Small Treasures: Rare Frogs of the New Zealand Rainforests
—Heather Moulton
Fig. 1 - Lifting a streamside rock revealed this adult Hochstetter’s frog.
Frogs are those little green guys who leap between lily pads in ponds, who croak “Bud-weis-er,” and who are kissed by young girls in pursuit of a prince, right? This is what I believed until I was fortunate enough to study the rare Hochstetter’s frogs (Leiopelma hochstetteri) of New Zealand (Fig. 1).
From May 31 to July 15, 2005, Janelle Bosse (another UNH student) and I conducted research in the Hunua Catchment Parkland, which is about 42,000 acres of native forest near the city of Auckland on the North Island of New Zealand. The specific area where we studied frogs was the 1,500-acre Kokako Management Area, which receives intensive predator control to help conserve the endangered Kokako birds (Fig. 2)
Fig. 2 - A typical stream in the native forest of the Kokako Management Area.
These little Hochstetter’s frogs are important to study and monitor for several reasons. Primarily, their unique characteristics contribute to our global diversity, which is declining at an alarming rate. Amphibians have permeable skin, which leaves them extremely vulnerable to pollutants and disease-causing organisms (1). Thus, a change in amphibian populations indicates a change in the surrounding environment.
New Zealand, in particular, is an excellent place to study ecological conservation: Its extended isolation from other continents has created countless numbers of species and ecosystems found nowhere else in the world (2). It is, therefore, crucial that these special creatures and unique places be protected from the negative impacts of humans and the mammals they bring with them.
The Hochstetter’s frogs are only found in New Zealand; they are currently endangered and highly valued for conservation purposes (3). They have existed since before the dinosaurs and have retained unique and ancient characteristics, quite different from our more modern species (4). These secretive frogs grow to a maximum of about 45 mm (less than the length of your pinky), and they do not croak or go through a tadpole stage (5). They can be various colors but are mostly shades of brown and green, to fit in with their surroundings. They live in wet habitats alongside streams, where they tend to shelter beneath rocks and logs in cool, shaded regions of the native rainforests (6). These habitats, however, are being fragmented and otherwise modified by human-induced disturbances and the effects of introduced mammals, including rats, stoats (animals much like our weasels), and possums. As a result, effective monitoring techniques are needed to conserve and manage this rare species.
Monitoring Programs
Monitoring programs are particularly important for obtaining information regarding the status of rare species, including where and how they spend their time and how abundant they are. It can be difficult, however, to acquire this information, for rare species occur either at low densities across a broad range or are locally abundant but not widely distributed (7).
I monitored the Hochstetter’s frogs using a “site occupancy” method, which is relatively new and has never been conducted on this species before. This method refers more to the presence and absence of a species at a site than to how many individuals are found. Site occupancy is a desirable way to examine a rare species for many reasons. It is non-invasive, whereas other techniques, such as toe-clipping used for capture-recapture, may be harmful to the animal. Site occupancy can also include a variety of habitats and multiple sites over a large area (8). Simply counting animals is not a very accurate way to estimate the abundance and status of a species because individuals may be present at a location but remain undetected (7). Therefore, patterns of detection and non-detection over multiple visits to the same site are vital to obtain non-biased results (9). The results can then be used to estimate the number of sites (or percentage) of a certain area that a species occupies. An amphibian ecologist from the Waikato Department of Conservation, Dr. Michael Crossland, introduced me to this methodology and aided me throughout the research and analysis.
I had many objectives that I wanted to accomplish when I began this exciting endeavor. I aimed to (A) estimate the detection probabilities and the site occupancy of Hochstetter’s frogs in the Kokako Management Area of the Hunua Ranges; (B) separate the ages of the frogs on the basis of size; (C) assess the current status of the frogs in the predator-controlled management area; and (D) conduct the first comprehensive site-occupancy monitoring of Hochstetter’s frogs. Through these objectives, I hoped to increase the known information regarding the distribution and abundance of Hochstetter’s frogs in the Hunuas and to facilitate future monitoring, habitat restoration and conservation of this important species.
A Typical Day
Janelle and I would get up around 6:30 a.m., make our lunches of PB & J, bundle up with layers of long underwear, put on wet boots and dirty coats, and pack our data sheets and equipment. We then drove about ten minutes up a narrow, winding logging road to the main trail. From there, we would hike anywhere from forty-five minutes to three and a half hours, depending on where we were surveying that day. The hikes were quite difficult, going up and down steep ridges, maneuvering through thick vines and flowing streams. What an adventure!
We searched three to four sites per day, depending on travel times and weather. At each site we conducted three searches, separated by a minimum of twenty minutes to a maximum of three days.
Most of the streams in the Kokako Management Area are divided into 120 sections, each 100 meters long. We randomly chose 55 of these sections without prior knowledge of whether frogs lived there. If the section was deemed searchable, meaning that it was not dangerous, dry, or with dense vegetation, we set up a random 20-meter transect (linear sampling area) within the section.
Fig. 3 - I am searching for Hochstetter’s frogs in the dense streamside vegetation.
We recorded the environmental factors of each 20-meter transect and carefully turned all cover objects (usually rocks) on both sides of the stream, using headlamps to look for frogs (Fig. 3). We were extremely cautious to walk either in the stream or five meters from the edge so we would not hurt the frogs or disturb their habitat. When a frog was found, we would (besides rejoice) record the micro-habitat features of the surrounding area. We measured the size of each frog and then classified it into an age group: <18 mm was considered a juvenile, 18–29 mm a sub-adult, and >30 mm an adult. If one frog of each age class was found prior to the 10 meter mark on the transect, we could stop at 10 meters; otherwise we would continue until either all three size/age classes were found or we reached 20 meters.
After all our searches were complete for the day, we would hike back through the woods (hopefully before dark) and drive to our home that was secluded from all civilization. The SkyTower of Auckland, however, could be seen vaguely in the distance and was a great complement to the gorgeous sunsets. We would lazily make ourselves dinner, enter some data, and crash either on the couches or in our beds. The rewards of such research and long days in a beautiful place were incredible.
Fig. 4 - A pencil displays the size of a well-camouflaged sub-adult frog
Outcomes
We found a total of 566 frogs, which is an impressive number of individuals for a rare species. Out of 165 searches (fifty-five sites searched three times), there were only thirty searches during which we did not detect any frogs. This means that we found at least one frog during 82 percent of our searches! This is astonishing, but what is even more astonishing is that there were only three sites out of the 55 total where we did not find frogs during any of the three searches. This indicates that 95 percent of the sites we searched were occupied by frogs! Frogs may have been absent from sites due to moisture gradients, preferred stream type, temperature, or the effects of introduced mammals. Some frogs may have been present but not detected due to our inexperience as searchers; the limitations of large rocks, crevices, and vegetation; or their size and cryptic coloring (Fig. 4).
Fig. 5 - My finger is pointing to a tiny green juvenile frog
The distribution of sizes/ages was quite even between the three classes; this distribution indicates a relatively healthy population (Graph 1). There were often all three size/age classes in the same site, resulting in 42 percent of our searches ending before the 20-meter mark. Some of the frogs we found moved before we could accurately measure them, but we could usually estimate their size and age. Juveniles tended to remain motionless, whereas many adults hopped away when disturbed (Fig. 5).
Another interesting aspect was that when Janelle and I searched only 20 minutes apart, we found that some of our data was so close that it could have been a repeat sighting of the same frog (based on approximate size and location). To our surprise, however, only 10 percent of the total frogs found were potential repeats, indicating that there were probably many more frogs present than we found. Of the 10 percent possible repeats, there was a fairly even distribution of sizes/ages. This could be due to the fact that, although juveniles were much smaller, they did not move, whereas adults were larger and easier to detect but often hopped away after the first searcher passed by. Consequently, there seemed to be similar opportunities to have a repeat sighting of any size/age frog. The differences in searching techniques between Janelle and myself accounted for a large observer bias. She turned more rocks and found more juveniles, while I took longer and found more sub-adults and adults. We made a greatteam!
Graph 1 - Hochstetter’s Frog Size/Age Distribution
The results will be analyzed with the PRESENCE statistical program, designed by Darryl MacKenzie. The program will allow me to adjust my naïve site occupancy estimates by accounting for different covariates, such as temperature; the number and size of frogs found during searches one, two, or three; the date; the assumption that we became better searchers as time went on; the time of day during which the search was conducted; and which person was searching.
Site occupancy is clearly an adequate monitoring technique for this rare species. It is statistically adequate, allows researchers to study differences in size/age classes, and adjusts raw counts by accounting for biases. Also, the short transects, non-invasive monitoring, and precautions of observers minimize the damage to frogs and the surrounding habitat. The one disadvantage of this method is that it is resource intensive. A great deal of time and money would be required to continue monitoring. Despite this limitation, however, I strongly encourage the continuation of monitoring, perhaps with the assistance of other students and volunteers. The results of this study have shown that the Hochstetter’s frogs are thriving in an area where predators are being controlled, and this is important information for not only amphibians, but perhaps for other rare species as well. The advantages and information obtained through the method of site occupancy should be taken strongly into consideration when monitoring other rare species.
With regards to this particular study and to Hochstetter’s frogs, I would recommend shortening the transect to obtain a site occupancy between 30 percent and 70 percent. I would also encourage sampling of more sites, and for the monitoring to be done over multiple seasons to compare results. Moreover, I would suggest decreasing the number of factors recorded when a frog is found to only those covariates that were used in the PRESENCE program, for the other factors appeared not to have a significant effect and were, therefore, not useful to record.
I would like to thank my mentors, Drs. Kimberly Babbitt and Matthew Baber, for their organization, assistance, and friendship. Thanks also to Michael Crossland, who offered a great deal of his time and effort to this study, and to the Waikato Department of Conservation and the Auckland Regional Council for use of the Kokako House, maintenance of trails, and their dedication to the restoration of native forest and the conservation of native species. Thanks to Donna Dowal, the EcoQuest program, Ria Brejaart and Jono Clark for their support and preparation, and, of course, thanks to Janelle Bosse for her dedication and camaraderie. This study was funded by a Summer Undergraduate Research Fellowship awarded by the University of New Hampshire’s Undergraduate Research Opportunities Program.
References
- Blaustein, A.R., Wake, D., Sousa, W.P. Amphibian declines: judging stability, persistence, and susceptibility of populations to local and global extinctions. Conservation Biology 8, 60-71 (1994).
- Mansfield, B. Moving from successful restoration of islands to ecosystem restoration on mainland New Zealand. Paper delivered at IUCN World Conservation Congress, Montreal, Canada (1996).
- Bell, B.D., Carver, S., Mitchell, N.J. and Pledger, S. The recent decline of a New Zealand endemic: how and why did population of Archey’s frog (Leiopelma archeyi) crash over 1996-2001? Biological Conservation 120, 189-199 (2004).
- Bell, B.D. A review of the status of New Zealand Leiopelma species (Anura: Leiopelmatidae), including a summary of demographic studies in Coromandel and on Maud Island. New Zealand Journal of Zoology 21, 341-349 (1994).
- Wilson, K-J. Flight of the Huia (Canterbury University Press, Christchurch, 2004).
- Bell, B.D. The amphibian fauna of New Zealand. In: Newman D.G. Ed. New Zealand Herpetology. New Zealand Wildlife Service occasional publication 2, 27-89 (1982).
- MacKenzie, D.I., Nichols, J.D., Sutton, N., Kawanishi, K., and Bailey, L.L. Improving inferences in population studies of rare species that are detected imperfectly. Ecology 86, 1101-1113 (2005).
- Crossland, M.R., MacKenzie, D.I., and Holzapfel, A. Assessment of site occupancy as a technique to monitor Hochstetter’s frog (Leiopelma hochstetteri) populations. Draft copy
- Pellet, J., Schmidt, B.R. Monitoring distributions using call surveys: estimating site occupancy, detection probabilities and inferring absence, Biological Conservation 123, 27-35 (2005).
