Control of the Invasive Black Rat (Rattus rattus) and Pacific Rat (Rattus exulans) Using a Large Scale Trapping Grid for Endangered Tree Snail and Plant Conservation in Hawaii Stephen M. Mosher,1,2 Darren Peters,3 Lindsay Wilson4, Jobriath L. Rohrer1,2, and Aaron Shiels5

1Pacific Cooperative Studies Unit, Botany Department, University of Hawaii at Manoa, Honolulu HI 96822 USA 2U.S. Army Garrison, Environmental Division, Directorate of Public Works, Schofield Barracks, HI 96857 USA 3Department of Conservation, National Office Predator Control, Wellington NZ 4Department of Conservation, Te Anau NZ. 5Botany Department, University of Hawaii at Manoa, Honolulu HI 96822 USA

INTRODUCTION Introduced rats (Rattus sp.) in Hawaii are known predators of birds, tree snails, and plants. Since 1997, the Oahu Army Natural Resources Program has been controlling rats through the use of diphacinone rodenticide in bait stations and snap traps on a relatively small scale at multiple sites for the protection of endangered plants and animals. In May 2009, as an alternative method of rodent control that did not require toxicants, a large scale trapping grid was initiated over a 26-ha forested management unit with 440 snap traps on the island of Oahu. New Zealand Department of Conservation current Achatinella mustelina best practice rat trap technology was utilized for the first time in Hawaii with this trapping effort.

Goals of this study: 1) to reduce rat numbers to a level that significantly benefits endangered and other native species 2) to monitor endangered resources and forest health 3) to determine whether rats may be suppressing invasive slugs and predatory snails (Euglandina rosea) Diospyros sp. Cyanea superba subsp. superba

METHODS .Study Sites: Rodent control was conducted at the Kahanahaiki Management Unit in the northern Waianae Mts. (Figure 1). Resources were monitored at Kahanahaiki, and at an adjacent site in the Pahole Natural Area Reserve where there was no rat control (Table 1). .Rodent Control & Monitoring: . 440 Victor snap traps in wooden boxes (Traps: 234 perimeter (12.5m) & 206 interior (25m). . Traps initially checked daily for 2 weeks then every 2 weeks thereafter. . Traps baited with peanut butter or FeraFeed with a half of a Macadamia nut. . Rat activity was measured monthly with 38 tracking tunnels divided among 3 zones at Kahanahaiki and quarterly with 30 tracking tunnels along two drainages at Pahole (Figure 1). .Cyanea superba subsp. superba Fruit Predation Monitoring: . Fruit predation monitored approximately every 2 days, December through January. . Achatinella mustelina Monitoring: . Annual census of live snails & quarterly ground shell plot monitoring for predated shells. . Seed Rain Bucket Monitoring: (48 randomly placed & 10 under native Diospyros sp.) . Buckets checked every 2 weeks.

. Seedling Plot Monitoring: (Kahanahaiki 80 plots & Pahole 65 plots) Figure 1. Kahanahaiki Management Unit . Plots read at 6 month intervals Table 1. Management Actions Kahanahaiki Pahole .Arthropod Composition & Abundance Monitoring: (16 pitfall traps & 32 vegetation beating samples) Rat Control Yes No . Sampling every 6 months. Rat Tracking Tunnel Monitoring Yes Yes . Slug & Euglandina rosea Monitoring: (Slug: 40 pitfall traps & Euglandina: 10 timed search plots) Oahu Tree Snail (Achatinella mustelina) Monitoring Yes No . Sampling every 3 months. Cyanea superba subsp. superba Fruit Predation Monitoring Yes Yes Seed Rain Bucket Monitoring Yes No Seedling Plot Monitoring Yes Yes Euglandina rosea Monitoring Yes Yes Slug Monitoring Yes Yes

RESULTS

Rodent Control & Monitoring: Cyanea superba subsp. superba Fruit Predation Slug & Euglandina rosea Monitoring . 562 rats caught (May 2009 - January 2010). Monitoring . Slug monitoring suggests that large scale rat control .Rat activity in tracking tunnels peaked in late fall at . 30+ plants monitored per site. does not increase slugs (Figure 6). Kahanahaiki (Figure 2). . Fruit predation was statistically significant. More fruit . Timed search survey data for Euglandina showed no .There was no difference in tunnel activity during the were predated at Pahole (no rat control) (Figure 4). significant difference between sites (Figure 7). two sampling periods for Kahanahaiki and Pahole

(Figure 3). χ2 = 97.786, p = 0.000

Figure 4. Fruit predation at Kahanahaiki and Pahole. Figure 6. Pitfall trap data for slugs.

Figure 2. Kahanahaiki rat captures & tracking tunnel activity May 2009 – January 2010 Seed Rain Bucket Monitoring P>0.05,(P > 0.05)T-Test . Seed predation of introduced Psidium cattleianum (Strawberry guava) generally reflected rat activity (Figure 2, 5). No observed predation on native Diospyros sp. (Lama) (Figure 5).

Invasive Psidium cattleianum (Strawberry Guava) Native Diosyros sp. (Lama) 100 100 A 26 B 80 80

43 60 60 2 98 40 132 40 Figure 7. Timed ground search survey data for Euglandia rosea

151 79 20 62 20 110 245 144 63 64 497 1003 492 2 1 5 3 15 5 24 6 10 1 2 0 0 2 0 0 Arthropod Composition & Abundance Percentage of rodent-chewed seeds rodent-chewed of Percentage Figure 3. Rat activity in tracking tunnels during two quarters (August seeds rodent-chewed of Percentage & December) at Kahanahaiki and Pahole. Monitoring 1-Jul-09 6-Oct-09 1-Jul-09 6-Oct-09 14-May-0928-May-0910-Jun-0917-Jun-09 14-Jul-0921-Aug-0931-Aug-0922-Sep-0920-Oct-093-Nov-0917-Nov-091-Dec-0915-Dec-0912-Jan-10 10-Jun-0917-Jun-09 14-Jul-0921-Aug-0931-Aug-0922-Sep-0920-Oct-093-Nov-0917-Nov-091-Dec-0915-Dec-0912-Jan-10 . Arthropod samples are in the processing stage with 61 Sampling date Sampling date Achatinella mustelina Monitoring Figure 5. Seed bucket results from Kahanahaiki for (A) Strawberry Guava (48 buckets) and families in 17 orders identified at present (May-June). . ~11 ha were surveyed with 212 snails (B) Lama (10 buckets). Total number of seeds (chewed and unchewed) above each point. counted in August 2009. Seedling Plot Monitoring . Ground Shell Plots revealed no predated . 313 seedlings were marked in 80 plots at Kahanahaiki and 140 seedlings marked in 65 plots at shells. Pahole. Plots will be read for the second time in February-March 2010.

CONCLUSIONS and FUTURE DIRECTIONS  The rat control grid appeared to be effective in reducing the amount of predation on Cyanea superba subsp. superba fruits at Kahanahaiki (Year 1). Rats were frequently captured, and tunnels tracked, in interior locations of the trapping grid when distances from edge to interior were less than 100m. Tunnels appeared to track the natural cycle of rat activity outside of the grid. A more frequent trap check interval may be needed.  Over time, data collection of annual tree snail counts, seedling plots, arthropods and Cyanea superba subsp. superba fruit predation will give us a better understanding of what rat activity thresholds must be met to maintain and increase rare and common native species. Data collected on slug and Euglandina rosea numbers suggests that rat removal does not cause increases in these highly invasive species.  Bait consumption by invasive slugs poses a hurdle that still needs to be overcome. Alternative baits are currently being pursued (wax baits and scented lures).

ACKNOWLEDGMENTS We would like to thank the many members of the Army Natural Resources Staff that have help with checking trap boxes, as well as, Stephanie Joe for slug and Euglandia rosea data collection, Richard Pender and Lalasia Bialic-Murphy for Cyanea superba subsp. superba data collection, Vincent Costello for Achatinella mustelina data collection, and Paul Krushelnycky for arthropod data collection. We thank Talbert Takahama for permission to collect data in the Pahole Natural Area Reserve.