1 Oophaga pumilio (strawberry poison dart frog) sex dependent microhabitat selection University of Manitoba Adam Borton April 07, 2016 (Photo: Kevin McRae)
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Oophaga pumilio (strawberry poison dart frog) sex
dependent microhabitat selection
University of Manitoba
Adam Borton
April 07, 2016
(Photo: Kevin McRae)
2
1.0 Introduction
Widespread population decline of various animal species can often times be
attributed to increased human activity. Influences such as agricultural or urban
development encroach upon species habitat, causing fragmentation of once continuous
environments (Nowakoski et al., 2012). Such influences can be detrimental to
populations, as individuals cope with the changing landscapes. Amphibian populations
have been declining over the years as essential habitat is lessened by humans. In fact, in
comparison to other vertebrate groups, amphibians are the most threatened group of
animal species in the world with much of the decline being attributed to human influence
and land-use (Murasaki, 2010). In order to prevent further destruction of critical habitat
of amphibians, it is important to understand the geographic range in which amphibians
live within. An amphibian’s home range can be defined as the area which includes the
space required for the individual to accomplish its daily movement and activity
(Murasaki, 2010). Understanding a species home range is extremely important from both
an ecological and conservation standpoint as it will highly influence distribution of the
species, population dynamics, resource use, as well as inter and intraspecific interactions
(Murasaki, 2010). Through focusing on factors which affect how and why a species
occupies a specific site, conservation efforts can begin to be implemented and further loss
of habitat avoided.
1.1 Oophaga pumilio – Strawberry Poison Dart Frog
A species which has begun to be affected by habitat loss is Oophaga pumilio.
Oophaga pumilio (O. pumilio), commonly known as the Strawberry Poison Frog, is a
diurnal species of frog belonging to the Dendrobitadae family (Murasaki, 2010). The
3
species is often found in tropical forested environments in Nicaragua, Costa Rica and
Panama (Meuche, Linsenmair & Prohl, 2011). Dendrobatid frogs have evolved a defense
mechanism to negate predation, through obtaining chemicals through their diet of
alkaloid-containing organisms (Stynoski et al., 2014). After consumption, poisonous
chemicals can then be secreted through specific glands found on their skin, enlarging as
the organism transitions from juvenile to mature life stages (Stympski et al., 2014). The
species typically occupies moist, dense, lowlands or swamps and avoids open, exposed,
pasture-like areas in order to avoid dehydration (Nowakowski et al., 2013). Oophaga
pumilio is said to be one of the most phenotypically diverse frogs observed, exhibiting a
range of colours varying with location. While in Nicaragua and Costa Rica the frog
typically appears red-orange with blue-black appendages, it is common to find the frog
appearing pure orange or green in other locations (Galeano & Harms, 2016). At maturity,
Oophaga pumilio grow to be anywhere from 17.5 to 24 millimetres long (Nowakowski et
al., 2013).
While current knowledge on amphibian land use is limited in some regards, much
research has gone into understanding behavioral and reproductive interactions between
O. pumilio. Using such knowledge can be beneficial from a conservation standpoint, as
understanding what environments or physiographic features a frog may occupy could
benefit protection efforts. Territorial ranges differ between male and female O. pumilio
with female frogs maintaining larger territories than males (Meuche, Linsenmair & Prohl,
2011). In general, males compete for dominance against one another in order to defend
their home range, attract females and secure reproductive sites (Murasaki, 2010). Males
typically occupy smaller home ranges than females as smaller areas are easier to defend
4
from other male frogs (Murasaki, 2010). Male O. pumilio territory typically includes
locations which are conducive to attracting or courting females, namely oviposition and
courting sites located on dry leaves in the organic litter of the forest floor (Meuche,
Linsenmair & Prohl, 2011). Male O. pumilio distribution is often influenced by high
densities of female frogs and territory is selected for areas with an absence of competing
males (Murasaki, 2010). Female territorial ranges are typically larger with movement
within the home range occurring much more frequently than males (Murasaki, 2010).
Females are involved with much of the parental responsibility, securing territory for
better access to clean water or food sources (Meuche, Linsenmair & Prohl, 2011). Since
female O. pumilio are involved with raising young individuals, it is common to find
female home ranges next to tadpole-rearing sites such as water filled axils of plants or
small, calm bodies of water (Meuche, Linsenmair & Prohl, 2012). The axils of bananas or
heliconias provide suitable sites for such rearing locations as well (Meuche, Linsenmair
& Prohl, 2011). Once females have established their home range, they commonly spend
much of their time within that area to tend to their young and secure food. Female O.
pumilio home ranges are typically located in locations which allow access for multiple
male mating (Murasaki, 2010). While the home range of male and female O. pumilio are
undoubtedly distinct from one another, overlap does occur. Due to the frogs preferring
moisture to prevent dehydration, it is quite common to find both male and female frogs in
cool, damp areas beneath the organic leaf litter or within the roots of plants (Meuche,
Linsenmair & Prohl, 2011).
5
The objective of this analysis was to determine whether there was an association
between the gender of O. pumilio and common locations in which it is found. By
analyzing if such an association is present, valuable insight on species site preference
could be attained. Research has shown that the roles and responsibilities of O. pumilio are
dependent upon gender, with both male and female individuals being involved with
different aspects (Murasaki, 2010). Could gender responsibilities and habits then be used
to explain why individuals occupy certain features of the habitat and distributions are
situated the way they are? Perhaps females are more commonly found in trees or on leaf
axils due to their parental and offspring rearing responsibilities. Perhaps male
distributions are often found on the ground or within the leaf litter due to their
competitive and reproductive habits. Research conducted in the past by Seiichi Murasaki,
looked to determine rates of movement as well as how both genders of O. pumilio
distributed themselves within environments. This particular study determined that
females and juveniles occupied significantly larger home ranges then males (Murasaki,
2010). However, the objective of our analysis focuses on the actual natural features the
frog was found upon (tree, ground or leaf) and the association that factor has with gender.
Size of home range of male and female frogs was not studied in this analysis.
Understanding how O. pumilio situates itself and operates within its home range would
be beneficial when looking to protect and conserve the species. If such an association
with gender is present, data collected could be used to better mitigate and avoid human-
related damage to specific preferred O. pumilio habitat based on gender. Location
information would also increase efficiency when conducting studies on O. pumilio as less
time would theoretically be spent locating the species.
6
2.0 Methods
2.1 Study Region
The study region was located on Isla Colon, Bocas Del Toro, Panama. Isla Colon
is the northernmost Island in the Bocas Del Toro Archipelago, located east of mainland
Panama in the Caribbean Sea. Bocas del Toro lies 9 degrees above the equator in the wet
tropical zone. The average annual rainfall and temperature are 4000 mm and 30°C (86°F)
respectively (ITEC, 2015). The island is highly diverse and offers habitat for several
thousand different organisms. Many ecosystems occur within the region such as tropical
lowland rainforest, which was the focus of this study. We collected our study samples
from one population located in Dixon Forest, near the Bocas Del Toro Biological Station.
2.2 Dixon Forest
Dixon Forest is approximately 6km2 and is comprised of tropical lowland
rainforest; a primary growth forest located on the northwest corner of Isla Colon. These
forests are home to many large ficus trees, which are more than 100 feet tall. These trees
are a keystone species in the area, and their buttresses create micro ecosystems for many
species including O. pumilio. The ground is layered in leaf litter with many species of
low-lying vegetation such as philodendrons and dieffenbachia. Nearby areas are cattle
pastures, secondary forest, marshes and human settlement.
7
Fig 1. Dixon Forest study location, ITEC, 2015
2.3 Study Site
Two survey locations (Plot A and B) were randomly selected within Dixon
Forest. Plots were located approximately 400 meters apart. Each plot was designed as a
20x20 meter quadrat. Quadrats were created using a tape measure and a compass to
ensure that the plot remained square. Once the quadrat was created, surveying flags were
placed to form the shape for the remainder of the study.
2.4 Sampling Methods
Plots were surveyed for 60 minutes every second day, for a total of three times
each. When a frog was located, the searcher would place a piece of surveying tape to
mark the location. The searcher would than capture the frog, place the frog in a zip lock
8
bag and place the bag at the first observed location. Careful consideration were given that
the frogs were not injured during the process. The frogs were only handled for a short
amount of time, as amphibians have permeable skin and should not be touched with dry
hands for long periods. Once the 60 minute period was complete, we began to record the
data. Each frog was measured, sexed and its first observed location documented on a field
message pad. Each frog was photographed with a digital camera to ensure that is was not
a recaptured frog. O. pumilio on Isla Colon all have different spotted patterns making
individuals distinguishable. Recaptured frog data observations were not included in this
study.
Fig 2. O. pumilio on Isla Colon displaying the variation of pattern. McRae, 2013
O. pumilio are easily sexed on Isla Colon as male frogs have a dark colouration
on their throats while females have lime green colouration. Frogs were measured with the
use of a pair of calipers. When we returned to the biological station, data recorded on the
field message pad was transferred to a Microsoft Excel spreadsheet.
9
Table 1: O. pumilio contingency table, McRae, 2016.
Tree Leaf Ground Total
Male 11 5 15 31
Female 16 2 13 31
Total 27 7 28 62
2.5 Statistical Analysis
A chi-square analysis was used to determine significant differences in observed
location in relation to frog gender. All assumptions were met to perform this test; the
sampling method is simple random sampling, the study variables are categorical.
The value of the number of sample observations for each variable is at least 5. A 2x3
contingency table was used to determine gender and perch. Frogs from both samples
plots were separated into two categories: male and female. The second variable, first
observed location was separated into three categories: trees, leaf and ground. The data
was compiled in a contingency table seen in Table 1. All data was analyzed in IBM SPSS
Statistics 22.
3.0 Results
It was determined that there is sufficient evidence to support the null hypothesis
and that the locations of sighted O. pumilio does in fact occur independent of their
gender.
10
3.1 Descriptive Statistics
Prior to initiating the inferential analysis of the O. pumilio information, the
descriptive statistics should be collected, measured, and analyzed. The three types of
measures within descriptive statistics to be analyzed consist of central tendency, measure
of dispersion, and measure of shape. The purpose of utilizing the method of measuring
descriptive statistics within the scope of the research at hand is to efficiently and
concisely summarize the collected information. The descriptive data collected for the O.
pumilio frogs and their respected locations are represented below in Tables 2 and 3.
Table 2: O. pumilio descriptive statistics, Elphick, 2016
Table 3: O. pumilio location frequencies, Elphick, 2016
Location Gender
N Valid 62 62
Missing 0 0
Mode 3.00 1.00a
Sum 125.00 93.00
a. Multiple modes exist. The smallest value is
shown
Frequency Percent Valid Percent Cumulative Percent
Tree 27 43.5 43.5 43.5
Leaf 7 11.3 11.3 54.8
Ground 28 45.2 45.2 100.0
Total 62 100.0 100.0
11
With the information utilized in this study consisting of nominal data, expanding
upon the many aspects of descriptive analysis proves challenging. Nevertheless, upon
examining the central tendency measures affixed with the data in Table 2, it can be seen
that the mode is 3.00, which throughout testing represents the ground. Taking the modal
unit into account it may be considered that perhaps no study is needed as the data shows
that the frogs are most observed upon the ground. However, this value considers both
genders of frogs whereas it is the locational difference between genders being evaluated.
Furthermore, Table 3 shows that the ground location does not far surpass the amount of
strawberry poison dart frogs observed amongst the trees. Lastly, considering the
variables are comprised of gender and location of the O. pumilio frog it is clear that a
median as well as a mean cannot be calculated as the variables cannot be ranked nor does
an average value in the data set exist. As all other aspects beyond mode in descriptive
statistics are based upon the mean, the measure of dispersion and shape also cannot be
further analyzed. Due to the nature of the variables allowing only a limited analysis,
Figure 3 provides an additional and potentially more beneficial representation of the data
in which to infer further information as it provides a visual representation of the
distribution and observed frequency of male and female O. pumilio frogs. It is evident
that for both genders, frogs found on leaves were observed at the lowest frequency, while
ground and tree location data was observed at two to three times larger frequencies.
Though when focusing on gender, it is apparent that males were most commonly
observed on the ground while females were most often found in trees.
12
Fig 3. O. pumilio double bar graph of distributions. Borton, 2016
3.2 Chi-Square Assumptions
In proving whether or not the location of where poison dart frogs are found is
dependent upon gender, determination resides upon performing a Chi-Square test on the
observed data obtained within the Dixon forest, Panama. In conducting the Chi-square
test of Independence three assumptions required to carry out such a test must to be met.
The assumptions in this case include that the data come from a single random sample,
that the variables be organized by nominal or ordinal categories, and lastly that no more
than one fifth of the categories contain less than five observations.
To initiate the hypothesis testing, data was collected using a single random
sample of sixty-two O. pumilio frogs. Sixteen females and eleven males of which found
in trees, two females and five males found on leaves, and thirteen females with fifteen
males found on the ground. Due to the data being a single random sample, as discussed
13
previously within the methods section, it is clear that the first assumption is met.
Secondly, based on the variables being both the strawberry poison dart frog’s gender
along with their quantifiable locations, the condition of nominal data can be easily
assumed. Lastly, as only one variable category, being the leaf location, contains less than
five observations it can be ascertained that this last crucial requirement is also fulfilled.
For additional clarification, Table 4 below further illustrates the data discussed and
assumptions verified by such.
Table 4: O. pumilio observational statistics, Elphick, 2016
Gender Measure Location Total
Tree Leaf Ground
Male Count 11 5 15 31
% within Gender 35.5% 16.1% 48.4% 100.0%
% within Location 40.7% 71.4% 53.6% 50.0%
% of Total 17.7% 8.1% 24.2% 50.0%
Female Count 16 2 13 31
% within Gender 51.6% 6.5% 41.9% 100.0%
% within Location 59.3% 28.6% 46.4% 50.0%
% of Total 25.8% 3.2% 21.0% 50.0%
Total Count 27 7 28 62
% within Gender 43.5% 11.3% 45.2% 100.0%
% within Location 100.0% 100.0% 100.0% 100.0%
% of Total 43.5% 11.3% 45.2% 100.0%
14
3.3 Chi-Square Test of Independence
With the assumptions shown to be satisfied, commencing the Pearson’s Chi-
Square test of independence can be perceived as viable. Based on the characteristics of
the study being conducted, determination will be established using a significance value of
0.05 or 5% using a two tailed analysis. Though the final outcome of the test will be
decided using the p-value approach, to further verify the conclusion it is helpful to
establish the critical statistic relevant to the data. With the degrees of freedom equal to
two and the significance level set at 0.05, the Chi-Square table provides a critical statistic
value of 5.991. Furthermore, as it is the independence of the O. pumilio frog’s location
contingent upon gender being investigated, the null and alternative hypothesis in this case
are as follows:
Ho: The locations of sighted O. pumilio frogs occurs independent of their gender
Ha: The locations of sighted O. pumilio frogs are not independent of their gender
With all prior necessary steps taken in order to perform the hypothesis test, the O.
pumilio data calculated through SPSS provide the below statistics in Table 5.
Table 5: Chi-Squared Test Results, Elphick, 2016.
Value Degrees of
Freedom
Asymp. Sig.
(2-sided)
Pearson Chi-
Square
2.354 2 0.308
Likelihood Ratio 2.403 2 0.301
Linear-by-Linear
Association
0.877 1 0.349
N of Valid Cases 62 - -
15
As it is the Pearson test of independence being conducted, the results being
considered will entail the Pearson Chi-Square row of values as they correspond with the
hypothesis test at hand. Furthermore, as standard with hypothesis testing, if the p-value is
greater than the value of alpha, one would accept the null hypothesis. With a calculated
p-value of .308 seen in Table 5 as the Pearson’s Chi-Square Asymp. Sig. it can determine
to not reject the null in the case of independence concerning O. pumilio frog’s gender and
their identified locations. This conclusion can be inferred from the prior established
alpha value of .05 being less than the p-value of .308. To further ascertain this
conclusion one can look to the critical value and test statistic of 2.354 and 5.991
respectively, additionally confirming the stated conclusion. In decisive summation of the
Pearson’s Chi-Square test of independence it can be deduced that at the 5% level of
significance there is sufficient evidence to support the null hypothesis and that the
locations of sighted O. pumilio frogs does in fact occur independent of their gender.
4.0 Conclusion
4.1 Discussion
In concluding that locations of sighted O. pumilio is independent of gender,
possible solutions for this occurrence should be examined. Based on the data collected,
we concluded that there is no association between gender and where the frogs are found,
leading us to analyze why both genders of the frog are distributed the way they are. Our
samples show that male and female frogs were both found more often in trees and on the
ground then on leaves. This could potentially be due to leaves representing far less
surface area within O. pumilio’s home range than trees or ground. It is interesting to note
16
however, that while frog location was independent of gender, both male and female frogs
appeared to occupy the same locations in relatively similar frequencies. This could
potentially be explained by how male distributions are influenced by female distributions,
often times with males being drawn to areas of high female density (Meuche, Linsenmair
& Prohl, 2012). To better understand why gender and locational distributions are situated
the way they are, it is important to analyze parental responsibilities as well as
reproductive and behavioral characteristics of the frog.
Location of frogs may be the result of many different stimuli. It may be related to
gender, which would also help explain why frogs are more frequently found at these sites.
This could help explain the duration male frogs spend courting females or the duration
female frogs spend transporting and caring for their tadpoles. Frogs use bromeliads axils
to deposit their tadpoles and when they return to feed their young. Therefore, it would be
a feasible hypothesis to believe that female frogs spend more time situated near
bromeliads than male frogs. However, frog location may be due to a predator response or
feeding, both, which are independent of gender. If behaviour was documented, it may
help explain why a frog is perched in a specific area. A better understanding of O.
pumilio home range and territory may also help us better understand perch location.
In total, male and female frogs were most often found on ground locations. This
concept could potentially be explained by some of the parental responsibilities the frog
possesses. Both sexes are involved with aspects of rearing young O. pumilio individuals.
Male and female frogs consistently return to tadpole rearing sites, often times located on
the ground next to small bodies of water, to feed young tadpoles nutritive material
(Murasaki, 2010). Consistent visitation to ground locations could potentially explain why
17
many of the frogs were found on ground locations. Another reason why both sexes were
commonly found on the ground could be due to reproductive factors. Oviposition sites
are mainly located within the dry leaf litter on the ground (Meuche, Linsenmair & Prohl,
2012). If the time frame happened to line up with the time courtship activities were taking
place between male and female frogs, the higher frequency ground observations make
sense, as both genders would be spending more time on the ground. Lastly, another
reason why ground observations were higher may have to do with the fact that O.pumilio
frogs require consistent moisture to survive. Many ground locations, such as beneath the
leaf litter or within the roots of trees provide such ideal locations (Meuche, Linsenmair &
Prohl, 2011). Perhaps frogs of both genders were found more often on the ground, as they
were seeking refuge from hot, dehydrating conditions.
Leaves were likely selected for the least by both genders due to their lack of
benefits for the frog. At times, it has been noted that females will raise tadpoles in the
axils of heliconias and other plants (Meuche, Linsenmair & Prohl, 2011). However, if
there was a more suitable body of water within our sampling location, it could be
suggested that the frogs would aggregate there instead. While gender was concluded to be
independent of location, female frogs were still found at the highest frequency when
located on trees. It has been suggested that female frogs move much more frequently
within their home range than do male frogs (Murasaki, 2010). Furthermore, movement
within female O. pumilio home range is typically to accomplish food collection or to
access refugium (Meuche, Linsenmair & Prohl, 2012). A common food source in which
O. pumilio feeds upon are ants, which are typically found in permanently humid
microhabitats. (Meuche, Linsenmair & Prohl, 2011) Such microhabitats can be found
18
around tree areas and could present a solution on why female distributions are situated
near trees. Another reason could be female frogs occupy tree sites as a way to escape
unfavorable conditions. This in turn could explain why male distributions were lower
around trees, as male frogs tend to remain more sedentary within their home range
(Murasaki, 2010). Perhaps the added energy needed to ascend trees are not worth the
benefit occupying tree sites provides.
4.2 Problems with current research
The observed location variable in this study vaguely describes O. pumilio’s
habitat. In tropical rainforests, the ground and trees both create their own microhabitats.
Better variables could have been used such as observations below 1m, between 1-3m and
above 3m. Large trees such as ficus trees where O. pumilio frequent are often over 100’
tall and the trunk buttress can be over 12’ wide. Frogs use trees for calling and to
transport tadpoles to water sources such as bromeliads. The ground offers a microhabitat
with leaf litter for them to lay their eggs, fallen trees and a wide variety of vegetation to
hide and feed on small insects.
4.3 Research Improvements
Further research should be conducted in the region and sampling frequency
should be increased to allow for in order to gather a larger data set. A larger sample size
is more representative of the population and it limits the risk of outliers. More plots could
be established in the Dixon forest and analyzed. Nearby islands could be surveyed in
order to compare results between different populations within the archipelago.
Radio telemetry would allow for researchers to monitor the frequency of time that
a frog spends at each location. Telemetry has been used in the past to research terrestrial
19
herps, including frogs. Small radio transmitters have been used to monitor O. pumilio
successfully (Murasaki, 2010). Recently internal radio transmitters have been
successfully used to study the home range of Hyla versicolour (Johnson et al, 2007). This
would allow for more consistent and accurate data collection as many observations are
missed during visual surveying (McGarrity et al, 2016). With the use of these
transmitters, researchers could monitor long term surveys, in order to determine why
frogs choose these perches. Observation of O. pumilio behaviour could also be
documented, such as feeding, courting (calling) or tadpole transporting. Recording
behaviour may give biologists a better understanding of why frog gender may be related
to perch location. Telemetry would also help gain valuable data on territory size and the
dispersal of metamorphosed tadpoles. However, due to the small size of these frogs,
battery lives will be greatly reduced and will have to be serviced more frequently.
20
5.0 Literature Cited
Galeano, SP., Harms, KE. 2016. Coloration in the polymorphic frog Oophaga pumilio
associates with level of aggressiveness in intraspecific and interspecific interactions.
Behavioural Ecology and Sociobiology 70(1): 83-97.
Institute of Tropical Ecology and Conservation. Bocas del Toro Biological Station. Map
2.
Lund, M. & Lund, A. (2013). Testing for Normality using SPSS Statistics. Retrieved
from https://statistics.laerd.com/spss-tutorials/testing-for-normality-using-spss-
statistics.php
McGarrity, Monica, Johnson Steven, 2016. Radio telemetry study of invasive Cuban tree
frogs. Gulf Coast research and Education Center, University of Florida.
McGrew, Jr. C., Lembo, Jr. A., & Monroe, C. B. (2014). An Introduction to Statistical
Problem Solving in Geography. Long Grove, IL: Waveland Press.
Meuche, I., Linsenmair, KE., Prohl, H. 2011. Female territoriality in the Strawberry
Poison Frog (Oophaga pumilio). Copeia (3): 351-356.
Meuche, I., Linsenmair, KE., Prohl, H. 2012. Intrasexual competition, territoriality and
acoustic communication in male strawberry poison frogs (Oophaga pumilio). Behavioural
Ecology and Sociobiology 66(4): 613-621.
Murasaki, S. 2010. Sex specific patterns of movement and space use in the strawberry
poison frog, Oophaga pumilio. FIU Electronic Theses and Dissertations. Paper 226.
Murasaki, 2010. Sex Specific Movements and Space use in Strawberry Poison Dart Frog
Oophaga pumilio. Florida International University.
Nowakowski, AJ., Jimenez, BO., Allen, M., Diaz-Escobar, M., Donelly, MA. 2013.
Landscape resistance to movement of the poison frog, Ophaga pumilio, in the lowlands
of northeastern Costa Rica. Animal Conservation 16(2): 188-197.
Related Documents
