ASSESSING FLORISTIC CHANGE FROM THE BOTANICAL LITERATURE. A WORKED EXAMPLE : CHANGE IN FABACEAE (PEA FAMILY) 1867 - 2000 (PRELIMINARY RESULTS)

J J Day

Edwin Lees, in Botany of Worcestershire 1867, presents his systematic list in the form of a table divided into four Botanical Divisions: The Table of Plants. For each species, an assessment of abundance is given for each division. He gives 13 categories of abundance/distribution which are applied to all species. In addition, three relating to habitat and four relating to status are noted when appropriate. Interpretation of the 13 categories relating to abundance/distribution present a number of difficulties. These relate mainly to the two parameters of geographical distribution and abundance being insufficiently defined and combined or confused by Lees. Furthermore, their application is not always consistent. Despite this the 13 categories seemed too close to the ten frequency classes defined in the Checklist of the Worcestershire Flora (Day, J.J. 2000) to be ignored. Having come to an understanding of Lees' categories after entering the whole of Lees' Table of Plants (about 3500 records) onto the database the Lees categories were aligned to the ten point scale in the Checklist.

Frequency class

Number of monads

Percentage of monads

Score

       

Very common

1501+

>75-100%

1

Common

1001-1500

>50-75%

2

Widespread

501-1000

>25-50%

3

Widespread but localised

201-500

>10-25%

4

Uncommon

101-200

>5-10%

5

Scarce

51-100

>2.5-5%

6

Very scarce

26-50

>1.25-2.5%

7

Rare

13-25

>0.6-1.25%

8

Very Rare

1-12

0.05-0.6%

9

Extinct

0

0%

10

Above: Table of Checklist Frequency Classes (Day, J.J. 2000)

Leesí Classes

Assigned to checklist class

   

General

1

Plentiful

2

Many places

3

Abundant

3

Not uncommon

4

Not common

5

Various places

5

Several places

6

Rather uncommon

7

Occasional

7

Very uncommon

8

Numbered sites (1-4 localities)

9

Extinct

10

Above: Table of Leesí classes and assignment to Checklist class

The Pea Family, Fabaceae were selected to test the methodology .

The assigned Checklist values were applied to the Lees categories. This for each species and within each district. The values were summed and divided by four to give a county average. Where a species was not present in a district in was assigned a rarity value of 10.

The Lees county averages were compared to the Checklist frequency class and the species ranked according to degree of change.

These are first draft results and the following notes should be taken into account before drawing conclusions.

they are not a direct comparison, the modern assessment is based on geographical distribution as a measure of frequency, Lees includes local and relative abundance.
the modern survey has covered the ground more systematically and thoroughly (see distribution map elsewhere of all records to 1875); it will be likely to overestimate uncommon species in relation to Lees.
2001 assessment is strictly numerical, Lees is an estimation from memory, his method is not good at ranking fairly common and widespread species
Lees' evaluations may need minor refinements
Lees has a tendency to inconsistency and error
Lees assessments change as he works through list (but not consistently)
Lees values have not been adjusted for the different areas of Botanical Districts, this is important as the 2001 classes are based on area.
Lees missed species (the Vicia orobus record is a herbarium specimen of Lees)
Lees did not know either the Lickey Division or the upper Teme valley well; there are under assessments here
loss within last 30 years not taken into account

The results are presented below. They indicate a number of trends but should not be taken as a template for the flora as a whole. It is not a random sample and is highly unlikely to be representative of the entire flora. This is a single family with biological unity, especially in regard to physiology and breeding strategies.

Those species showing no apparent change or declines are similar to expected. This gives credence to the methodology as an agent for measuring change. There is a surprise, however, with the large group of species exhibiting apparent increases. This family, as a unit appears to have done well out of the land use changes of the past 150 years. The results seem reliable with regard to the spread of non-native species. However, there are some features which suggest that a further refinement of the method may be needed. Some anomalies may be due to Lees' inconsistencies. For instance, the surge in Medicago lupulina was entirely unexpected and will need verification against other sources. There is also a group of uncommon natives which indicate a low level of increase. I suspect this may be an artefact of recent survey effort and they are species in decline, having been under-recorded in the past.

In summary, the overall trends and patterns appear to be real, in the majority of cases, but the levels of change should not be taken as absolute values.

CHANGE 1867 - 2000

In the list that follows the First number = frequency based on Lees 1867 and the Second number = frequency based on Day 2001. X = not recorded in Lees Table of Plants

Species with No Apparent Change

tend to be :
very common species
very rare species (where suitable habitat has survived)
Genista anglica 9 9
Genista pilosa 10 10
Hippocrepis comosa 9 9
Lathyrus aphaca 9 9
Lathyrus pratensis 1 1
Lotus corniculatus 1 1
Ononis spinosa 7 7
Securigera varia 9 9
Trifolium dubium 1 1
Trifolium incarnatum 9 9
Trifolium ochroleucon 9 9
Trifolium pratense 1 1
Trifolium repens 1 1
Vicia bithynica 9 9
Vicia cracca 2 2
Vicia lathyroides 9 9
Vicia parviflora 9 9
Vicia tetrasperma 3 3

Species with Apparent Declines

tend to be:
scarce natives confined by habitat
light demanding perennials (grassland / heath & saum communities)
One order decline
Vicia sepium 1 2
Ulex europaeus 2 3
Anthyllis vulneraria 7 8
Vicia sylvatica 7 8


Two order decline
Onobrychis viciifolia 7 9
Trifolium campestre 1 3
Ulex gallii 3 5


Three order decline
Lathyrus linifolius 1 4
Genista tinctoria 1 5

Extinctions
Lathyrus palustris 9 10 Extinction
Trifolium scabrum X(9) 10 Extinction
Vicia orobus X(9) 10 Extinction

Species with Apparent Increases
tend to be:
non-native taxa
species associated with specific grassland management regimes
plants of tall grassland (climbers)
plants of tight grazed grassland (impact of motor mowing)
annuals
colonists, plants of disturbed ground and arable habitats
One order increase
Vicia sativa 3 2
Cytisus scoparius 4 3
Ononis repens 5 4
Trifolium fragiferum 6 5
Ornithopus perpusillus 7 6
Trifolium striatum 7 6
Astragalus glycyphyllos 8 7 [but recent decline]
Lotus glaber 8 7


Two order increase
Vicia hirsuta 4 2
Trifolium medium 5 3
Melilotus altissimus 6 4
Medicago sativa subsp. sativa 7 5
Trifolium arvense 7 5
Trifolium micranthum 7 5
Lathyrus sylvestris 9 7
Medicago polymorpha 9 7
Medicago minima X 9
Medicago sativa subsp.falcata X 9
Melilotus indicus X 9
Melilotus sulcatus X 9
Trifolium ornithopodioides X 9
Trifolium subterraneum X 9
Trifolium aureum X 9
Trifolium resupinatum X 9
Vicia lutea X 9


Three order increase
Lotus pendunculatus 6 3
Lathyrus nissolia 7 4
Lathyrus grandiflorus X 8


Four order increase
Medicago lupulina 5 1
Melilotus officinalis 9 5


Five order increase
Lathyrus latifolius 10 5
Medicago arabica 9 4
Melilotus albus X 6
Vicia faba X 6


Eight order increase
Trifolium hybridum X 3

References:

DAY, J.J., Checklist of the Worcestershire Flora, 2001.
LEES, E., Botany of Worcestershire, 1867.

WBRC Home

Worcs Record Listing by Issue

Worcs Record Listing by Subject