Spiny Rats

Spiny Rats (Proechimys Semispinosus)

by Gregory H. Adler, Ph.D.

Spiny rats (genus Proechimys) occur in virtually every type of lowland forest in tropical Central and South America. These rodents are not spiny like porcupines or hedgehogs, but instead have much smaller spines on their backs that are hidden by soft guard hairs. The spines are evident only when the fur is brushed toward the head. The function of these spines is not known, but they may aid in repelling water in very rainy tropical forests.

The Central American spiny rat (Proechimys semispinosus), the only species found on Barro Colorado Island, is distributed from southeastern Honduras to the Pacific slope of Colombia. This large rat (up to 600 g) is generally a rich reddish-brown above and immaculate white under the belly. Juveniles are a darker brown and have only a few poorly developed spines. Spiny rats eat mostly fruits and seeds but may also eat fungi and insects. Because they are abundant and widely distributed, they are important seed dispersers and predators of a wide variety of forest plants and hence may affect forest regeneration. They also serve as important prey for a diverse array of predators, including ocelots, margays, jaguarundis, owls, boa constrictors, fer-de-lances, and bushmasters.

Spiny rats are often the most abundant rodent within their large geographic range. They are found in both young and old forests, but they are generally most abundant in forests of intermediate age with many fruit trees such as palms and figs. In many forests, spiny rats appear to be most abundant along small streams. Spiny rats are mostly nocturnal and spend the day in underground dens. They emerge from their burrows just after dark to forage on the forest floor. Although they are largely nocturnal, they may sometimes be seen sitting at the entrance to their burrows during the day. They are also readily seen at night if one searches the forest floor with a spotlight. On Barro Colorado Island, Shannon Stream, or along the many forest trails, are good places to search for spiny rats.

An interesting feature of spiny rats is the ease with which their tails break off. If grasped by the tail, the tail will easily break off between the fourth and fifth vertebrae. The rat will then flee to its burrow unharmed, except for the loss of its tail. This characteristic may be an anti-predator adaptation, but it will work only once since the tail will not grow back. In central Panama, about 15-20% of all adult spiny rats are lacking tails.

Recent Research

For the past five years, I have been using 12 small islands in the Panama Canal near BCI as experimental systems to study populations of spiny rats. I had numerous questions about how spiny rat populations vary with food supply and how their reproductive efforts and social behaviors might change as islands become more crowded or less. Specifically, I have tested the following hypotheses:

Hypothesis 1

Spiny rats, which eat primarily fruits and seeds, are not food-limited during the season of greatest food abundance. This was tested by provisioning four island populations with native fruits for six months from May through October 1992, when natural fruit crops are at their greatest abundance. Rats were trapped and marked by toe-clipping and recaptured to determine their population density (number of rats per unit area).

Hypothesis 2

Both sexes of adult spiny rats regulate population densities by their aggressive interactions with young and socially subordinate adult individuals. This was tested by:

  1. Experimentally decreasing densities of adult males on four island in June 1993 and June 1994
  2. Concurrently decreasing densities of adult females on four other islands
  3. Experimentally increasing densities of adult males on four islands in June 1994
  4. Concurrently increasing densities of adult females on four other islands

Hypothesis 3

Adult female spiny rats regulate population densities by adjusting their reproductive effort in relation to changes in density. This was tested by the same experimental procedures outlined in hypothesis 2.

Analysis of the data thus far indicates that population densities generally were not limited by food availability during the season of resource abundance. Adding food to an island did not necessarily result in an increase in rat population density. Hypothesis 1 appears to be correct. Although reproductive output of females was greater in all four experimental populations when compared with four control populations, this greater reproductive effort resulted in higher densities in only one population.

However, seasonal availability of food resources varied among islands, and one population showed much higher densities relative to those expected based on the abundance of naturally available fruit. Thus, this population was not provisioned solely during the period of greatest resource abundance. Survival rates of young and adults did not vary between control or experimental populations. Results of the density manipulations suggested an important role for adult males in regulating population density, presumably as a consequence of aggressive interactions with young and subordinate individuals and consequent effects on recruitment. When the density of adult males was reduced, more young matured into adulthood and became permanent residents in the area where they were born. Results also suggested that adult females did not play a similar role in regulating densities by aggression. Therefore, Hypothesis 2 is correct for males only. However, adult females apparently increased their reproductive output when population densities were reduced, which indicated that females also played an important role in regulating population density, presumably as a consequence of such reproductive adjustments, confirming Hypotheses 3. Thus it appears that when populations dip low, females tend to increase their reproductive effort (if food is also sufficient for reproduction), but if the population swings high, then aggression by the males prevents young rats from settling down and increasing local densities. Future work will help to fine-tune these conclusions and to relate them to the broader understandings of the dynamic interactions between forest fruit production and animal population fluctuations.
© Gregory H. Adler 1997