8/16/2019 TQ-40(1)-9

1/10

fEPIDEMIOLOGY OF TEA BLISTER BLIGHT

{EXOBASIDIUM VEXANS)

•A. Kerr & R. L. de Silva

We may be biased, but we believe that Blister Blight of tea is the ideal diseasefor epidemiological studies. M any of the advantages are inherent in the host andin the way it is cultivated. Tea is an evergreen perennial cro p main tained in thevegetative state and pruned once every four years or so, and because of this, thereis no marked change in susceptibility from season to season as occurs with mostcrops. Only young leaves are susceptible and tea is pruned and trained in such away that all the young leaves are confined to the top of the bush where they areexposed to the atmosphere and there is no complication of micro-environment.Tea is plucked every week and consequently, disease incidence can be readily assessedevery week. So with Blister Blight we can have 52 assessments per year, wherea swith many other diseases we have to be content with one. As temp erature is relatively uniform throughout the year, it can be almost ignored and in fact, only twomajor factors appear to be important in the epidemiology of the disease—the numberof spores land ing on susceptible leaves and th e dur atio n of leaf wetness. Beforedealing with the epidemiology, however, we should like to introduce you to the disease and describe briefly its method of spread and control.

T H E D I S E A S E

Figu re 1 illustrates the disease. Blisters are appro xim ately one centim etre indiameter, generally convex on the under surface of a leaf and concave on the upper,although occasionally the reverse occurs. Spores are born e on the convex surface and spore trapping in infected tea, reveals a marked diurnal periodicity with amax imum catch aroun d midnight (Figure 2). The reason for this was investigatedby considering the two mo st likely factors, tem perature and hum idity. Fo r theseexperiments, cut shoots were used, because it was found that there was no differencein the pattern of spore production between cut shoots and intact shoots (Figure 3).W hen the relative humidity aroun d infected leaves is kept at abo ut 100 and tem peratu re varies, spores are liberated continuously, and when tem perature is kept constant, but humidity varies, spore liberation is directly correlated with high humidity(Figure 4). This applies whether or not infected sho ots are removed from bushes.Diu rnal spore liberation is, therefore, almost entirely due to diurnal fluctuations ofrelative humidity.

D A M A G E

Before studying the epidemiology of a disease it is necessary to determine whatdamage it causes and what is the critical level above which appreciable crop lossresults. Th e incidence of Blister Blight is norm ally assessed by determ ining p ercentage infection on the third leaf and 3 5 infection has long been accepted as thecritical level. W hen disease incidence is below 3 5 mo st blisters occur on leaves,but as infection rises above this level, an increasing number of blisters occur on leaf

+ This paper wa s presented at a symp osium on The epidemio logy of leaf diseases in thehum id tropics at the First International Congress of Plant Patholo gy held in Lon don inJuly 1968.

* Present address : Wa ite Agricultural Institute, Gle n Osm on d, Sou th Australia.

8/16/2019 TQ-40(1)-9

2/10

loo

9 0

8 0

8 7 0

VI *

o 6 0

uX IEzc

5 0

•S 4 0

3 0

2 0

1 0

13 15 17 (9

M I D N I G H T

I21 2 3 1 3 5 7 9

Time (Hours)

F I G U R E 2 -Mean diurnal periodicity of spores ofE. vexans present in the atmosphaiiring a monsoon

II

ere

_ Intact shoo t s - ield conditionsCut shoots - ield conditionsCut shoots - Laboratory conditions

18 20 22 • 0 2

Time (Hours)F I G U R E 3 — ean diurnal periodicity of spore production by E . vexans from blisters

pnmti on cut a nd intact slwots in the field an d in the laboratory ZVtkesaZ

10

8/16/2019 TQ-40(1)-9

3/10

*•••..'

R e l a t i v e h u m i d i t yT e m p e r a t u r e

/ . y \ 'A

V\ -\

0 H o u r d a y12 H o u r d a y [ 06 1 8 hours]2 4 H o u r d a y

2 5

2 0 H3

151

TO-r>

5

0

8

7 x»

6 I#

5o

4 5

8/16/2019 TQ-40(1)-9

4/10

8/16/2019 TQ-40(1)-9

5/10

petioles (Figure 5) which are completely girdled, leading to leaf fall and crop loss.So the purpose of any control measure is to prevent disease incidence rising above3 5 on the third leaf Even very low rates of fungicide can often achieve this,giving a marke d increase in cro p (Figure 6). It was further found tha t disease control was directly proportional to the log. of the dose of fungicide used (Figure 7).

20i>

OII

Metal l ic copper per acre (oz)

F I G U R E 6 —Effect of the dose of meta llic coppe r on tea crops as a resu lt of the contro lof Blister Blight

Disease is controlled by weekly sprays during monso on periods. Thirty sprayrounds per year are not uncommon, but theoretically, as few as four rounds maybe adequ ate in some years. By accurate disease prediction and forecasting, considerable saving in the cost' of spraying should be possible .

I00r

Fungicide dosage (oz per acre)

F I G U R E 7 —The relation between the dosage of fungicide applied and the resulting degree of protection against Blister Blight

13

8/16/2019 TQ-40(1)-9

6/10

700

600au

a 5 0 0 -

§ 4 0 0 -La

o°- 300

200

IO 0

p 0 h i n c i l ( wt tt * i u

• • ^ t IP 0411 J1

"JSJi 1 3uT 1 M A 1 S E P I NOV.. • ' D EC,

30t n

IB O1

16 0M

14 • D

O112 3O

6 XNO

4 ^ :o

2

F I G U R E 8 —The relation between the number of spores in the atmosphere and the number of blisters per unit area of tea

S P O R E N U M B E R S

In studying the epidemiology of Blister Blight, spore numbers were first considered and it was assumed that they would be simply and directly related to thenu mb er of blisters per unit area of crop . M uch to our astonishm ent, this was notthe case (Figu re 8). Yo u w ill notice that a s disease incidence is increasing, spore

20001-

APR. 1 MAV- X JUN, JUL AUG. 1 S E P. O C T. . . N O V. OEC.

F I G U R E 9 —The estimated and measured num ber of spores per cu bic metrejof air from9 6 2 t o X964

14

8/16/2019 TQ-40(1)-9

7/10

numbers are higher than expected and vice versa. Th e reason for this is obscure,but it has to be allowed for when attempting to estimate spore numbers in the atmosphere , as will be seen later. W e do not want to go into the mathe ma tical relationships toda y. It would suffice to say that spore num bers can be accurately estimated(Figure 9). These estimates are for spore num bers in the atmo sphere bu t, as might

be expected, they are directly correlated with numbers deposited on susceptible leaves.

L E A F W E T N E S S

The relationship between disease incidence and leaf wetness is relatively simple. As dur atio n of leaf wetness increases, so does disease incidence and th is canbe expressed mathem atically. Du ratio n of leaf wetness, however, is very stronglynegatively correlated with duration of sunshine, and as the latter is more convenientfor superintendents of tea estates to measure, we have used sunshine records in allour calculations to predict disease incidence.

PREDICTION OF DISEASE INCIDENCE AND DISEASE FORECASTING

By com bining our estimates of spore numbers with sunshine record s, we canpred ict disease incidence three weeks later fairly accurately. Un fortu nate ly, diseaseforecasts issued from the Tea Research Institute would serve little purpose, becausethe tea country in Ceylon is mountainous with very variable climate over relativelysho rt distance s. An alterna tive is for disease incidence to be forecast on individualestates. Altho ugh the data required for this could be obtained o n tea estates, severalmeasurements are required and the calculations involved are rather too tedious forsuperintendents to adop t. These have been considerably simplified.

In ou r initial calculations, the num ber of blisters pe r 100 shoo ts w as used toestimate spore num bers, bu t the relationship is nonlinear. If, however, the square

root of percentage infection is plotted against spore numbers, the relationship is

infection (third leaf)

F I G U R E 10— he relation between percentage infection of Blister Blight assessed on thethird leaf and on two leaves and a bud

15

8/16/2019 TQ-40(1)-9

8/10

practically linear (Figure 10). In other words the square roo t of percentage infection can give an estimate of spore numbers, but the accuracy of the estimatecan be markedly increased if the rate of increase of estimated spore numbers is alsomea sured. Th e most convenient way of expressing this is by the part ial regressionequation

Y = a + x + b x (1)

where Y = log. estimated spore num bers

jfx = log. V ( infection) t t

x2 = log. V ( infection) f2 - log. V ( infection f,)

/ a — *i = three weeks and

a , bx and b 2 constants.

The only measurement required to estimate spore numbers is percentage infection. No rm ally this is assessed on the third leaf bu t on comm ercial tea estatesonly the youngest two leaves and a bud are plucked for manufacture and it wouldbe mu ch m ore conv enient for plant ers, if percentage infection could be assessed onleaf broug ht to the factory. The relationship between the two method s of assessment is shown in Figure 11. Using this graph, data obtained by either method canbe readily interconverted.

Square root percentage infect ion

Glisters pe r 100 shoots

F I G U R E 11—T he relation between the num ber of spores in the atmosphere, the number of

blisters per IO O shoots an d the square root of percentage infection

16

8/16/2019 TQ-40(1)-9

9/10

It only remains to incorporate sunshine records into the partial regression anddisease incidence, three weeks later, can be predicted. This can be represented by

Y = a + bjXi + * 2 * 8 - M s (2)

where Y = predicte d disease incidence at time t2 + 3 weeks,

X and x 2 , are as in equation (1) and

x 8 = me an daily sunshine.

Values have been calculated for all the constants (a, b t ba and ba and bothpredicted and m easured incidence are shown in Figure 12.

8/16/2019 TQ-40(1)-9

10/10

To avoid square root and logari thmic transformations and ari thmetic calculations, a simple calculating device has been constructed and is illustrated in Figure13. Th e calculating device represents equa tion (2).

F I G U R E 1 3 —Calculating device for disease forecasting

W e believe a simple practical me thod of disease forecasting can be developedand applied by superinten dents on a ny tea estate in C eylon where Blister Blight isan im po rta nt disease. Th e m etho d is still being tested in the field, bu t we believethat it can be safely applied to mature tea, one year or more from pruning and shouldsave many rounds of spraying each year.

18