Maliau rotting fruit project

From BiodivBorneo09

Jump to: navigation, search

Colonization Ecology of Rotting Fruit Resources

Jessica Lavash, Shana Caro, and Kristina Prus



Fruit resources are sporadic in the rain forest, so any new fruit resource will be quickly colonized and exploited by frugivores. This experiment aimed to study colonization patterns of differently sized fruit patches in the forest, looking for variation in number of species, community composition, and early successional patterns of species. We tested this by placing fruit resources (a mixture of rotting apples and eggplant) of varying size classes along transects and watching the colonizers over both twenty-four and two hour periods. After using the analysis of similarities test, we did not find significant variation among communities between different patch sizes. Furthermore, there was not an obvious successional pattern within short term study data; flies and ants colonized the patch at similar numbers throughout the six hour study, though there was a slight increase in both lepidoptera and coleoptera over time. We did find a significant pattern in nocturnal consumer foraging: smaller patches were preferentially consumed over larger. This study can be used as a pilot project to further investigate colonization patterns in the forest, such as a longer study with twenty-four hour observation and nocturnal frugivore exclusion to investigate arthropod community composition over a longer and more precise time frame.


In the rain forest, fruit resources are distributed in patches that are spatially and temporally distributed due to a general lack of fruiting seasonality and random spatial distribution of fruiting trees (Saracco, Collazo, and Groom, 2004). Subsequently, terrestrial frugivorous insects must constantly forage for food sources such as fallen fruits in the forest. Patch size, or abundance of decomposing fruits in one area, is variable and so it is possible that decomposers selectively forage for larger resources that would provide more food, according to optimal foraging theory (MacArthur and Pianka, 1966). Smaller patches of food could also be less visited since they would not be as easily detected as the larger fruit falls. Larger organisms such as civets and squirrels scavenge for food on the forest floor, while several insect species specialize on consuming fruit resources. Several species of fruit flies, ants, butterflies, and beetles are frequently found on fruit falls and other arthropods place their larva inside for nutrients. This system of frugivorous organisms foraging for decaying food provides an insight into community dynamics dependent on limiting resources in the forest.

The question

  • How does patch size influence the process of colonization and community assembly at new fruit resource sites?

Anticipating that patch size is a factor in frugivore detection and selective exploitation of the new resource, understanding these community colonization patterns allows us to better understand tropical forest decomposition processes and nutrient cycles.


Larger resources will likely have the fastest colonization with the greatest biodiversity because they will be easier to detect by frugivores. This hypothesis is based on the optimal foraging theory, which suggests that foragers seek out larger resources since they provide more food per unit effort. Consequently, smaller food resources will likely be colonized less rapidly since decomposers will preferentially seek out larger resources, only utilizing smaller resources when competition is too intense at the higher quality patches.

Community composition is anticipated to vary between the patch sizes as well. Larger patches should host more species of frugivores than smaller patches. The larger surface area available for feeding on larger patches provides more opportunities for feeding. Additionally, smaller patches might be subject to higher intraspecies competition. For this reason, we anticipate that organisms will avoid competition by exploiting different resource patch sizes. This will result in variation in types of organisms that colonize each patch size. This pattern of niche differentiation is often seen in ecological studies: species will specialize on different resource types and patch sizes.


Field Methods

Two experiments were performed in primary rain forest at Maliau Basin from July 13-17, 2009. We created standardized fruit resources by concocting a uniform paste comprised of rotting apples and eggplants and separated this paste into three size patches: five grams (small), ten grams (medium), and twenty grams (large). Fruit resource patches were distributed randomly throughout the forest using transect lines with either twelve or fifteen samples along each line, separated by five meters each. Transect lines were laid in uniform environments of undisturbed, shaded lowland mixed Dipterocarp rain forest. To determine community composition of each sample, we calculated percent cover on each resource and collected presence/absence data for all arthropod species found on the resource.

Experiments were of long term and short term duration. The long term study consisted of four transect lines of twelve samples, with four of each size type, small through large, for a total of 48 samples. We observed these patches every 24 hours for three days to determine community composition and time elapsed for total resource consumption. For the short term experiment, two transect lines of fifteen samples were laid with five of each size type, small through large, for a total of 30 samples. To determine initial colonization patterns and community ecology on fruit resources, we observed these patches on two hour intervals for six hours over two days.

Statistical Analysis

Our statistical hypothesis is that there will be a significant difference between the number and type of species on different resource patch sizes. Using the statistical software R, we will test this hypothesis with an analysis of similarities test (ANOSIM) and a dissimilarity index, both designed to quantify similarities within community composition. We expect to see a significant difference in biodiversity and community composition between the small, medium, and large resource types. The null hypothesis is that communities on the different resources patches are statistically indistinguishable from one another. We will analyze the long term data using linear models to see whether resources that went missing overnight were related to resource size, and we expect to a significant relationship between these two factors. The null hypothesis is that no such relationship exists.


Figure 1:The analysis of similarity test (ANOSIM) demonstrates that there was no difference in community composition between the three patch sizes in the short term experiment. This holds for the communities at two hours, four hours, six hours, and overall.

Patch size influenced which samples disappeared overnight in the long term experiment: 41% of missing samples were small, 32% were medium, and 26% were large. Smaller patches went missing sooner than larger samples as well; over the first 24 hour period, 13 small samples disappeared while only 9 medium and 7 large samples disappeared. The relationship between size and whether or not a sample went missing was highly significant when fit to a linear model (p=0.00297).

Greater alpha diversity was observed at large resource patches (average 3.03 species), then medium patches (average 2.69 species), and lowest at small patches (average 2.20 species) in the short term experiment. However a statistical analysis of community composition (ANOSIM) did not show a significant difference between communities on differently sized resources. The observed greater alpha diversity was most likely caused by increased surface area on lager patches: more individuals are able to utilize larger resources. This pattern was not observed in the long term experiment, possibly because biodiversity counts were confounded since small patches disappeared at a faster rate than medium and large patches.

The short term study did not yield any clear successional patterns. Formicidae and diptera were equally present at all times. Coleoptera and lepidoptera did show a moderate increase in presence over time; after four hours coleoptera increased in number from two species to five species present, while lepidoptera took six hours to plateau at six species.


Our data did not support our hypotheses.

Data from the long term study showed that patches are consumed preferentially based on size. Nocturnal consumers of the fruit patches, most likely civets and other small mammals, showed a preference for smaller patches; these were fully consumed overnight more frequently than medium patches, with large patches disappearing the least. Perhaps resource patches of five grams have a shorter processing time than patches of twenty grams, making the smallest patches the easiest for decomposers to consume quickly and thus safely.

Testing data from the short term experiment resulted in no apparent difference in arthropod colonization rates between the different fruit resource patch sizes. This indicates that arthropods are not limited by patch size in detection of fruit resources. This is reasonable given the variable spatial distribution of fruit falls in the forest, so insects should be able to detect even small resources in order to exploit the environment fully.

Furthermore, there were no significant differences in community composition over time; this demonstrates that within the first two hours that a resource is present on the forest floor, it has been colonized by nearly all of the frugivorous consumer types. We were unable to determine successional patterns in insects that utilize fruit falls for laying eggs or developing larva. However, based on data from the long term study, the length of time before resource patches are completely consumed is too short for many arthropods to have the opportunity to utilize fruit falls in this way before the resource disappears.

An additional experiment to further illuminate ecological processes of arthropod resource colonization would be collecting 24-hour data with a camera to measure the entire consumption process. Since this study focuses on arthropods, exclusion netting should be set up around the resource to prevent small mammals or other large decomposers from consuming the material overnight. These changes would minimize confounding factors and measure for arthropod utilization of fruit falls. Additionally, a similar experiment could be conducted with whole fruits as a resource rather than fruit paste. Whole fruits have skins that mask scent and protect against some types of consumers. This might allow for more burrowing consumers to colonize the resource while other arthropods might not be able to detect it as rapidly.

In conclusion, arthropod species do not show a preference for larger fruit resource patches and do not detect differently sized patches at different rates according to the time scale of our experiment. Nocturnal consumers, conversely, showed a clear preference for smaller patches over the long term experiment.


  • MacArthur, R. H. and Pianka, E. R. (1966). On the optimal use of a patchy environment. American Naturalist, 100.
  • Saracco, J.F., Collazo, J.A., and Groom, M.J. (2004) How do frugivores track resources? Insights from spatial analyses of bird foraging in a tropical forest. Plant Animal Interactions, Oecologia 139: 235–245.