Two species of Arceuthobium are known from Pakistan; A. minutissimum and A. oxycedri. Arceuthobium minutissimum is regarded as having the smallest aerial shoots of any of the dwarf mistletoes (5-10 mm). It is known from the Himalayan Region of Bhutan, India, Nepal and Pakistan. Primary host is Pinus wallichiana (Hawksworth & Wiens 1996). In Pakistan, A. oxycedri is presently known only from a single location; two adjoining valleys in the Province of Balochistan (Beg 1973, Zakaullah 1977, Ciesla 1993).

4.1.  DISCOVERY

The first "official" report of the occurrence of A. oxycedri in Pakistan is by Beg (1973) who detected the parasite in juniper forests of the Sasnamana Valley, near the community of Ziarat, as part of a forest disease survey. At the time of discovery, 50% tree mortality was attributed to dwarf mistletoe infection. A second report (Jamal & Beg 1974) suggests that the parasite had been present in the area for some time but had been overlooked because the aerial shoots tend to mimic the foliage of the host plant. Discussions that the author had with Mr. Muhammed Yousuf Khan, who was the District Forest Officer in Ziarat in 1993, indicate that local people were aware of the occurrence of dwarf mistletoe in the area at least as early as 1920 and routinely collected the aerial shoots to feed their livestock.They were not aware, however, of the damaging nature of this plant (Ciesla 1993). Reference to the practice of using shoots of A. oxycedri as fodder for goats and sheep in this area is also mentioned by Zakhaulla and Badshah (1977) who expressed concern that this practice might spread the parasite.

In 1977, a survey by the Pakistan Forest Research Institute in Peshawar detected infection by A. oxycedri in the juniper forests of the Chasnak Valley which lies immediately to the north of the Sasnamana Valley (Zakhaullah 1977).

There is no evidence to indicate that this dwarf mistletoe infection is the result of an accidental introduction.

4.2.  HOST TREE

Apparently, the identity of the host tree in Balochistan is less clear than the identity of the parasite. The species of Juniperus found in the region has been alternately referred to as J. excelsa, J. macropoda and J. polycarpos. The first reports of the discovery of A. oxycedri in Balochistan refer to the host tree as J. polycarpos (Beg 1973, Jamal & Beg 1974). Baloch (1974, 1977) studied insects associated with A. oxycedri in the Ziarat area and initially referred to the host tree as J. macropoda and later as J. excelsa (J. macropoda). Reports from Pakistan, prepared after 1979, refer to this juniper as J. excelsa (Chaudry & Wali-ur-Rehamen 1979, Sheikh 1985, Ciesla 1993).

Collections of A. oxycedri from juniper in the Himalayan Region of neighboring India are equally confusing. Duthie (1885) refers to the host tree as J. excelsa while a report by Brandis (1907) refers to it as J. macropoda. Later reports, Rau (1975) and Bhattacharyya & Uniyal (1982) refer to Juniperus spp. and J. polycarpos respectively.

Vidakovic (1991) describes J. excelsa as a monoecious tree, up to 20 m in height with fruits containing 4-6 seeds. He gives the range of this species as Asia Minor to the Caucasus, Crimea and the eastern and southern portions of the Balkan Peninsula. He describes J. macropoda as a shrub or tree up 10-12 m high with foliage similar to J. excelsa and having 2-4 seeds per fruit. He includes a photo of a stand of J. macropoda from Quetta, Pakistan. This photo is undoubtedly taken in the high mountains somewhere in northern Balochistan. The elevation of Quetta is too low and the climate is too arid to support natural forests of Juniperus. Range of J. macropoda is given as Iran, Afghanistan and Pakistan.

Rushforth (1987) describes J. excelsa as monoecious or dioecious, up to 20 cm in height and containing from 4-6 seeds per fruit. The natural range of this species is given as the Balkan Peninsula across Turkey to southern Russia, Iran and Lebanon. He recognizes J. macropoda as being closely related to J. excelsa and is sometimes treated as a variety (J. excelsa var. polycarpos). He indicates that the two species can be separated by the presence of light, open foliage with spreading, sharp pointed needles to 0.8 cm on the lower foliage of J. macropoda and with dark green scale leaves in the upper crown being very closely pressed to the stem. The fruits of J. macropoda are said to bear 2-5 seeds per fruit. The natural range of this species is given as Iran through Afghanistan to Uttar Pradesh, India and also parts of Turkey and southern Russia.

Brandis (1907) refers to the juniper population in Balochistan as J. macropoda and suggests J. excelsa as a synonym of J. macropoda.

Sheikh (1985) indicated that the population of Juniperus in Balochistan has both male and female trees. Field observations by the author in 1993 confirmed that the population of Juniperus in the vicinity of Ziarat is clearly dioecious with female trees bearing fruits which consistently bear 4 seeds per fruit.

These papers suggest that J. polycarpos is an early synonym of both J. excelsa and J. macropoda and that the two taxa are either two separate, closely related species or a single, variable species. For purposes of this report, the designation J. excelsa is used to refer to the population of Juniperus in the Ziarat area.

4.3.  BIOLOGY

Information on the local biology and mechanisms of dispersal of A. oxycedri in Balochistan is sketchy. Zakaullah and Badshah (1977) and Zakaullah (1988) report that the primary means of the spread of infection is from tree to tree although they also suggest that birds and mammals are agents of dispersal. They observed a bird, referred to as a "mistle thrush" to be common during the period of seed maturation and dispersal. This bird feeds on juniper seeds and might carry dwarf mistletoe seeds. A similar report exists for France where a bird identified as Turdus musicus (possible synonym - T. viscivorus, the mistle thrush [Flegg and Hosking 1990]), was reported to be associated with dwarf mistletoe infected junipers (Gercer and Cotte 1908). Zakaullah and Badshah (1977) also suggest that field rats and shepherds, collecting mistletoe plants for livestock fodder, especially during the period from mid-August to November could introduce the parasite to new locations. Considering the relatively localized distribution of A. oxycedri in Balochistan, it is unlikely, however, that birds, small mammals or humans are important vectors of this parasite.

Hawksworth and Wiens (1976) report that flowering of A. oxycedri is from September to October and that fruits typically mature between October and November. Kiu (1984a) reports that A. oxycedri flowers between August and September in Xizang Province, China but does not give a period of seed maturation and dispersal. During the evaluation made in 1993 (Ciesla 1993), local foresters reported that in Balochistan, seeds of A. oxycedri matured in September. There was no evidence of seeds or flowers when field observations were made in October 1993 and the dwarf mistletoe plants appeared as though all seeds had dispersed.

4.4.  LOCATION AND INTENSITY OF INFECTION

The only known area of dwarf mistletoe infection in Pakistan is a single, more or less contiguous area which includes the Sasnamana and Chasnak Valleys near Ziarat. As of 1993, this infected area encompassed ca 4,000 ha and is equivalent to less than 5% of the total area of juniper forest in the Ziarat and Loralai Forest Districts (Ciesla 1993)

When A. oxycedri was first detected in the Sasnamana Valley in 1973, 50% host tree mortality was attributed to the parasite (Beg 1973). A later survey of the infected area indicated that 36% of the trees were infected and 13% had died. Average infection rating, based on the 6-class dwarf mistletoe rating system (Hawksworth 1977) was 4.6 (Zakaullah and Badash 1977). Additional analysis indicated that incidence of infection decreased with increasing elevation although infection rating remained relatively constant, the highest incidence of infection was on SW facing slopes and the larger trees, both in terms of diameter and height had the highest incidence and intensity of infection (Zakaulla 1987). Another study established that the highest incidence of infection and highest dwarf mistletoe ratings were on gentle slopes (< 10%) (Table 9) (Zakaullah 1979).

When dwarf mistletoe infections were detected in the adjoining Chasnak Valley, a survey indicated that the highest incidence of infection (35.2%) and the highest infection ratings (4.5) occurred on the ridge separating the two valleys (Table 10).

An evaluation of the dwarf mistletoe infection in the Sasnamana and Chasnak Valleys conducted in 1993 by FAO (Ciesla 1993) indicated that the area of heaviest infection was the north facing slope of the Chasnak Valley. Many trees had large, conspicuous witches brooms

and profuse growth of aerial shoots (Fig 17-18). There was a wide range in the age of infections with some being very young and barely detectable and others up to 20 to 25 years old. Heavily infected trees had thin crowns and were in a state of decline In the most heavily infected trees, the only host foliage remaining was in witches brooms. Some trees had died from dwarf mistletoe infection. It was difficult to determine when death occurred because the rate of decay is slow under the arid conditions typical of the area and snags remain standing for many years after death. Several standing snags, with the remains of witches

Figure 17 - A. oxycedri, large plant, Sasnamana Valley, Balochistan, Pakistan.	Figure 18 - Juniper with heavy dwarf mistletoe Infection, Chasnak Valley, Balochistan, Pakistan.
Figure 19 - Open juniper forest near Ziarat, Balochistan, Pakistan.	Figure 20 - Old juniper near Mt. Khalifat, Balochistan, Pakistan.

4.7.  PAST CONTROL EFFORTS

A dwarf mistletoe control operation was begun in 1978 with the objective of eradicating the parasite. This project was funded for a 5 year period and included cutting of heavily infected trees and pruning infected branches from the less severely infected trees (Sheikh 1985). Work was done between mid-April and mid-August. During the final two years, the number of trees treated decreased (Table 12) but a reconnaisance survey conducted during 1985-86 indicated that latent infections remained in the area and additional follow-up work was needed (Zakaullah 1988). The project ended in 1983 primarily because of lack of funds. In addition, there was concern that cutting of infected trees would have adverse effects on watershed and other forest values.

TABLE 12

DWARF MISTLETOE CONTROL EFFORTS - BALOCHISTAN
1978-83

YEAR	NUMBER OF TREES FELLED	NUMBER OF TREES PRUNED	TOTAL TREES TREATED
1978-79 1979-80 1980-81 1981-82 1982-83	1000 2194 5092 1530 1075	4000 1592 5061 2335 850	5000 3786 10153 3865 1925
Total	10891	13838	24729