Cyclic population dynamics of the autumnal moth: biology and importance of parasitic wasps
​Host (Epirrita autumnata) and its enemy (Agrypon flaveolatum)

Study sites

Our main field sites are located in the northernmost Finland and Norway, where the Kevo Subarctic Research Station (administered by the University of Turku) provides good facilities. Southern, non-outbreaking populations of the autumnal moth are studied in the vicinity of Turku, SW Finland.

The Species

The autumnal moth, Epirrita autumnata (Lepidoptera: Geometridae), is a widely distributed and common geometrid throughout most of the Holarctic region. In northern and mountainous parts of Fennoscandia, autumnal moth populations have been shown to display cyclic (with a statistically significant 9–10-year periodicity), high-amplitude fluctuations in densities, which may culminate in devastating outbreak densities for 1–3 successive years. In these areas, the mountain birch (Betula pubescens ssp. czerepanovii) is the main host for larvae, and vast areas of the subarctic mountain birch zone can be severely damaged, or even killed, during outbreaks. Autumnal moth populations in southern Fennoscandia remain generally at low and stable densities. Recently (2006-2009), winter moth (Operophtera brumata) populations have also reached outbreak densities, for the first time, in Finnish Lapland. Both autumnal and winter moths are obligatorily univoltine. Larvae hatch in the spring simultaneously with the budding of a host plant. Leaf-chewing larvae feed on foliage through five larval instars, descending from host trees by mid-summer and pupating in the soil. Adults fly in autumn and eggs overwinter.


Our project, focusing on cyclic geometrid defoliators, continues rich tradition of the Section of Ecology at the University of Turku. Earlier projects, ever since 1970s, have looked for an explanation for cyclic population dynamics e.g. from intrinsic changes of autumnal moth populations at successive cycle phases and particularly from multifaceted interactions with the main host plant, the mountain birch. However, the working hypothesis of our current project suggests that delayed density-dependent mortality due to egg, larval and/or pupal parasitism (introduced by parasitic wasps of the order Hymenoptera) generates the cyclic population dynamics of the autumnal moth. Thus, we concentrate on the parasitism in detail to obtain understanding of the biology of parasitoid species involved in the system. Their two- and three- trophic-level interactions with autumnal moths and mountain birches are examined by large-scale field samplings and by laboratory and field experiments. Together with taxonomists, we also explore taxonomy of these relative poorly known insects.

Autumnal moth populations exhibit cyclic and (fairly) synchronous fluctuations in densities across Fennoscandia

Our statistical analyses have shown that density fluctuations of autumnal moth populations are cyclic in the whole of Fennoscandia (Klemola et al. 2002, Oikos 99:83–94; Klemola et al. 2006, Oikos 114:349–359). The longest time series available on larval counts revealed cycle lengths from 9 to 10 years. Shorter time series displayed higher variation in the cycle length, but the 10-year cycle seemed to prevail also in these. In northern and mountainous parts of Fennoscandia, the difference between peak and bottom phases in autumnal moth densities is >10 000-fold, while that observed in southern Finland is only ten-fold at maximum.
Large-scale synchrony characterises density fluctuations of distinct autumnal moth populations. Peak densities and troughs have been observed to be spatially synchronous over very large areas, at least within a time window of 2–3 years (see also Tenow et al. 2007, J. Anim. Ecol. 76:258–268). However, a closer look to dynamics in northern Fennoscandia revealed that three separate clusters exist; cyclic peaks in autumnal moth abundance coincided within a given cluster but lagged 1–2 years in the adjacent cluster (Klemola et al. 2006, Oikos 114:349–359). This cluster partitioning was supported by the geographic and climatic structuring of northern Fennoscandia. Within regional clusters, moth populations were probably exposed to the synchronizing effects of common, spatially autocorrelated environmental conditions, i.e., a Moran effect. Thus, we concluded that a high amount of environmental variation may result in a clear structuring of spatial synchrony across even relatively small scales. In Fennoscandia, the Gulf Stream, together with a northern location, could cause such large environmental and climatic variation at scales of only a few hundred kilometres.

Parasitoid assemblage attacking the autumnal moth

According to our studies, eggs, larvae and pupae of the autumnal moth are all attacked by hymenopteran parasitoids. We know one true egg parasitoid, belonging to the genus Telenomus (family Scelionidae). Adults of this species are only 0.6–0.7 mm in length. The species was common during the decline phase of the latest population cycle. In addition, we have occasionally sampled specimens of a polyembryonic egg-prepupal parasitoid, Copidosoma chalconotum (Encyrtidae). However, this species seems to be rare in outbreaking, northern populations.

Larvae of the autumnal moth are attacked by approximately 15 different species. From larval parasitoids, endoparasitoids belonging to families Ichneumonidae and Braconidae are the most commonly met ones. Early-larval parasitoid species, such as braconids Cotesia salebrosa, C. autumnatae, Protapanteles anchisiades, P. immunis and Aleiodes gastritor as well as ichneumonids Phobocampe sp. typically parasitize on larvae during the third instar and onwards. Late-larval parasitoid species, such as Zele deceptor (Braconidae) and Campoletis varians Ichneumonidae), attack autumnal moth larvae on their ultimate (5th) or penultimate (4th) instars. As far as we know, the only ectoparasitoid attacking autumnal moth larvae (during 4th and 5th instars) is Eulophus larvarum (Eulophidae). Furthermore, a larval-pupal parasitoid Agrypon flaveolatum (Ichneumonidae) has been common during the decline phase of the latest population cycle.
The pupal stage is attacked by approximately 5 different species. During the last couple of years, the most commonly met species have been ichneumonids Cratichneumon viator, C. rutifrons and Pimpla flavicoxis.
Most parasitoids are potentially generalists, but such a potential may become realized only in southern communities with high enough density of alternative hosts; that is, in the north they are functional specialists having autumnal (and winter) moths as the main hosts. As concluded many times, some sort of delayed density dependence is a prerequisite for a population cycle. Parasitoids as consumers have inevitably a built-in time delay in their numerical response to changes in the host density, fulfilling the condition of delayed density dependence. Our data reveal that during the last outbreak year and especially in the decline phase of a cycle, percentage parasitism of both egg, larval and pupal parasitism can be very high (often > 50%, sometimes > 90%). When the parasitism rates of different developmental stages are calculated as the total percentage parasitism, it is clear that parasitoids must have very strong effect on autumnal moth mortality and population density.

Empirical evidence

Recently, we experimentally tested the parasitism hypothesis of moth population cycles by establishing a four-year parasitoid exclusion experiment, with parasitoid-proof exclosures, parasitoid-permeable exclosures and control plots (Klemola N. et al. 2010, Ecology 91: 2506–2513). The exclusion of parasitoids led to high autumnal moth abundances, while the declining abundance in both the parasitoid-permeable exclosures and the control plots paralleled the naturally declining density in the study area and could be explained by high rates of parasitism. Our results provide firm experimental support for the hypothesis that hymenopteran parasitoids have a causal relationship with the delayed density-dependent component required in the generation of autumnal moth population cycles (Klemola N. et al. 2010, Ecology 91: 2506–2513).
In sum, observations on parasitism, together with modeling approaches, have already provided a framework whereby delayed density-dependent mortality due to parasitism is thought of as a driver of regular population cycles of the autumnal moth in northern Fennoscandia. A similar state of affairs prevails for many other cyclic forest lepidopterans. As these previous observations and theoretical studies can now be combined with the first experimental results (Klemola N. et al. 2010, Ecology 91: 2506–2513), the idea of the importance of parasitoids as causal agents in population cycles is further strengthened. We conclude that parasitoids do not merely track autumnal moth densities, but can actually stimulate a dynamic feedback process with their host species. Thus, parasitoids are responsible for the delayed density dependent (i.e., second-order) component of autumnal moth cycles and drive cyclic population dynamics of this forest pest in continental Finnish Lapland.
Our published articles provide more interesting results on autumnal and winter moth population dynamics, see page "Articles".


Senior researchers

PhD students and post docs

  • PhD Netta Klemola, 2005-2009
  • PhD Elina Mäntylä, 2005-2009
  • PhD Annette Scheiner, 2007-2009
  • PhD Tea Ammunét, 2008-2011


Publications of the project

Contact information

  • Section of Ecology, Department of Biology, University of Turku, FI-20014 Turku, Finland
  • Tero Klemola, +358-2-3335769, tero.klemola[at]
  • Kai Ruohomäki, +358-2-3335766, kai.ruohomaki[at]