Since 1989, quarantine plant pathologists from PNG and Australia have regularly jointly surveyed PNG’s border regions. These surveys (conducted in recent years by PNG’s National Agricultural Quarantine and Inspection Authority (NAQIA) and Australia’s Department of Agriculture, Fisheries and Forestry (DAFF)) have demonstrated freedom from some of the world’s most devastating banana diseases. Investigating wilt—like problems of banana has always been a high priority and this is done by destructive sampling of any plants showing yellowing and leaf death followed by examination of internal psudostem and other tissues for characteristic symptoms of fusarium or bacterial wilt. Many cooking banana plants (ABB genome) obviously yellowing or dying (Fig. 1), have been investigated in this way and the symptoms seen were clearly not those of fusarium or bacterial wilt. However, no other clear explanation for decline could be found in these plants. Inside these pseudostems, discontinuous streaking was present, appearing as small regions of brown or black vascular tissues, usually with wet, necrotic pockets (Figs. 2, 3). In 2008, a sample of such material tested PCR positive for phytoplasma in the laboratory of PNG’s Oil Palm Research Association (Carmel Pilotti OPRA, personal communication).

Fig. 1
figure 1

External symptoms (leaf yellowing and leaf death) associated with BWAP isolates RID5519 in East Sepik Province (left) and RID5837 in Western Province (centre), and with unidentified phytoplasma isolate RID5861 in North Solomons Province (right)

Full size image
Fig. 2
figure 2

Internal symptoms (discontinuous necrotic vascular streaks and pockets of rot and discolouration) associated with BWAP isolate RID5546 at Taman, Morobe Province, PNG

Full size image
Fig. 3
figure 3

Internal symptoms (discontinuous necrotic vascular streaks and adjacent necrosis) associated with unidentified phytoplasma isolate RID5545 at Taman, Morobe Province, PNG

Full size image

In response, NAQIA/DAFF conducted disease surveys of banana plants in Madang Province (MaP), Morobe Province (MoP), Milne Bay Province (MBP) and East Sepik Province (ESP) in 2009, and in Western Province (WP) and North Solomons Province (NSP) in 2010. Vascular samples were taken from the lower pseudostems of 16 yellowing and dying banana plants showing internal symptoms as described above (suspect phytoplasmas) and two associated suckers as well as from banana plants not showing such internal symptoms (controls). Some controls were symptomless and some were showing yellowing or leaf death. However, some problems were experienced with returning vascular material in good condition to Australia, especially from locations where refrigeration was not available. Only control DNA extracts from PNG that were known, from using 16S rDNA internal control primers, to be of excellent PCR integrity were included. Therefore, seven samples with no internal symptoms from PNG and five additional symptomless plants sourced from Australia served as controls. All details on test plants, tissue sample preparation and treatment, DNA extraction and PCR are provided in Table 1.

Table 1 Banana plants sampled in 2009 and 2010, and phytoplasma test results obtained
Full size table

All 16 symptomatic banana plants plus the two associated suckers tested phytoplasma positive (Table 1). The seven control samples, collected from PNG with no internal symptoms, plus the five symptomless Australian controls, tested negative (Table 1). This suggests that phytoplasmas are consistently association with banana wilt disease, rather than just being opportunistic phloem inhabitants, sometimes present in diseased plants.

The 16S rRNA gene, 16S-23S spacer region and a part of the 23S rRNA gene and the ribosomal protein (rp) S19 (rps19), ribosomal protein L22 (rpl22), and ribosomal protein S3 (rps3) genes of the phytoplasmas from samples RID5365 and RID5376 were amplified using P1/P7 and the rpL2F3/rp(I)R1A primer pair (Martini et al. 2007), respectively. Platinum® Taq DNA Polymerase (Invitrogen, USA) was used according to the manufacturer’s instructions except that the total reaction volume was 25 μl. Amplicons of expected size were obtained from both samples and were purified and cloned using the pGEM-T Easy Vector system according to the manufacturer’s protocol (Promega, USA). Transformants were screened and selected using standard protocols. Three clones of each gene region were sequenced using primers SP6 and T7, with an ABI PRISM® BIGDYE™ Terminator Cycle Sequencing Kit Version 3.1 (Applied Biosystems, USA). Sequences of 16S rRNA gene, 16S-23S spacer region and a part of the 23S rRNA gene and the rp gene region of both samples were identical and have been deposited in Genbank (JN872863 and JN872864 Banana wilt associated phytoplasma: BWAP).

Phylogenetic analyses of the 16S rRNA gene using ‘Cocos nucifera’ lethal yellowing phytoplasma1 (GQ227853) and 44 formally recognised phytoplasma species (Wei et al. 2007; http://plantpathology.ba.ars.usda.gov/pclass/pclass_phytoplasmaclassification_system2.html, Fig. 4) and the rp gene regions using 75 phytoplasma species or strains (Fig. 5) were conducted using MEGA4 (Tamura et al. 2007). These analyses showed that the phytoplasma from samples RID5365 and RID5376 clusters most closely with phytoplasmas associated with lethal yellows diseases of coconut in PNG and other countries but do form a distinct lineage from all other phytoplasma groups (Figs. 4 and 5). Sequence similarity matrices were generated for the 16S rRNA gene and the rp gene regions using BioEdit (Hall 1999). Based on the 16S rRNA gene the banana phytoplasma is identical to the ‘Cocos nucifera’ lethal yellowing phytoplasma1 (GQ227853) recently reported from coconut trees in PNG (Kelly et al. 2011). 16S rRNA gene sequence similarities ranged from 90–95.4 % and rp gene region similarities ranged from 70–75.2 % for the remaining phytoplasma species analysed, suggesting the BWAP is a unique species.

Fig. 4
figure 4

A phylogenetic tree representing the evolutionary relationships of 47 phytoplasma species based on the 16 S ribosomal protein gene using the Maximum Parsimony method. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches. *16Sr UD indicates that the 16Sr group of the phytoplasma has not been determined. The tree is drawn to scale, with branch lengths calculated using the average pathway method (Nei and Kumar 2000) and are in the units of the number of changes over the whole sequence

Full size image
Fig. 5
figure 5

A phylogenetic tree representing the evolutionary relationships of 70 phytoplasma species based on the ribosomal protein S19 (rps19), ribosomal protein L22 (rpl22), and ribosomal protein S3 (rps3) genes using the Maximum Parsimony method. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches. The tree is drawn to scale, with branch lengths calculated using the average pathway method (Nei and Kumar 2000) and are in the units of the number of changes over the whole sequence

Full size image

To the author’s knowledge, this may be the first description of a completely new wilt disease of banana. The only other records of phytoplasma detection in banana plants are two phytoplasmas in the 16SrI (Aster yellows group) found in association with disease symptoms similar to banana bunchy top disease. One is from China (Li et al. 1999) and one from India (Singh et al. 2009).

Almost all PNG phytoplasma records so far are from herbaceous dicotyledonous hosts (Davis and Ruabete 2010). An important exception being the recent find of a phytoplasma related to the 16SrIV (coconut lethal yellows) group associated with severe disease in coconut palms in MaP (Kelly et al. 2011). RID5365 and RID5376 isolates were obtained from active coconut phytoplasma disease outbreaks in MaP (Table  1 ), suggesting that banana should be investigated as a possible alternative host in PNG’s new coconut disease epidemic. Little is known of what species from other plant families may act as inoculum reservoirs in coconut phytoplasma disease epidemics. Recently, three herbaceous weeds were identified as alternative hosts of phytoplasmas in the 16SrIV group in Jamaica (Brown and McLaughlin 2011), and a fourth was found to be a host of a related coconut disease associated phytoplasma in Ghana (Danyo 2011). However, based on PCR detection, cloning and sequencing as described above, the BWAP was also found in banana plants from WP, MoP and ESP (Table 1, Fig. 6), where coconuts were abundant and showing no signs of phytoplasma—like disease. This apparent lack of transfer between host species might be explained by so far unidentified differences within the BWAP, or because of differences in the feeding behaviour of insect vectors present. Further work on the phytoplasmas involved is clearly needed. A 16SrII group phytoplasma was detected in one banana plant showing the disease symptoms described above in MaP and a 16SrVIII group phytoplasma was detected in another banana plant not showing discontinuous vascular streaking in MoP (Table 1, Fig. 6). The latter could be a slightly different disease relationship, or alternatively a chance ‘infection’, which should be expected to occasionally occur in perennial crops like banana. This is the first record of a 16SrVIII phytoplasma from PNG.

Fig. 6
figure 6

Map of PNG, indicating the approximate location of banana plants from which the following phytoplasmas were obtained. BWAP, in a location where active coconut disease epidemics were in progress (black four pointed star); BWAP, in a location where no coconut disease problems were apparent (black star); as yet unidentified phytoplasmas (black up-pointing triangle); and phytoplasmas most closely related to phytoplasmas in the 16SrII (black diamond suit) and 16SrVIII (black circle) groups

Full size image

Further investigations into the phytoplasma disease status of monocotyledonous crops and weeds, plus studies to determine what are the insect vectors in these areas of PNG are essential. This information would underpin sustainable disease management strategies for banana and possibly also coconut growers of PNG.