A Biocontrol Plan for the Cook Islands

Gerald McCormack and Quentin Paynter looking at peltate morning glory on the edge of a coffee plantation field.

A heightened awareness of the seriousness of weed problems in the Pacific region, following the Pacific Biocontrol Strategy Workshop that we hosted in 2009, has led Sarah Dodd and Quentin Paynter to scope out the opportunities for boosting weed biocontrol in the Cook Islands.

The Cook Islands comprise 15 islands, the largest being the well-known holiday destination, Rarotonga. Rarotonga is known for its lush tropical vegetation that contains many plant species indigenous to the region.

Sadly, much of the Cook Islands’ natural habitats and agricultural land are being threatened by invasive weeds, many of which were brought to the island for their ornamental value, as edible fruit, or as a timber species. Like here in New Zealand, there are now more introduced plant species in the Cook Islands than indigenous species. Not only are the weeds threatening native biodiversity and many traditional cultural practices, but also the sustainable development of the island group. In particular, a suite of woody vines (Cardiospermum grandiflorum, Merremia peltata, Mikania micrantha and Passiflora rubra) are smothering trees, causing massive deforestation and replacing the native forest with impenetrable vine thickets. There are concerns that this may have a devastating impact on natural watershed systems and consequently on the economy and quality of life of the islands. Unfortunately increasing tourism and global trade are likely to result in even more unwanted introductions down the track.

With funds provided by New Zealand’s Ministry of Foreign Affairs and Trade (MFAT), Sarah and Quentin have recently scoped out a 5-year plan for developing weed biocontrol in the Cook Islands. They travelled to the Cook Islands to meet with regional experts involved in agriculture (forestry, horticulture, livestock), biodiversity conservation and biosecurity. At a workshop these experts provided key information needed to identify the best and worst biocontrol targets using a prioritisation framework we developed initially for Australia.

Fifty-two weeds were discussed at the workshop. Nine were rejected from further consideration because they had some desirable properties (e.g. edible fruit or cultural importance) so there would likely be opposition to any biocontrol attempts. A further two species, the giant sensitive plant (Mimosa diplotricha) and lantana (Lantana camara), were also removed from the list because they are already well controlled by biocontrol agents that have been introduced to the Cook Islands. “Successful biocontrol projects like these really help to pave the way for other biocontrol efforts,” noted Sarah.

The panel of experts then ranked the remaining 41 plants as either “hot”, “warm” or “cold” depending on how important they thought it was to control the plant. This was used to develop a weighted weed importance score that was combined with a biocontrol feasibility score and offset against the effort that it would take to complete the programme, to come up with an overall score. “Plants with existing biocontrol programmes in other countries are obviously cheaper targets because much of developmental work has already been done,” explained Quent.

The 15 weeds with the highest total scores and which are therefore the top biocontrol targets are listed in the table. Quentin notes, “This scoring scheme still allows some species of lower importance, such as broom weed (Sida rhombifolia), to rank quite highly.” This highlights the need to trade off the minor benefits of targeting weeds that are of relatively low importance, but have a high chance of successful control, with targeting weeds that are of high importance (potentially major benefits) but a lower chance of successful control. This was a cause for much discussion and the panel concluded that (assuming costs were comparable between plant species) they preferred to work on plants that were of higher importance but had a lower feasibility of successful biocontrol.

The top 15 weeds from a potential of 41 weed species selected from the Cook Islands as candidates for biocontrol (where: F =Feasibility; E = Effort; I = Weed Importance). The weeds included in the 5-year plan are highlighted. Note the total score is I+ F – E.

Rank	Weed species	Common name	F	E	I	Total Score
1	Arundo donax	Giant reed	93	13	100.00	180.00
2	Xanthium pungens	Cockleburr	85	7	95.83	173.83
3	Mikania micrantha	Mile a minute	50	12	100.00	138.00
4	Cardiospermum grandiflorum	Grand balloon vine	50	14	100.00	136.00
5	Passiflora rubra	Red passionfruit	50	19	100.00	131.00
6	Spathodea campanulata	African tulip tree	50	21	100.00	129.00
7	Psidium cattleianum	Strawberry guava	50	12	85.00	123.00
8	Sida rhombifolia	Broom weed	99	8	30.00	121.00
9	Clerodendrum chinense	Honolulu rose	80	29	52.50	103.50
10	Ludwigia octovalvis	Willow primrose	77	25	46.67	98.67
11	Sorghum bicolor subsp. drummondii	Sudan grass	50	28	66.67	88.67
12	Senna obtusifolia	Sicklepod	15	17	87.50	85.50
13	Phyllostachys bissetii	Bisset’s bamboo	50	33	68.33	85.33
14	Hedychium coronarium	White ginger	50	25	60.00	85.00
15	Merremia peltata	Peltate morning glory	15	32	100.00	83.00

For this reason peltate morning glory (Merremia peltata), initially ranked only 15th, has made the final eight species chosen for inclusion in the 5-year plan. This vine is a highly problematic weed that can climb over and smother trees up to 20 m tall, but there is uncertainty regarding its status in Rarotonga and the wider Pacific region. Peltate morning glory was present in Rarotonga when the first European botanists began documenting the flora but may have been an early introduction rather than a native species. We have recommended that a study of the genetics of peltate morning glory should be undertaken to determine, if possible, how and when this plant colonised the Pacific region. Only if this study can provide convincing evidence that the plant is a recent introduction to the Cook Islands should biocontrol proceed.

A number of high importance weeds are excellent prospective biocontrol targets in the Cook Islands. These include two vines: grand balloon vine (Cardiospermum grandiflorum) and mile-a-minute (Mikania micrantha); a woody shrub, strawberry guava (Psidium cattleianum); a grass, giant reed (Arundo donax); and an annual herb, cockleburr (Xanthium pungens). Biocontrol agents are already available for all these weeds that could be released at relatively little cost.

Red passionfruit (Passiflora rubra) could be a more difficult biocontrol target due to the risk of a non-target attack affecting closely related Passiflora species that are cultivated for their edible fruit. However, the edible species of Passifl ora grown in the Cook Islands all belong to the subgenus Passiflora, while P. rubra belongs to the subgenus Decaloba. Attractive Heliconius butterflies are a well-studied group and there is strong evidence that some species only utilise plants that belong to the subgenus Decaloba. Many species of Heliconius are readily available from suppliers for butterfly houses. So it seems likely that biocontrol of red passionfruit could proceed fairly rapidly and cheaply, without the need for extensive native-range surveys, although some host-range testing would be required.

African tulip tree (Spathodea campanulata) is a major invasive weed throughout the Pacific Region, for which biocontrol is predicted to have a medium feasibility. Preliminary surveys for biocontrol agents have been conducted in Ghana where this tree is native, funded by the Secretariat of the Pacific Community. Considerable investment would be required before any agents could be released in the Cook Islands but we recommend that this work should be tackled anyway.

We do not recommend starting work on any medium importance weeds within the next five years, with the possible exception of tree marigold (Tithonia diversifolia). This is currently a biocontrol target in South Africa and it may become a higher priority if researchers demonstrate that the agents currently held in containment in South Africa are adequately specific for release in the Cook Islands. Two other “medium importance” weeds (Clerodendrum chinense, Ludwigia octovalvis) are predicted to be highly feasible targets for biocontrol, but native-range surveys and host-range testing would have to be done, so they are both relatively “high effort” targets.

We also do not recommend starting work on any of the low importance weeds within the next five years, with the possible exception of broom weed (Sida rhombifolia). A biocontrol agent for this species (Calligrapha pantherina) could be collected in Australia at the same time as agents for cockleburr (Xanthium pungens) at very little extra cost, and biocontrol would almost certainly succeed.

As part of this scoping project Dr Maja Poeschko, of the Ministry for Agriculture in Rarotonga, travelled to New Zealand and spent time with our staff learning more about weed biocontrol. Maja has previously worked mostly on biocontrol for insect pests and is keen to be involved more on developing solutions for weeds. So everything is now in place to implement an exciting new weed biocontrol programme for the Cook Islands, just as soon as the necessary funding can be found, and that will be our next task.

Many thanks to MFAT for providing the funds for this scoping project and to all those who attended the workshop and provided valuable information, especially Gerald McCormack and Maja Poeschko.