The Effects and Importance of Predation on Immature Monarch Populations

Alma De Anda
University of Minnesota, St Paul, MN

Project Summary

Understanding predator-prey interactions and its impact on population dynamics has long been a focus of ecological studies. Monarch butterflies (Danaus plexippus) are excellent candidates for predation studies because of their multiple overlapping summer breeding generations. Eggs and larvae of these generations have low survival rates and their immobile or highly immobile early stages allows for detailed tracking of individuals and specific predators. After three summers of detailed tracking, I present preliminary data on factors affecting monarch survival and survival rates in the wild. These small-scale interactions and identifying how predators respond to changing prey densities will help to accurately estimate monarch survival in the wild, which can assist in conservation efforts. Also, quantifying interactions between predators and prey will allow for more accurate parameter estimation in predation models, the results of which can help our understanding of how models portray survival compared to what is observed in nature. My involvement with the Monarch Larvae Monitoring Project, a citizen science program that provides date and location-specific occurrence and abundance data for immature monarchs in 32 states and two Canadian provinces, will allow for my research to reach a broader audience and citizens could potentially contribute to my research.

Field Study

I tracked individual immature monarchs on common milkweed (Asclepias syriaca) and swamp milkweed (A. incarnata) plants during the 2004, 2006 and 2007 breeding seasons to determine causes of mortality. Here, I focus on the 2006 season, during which I visited a site in St. Paul, MN five days a week from 31-May through 28-July. This site is a community park that maintains a restored prairie area with naturally-occurring milkweed plants.

Methods

I selected three different areas within the park from which to collect data. My research team and I began monitoring at approximately the same time each day, and used aluminum tags to identify milkweed plants on which we observed naturally-laid monarchs. We monitored 237 milkweed plants for the presence of monarch individuals for the rest of the season, with a total of 2,526 observations of individual eggs and larvae.

We assigned a fate to each monarch observed during each observational period. Because we only observed on weekdays, I separated data by the time interval since the last observational period: either one day (Tuesday – Friday) or three days (Monday). Fates included:

still present on the milkweed plant
missing from the milkweed plant
"sucked out" (only chorion or exoskeleton remaining)
chewed
larva dead by miring in latex or with mouthparts glued shut by plant latex
observed predation event

The short time intervals between observations allowed us to assign fates to most "missing" individuals, because eggs and first instar larvae do not move off milkweed plants on their own. I assumed that missing eggs and first instar larvae were removed by predators that take the entire organism, such as ants. Eggs or larvae that had been sucked out were assumed to have been preyed on by organisms with sucking mouthparts, insects in the orders Hemiptera or Homoptera, or arachnids in the order Araneae. Chewed eggs or larvae were assumed to have been preyed on by insects in the orders Neuroptera, Coleoptera and possibly Hymenoptera. The spiders and insects mentioned above were all observed on milkweed plants throughout this breeding season as mentioned below.

In addition to monarch survival, I monitored potential predators on the plants, including spiders (Araneae), ants (Formica), lacewing larvae (Chrysopa spp), and larvae and adults of the exotic coccinellid, Harmonia axyridis. Figure 1 shows the total arthropod presence during the observational period, with Hymenoptera being the most abundant.

Figure 1. Total arthropod abundance on milkweed plants observed over the 2006 breeding season. Hymenoptera, specifically ants, were the most abundant arthropods. Ants are known predators of eggs and larvae and take the entire prey back to their brood.

I also recorded rough estimates of aphid density (mostly Aphis nerii) on the plants. Aphid presence may affect monarch survival in three ways. They are an alternative prey for many of the monarch predators that we observed, and could thus possibly provide some relief for monarch eggs and larvae on the same plants. An alternative effect may be due to A. nerii’s mutualistic relationship with ants (pers. obs.); in return for the honeydew that aphids provide to ants, the ants provide protection from predators. Thus, aphid presence may indirectly affect monarchs by increasing the presence of ant predators. Aphids may also have a negative impact on monarchs by decreasing the nutritional value of the milkweed plant.

I used a logistic regression analysis to determine the effects of 16 predictors on the fate of an individual monarch over 24 hours: date, start stage, position of an individual on a plant, number of other monarchs on the same plant, aphid density, herbivory, plant condition, milkweed density, average air temperature, average dew point, average precipitation and four categories of predator presence. Here, I focus on the analysis that combines eggs and larvae.

Preliminary Results

In 2006, we made 955 observations of eggs over 24-hour intervals. Of these, 364 had died, for a 24-hour survival rate of 62%. The egg stage of a monarch typically lasts about 4 days; extrapolating this 24 hour survival rate over a 4 day period gives a 15% survival rate from the time at which an egg is laid until the first instar. We observed 346 first instar larvae over 24-hour intervals, and found a 60% survival rate. Extrapolating this rate over the average two day first instar period results in a 36% chance of surviving from the first to the second instar, and a 5% survival rate from oviposition to the second instar. Movement after the second instar made similar calculations difficult for older larvae; I will address this problem with an observational and experimental study this summer. In all of our observations of larval mortality, only one miring death (a larva's mouthparts glued shut/glued to the plant) was observed throughout the breeding season, suggesting that this is not an important source of mortality in this population.

Five of the 16 predictors we tested have significant explanatory power for egg and all larval instars (Table 1 below).

Table 1. Logistic regression summary with the response survival and important predictors for monarch egg and larval survival.
Predictor	Estimate ± Standard Error	p value
Constant	-1.44 ± 0.481	0.0028
Air °C	0.0854 ± 0.0219	0.0001
Precipitation (mm)	-0.598 ± 0.108	<0.0001
Flowering Plant	-0.214 ± 0.0552	0.0001
log(# other monarchs)	-0.150 ± 0.0568	0.0082
Start Stage	-0.433 ± 0.0393	<0.0001

P-values are all < 0.05, indicating these predictors significantly affect survival and is not due to chance.

Survival is affected by average air temperature, precipitation, whether the milkweed plant is flowering or not, the number of other monarchs on the same plant, and the stage of the individual. The signs of the coefficients in Table 1 suggest several interesting interpretations.

More eggs and larvae survive on days with warmer temperatures.
Survival is negatively affected by rain, possibly because rain knocks immature monarchs off of milkweed.
The negative effect of flower presence may have two explanations, one biological and one due to observational errors:
1. Ants and other insects visiting flowering plants for nectar could increase the chance of predation.
2. It is difficult to find eggs and small larvae in the umbels containing 20-130 flowers (Rea et.al, 2003).
Survival is negatively affected by the presence of other monarchs on the same plant, which suggests that predation may be density dependent.
Younger individuals die more often than older individuals, possibly because they are less mobile and cannot escape predation, or simply because there are more predators that are capable of consuming smaller prey.

Identifying and understanding important factors that affect the survival of monarch butterflies is important for conservation efforts of this charismatic and well-known species. Looking at smaller scale interactions between insects and their environment can provide important information that could potentially be carefully extrapolated to improve the ability to predict the outcomes of alternative management strategies for organisms and their habitats.

Parasitism Study

Our lab has studied mortality rates due to a Dipteran parasitoid in the family Tachinidae, but there are few data on other monarch parasitoids, especially wasps, which often attack Lepidoptera as prepupae and pupae (Henneman et.al. 2001). Two superfamilies in the order Hymenoptera, the Chalicidoidea and the Ichneumonoidea (including the Braconidae and Ichneumonidae families), are known to parasitize lepidopteran larvae and pupae (Grimaldi and Engel, 2005), but there are only anecdotal reports of their importance to monarchs. I exposed lab-reared monarchs in their prepupal "J" stage, when they stop eating and find a pupation spot, to the natural environment near a milkweed patch and am currently quantifying the rate of parasitism by tachinid flies and other parasitoids.

Methods

I raised 400 monarch larvae from eggs laid in the laboratory to rule out the possibility of an individual becoming parasitized. I conducted this experiment during the mid and late breeding season in order to detect temporal differences in the rate of parasitism. They were allowed to "J" inside plastic containers and later fully exposed to natural conditions for 1 1/2 weeks, the average time it takes for pupae to become adults for the first experimental round (mid season). For the late season experimental round, I changed the methodology a bit, and instead of exposing them only when they reach the "J" stage, I allowed monarchs to be reared outside in mesh cages. This was done to give parasitoids possible cues of monarch presence, such as the detection of frass or plant chemical cues after it has incurred damage due to herbivory. Once the monarch larvae were ready to "J", I opened the cage and allowed them to be exposed to the environment. At each time period, I placed on average 100 larvae in each of four different locations to increase the chances of exposure to natural populations of parasitoids, assuming that the parasitoids may be patchily distributed. After the exposure time, I placed each pupa inside an individual container, labeled it and housed all containers in the laboratory at 25°C with a photoperiod of 16:8 (L:D) hours. Emerging parasitoids of monarch pupae will be identified and pinned for a laboratory inventory of monarch parasitoids.

Preliminary Results

The first round of the experiment resulted in 3 cases of parasitism: 1 unidentified fly species and 2 unidentified wasp species. The second round of the experiment had approximately 60 cases of parasitism, mostly due to a wasp species thought to be Pteromalus puparum (Pteromalidae), though proper identification is needed to verify this. I will quantify these results and perform statistical tests soon.

cages