Friday, October 19, 2012

Rabbits versus flowers on Middleton Island, Gulf of Alaska

Middleton Island is currently inhabited by an introduced European feral Rabbit Oryctolagus cuniculus population, which originates from four domestic rabbits (3 females and 1 male) that were purposely liberated on the island in the fall of 1954. An earlier (but unsuccessful) introduction of feral rabbits took place during the fox farm era in the beginning of the twentieth century. An interesting account of the first few years of the successful establishment of the rabbits on Middleton has been written by O'Farrell (1965). In the present these rabbits are the only land mammals on the island and during my four visits between 2005 and 2010 they were abundant. This blog entry presents and discusses a few of the findings of a study that focuses on the effects the introduced rabbits have on the remaining wildlife of Middleton Island. It was initiated by me in 2009.

Armand Biron (720th A.C.&W. Squadron) holding one of the earliest of Middleton’s rabbits (photo: Armand Biron, 1960).

Introductions of non-native species into island ecosystems are known to be a reason for concern. The presence of Arctic Foxes on Middleton Island between 1890 and the 1930’s, for example, certainly was disastrous for the island’s wildlife, as it clearly reduced the possibilities for many species. The effects of fox introductions to Alaskan islands have been studied before (see i.e. Bailey 1993 for a review) and eradication projects, that have been taken place as part of wildlife conservation and habitat restoration, have been very successful. Thus far, however, the effects of rabbit introductions on Alaskan islands have been poorly studied. Currently, introduced rabbits occur only on a few Alaskan islands, as most other introduction attempts in the past have been unsuccessful. Only recently have people started to eradicate introduced rabbits from an Alaskan island as an ideological way of nature conservation ( This action’s proposal consisted of not a single scientific field study from the relevant area, but was largely based on references from the southern hemisphere (the other side of the world…).
The reason for me to investigate Middleton’s rabbit population has been that, during my first few visits, I did not notice many negative effects of their presence on the remaining wildlife. Instead, I noticed several what seem to be some positive effects of their occurrence on the island. This made me very curious.

One of the things I wanted to find out was how many rabbits were present on Middleton during my stays. For this I conducted regular counts throughout the summer (early April – end of August) in all of the different habitat types that can be distinguished on Middleton, at different times of the day. The rabbits produce litters usually from the end of April until the beginning of August, but their timing depends on climatic conditions and can vary notably.
Preliminary results indicate that in 2009 (after the species dealt with a relatively mild winter) their population contained about 3000 individuals at the start of the summer, and after 2 - 3 litters about 12,000 individuals inhabited the island by the end of August (Van Nus, personal records; results of the following year’s count (2010) still need to be analyzed). O’Farrell estimated the population to be between 3,600 and 3,700 rabbits in February 1961 and noted that winter mortality can be high (O’Farrell 1965). In 2006 and probably to an even greater extend in 2012, it also seemed that a previous winter had taken its toll, decreasing rabbit numbers to a large extend and resulting in temporary disappearance of the rabbits from parts of the island.

I also wanted to find out more about the interaction of the rabbits with other animals. For example the influence of the introduced rabbits on the predators: which of the present species feeds on rabbits and to what extend are these dependant on them? And what other prey items do these particular predators feed on? For this, I conducted several diet studies among the main predatory species, of which the Bald Eagle Haliaeetus leucocephalus and the Snowy Owl Bubo scandiacus are the most important, but a small number of other species is known to take rabbits on Middleton (i.e. Roberts 1985 (for the Glaucous-winged Gull Larus glaucescens) and Van Nus 2006 (for the Northern Harrier Circus cyaneus) and these predators’ diets were investigated as much as possible.

A poor photo of an immature Gyrfalcon Falco rusticolus on its recently captured prey. It got flushed from its prey by the photographer, revealing the remains of an adult rabbit (photo: Tim van Nus, April 2009).

Preliminary results show that predation makes up an important part of mortality, but also climatic conditions during the winter are of large importance to the annual rabbit survival, probably even more important than predation. The island gets sometimes visited by hunters that shoot an unknown number of rabbits and the impact of their actions remains unknown.

Rabbits are often accused of preventing some burrow-nesting seabird species from nesting on the island, such as Storm-petrels Oceanodroma sp. or Ancient Murrelets Synthliboramphus antiquus (i.e. see Audubon’s site description for Middleton Island: However, there seems to be no scientific data to support this.
I am studying this by using an extensive database of historical bird records, which covers pretty much the whole of Middleton’s recorded history. In combination with my recent inventory work and the work of others, by now I know relatively well which species nest or have nested on the island, which are seen on or around the island and when, and which are not seen.  Neither Storm-petrels nor Ancient Murrelets are known to nest or have nested on Middleton but both have been recorded irregularly around the island in summer. These species nest in extensive colonies in the surrounding area of Middleton, and are abundant.

A beached Ancient Murrelet in 1960. This is the second oldest record of this species on or around Middleton that I could find (photo: Armand Biron).

Besides these aspects, there are several other things concerning Middleton’s rabbit population that I’m studying, which I will not go into detail about here, except for one. Introduced rabbits are often thought to cause habitat degradation and negatively affect the indigenous flora, indirectly affecting other wildlife. Therefore I took the opportunity to study the rabbit’s influence on the island’s vegetation, which I thought would be interesting enough to present some parts of in this post.

Middleton Island’s vegetation has been studied before by Thomas (1957) who identified 116 plant species. Thomas’ study was part of the 1956 biological reconnaissance of Middleton. Thomas conducted his study in June 1956, less than two years after the introduction of the first four rabbits. Because of winter mortality, in the spring of 1956 their population was reduced to five females and two males. The summer population in 1956 was estimated at 200 animals (O'Farrell 1965). I therefore suspect the influence these rabbits had on the vegetation at that time was still minimal and Thomas’ work still largely represents the pre-rabbit situation.

In 2010 I conducted a vegetation inventory and managed to identify 105 plant species, excluding a number of species (about 20) which up to this point have remained unidentified. Now, this would seem to indicate that the rabbits grazing activity has probably not decreased the island’s vegetational diversity, but unfortunately things are a bit more complicated. Due to the island’s uplift during the 1964 Earthquake (which created a whole new supra-tidal area around the island) a number of new species can be expected to have shown up on the island during the following decades. At the same time, the uplift may have influenced the vegetation of the older island parts, as this got further out of the reach of the ocean. When not taking into account the newly created formations, the number of species identified on the pre-1964 parts of the island (98) is lower than was found by Thomas in 1956. Some species that I could not relocate on the older formations, however, were found in the post 1964 areas instead, indicating that ‘succession’ (or the uplift) may be an issue. Additionally, it appears that the number of invasive species has increased during the recent decades. And finally, I must point out that I probably was not carrying the same expertise as Thomas’ in 1956 (but certainly not unusual with vegetation inventories). None the less, at this point I believe it would be safe to say that the rabbits have NOT caused any serious decrease in the diversity of the vegetation of Middleton Island, about 57 years after their introduction.

But what are the effects of the rabbit’s grazing activity and what would the island look like without them? In order to study this I erected several exclosures:  small fenced off areas that are inaccessible to the rabbits. The idea was (and still is) to keep the rabbits out for a period of many years in order to be able to monitor vegetation development without the presence of rabbits. These exclosures were made from metal poles and 50 cm high chicken wire that was pinned into the ground, and they vary in size from about 2 x 2 m for the single smallest, to about 4.5 x 4.5 m for the largest. Three exclosures were erected in the summer of 2009 and another two during the following summer. The exclosures are distributed over different vegetation types and/or formations.

Exclosure 1 in the north-end of the island in the spring of its fourth year in 2012 (photo: USGS crew 2012)

Rabbits are not the only herbivore on the island. There is a large population of Canada Geese Branta canadensis present on the island throughout most of the year, numbering just over 2000 birds (adults and young) in late June 2008. After a fast increase from the 1970's, their population stabilized in the mid 1990's around this number (Petrula et al 2008). These geese will also influence the vegetation, but could technically have access to the vegetation within the exclosures if they want to (and they are allowed to). Due to the exclosures’ sizes, however, I doubt they ever did thus far. The geese are not very common around three of the five exclosures and I did not see any signs that these had regularly been feeding in or in the vicinity of the remaining two (no, or hardly any, fecal pellets of them were found there during my visits).

Already within a year after the first exclosures had been placed it became notable that, during the summer, the rabbit’s grazing activity was aimed primarily at flowers, but also at seeds and fruits of the plants. At closer examination in 2010 it appeared that all of the flowering plants that were found within the exclosures could also be found elsewhere on the island. Some of them were very common and others relatively rare.

An abundance of flowering Yellow Monkey-flowers Mimulus guttatus within exclosure 3. The Yellow Monkey-flower grows abundant just around the exclosure too, but there their flowers get eaten by rabbits (photo: Tim van Nus, June 2010).

Strangely, in a few cases it got noticed that some of the plants that were found flowering within the exclosures were only incidentally found flowering elsewhere, but notably only in the middle of the area containing the highest rabbit density! This was the case, for example, with a single Yellow Paintbrush Castilleja unalaschensis growing in exclosure 1 at the island’s north end. Though this species is stated as being fairly common by Thomas (1957), this was only the second time I observed it on the island. Remarkably, a few days after this finding, I also observed a small stand (3 - 4 plants) growing in the center of the island, at a location where rabbits are extremely abundant, and mainly feed on Annual Bluegrass Poa annua.

Since the summer of 2010 there is a single Yellow Paintbrush to be found flowering in the middle of exclosure 1 (photo: USGS crew 2012).

In 2010 Yellow Paintbrush was also noticed growing in the A.C.&W. station, the area containing the island’s highest rabbit density (photo: Tim van Nus, July 2010).

During 2011 and 2012, when I did not visit Middleton, the exclosures were monitored by others, and a number of photos got taken of each exclosure, once during spring and once during summer (the beginning of July). Thus far, not all the exclosures have shown such a dramatic difference. Two exclosures placed in areas with a relatively low rabbit density, up till now have shown only little differences with their surrounding vegetation.

Three years after its construction, the largest exclosure (#5), in the southern upland of Middleton Island, still shows no clear difference in vegetation structure compared to the surrounding vegetation (photo: USGS crew 2012).

After analyzing these photos as much as possible, I noticed a similar pattern taking place as during the first two growing seasons of the project. However, one thing that stands out at the moment is the presence of a yet unidentified purple flower that appears to be growing well within exclosure 3. Unfortunately, from the photos I can’t manage to identify it, and I don’t know whether I’ve listed it before in a ‘non-flowering state’.

Suddenly there were these yet unidentified purple flowers occurring in exclosure 3. A Shooting-star sp?? (photo: USGS crew 2012).

But, with only four years after the first exclosures were placed, there are still new developments expected to be taking place within them, so what is presented here will be very preliminary. A continuation in monitoring the exclosures would be very helpful, and so would a further increase in the number of exclosures be. Thus far, the observations indicate that:
  • Although already a colorful place, without the presence of the rabbits Middleton would probably host a lot more flowers.
  • The rabbits’ occurrence probably changed abundances of some plant species, but yet there is no clear evidence that plant species disappeared due to the rabbits' presence.
  • The rabbit’s grazing pressure varies along the location and is absent in some locations.
  • The rabbit’s grazing pressure varies between years and is absent in some locations in some years (probably allowing some plant species to flower irregularly or incidentally).
  • Due to repeated grazing, a large number of plant species occurs in a sort of ‘dwarf state’; they remain low and will never flower, yet they can be very abundant.
  • Certain plant species are preferred by the rabbits and when these are growing abundant they ignore certain other plants that would readily be eaten in other locations on the island.
Throughout the summer the introduced Annual Bluegrass Poa annua appeared to be the rabbits’ favored food item and the highest rabbit densities are found on and in the vicinity of short grazed foraging patches containing this species, primarily on and near the island’s infrastructure. These areas are also the place where the highest winter survival occurs, possibly allowing the species to remain on the island during the more severe winters. It remains unclear whether the establishment of humans on the island ultimately created a life-line for the rabbit population throughout the following decades, but I am more and more getting that impression.

Rabbits are most abundant on or near the island’s infrastructure, where they maintain short-grazed foraging patches. Here they are also most easily seen (photo: Tim van Nus, August 2009).

Now, does this mean that the rabbits are causing habitat degradation and are negatively affecting the indigenous flora? I don’t believe so. Thus far it seems their presence results only in changes in the vegetation, and from a vegetational point of view I believe these changes are not necessarily a bad thing.

And whether the rabbits are indirectly affecting other wildlife… Yes, certainly they do! The reduction of flowers, seeds and fruits will likely have an effect on the island’s insect world and therefore indirectly affect food supply of some birds species. According the Audubon the “Heavy grazing by the hares must also make the island less attractive as nesting habitat for some shorebirds, such as Short-billed Dowitchers Limnodromus griseus” (, but again I did not managed to find any scientific data that can support this. (A single pair of this species is recorded to have likely nested on Middleton Island for a period of several years during the late 1990’s (Van Nus, unpublished data)).
Whether the overall effect of the rabbits’ grazing is negative remains yet to be seen; on Middleton the locations with the highest rabbit densities coincide with the highest densities (record densities!) of savannah sparrows, which, in combination with the presence of juvenile rabbits in late summer, seems to allow a pair of Northern Harriers to nest there irregularly (Van Nus 2006).

This blog entry would get a bit too complex to present and discuss other findings, but I can ensure you that, after examination, at this point for every negative aspect of the rabbit’s occurrence I can come up with a positive. In my opinion, the assumed but not proven absence of Storm-petrels and/or Ancient Murrelet on Middleton, due to the presence of rabbits, gets easily compensated by the remarkable appearance of Snowy Owls on Middleton. During each winter, a number (1 – 35+) of these owls resides on Middleton and 99% of their diet consists of rabbits (Rausch 1958, O’Farrell 1965, Van Nus, unpublished data 2005 – 2006). Immature birds, usually 1 - 2, are resident to the island in the summer (well away from their breeding range), although irregularly. In 2005, 96% of the diet samples that were collected from two immature individuals throughout the summer, contained the remains of only rabbits (Van Nus 2006).
I remember that the birdwatchers that paid organized summer-day visits to the island during my stays were all pretty excited about putting the Snowy Owl on their list. I believe that these owls are one of the many things that make Middleton Island unique, and this could also count for the rabbits, which certainly deserve more attention in the future, and not from hunters but from scientists, before the opportunity is lost.

Thanks to the rabbits Middleton Island can contain a Snowy Owl population year-round. Unique for southern Alaska! (photo: Brian Guzzetti, May 2005).

I want to thank everybody that helped out with constructing and placing the exclosures in 2009 and/or 2010, as well as the people that helped monitoring the exclosures during the following two years. I thank Scott Hatch (USGS) for letting me use the chicken wire and a few stakes, and the FAA for the leftover metal poles. 

Bailey, E.P. 1993. Introduction of foxes to Alaskan Islands – History, effects on Avifauna, and eradication. United States Department of the Interior, Fish and Wildlife Service. Resource publication 193. Washington DC.
O'Farrell, T.P. 1965. The Rabbits of Middleton Island, Alaska. Journal of Mammalogy, Vol. 46, No. 3, pp. 525-527.
Thomas, J.H. 1957. The vascular flora of Middleton Island, Alaska Contributions from the Dudley Herbarium, 5:39-56.
Rausch R. 1958. The occurrence and distribution of birds on Middleton Island, Alaska. The Condor 60: 227-241. 
Roberts, B. 1985. Glaucous-winged Gulls prey on feral rabbits on Middleton Island, Alaska. Murrelet 66:24. 
Petrula, M.J., Rothe, T., Rosenberg, D. & Crowley, D. 2008. Canada Goose Survey on Middleton Island – 2008. Alaska Department of Fish & Game. Unpublished report.
Van Nus, T. 2006. The simple diet of a Northern Harrier Circus cyaneus hudsonius pair at Middleton Island in the Gulf of Alaska. De Takkeling 14: 68-77.