Fewer infections found in forest fragments

As human populations expand, we are converting ecosystems from one state to another.  In the case of tropical forests, conversion of forest to cropland may leave behind fragments of relatively undisturbed forest surrounded by a matrix of cropland or other forms of development.  Conservation ecologists are exploring whether ecological processes and ecosystem structure in these fragments work pretty much like normal forested regions, or whether fragments behave differently.  To do this, in a few locations around the world such as the Wog Wog Fragmentation Experiment in New South Wales, Australia, researchers have systematically created forest fragments of various sizes.  They can then ask a variety of questions comparing fragments vs. intact forest. For example,  how does species diversity, or how do processes such as competition, predation and mutualism differ in the two landscapes?

aerial photo2

Aerial photo of Wog Wog Fragmentation Experiment at the time the experiment began in 1987. Credit: Chris Margules.

Julian Resasco was working as a postdoctoral associate in Kendi Davies’ lab at the University of Colorado on a study that looked at changes in invertebrate communities in response to fragmentation at Wog Wog. Beginning in 1985, researchers had set up a network of pitfall traps, which are cups that are buried with their tops level to the ground, so that any careless organism that wanders in will be trapped in the cup.  Some pale-flecked garden skinks, Lampropholis guichenoti, also had the misfortune to become entrapped and became subjects for the study. The invertebrates, and the 186 unfortunate skinks were preserved in alcohol and stored as part of the Australian National Wildlife Collection.

Museumspec

Skink museum specimens at the Australian National Wildlife Collection. Credit: Julian Resasco.

Much later, Resasco arrived and began dissecting skink guts to analyze the prey items for a study that looked at how the skinks shifted prey consumption (their feeding niche) in response to fragmentation. While dissecting the skink guts, he noticed that some of the skinks had worms (nematodes) inside their guts.  These nematodes were relatively common among skinks from continuous eucalypt forests, rare among skinks from eucalypt fragments, and absent from skinks in the cleared, pine plantation matrix.

ResascoFig1

Top. The study area in southeast Australia, showing location of continuous forest, forest fragments and surrounding matrix.  Dots indicate locations of pitfall traps. The matrix was planted in pine seedlings soon after fragmentation.  Bottom. The pale-flecked garden sunskink Lampropholis guichenotti. Credit: Jules Farquar

As it turned out, the nematode was a new species, which Resasco and a colleague (Hugh Jones) named Hedruris wogwogensis. Nematodes in the genus Hedruris use crustaceans as intermediate hosts, which alerted Resasco and his colleagues that the terrestrial amphipod Arcitalitrus sylvaticus, which was very common in the pitfall traps, was probably an important intermediate host.  When amphipods from pitfall traps were examined microscopically, a small portion of them were infected with Hedruris wogwogensis. The researchers concluded that amphipods became infected when they ate plants that harbored nematode eggs or young nematodes, which then developed in amphipod guts, and were passed on to skinks that ate the amphipods.  Thus somewhat inadvertently, one aspect of the study transitioned into the question of how fragmentation can influence the transmission of parasites.

After concluding their skink dissections, Resasco and his colleagues discovered that skinks in continuous forest had five times the infection rate as did skinks in fragmented forest.  In addition, no skinks collected in the matrix were infected. Infected skinks harbored a similar number of nematodes, whether they lived in continuous forest or fragments (see the Table below). Lastly, amphipods were considerably more common in skink guts and pitfall traps from continuous forest, less so in fragments, and least in the matrix.

ResascoTab1good

Summary of data collected by Resasco and his colleagues. Nematode intensity is the mean number of nematodes per infected skink. Nematode abundance is the mean number of nematodes per skink (infected and uninfected). 

The researchers put these findings together in a structural equation model.  The boxes in the model below represent the variables, while the numbers in smaller boxes over the arrows are the regression coefficients, with larger positive numbers (in black) indicating stronger positive effects, and larger negative numbers (in red) indicating stronger negative effects.  The model revealed three important findings.  First, habitat fragmentation strongly reduced amphipod abundance.  High amphipod abundance was associated with high nematode abundance (that is the +0.20 in the model), so lower amphipod abundance from fragmentation reduced nematode abundance. Second, habitat fragmentation positively affected skink abundance – more skinks were captured in fragments than in intact forest, but this increase had no effect on nematode abundance in skinks.  Finally, note the direct arrows connecting “Fragmentation” to “Log nematode abundance in skinks”.  This indicates that other variables (beside amphipod abundance) are reducing infection rates in skinks that live in fragments and the matrix.

ResascoFig3

Structural equation model showing effects of fragmentation on nematode infection in skinks. Amphipods are the intermediate host.  Black arrows indicate significant positive effects of one variable on the other, while red arrows indicate significant negative effects. Solid lines represent fragments compared to controls and dashed lines represent the matrix compared to controls. Thicker lines are stronger effects.

At this point, we still have an incomplete understanding of the system.  We know that fragmentation reduces amphipods, which require a moist and shaded environment to thrive.  Reduced amphipod abundance leads to lower nematode infection rates in skinks.  But we know that other variables are important as well; perhaps nematodes survive more poorly in fragment and matrix soils. Interestingly, pine trees were planted in the matrix and are beginning to mature and shade out the matrix environment. Amphipod abundances are on the rise, so the researchers predict that nematode infection rates will begin to increase accordingly.  Those studies have begun.

IMG_2646

Eucalypt forest canopy at Wog Wog. Credit: Julian Resasco.

Looking at the bigger picture, it is clear that fragmentation may decrease (as in this study) or increase the abundance of an intermediate host. As an example of fragmentation increasing intermediate host abundance, the researchers describe a study in which fragmentation increased the abundance of the white footed mouse, an intermediate host for black-legged ticks (that host the bacteria that causes Lyme disease). We need to unravel the connections between landscape factors and the various species they influence, so we can begin to understand how human changes to the landscape can influence the transmission of diseases.

note: the paper that describes this research is from the journal Ecology. The reference is Resasco, J.,  Bitters, M. E.,  Cunningham, S. A.,  Jones, H. I.,  McKenzie, V. J., and  Davies, K. F..  2019. Experimental habitat fragmentation disrupts nematode infections in Australian skinks. Ecology  100( 1):e02547. 10.1002/ecy.2547. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2019 by the Ecological Society of America. All rights reserved.

The complex life of the pea

As a behavioral ecologist, I’m spending a surprising amount of time reading and writing about plants these days.  It turns out that plants are amazingly complex and interactive; you just need to know where and how to look.  Today we will discuss the humble pea plant, how it is infected with a virus that is carried by an aphid that sucks its xylem, and how a herbivorous weevil fits into the whole system.   The virus is the pea enation mosaic virus (PEMV), which causes pea leaves to yellow and wither, and also creates enations (scaly tissue) to develop on a leaf’s undersides. The aphid vector (a vector is the organism that carries a disease) is Acrythosiphon pisum, while the herbivorous weevil is Sitona lineatus.

Picture1

Pea plant infected with pea enation mosaic virus. Credit Paul Chisholm.

David Crowder has been studying plant/insect interactions for many years, and knew that most researchers who studied interactions between plants and insect vectors focused their attention on the plants, insects and the disease, but did not consider how other species in the community might affect this relationship. Paul Chisholm was a PhD student in Crowder’s lab; and working with two other researchers, they explored whether S. lineatus, an abundant herbivore of peas, influenced viral transmission.  They expected that if the pea was first attacked by the weevil, it might be more susceptible to subsequent viral infection.  Conversely, if the pea was infected by the virus, it might be less able to chemically defend itself against subsequent herbivory by the weevil.

ChisholmFigAppendixBC

(Left) Pea aphids. Credit: Shipher Wu under a Creative Commons Attribution 4.0 International Public License. (Right) Very adult pea leaf weevils, Sitona lineatus. Credit: Gail Hampshire under a Creative Commons Attribution 4.0 International Public License

It is easy to visually distinguish between PEMV-infected and uninfected plants, so the researchers could assess whether infected plants tended to suffer more defoliation by weevils than did uninfected plants.  They visited 12 different fields in northern Washington and western Idaho, USA, and measured defoliation by counting the number of feeding notches left by the weevils after feeding on 3 – 10 infected pea plants and an equal number of nearby uninfected plants on each field (more feeding notches = more defoliation).  They discovered that PEMV-infected plants tended to suffer substantially higher herbivory than did uninfected plants

ChisholmFig1A

Herbivory (as measured by number of feeding notches) caused by weevils on paired uninfected (black bars) and PEMV-infected (blue bars) pea plants sampled from 12 different fields. Error bars for all figures represent 1 SE.

Given the correlation between herbivory and infection, Chisholm and his colleagues then explored whether (1) the weevil preferred to feed on infected plants, and/or whether (2) infective aphids preferred to feed on plants that had been damaged by herbivorous weevils.  Both questions were answered with behavioral choice assays done in a greenhouse. First, the researchers created two groups of pea plants.  The first group, sham infected, were fed on by aphids not carrying PEMV for 48 hours, while the second group of plants were fed on by PEMV-infected aphids for the same duration. Aphids were removed and PEMV infection developed within 15 days in the PEMV infected plants.  The researchers then set up one sham infected and one PEMV-infected plant in a test cage, and released two weevils equidistant from the plants, allowing them to feed for six days.  They discovered that aphids fed much more voraciously on the PEMV-infected plants.

ChisholmFig1B

Mean leaf area removed from sham infected and PEMV-infected pea plants.

For the second experiment, the researchers again created two types of plants: undamaged – no herbivory, and damaged – 48 hours of weevil herbivory.  Weevils were then removed, and one leaf from each plant was connected to each end of a tube, while still attached to each plant.

Picture9

Experimental setup with tube attached to one leaf of each experimental plant.  Aphids were introduced into the central tube. Credit Paul Chisholm. 

The researchers added either 25 infectious or 25 non-infectious aphids, and allowed them 3 hours to choose a leaf.  PEMV-infected aphids preferred damaged leaves, while uninfected aphids showed no preference.

ChisholmFig1C

PEMV-infected aphid preference for undamaged (no herbivory) or damaged (weevil herbivory) leaves.

Chisholm and his colleagues then turned their attention to whether weevil herbivory made pea plants more susceptible to PEMV infection.  In one experiment they allowed PEMV-infected aphids to feed on plants for 3 days, and then introduced 0, 1 or 3 weevils who fed on the plants for another 6 days.  They used a protein assay to estimate the PEMV-titer (concentration) of each plant and discovered that the plants that were exposed to greatest herbivory had the highest PEMV titer (see graph below).  In a second experiment the researchers allowed weevil herbivory before adding the aphids, and found no effect of prior herbivory on PEMV titer.

ChisholmFig2A

Relative PEMV-titer of infected leaves after they were subjected to herbivory by zero, one or three weevils for six days. different letters above bars indicate significant differences between treatments.

What causes these plant responses to challenges by PEMV and weevils?  The researchers discovered that levels of three important plant hormones increased either in response to PEMV infection, weevil herbivory or both.  At this point it is not clear how these different hormone levels interact to bring about the changes we’ve described.

The researchers conclude that weevil behavior has a profound influence on the interactions between aphids, the viruses they carry and the pea plants they feed on (and infect).  The weevil is not a vector for the virus, yet it affects the virus directly by altering plant behavior and physiology and indirectly by altering the behavior of the vector (the aphids).  PEMV outbreaks are more likely when weevils are abundant, as aphids prefer damaged plants, and feeding by weevils increased the PEMV titer in infected plants.  Crowder argues that interactions in which a non-vector species influences the relationship between a host and its vector (and the pathogen it carries) are probably extraordinarily common in crop systems.  So if we want to understand crop susceptibility to pathogens we need to cast a broad net and consider both the direct and indirect effects of a community of species that can influence how the crop responds to infection.

note: the paper that describes this research is from the journal Ecology. The reference is Chisholm, P. J., Sertsuvalkul, N. , Casteel, C. L. and Crowder, D. W. (2018), Reciprocal plant‐mediated interactions between a virus and a non‐vector herbivore. Ecology, 99: 2139-2144. doi:10.1002/ecy.2449. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2018 by the Ecological Society of America. All rights reserved.

Beautiful buds beset bumblebees with bad bugs

Sexual liaisons can be difficult to achieve without some type of purposeful motion.  Flowering plants, which are rooted to the ground, are particularly challenged to bring the male close enough to the female to have sex.  One awesome adaptation is pollen, technically the male gametophyte –  or gamete (sperm)-generating plant. These tiny males get to females either by floating through the air, or by being transferred by animal pollinators such as bees. Plants can lure bees to their flowers by producing nectar – a sugar rich fluid – which bees lap up and use as a carbohydrate source.  While nectaring, bees also collect pollen, either intentionally or inadvertently, which provides them with essential proteins. When bees travel to the next flower, they may inadvertently drop some of their pollen load near the female gametophyte – in this case a tiny egg-generating plant (though tiny, the female gametophyte is considerably larger than is the male gametophyte).  We call this process of “tiny boy meets tiny girl” pollination. Once the two gametophytes meet, the pollen produces one or more sperm, which it uses to fertilize an egg within the female gametophyte.  There is more to it, but this will hopefully clarify the difference between pollination and fertilization.

monardadidyma.jpg

Bumblebee forages on beebalm, Monarda didyma. Credit: Jonathan Giacomini.

All of this business takes place within the friendly confines of the flower.  The same flower may be visited by many different bees of many different species. While feeding, bees carry on other bodily functions, including defecation.  They are not careful about where they defecate; consequently a bee’s breakfast might also include feces from a previous bee visitor. Bumblebee (Bombus impatiens) feces carries many disease organisms, including the gut parasite Crithidia bombi, which can reduce learning, decrease colony reproduction and impair a queen’s ability to found new colonies. Because pollinators are so critical in ecosystems, Lynn Adler and her colleagues wondered whether certain types of flowers were better vectors for harboring and transmitting Crithidia bombi to other bumblebees.

Antirrhinummajus

Bumblebee forages on the snapdragon, Antirrhinum majus. Credit: Jonathan Giacomini.

The researchers chose 14 different flowering plant species, allowing uninfected bumblebees to forage on inflorescences (clusters of flowers) inoculated with a measured amount of Crithidia bombi parasites.  The bees were reared for seven days after exposure, and then were assessed for whether they had picked up the infection from their foraging experience, and if so, how intense the infection was. The researchers dissected each tested bee and counted the number of Crithidia cells within the gut.

researcher-photo.jpg

Researcher conducts foraging trial with Lobelia siphilitica inflorescence. Credit: Jonathan Giacomini.

Adler and her colleagues discovered that some plant species caused a much higher pathogen count (mean number of infected cells in the bee gut) than did other plant species.  For example bees that foraged on Asclepias incarnata (ASC) had four times as many pathogens, on average, than did bees that foraged on Digitalis purpurea (DIG) (top graph below). Bees foraging on Asclepias were much more likely to get infected (had greater susceptibility) than bees that foraged on several other species, most notably Linaria vulgaris (LIN) and Eupatorium perfoliatum (EUP) (middle graph). Lastly, if we limit our consideration to infected bees, the mean intensity of the infection was much greater for bees foraging on some species, such as Asclepias and Monarda didyma (MON) than on others, such as Digitalis and Antirrhinum majus (ANT) (bottom graph).

AdlerFig1

(Top graph) Mean number of Crithidia (2 microliter gut sample) hosted by bees after foraging on one of 14 different flowering plant species. This graph includes both infected and uninfected bees. (Middle graph) Susceptibility – the proportion of bees infected – after foraging trials on different plant species. (Bottom graph) Intensity of infection – Mean number of Crithidia for infected bees only. The capital letters below the graph are the first three letters of the plant genus. Numbers in bars are sample size.  Error bars indicate 1 standard error.

It would be impossible to repeat this experiment on the 369,000 known species of flowering plants (with many more still to be identified).  So Adler and her colleagues really wanted to know whether there were some flower characteristics or traits associated with plant species that served as the best vectors of disease.  The researchers measured and counted variables associated with the flowers, such as the size and shape of the corolla, the number of open flowers and the number of reproductive structures (flowers, flower buds and fruits) per inflorescence.

bluelobelia.png

Flower traits measured by Adler and colleagues (example for blue lobelia, Lobelia siphilitica). CL is corolla length. CW is corolla width. PL is petal length. PW is petal width. Credit: Melissa Ha.

The researchers also wanted to know whether any variables associated with the bees, such as bee size and bee behavior, would predict how likely it was that a bee would get infected.  Surprisingly, the number of reproductive structures per inflorescence stood out as the most important variable. In addition, smaller bees were somewhat more likely to get infected than larger bees, and bees that foraged for a longer time period were more prone to infection.

AdlerFig2

Mean susceptibility of bees to Crithidia infection after foraging on 14 different flowering plant species, in relation to the number of reproductive structures (flowers, buds and fruits) per inflorescence.

These findings are both surprising and exciting. Adler and her colleagues were surprised to find such big differences in the ability of plant species to transmit disease.  In addition, they were puzzled about the importance of number of reproductive structures per inflorescence.  At this point, they don’t have a favorite hypothesis for its overriding importance, speculating that some unmeasured aspect of floral architecture influencing disease transmission might be related to the number of reproductive structures per inflorescence.

Penstemondigitalis

Bumblebee forages on Penstemon digitalis. In addition to the open flowers, note the large number of unopened buds.  Each of these counted as a reproductive structure for the graph above. Credit: Jonathan Giacomini.

The world is losing pollinators at a rapid rate, and there are concerns that if present trends continue, there may not be enough pollinators to pollinate flowers of some of our most important food crops. Disease is implicated in many of these declines, so it behooves us to understand how plants can serve as vectors of diseases that affect pollinators. Identifying floral traits that influence disease transmission could guide the creation of pollinator-friendly habitats within plant communities, and help to maintain diverse pollinator communities within the world’s ecosystems.

note: the paper that describes this research is from the journal Ecology. The reference is Adler, L. S., Michaud, K. M., Ellner, S. P., McArt, S. H., Stevenson, P. C. and Irwin, R. E. (2018), Disease where you dine: plant species and floral traits associated with pathogen transmission in bumble bees. Ecology, 99: 2535-2545. doi:10.1002/ecy.2503. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2018 by the Ecological Society of America. All rights reserved.

A tale of too many ticks

Many people I know have had the unfortunate experience of a warm season bout with the following symptoms: fatigue, achy joints, headaches, dizziness, fever and night sweats. Some of these symptoms are part of the daily experience of someone who has reached my level of maturity (okay – age), but in combination they suggest infection by the bacterium Borrelia burgdorferi that is transmitted by Ixodes ticks, and causes Lyme disease.  So three years ago, when I experienced those symptoms, I went off to my doctor (after some prodding by my wife) who immediately prescribed a regime of antibiotics that is effective against Lyme. My region of the United States (southern Appalachians) is a center of Lyme infection, so the diagnosis was pretty easy, and thankfully, the antibiotics were effective.

Lyme 2016

Each dot represents one verified case of Lyme disease in the United States in 2016.  I live in the dark blotch in western Virginia.

Richard Ostfeld began investigating the ecology of Lyme disease as a result of a chance event.  About 26 years ago Ostfeld started a new project that explored how white-footed mice may control populations of the invasive forest pest, the gypsy moth.  Mice eat the moth pupae for a couple of weeks in mid-summer.  When he started trapping at the Cary Institute of Ecosystem Studies in New York, he was amazed to see tremendous burdens of larval blacklegged ticks attached to the white-footed mouse (Peromyscus leucopus).  At the field site there was a boom one year and a crash the following year in acorn abundance, which was followed, with a one year time lag, by a boom and a crash in mouse abundance.  Ostfeld wondered what role fluctuating mouse abundance might play in human risk of exposure to tick-borne disease, and how factors affecting mouse abundance might influence the system.

Mouse with 52 larval ticks closeup

This unfortunate mouse harbors 52 larval ticks. Credit: Ostfeld lab at Cary Institute.

Ixodes ticks have a two-year lifecycle, with eggs laid in the spring, six-legged larvae hatching out in summer, getting one blood meal from a rodent or bird host, and emerging as eight-legged nymphs the following spring.  Nymphs find themselves a second host in spring or summer, from which they suck more blood and ultimately metamorphose into adults during the fall season. Adults seek large mammalian hosts, such as white-tailed deer; females feed on the deer, mate with males (who generally don’t feed), lay eggs and die, usually the following spring.

img_7791.jpg

Human finger with (left to right) adult female, adult male, nymph and larval ticks. Credit: Ostfeld lab at Cary Institute.

What makes these ticks tick? Ostfeld, Taal Levi and their colleagues knew from previous work that biotic factors such as mice, acorns and deer were likely to be important, but that predators on mice might also play a role.  It also seemed likely that abiotic factors such as temperature, moisture and snow cover could also be important.  For 19 years, the researchers systematically collected data related to these factors from six large (2.25 ha) field plots at the Cary Institute. They used standard capture-mark-recapture methods to estimate rodent abundance, and data from the Cary Institute’s bow-hunting program to estimate deer abundance. They monitored the presence of carnivores with LED camera traps that were baited with cans of cat food.

coyote-camera-trap.jpg

Coyote captured on LED camera. Credit: Ostfeld lab at Cary Institute.

Lastly, the researchers needed to estimate tick abundance and the percentage of ticks that were infected with the Lyme disease bacterium, Borrelia burgdorferi.  To estimate tick abundance, the researchers systematically dragged 1-m2 white corduroy drag cloths across each plot every three weeks throughout the times of peak tick abundance. Ticks that are searching for a host (known as questing ticks) will grab onto the drag cloth, so in essence, drag cloth censuses provides an estimate of ticks that have not had a blood meal.  Tick infection rates were estimated by subjecting an average of 378 ticks per year to molecular analyses (initially direct immunofluorescence assay, and later quantitative PCR).

tick

Researchers sample for questing ticks by dragging a cloth across the forest floor. Credit: Ostfeld lab at Cary Institute.

Across the 19 years of the study, the density of infected nymphs was strongly correlated to mouse density the previous year, and weakly correlated with deer density two years previously.  Recall the details of the two-year life cycle; it takes a year to go from tick larva to nymph, and a second year to go from nymph to adult to eggs, so these time lags are not surprising. What is surprising is that the density of infected nymphs is negatively correlated with mouse density in the current year and with winter warmth.

OstfeldFig2

Density of infected ticks (x 100) per 100 m2 in relation to (far left) mouse density (per 2.25 ha) in the previous year, (2nd from left) mouse density in the current year, (2nd from right) winter warmth, and (far right) deer density two years previous.  Different color dots represent the six different field sites.

Ostfeld and his colleagues explain that during years of high mouse abundance, many nymphs were attached to rodent hosts, or had already had a blood meal, and thus were not collected on drag cloths. By using the abundant rodents as their secondary hosts, rather than people, high rodent abundance is actually decreasing the probability that the nymphs will infect a human. Infection of humans by adult ticks is less common than infection of humans by nymphs, because many nymphs don’t survive to adulthood, male adults do not feed, and adults are more likely than nymphs to be spotted and removed, due to their larger size.

Nymphal infection prevalence (NIP) measures the fraction or proportion of the nymphs within the community that are actually carrying the bacterium.  From a human perspective, a high NIP indicates that a tick bite is relatively likely to lead to Lyme disease. There was only a small relationship between rodent density the previous year and NIP, so the researchers decided to see if the composition of the predator community might influence NIP. They reasoned that foxes and bobcats were known to be major mouse predators, so by eating mice, they would be removing infected ticks from the population.  Raccoons and opossums have a double effect; they eat mice – though not as many as do foxes and bobcats.  In addition they are dilution hosts, in that they provide blood for nymphs, but do not serve as a vector to the bacterium.  Thus a community with all four of these predators was expected to reduce NIP. The effect of coyotes were more complex because they eat mice, which should reduce NIP, but they also eat or scare away other predators, such as foxes and opossums, which could increase NIP.

OstfeldFig4

Effect size of predator community structure on nymphal infection prevalence (NIP).  Top row animals are (left to right) fox, raccoon, opossum and bobcat.  Communities with coyotes (bottom five communities) tend to have higher NIP, particularly if they lack other predators.

In general, more diverse predator communities tended to have lower nymphal infection prevalence.  Communities with coyotes that also lacked some of the other predators tended to have the highest NIP values.

Ostfeld and his colleagues were surprised to discover that a warm and dry winter and spring season tended to depress tick abundance, while cold winters had little effect. Presumably, emerging nymphs can dry out under warm, dry conditions. The researchers were also surprised to observe the strong decrease in tick abundance associated with high mouse abundance in the current year. It is not uncommon for a boom in mouse abundance one year to be followed by a mouse population crash the next year.  When that occurs, there will be a large number of questing nymphs lurking in the vegetation for hosts, and thus the potential for a major outbreak of Lyme disease.

note: the paper that describes this research is from the journal Ecology. The reference is Ostfeld, R. S., Levi, T. , Keesing, F. , Oggenfuss, K. and Canham, C. D. (2018), Tick‐borne disease risk in a forest food web. Ecology, 99: 1562-1573. doi:10.1002/ecy.2386. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2018 by the Ecological Society of America. All rights reserved.

Fungi attack plants – insects respond!

As she was preparing to do her dissertation research on the interactions between the Asian chestnut gall wasp, the chestnut blight disease and the European chestnut, Pilar Fernandez-Conradi read a lot of papers about fungal-insect-plant interactions.  She was impressed by the diversity of outcomes that resulted when plants were attacked by both insects and fungi, and wondered whether there were any generalities to glean from these research findings. She asked two basic questions. First, if a plant is infected by a fungus, is it more or less likely to be attacked by insects than is an uninfected plant?  Second, does an insect that attacks a fungal-infected plant perform better or worse than it would have on an uninfected plant?

D. Kuriphilus+Gnomo

Three-way interaction between the chestnut tree, the chestnut gall wasp, and the fungus Gnomopsis castanea. Female wasps induce the plant to create galls, which house developing larvae. Green globular galls (with a hint of rose-color) have not been infected by a fungus, while the very dark tissue is the the remains of a gall that was attacked by the fungus. Credit: Pilar Fernandez-Conradi.

Fernandez-Conradi and her colleagues thought they were more likely to discover a negative effect of fungal infection on the preference and performance of herbivorous insects.  Several studies had shown that nutrient quantity and quality of host plants is reduced by fungal infection, so it makes sense that insects would avoid infected plants.  But the researchers also knew that fungal infection can, in some cases, actually increase the sugar concentration of some plants, so insects might prefer those plants and also develop more rapidly on them. In addition, fungal infection can induce chemical defenses in plants that might make them less palatable to insects, or alternatively, fungal infection could weaken plant defenses making them more palatable to attacking insects.

To resolve this conundrum, Fernandez-Conradi and her colleagues did a meta-analysis, of the existing literature, identifying 1113 case studies based on 101 papers.  To be considered in the meta-analysis, all of the studies had to meet the following criteria: (1) report insect preference or performance on fungal-infected vs. uninfected plants, (2) report the Genus or species of the plant, fungus and insect, (3) report the mean response and a measure of variation (standard error, standard deviation or variance). The measure of variation allows researchers to calculate the effect size, which calculates the strength of the relationship that is being explored. The researchers found that, in general, insects avoid and perform worse on infected plants than they do on uninfected plants.

Fernandez-conradi-myfig

Mean effect size of insect preference and performance (combined) in response to fungal infection infection.  Error bars are 95% confidence intervals (CIs).  In this graph, and the next two graphs as well, a solid data point indicates a statistically significant effect.  You can also visually test for statistical significance by noting that the error bar does not cross the dashed vertical line that represents no effect (at the 0.0 value). The negative value indicates that insects respond negatively to fungal infection.

Fernandez-Conradi and her colleagues then broke down the data to explore several questions in more detail. For example, they wondered if the type of fungus mattered.  For their meta-analysis, they considered three types of fungi with different lifestyles: (1) biotrophic pathogens that develop on and extract nutrients from living plant tissues, (2) necrotrophic pathogens that secrete enzymes that kill plant cells, so they can develop and feed on the dead tissue, and (3) endophytes that live inside living plant tissue without causing visible disease symptoms.

Fernandez-conradiFig1

Effect of fungus lifestyle on insect performance.  k = the number of studies.  Different letters to the right of CIs indicate significant differences among the variables (lifestyles).

The meta-analysis showed an important fungus-lifestyle effect (see the graph to your left).  Insect performance was strongly reduced in biotrophic pathogens and endophytes, but not in necrotrophic pathogens, where insect performance actually improved slightly (but not significantly). The researchers point out that biotrophic pathogens and endophytes both develop in living plant tissues, while necrotrophic pathogens release cell-wall degrading enzymes which can cause the plant to release sugars and other nutrients.  These nutrients obviously benefit the fungus, but can additionally benefit insects that feed on the plants.

To further explore this lifestyle effect, Fernandez-Conradi and her colleagues broke down insect response into performance and preference, focusing on chewing insects, for which there were the most data. Insects showed lower performance on and reduced preference (i.e. increased avoidance) for plants infected with biotrophic pathogens. They also performed equally poorly on endophyte-infected plants, but did not avoid endophyte-infected plants (see graph below). This was surprising since you would expect natural selection to favor insects that can choose the best plants to feed on. The problem for insects may be that endophytic infection is basically symptomless, so the insects may, in many cases, be unable to tell that the plant is infected, and likely to be less nutritionally rewarding.

Fernandez-conradiFig2

Effects of fungal infection on preference and performance of chewing insects.  k = the number of studies.  Different letters to the right of CIs indicate significant differences among the variables. Variables that share one letter have similar effect sizes.

Many ecological studies deal with two interacting species: a predator and a prey, or a parasite and its host.  Fernandez-Conradi and her colleagues remind us that though two-species interactions are much easier to study, many important real-world interactions involve three or more species. Their meta-analysis highlights that plant infection by pathogenic and endophytic fungi reduces the performance and preference of insects that feed on these plants. But fungus lifestyle plays an important role, and may have different effects on performance and preference. Their meta-analysis also suggests other related avenues for research.  For example, how are plant-fungus-insect interactions modified by other species, such as viruses, bacteria and parasitoids (an animal that lives on or inside an insect, and feeds on its tissues)? Or, what are the underlying molecular (hormonal) mechanisms that determine the response of the plant to fungal infection, and to insect attack?  Finally, how does time influence both plant and insect response?  If a plant is recently infected by a fungus, does it have a different effect on insect performance and preference than does a plant that has suffered from chronic infection.  There are very few data on these (and other) questions, but they are more likely to get pursued now that some basic relationships have been uncovered.

note: the paper that describes this research is from the journal Ecology. The reference is Fernandez‐Conradi, P., Jactel, H., Robin, C., Tack, A.J. and Castagneyrol, B., 2018. Fungi reduce preference and performance of insect herbivores on challenged plants. Ecology, 99(2), pp.300-311. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2018 by the Ecological Society of America. All rights reserved.

Powdery parasites pursue pedunculate oak

Studying disease transmission is tricky for many reasons. Most humans frown on what might seem like the easiest experimental protocol – release a disease into the environment and watch to see how it spreads. For his doctoral dissertation in 2006, Ayco Tack settled on a different experimental protocol – bring the potential hosts to the disease. In this study, staged in Finland, the hosts were pedunculate oak trees, Quercus robur, and the disease was the powdery mildew parasite, Erysiphe alphitoides. Almost 10 years later, Adam Ekholm continued research on the same system, with Tack as his co-supervisor.

Ayco Tack

Trees on the move. Credit: Ayco Tack.

But before moving trees around, the researchers first needed to see how the disease moved around under field conditions.  Within a tree stand, powdery mildew success will depend on how many trees it occupies, how many trees it colonizes in the future, and how many trees it disappears from (extinction rate). The researchers measured these rates over a four year period (2003 – 2006) on 1868 oak trees situated on the island of Wattkast in southwest Finland. They also measured spatial connectivity of each tree to others in the stand. In this case connectivity is a measure of the distance between a tree and other trees, weighted by the size of the other trees. So a tree that has many large neighbors nearby has high connectivity, while a tree with a few distant and mostly small neighbors has low connectivity. Results varied from year-to-year, but in general, the researchers found higher infection rates, lower extinction rates, and some evidence of higher colonization rates in trees with high connectivity.

Mildew_Adam Ekholm

Oak leaf infected with powdery mildew parasite. Credit: Adam Ekholm.

The importance of connectivity indicated that the parasites simply could not disperse efficiently to distant trees. But perhaps the environment might play a role in colonization rates as well. For example, fungi like powdery mildew tend to thrive in shady and humid environments. Thus a tree out in the open might resist colonization by powdery mildew more effectively than would a tree deep in the forest. To test this hypothesis, Tack and his colleagues placed 70 trees varying distances (up to 300 meters) from an infected oak stand. On one side of the oak stand was an open field, while the other side was closed forest. Thus two variables, distance and environment, could be investigated simultaneously.

Ayco Tack inspecting a potted tree_Tomas Roslin

Ayco Tack inspects an oak tree placed in an open field. Credit: Tomas Roslin.

The researchers collected infection data twice; once in the middle of the growing season (July) and a second time at the end of the growing season (September). Not surprisingly, infection rates were higher by the end of the growing season. In general, infection rates, and infection intensity (mildew abundance) were higher in the forest than in the field, indicating a strong environment effect. In the July survey, trees further from the oak stand had lower infection intensity, but as infection rates increased over the course of the season, the effects of distance diminished, particularly in the forest.

EkholmFig3

Upper two graphs show the impact of habitat type on (a) proportion of trees infected and (b) mildew abundance. The lower two graphs are the influence of distance from parasite source on mildew abundance of trees set in (c) a forest habitat and (d) an open field. Mildew abundance was scored on an ordinal scale with 0 = none and 4 = very abundant.

Ten years later, Adam Ekholm, as part of his PhD dissertation that studies the effect of climate on the insect community on oak trees, added a third element to the mix – the influence of genes on disease resistance. He wondered whether certain genotypes were more resistant to powdery mildew infection. The researchers grafted twigs from 12 large “mother” trees, creating 12 groups of trees, with between 2 – 27 trees per group (depending on grafting success). Each tree in a given group was thus genetically identical to all other trees within that group.

Ayco Tack

Oak tree placed in the forest. Credit: Ayco Tack.

The researchers chose a site that contained a dense stand of infected oaks, but was surrounded by a grassy matrix that contained only an occasional tree. To study the impact of early season exposure, Ekholm and his colleagues divided the trees into two groups; 128 trees were placed in the matrix at varying distances from the infected stand, while 58 trees were placed directly in the midst of the stand for about 50 days, and then moved varying distances away. The researchers scored trees for infection at the end of the growing season (mid-September).

 

Trees that spent 50 days within the oak stand had much higher infection frequency and intensity than trees that were initially placed in the matrix. Some genotypes (for example genotype I in graphs C and D below) were much more resistant to infection than others (such as genotypes D and J). Finally trees further from the source of infection were less susceptible to become colonized over the course of the summer (data not shown).

EkholmFig2

Proportion of trees infected (A) and proportion of leaves infected (B) in response to early season exposure to stand of oaks infected with the powdery mildew parasite (oak stand) or no early season exposure (matrix). Proportion of trees infected (C) and proportion of leaves infected (D) in relation to tree genotype. Genotypes are labeled A – L; numbers in parenthesis are sample size for each group.

These findings illustrate how dispersal, host genotype and the environment influence the spread of a parasite under natural conditions. The parasite exists as a metapopulation – a group of local populations inhabiting networks of somewhat discrete habitat patches. Some populations go extinct while others successfully colonize each year, depending on distance from a source, tree genotype and environment. Ekholm and his colleagues encourage researchers to use similar experimental approaches in other host-parasite systems to evaluate how general these findings are, and to explore how multiple factors interact to shape the dynamics of disease transmission.

note: the paper that describes this research is from the journal Ecology. The reference is Ekholm, Adam; Roslin, Tomas; Pulkkinen, Pertti and Tack, Ayco. J. M. (2017). Dispersal, host genotype and environment shape the spatial dynamics of a parasite in the wild. Ecology. doi:10.1002/ecy.1949. The paper should come out in print very soon. Meanwhile you can also link to Dr. Tack’s website at www.plantmicrobeinsect.com Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2017 by the Ecological Society of America. All rights reserved.

Frogs face fatal fungal foes

Pathogens are organisms that cause disease, and like all organisms, they obey evolutionary principles. Pathogens that survive and reproduce successfully in a particular environment will have more offspring than those that are less successful, thereby passing on those traits that promote successful reproduction to future generations. The problem is that many pathogens change their environment in a way that makes their environment less hospitable for their own survival or reproduction. For example, the fungal pathogen Batrachochytrium dendrobatidis (Bd) causes chytridiomycosis in its amphibian host, which may severely reduce the host population size to the point where few individuals survive. If the host population goes extinct, then there are no hosts for the fungal offspring to infect.

Scheele fungal spore

Scanning electron micrograph of Batrachochytrium denbdrobatidis spore. Credit: Dr. Alex Hyatt, CSIRO Livestock Industries’ Australian Animal Health Laboratory.

Fortunately for Bd, but unfortunately for amphibians, there are several ways out of this conundrum. One approach is a reduction in pathogenicity so that a pathogen’s host species is able to tolerate the infection (and of course, natural selection will at the same time favor an increase in the host species’ tolerance for the pathogen). A second approach is to broadcast a wide net by infecting many different species. That way if one host species goes extinct, there are always many other species to infect. Bd infects over 500 species of amphibians, and has been implicated in the extinction of over 100 amphibian species, and the severe decline of an additional 100 species.

Ben Scheele and his colleagues wanted to know why the endangered northern corroboree frog, Pseudophryne pengilleyi, was declining in southeastern Australia. Several previous studies showed that many corroboree frog populations declined or went extinct in that region over the past 20 years, while the abundant common eastern froglet, Crinia signifera, showed no signs of decline over the same time period. Pilot studies showed that eastern froglets were heavily and commonly infected with Bd. The researchers reasoned that eastern froglets could be acting as a reservoir for Bd, so that corroboree frog populations are being decimated by association with Bd-infected eastern froglets.

Female Ppen copy Hunter

Female Pseudophryne pengilleyi. Credit: David Hunter.

Preliminary surveys indicated that the decline of corroboree frogs was not uniform across the study site; in fact there were some newly discovered populations that were doing very well. The researchers defined three types of sites in their research area. Absent sites (40 in total) had corroboree frogs in 1998, but the population went extinct by 2012. Declined sites (17 in total) had a greater than 80% decrease in abundance since 2000. New sites (25 in total) were newly discovered since 2012, and had much higher population densities than declined sites.

scheele-fig-1.png

Study area in southeastern Australia, showing locations of Absent, Declined and New sites.

Unfortunately, it is impossible to visually distinguish an infected frog from an uninfected frog, at least until the few hours before death. But the researchers needed to be able to tell if a frog had chytridiomycosis. So they collected skin swabs from the frogs during the breeding season – only working at night to ensure cool humid conditions which minimized frog stress. They then did real time PCR on these samples to quantify the intensity of Bd infection.

Scheele and his colleagues had three important questions they were now prepared to answer. First, how prevalent is Bd in these two species? They found that infection rate was much higher in eastern froglets (79.4%) than in corroboree frogs (27.3%). The intensity of infection (measured by the number of fungal spores) was also much greater in eastern froglets than in corroboree frogs.

Second, do eastern froglets act as a reservoir for Bd, leading to infection and decline of corroboree frog populations? As we discussed earlier, the two species coexist at some sites, but not at others. If eastern froglets act as a reservoir for Bd, we would expect corroboree frogs to have higher infection rates at sites they share with eastern froglets, than they do at sites without eastern froglets. In support of this prediction, Bd prevalence in corroboree frogs was 41.4% at sites with eastern froglets, but only 2.6% at sites with no eastern froglets.

crinia and pengilleyi 3

C. signifera (left) and P. pengilleyi spending quality time together in a P. pengilleyi nest. Credit: David Hunter.

Finally, the researchers want to identify conditions that will promote corroboree frog recovery. They approached this quantitatively by modeling the probability of a site being classified as Absent, Declined or New, in relation to eastern froglet abundance. Based on their survey data of 81 sites, those sites with the highest eastern froglet abundance are most likely to be classified as Absent (corroboree frog extinction), while sites with very few eastern froglets are most likely to be classified as New (thriving corroboree frog populations).

scheele-figure-2.png

Probability of a site being classified as Absent, Declined or New, based on eastern froglet abundance. Data are log transformed. Dashed lines are 95% confidence intervals.

Scheele and his colleagues conclude that eastern froglets are a reservoir host for Bd, and have played a major role in the decline in corroboree frog populations. The researchers point out that, in general, areas lacking reservoir hosts may provide endangered species with refugia from infectious disease. For managing endangered species, conservation biologists should carefully monitor sites for the presence of reservoir hosts so they don’t reintroduce rare and endangered animals into locations where they will be attacked and killed by pathogens.

note: the paper that describes this research is from the journal Conservation Biology. The reference is Scheele, Ben C., David A. Hunter, Laura A. Brannelly, Lee F. Skerratt, and Don A. Driscoll. “Reservoir‐host amplification of disease impact in an endangered amphibian.” Conservation Biology 31, no. 3 (2017): 592-600. Thanks to the Society for Conservation Biology for allowing me to use figures from the paper. Copyright © 2017 by the Society for Conservation Biology. All rights reserved.