- Home
- Construction of pictographs of Karoo vegetation from quantitative survey data

Last update: August 18, 2011 11:11:07 AM |

**CONSTRUCTION OF PICTOGRAPHS OF KAROO VEGETATION FROM QUANTITATIVE SURVEY DATA**

P W Roux & F A Roux

Grootfontein College of Agriculture

**PREAMBLE**

A practical method of constructing pictographs for visual comparison of quantitative vegetational survey data is described. The pictograph consists of an oblique block diagram, 24 x 6cm, representing a field ground area of 6 x I,5m. A pictograph is constructed by using canopy spread cover data, average diameter and radius of species and the calculation of the number of plants per vegetation component. Suitable symbols are used to represent the main components of the vegetation, for example grass, palatable shrubs, unpalatable shrubs, pioneer species, herbaceous annuals and ephemerals. Additional symbols may be designed to accommodate extra components or individual species.

**1. Introduction**

Quantitative vegetation survey data are generally set out in the form of tables, which are probably the best and most comprehensive way of presenting such data. It is, however, at times required to present survey data in a more illustrative and digestible form such as in graphs, pie graphs, histograms and pictograms. Such figures are especially useful in educational demonstrations, agricultural extension work, and visual comparisons and publications. In all these depictions, however, one is not readily able to visualize comparisons where more than two surveys are concerned and which contain several components or species.

The pictograph described here, constructed from quantitative survey data, provides an excellent alternative to other possible graphic representations. Furthermore, a pictograph, which is a strict stylization of the vegetation and is statistically accountable, provides near accurate information on the number of plants per unit area. By comparing before and after pictographs, change in cover of the components, if any, can be readily seen. A lesser drawback is, however, that the species occurring at low percentages, less than approximately 0,5 %, cannot be accommodated successfully in the pictograph and recourse to the survey table becomes necessary. A minor disadvantage is also that symbols of plants can overlap on the pictograph. Figure 1 is a good illustration of the use of a pictograph in a publication.

**2. Calculations and Construction **

**2.1 The Survey**

A canopy spread cover survey, S-strikes, as defined by Roux (1963) is required. For this purpose the descending-point method, as devised by Roux (1963), is recommended. The chain-point method, as devised by Roux, Tidmarsh & Havenga (1955) can also be successfully employed. Five hundred or more sampling points per survey are usually of sufficient accuracy for constructing the pictograph. For rougher estimates fewer sampling points may be employed. For the construction of the pictograph, shown in Fig. 2, 1 100 samples were employed (see Table 1).

**2.2 Component categories**

The results of a survey should be tabulated by grouping the S-strikes into vegetation component category totals. For Karoo vegetation these components may be Perennial grass, Palatable shrubs, Unpalatable shrubs, Pioneers, and Opslag (annuals, ephemerals, herbs, weeds, etc.). For the grouping of species into component categories the classification of Blom (1980) may be used. If desired, individual dominant species may form a component on their own (as in Fig. 2) or different groupings may be made to suit a particular objective. In Table 1 the survey is classified into seven components; the two species, Eragrostis lehmanniana and Eriocephalus ericoides are in this case regarded as separate components on account of their dominance.

**2.3 Diameters of species**

The mean diameter, d, of species' canopies in a component must be determined. Measurements (cm) are made of the dominant species in each component i.e. those species contributing to the far greater percentage of a component (see Table 1). Two diameter measurements (d1 + d2) are made at right angles, usually in north-south and east-west directions, through the centre of a plant. The measurements are within the limits of canopy spread cover. Diameter measurements can be made simultaneously with the survey, or later at random. Usually fifty measurements per species per component are required. It is suggested that fiduciillirnits (Snedecor 1946) be calculated for the number of species to be measured if it is required to obtain a predetermined accuracy. In this case it was found that as few as 30 measurements for some species may suffice, and as many as 90 or more for other species to satisfy the p = 0,05 level of accuracy. _Where fifty plants are measured, the average diameter (d) per species is :

đ = Σd1 + Σd2

100 cm

where d1 and d2 are the sums of the north-south and east-west measurements respectively.

**2.4 Radius of species (f)**

The average radius of a species is calculated by:

ř = đ ¸ 2cm. In the calculation of the average radius of a component, weighted means may be employed instead of the straight average. Weighted means are where the percentage to which a species contributes to the radius measurements is used. Weighted means were used in this present exercise, ego Hertia pallens, Lycium sp. and Pteronia sordida, i.e. for component 5 :

đ = (6/22 x 42,4cm) ¸ (9/22 x 72,8cm) +(7/22 x 50,4cm) = 57,4cm.

Thus: ř = 57,4 ¸ 2 = 28,7cm (see Table 2)

**2.5 Number of plants**

The number of plants per component or specific species is calculated by:

Cn = |
Sn_{1 } |
X Cm^{2 } |

N | ||

Pr^{-2 } |

Where:

Cn = The number of plants to be depicted of the particular component e.g. C1 (Eragrostis lehmanniana or C_{4} Palatable shrubs)

Sn_{1} = Total S-strikes for the particular component or species.

N = Total number of sampling points (e.g. 1 100 as in Table 1)

Cm^{2} = Area of ground to be depicted viz. 6 x 1,5m (i.e. 9m^{2} or 90 000 cm^{2})

ř = Mean radius of average plant in component.

P = 22 ¸ 7 (i.e. 3,143)

For C_{1} (Eragrostis lehmanniana) this is:

(156 ¸ 1 100) x 90 000 cm^{2 } |
The number of plants = 219.6 (rnd 220) |

3.143 x (4.3 x 4.3)cm |

Calculate number of plants for all components accordingly (See Table 2).

**2.6 Scale reduction factor, F**

In order to depict an average plant per component, on the block diagram (24 x 6cm) the diameter of a plant must be reduced by factor F.

F= | 1 |

L |

Where:

1 = length of long side of block diagram (24cm)

L = length of long side of field area depicted (6 cm or 600cm)

For Fig. 2 this is

24 cm |
= 0.04 |

600 cm |

There are 220 individuals in Component 1, as calculated, which have an average diameter, đ, of 8,5cm (see Table 2). The reduction of an individual symbol to be represented on the figure is d x 0,04. This is 3,4mm (0,04 x 8,5cm) for Component 1.

The results of all calculations should be tabulated as shown in Table 2.

**2.7 Constructing the pictograph **

Suitable symbols are used to represent the different components (See Fig. 2). The base lengths of symbols are adjusted according to the calculated base length in mm (e.g. 3,4 mm for Component 1). From the data in the last two columns in Table 2 the pictograph can be finally constructed on the block diagram. In Figure 1 the angles of the block diagram are 35° and 125° to give it a pleasing perspective. In Figure 2 the angles are 60° and 120°. In this case, however, a larger area for depicting plants is available. The angles of figures can be suitably varied according to the dictates of the survey and number of plants, as long as the scale corresponds to the field area depicted.

For E. lehmanniana there are 220 individual "plants" to be entered on the diagram each with a base length of 3,4mm. Repeat for the rest of the components. In the drawings of the pictograph, the "plants" should be distributed more or less randomly over the surface of the block diagram, starting with the component with "plants" which have the longest base length. In the entering process the plant base lengths are arranged parallel to the base of the diagram. The bases can be entered accurately on the diagram by using a draughtsman's deviders set at the appropriate length ego 12,5mm for perennial grass. It is by no means essential that the lengths should be exact. In the key to the symbols (see Figure 2) the exact measurements can be indicated as obtained from base lengths (mm) in Table 2. By counting the number of plants per component on the figure, the number of S-strikes for the component and the S-percentage can be calculated. It may happen that the calculated number of plants is less than 1,0 as shown e.g. in Table 2 for Unpalatable shrubs (Component 5) which is 0,7 with base length of 23mm. The "plant" can be depicted on the edge of the diagram with an overlap of 6,9mm falling outside the effective area of the figure as shown in Figure 2. Alternately decimals for number of plants may be rounded off to whole numbers. Colour coded symbols for components greatly enchance the visual effect of the pictograph. To facilitate visual comparisons of before and after surveys, it is suggested that the positions of similar symbols, on the different pictographs, be matched as far as possible, e.g. where 20 and 12 similar symbols are to be drawn on two different pictographs the positions of 12 symbols should match. Figure 1 gives an example of the pairing of symbols. For publication purposes the pictographs can be suitably reduced in size. The field areas (e.g.6 x 1,5m) to be depicted can be changed as desired to suit a particular purpose, or when plants are too numerous or too large. The size of the pictographs can also be appropriately adjusted. The principles as set out above should, however, be strictly adhered to at all times.

**2.8 Calculating %s and S-strikes**

Count the number of "plants" per component on Figure 2 (e:8. Eragrostis lehmanniana = 220). Divide the base length, in this case 3,4mm, with the conversion factor (F = 0,04).

i.e. 3,4 ¸ 0,04 = 85mm = 8,5cm (This is equal to the mean diameter, d see Table 2).

By applying the following formula, S% can be determined.

S% = | Cn x P (r)^{2 } |
x | 100 |

A cm^{2 } |
1 |

Cn = 220 (number of plants in the particular comnent)

r = (see Table 2)

A = Area of ground depicted (6 x 1,5m or 90 000cm )

P = 3,143

e.g.

220 x P x (4.3 cm)^{2 } |
x | 100 | x | 1 278 500cm^{2 } |

90 000cm_{2 } |
1 | 90 000cm^{2 } |

Number of S-strikes = 14,2% of 1 100 samples = 156

**REFERENCES**

BLOM, C.D., 1980. Groepsklassifikasie van karooplante. Roneo (ongepub.). Direkteur: Karoostreek, Middelburg K.P.

ROUX, P.W., 1963. The descending-point method of vegetation survey. A point sampling method for the measurement of semi-open grasslands and Karoo vegetation in South Africa. S. Afr. J. Agric. Sci. 6,273 - 288.

ROUX, P.W., 1973. The Angora goat in successful veld management. Angora Goat and Mohair J. 15 (2),5-13.

SNEDECOR, G.W., 1946. Statistical methods. The Iowa State College Press. pp 485.

TIDMARSH, C.E.M. & HAYENGA, C.M., 1955. The wheelpoint method of survey and measurements of semi-open grasslands and Karoo vegetation in South Africa. Bot. Surv. S. Afr. Memoir No. 29, Gov. Printer, Pretoria.

**Published**

Karoo Agric 4 (2) 19-24