greater Worcestershire". For this survey the area is the whole of "Greater Worcestershire" in 10 kilometre National Grid squares SP03, SP13, SP14 plus that part of it south and east of the rivers Avon and Isbourne in SP04.

As historical recording was done by core county it means some old records in the survey area were included in with Gloucestershire or Warwickshire ones. The Flora Project required that records from areas in former counties were to be kept separate for comparative purposes. Consequently in this survey it was decided for simplicity to keep separate records in all parishes, or part parishes as appropriate, within the survey area. The parishes, or part parishes, are

Data were entered onto a computer database. From this was extracted the records for those trips in which fifty or more species had been recorded. (The figure of fifty was chosen to give a reasonable number of trips containing a reasonable number of species.) For each trip a score was assigned to the species as follows: first seen 100, second 98 and so on to the fiftieth which scored 2. Any further species were ignored. A "Density Index" was then calculated for a species by summing its scores in the tetrad or "Parish" and dividing by the number of trips to that tetrad or "Parish". Index values will thus lie between 0 and 100. In order for a species to have an Index of 100 it would need to be the first seen in every trip. Conversely, to have an Index of zero it would need to be never seen in the first fifty in any trip.

For most common herbs it was found from the records that they could be readily identified throughout the recording season (Weeks 12 to 46 inclusive). However, most common grasses and a few common herbs could only easily be identified in limited periods of the recording season. For these, only records from trips made during that period are appropriate. Only records from the full season and one limited period (Weeks 22 to 46 inclusive) are included in the analysis.

In addition to the "Density Index" the frequency of each species was calculated for comparative purposes. This was done by finding the percentage of the 833 or 636 trips in which a species was recorded in the first fifty species seen.

Table 1 lists the calculated Index by species and "Parish". Tables 2A and 2B lists it by species and tetrad. The tetrad number is the Grid Reference to its centre point (eg 137 is SP010370 and 1747 is SP170470). Table 2B also gives the percentage frequency for each species. Distribution maps for a selection of species are given showing the variations in calculated Index by tetrad in five ranges, 1 to 20 up to 81 to 100. Tetrads with small areas in the recording area and/or having a very limited number of trips are not included. These are 935, 1333, 1337, 1141, 1549 and 1747.

41650 (833 x 50) records were used in the analysis. A sample of 800 records was taken to carry out some rudimentary statistical analysis. This indicated that the greater the number of trips and the greater the value of the Index the more accurate the results would be. Figure 1 gives a rough estimate of the accuracy for different Index values and number of trips.

Figure 1 gives a rough estimate of the accuracy for different Index values and number of trips

Top line  -  70% probability to be within +/- 20%
 Middle line  -  95% probability to be within +/- 20%
 Bottom line-  95% probability to be within +/- 10%

"Density Index" and Frequency

For the 41 species given in the Tables a plot of the percentage Frequency against the "Density Index" is given in Figure 2. The points all lie very close to a curved line through the origin suggesting the species with either low or high Index values have a relatively lower frequency. (A very similar result was found by Walker, Preston and Pearman (2004) from their survey of the Island of Rhum using an average DAFOR abundance estimate instead of a "Density Index".) The curve shown in Figure 2 given by the equation

F = 2.54 x D - D x D/40.5 + D x D x D/9390

is a good fit through these points.

The above analysis was done using the first fifty species seen in each trip. A supplementary analysis was also done using the first one hundred species from 200 trips. For this the first seen was given the score of 100, the second 99, down to the one hundredth which was given 1. As would be expected, for each species the frequency was increased and also the "Density Index". For example the frequency and Index for Common Nettle increased from 96.0 and 76 to 99.0 and 85 respectively, Creeping Cinquefoil from 56.7 and 31 to 86.0 and 53, and Herb-robert from 30.1 and 13 to 63.0 and 27. However the graph of Frequency against "Density Index" altered little, a good fit curve being given by the equation

F = 2.82 x D - D x D/36 + D x D x D/10300.

Recording once a week through one recording season produces data for 35 trips. In the recording area this would be expected to give reasonably accurate Index values for about 35 of the most common species. The amount of effort required to carry out this exercise on its own probably is not justified for the amount of information gathered but as part of a larger recording programme the small additional effort is worth considering. If climate changes significantly as predicted, surveys of comparative plant abundance will become more important to assess the effects on our flora. Consideration should therefore be given to starting a programme of quantitative abundance surveys to give a baseline for the future. The method described provides one simple objective way in which this might be done.

The survey suggests that South-east Worcestershire has an impoverished flora typical of many lowland areas of England. Many of the plants having the higher index values are considered “weedy” species. Four of the five Injurious Weeds (Creeping Thistle, Spear Thistle, Broad-leaved Dock, and Curled Dock) feature in the list of most common species having average index values of 56, 29, 27, and 12 respectively. However, the other Injurious Weed (Common Ragwort) is much rarer. No truly woodland or wetland plants were found to be common. This is not surprising as the recording area is dominated by arable with most of the other land being intensively grazed, built-up or horticultural.

REID B, 2003 Worcestershire Flora Project Progress Report - Worcestershire Record 14; 25.

WALKER K, PRESTON C, and PEARMAN D, 2004 The DAFOR Scale and tetrad recording on the Island of Rhum, Inner Hebrides. BSBI News 97; 12.