Hungry Caterpillars: Massive Outbreaks of Achaea lienardi in Hluhluwe-iMfolozi Park, South Africa

Simple Summary

Understanding the conditions which facilitate massive outbreaks of insects is important to better understand the biology and conservation of the species, as well as to be able to predict future outbreaks. Some species target agricultural crops, whilst others target indigenous species and are considered part of the natural cycle. The climatic variables over the last five years were assessed to determine if there was any correlation between them and the Achaea lienardi outbreak that occurred during February and March of 2025 in the Hluhluwe-iMfolozi Park in KwaZulu-Natal, South Africa.

Abstract

Achaea lienardi is a polyphagous moth occurring in sub-Saharan Africa. It is a fruit-sucking moth, causing secondary damage to fruit such as citrus and peaches, while the larval stage can cause significant tree defoliation, including in several indigenous trees, wattle, Eucalyptus, and castor oil plants, amongst others. In February and March of 2025, a massive outbreak of the caterpillars was observed in the Hluhluwe-iMfolozi Park in South Africa, feeding primarily on Tamboti trees (Spirostachys africana). Satellite imagery from the previous five years was examined, but no similar large defoliation events were observed during this period. Climate data for the last five years (September 2019–March 2025) were collated and examined to determine the conditions supporting the outbreak. Above average winter rainfall, early spring rains, sustained rains, and high humidity in January and February, with warm nighttime temperatures, likely acted in concert to create favourable conditions for the caterpillar outbreak. This outbreak coincided with historic outbreaks of the African armyworm (Spodoptera exempta) in the summer rainfall areas of South Africa where precipitation, temperature, solar radiation, and humidity were found to be critical factors affecting armyworm outbreaks. Further research is required to determine specific criteria to enable predictions of future outbreaks.

1. Introduction

Achaea lienardi (Boisduval, 1833), or Lienard’s Achaea, is a Noctuid (Owlet) moth in the Erebidae family [1]. It is considered a pest as the caterpillar can defoliate plants, and the adult moths can damage fruit. It occurs in sub-Saharan Africa [2], including Madagascar, Mauritius, and the Comoros [1].

It is polyphagous and has been recorded on 44 host tree species [3], as well as on commercial crops such as castor oil plants, wattle, Eucalyptus, and various fruits including grapefruit, lime, mandarin, nartjie, cashew, mango, apple, apricot, plum, banana, fig, grape, guava, litchi, peach, pear, pomegranate, and prickly pear [4].

There are two categories of fruit-feeding moths—fruit-piercing and fruit-sucking. Fruit-piercing moths have the ability to damage healthy fruit, using their modified proboscis (primary damage), whereas fruit-sucking moths can only feed on fruit that is already damaged (secondary damage) [5,6]. A. lienardi is classified as a fruit-sucking moth [6]. Hence, their outbreaks often follow outbreaks of fruit-piercing moths such as Serrodes partita (catapult moth) and Eudocima divitiosa (Spotless Underwing moth) [6]. Historically, previous authors have recorded them as fruit-piercing, despite observing their preference for split and over-ripe fruit to sound fruit [7]. However, recent scanning electron micrographs of proboscides confirmed them to be fruit-sucking moths [6].

A. lienardi moths are of interest due to their occasional mass outbreaks, being recorded as early as the 1890s [7]. Taylor [7] noted severe outbreaks in the Eastern Cape, South Africa, in 1952, 1953, 1959, and 1963, suggesting that outbreaks are linked to suitable climatic conditions. Webb [4] refers to outbreaks in KwaZulu-Natal (KZN), South Africa (SA), in 1935/36, 1941/42, and 1946/47, which correlated to early season droughts with late summer rains. The 1942 outbreak defoliated more than 10,000 ha of wattle plantations in KZN. In Accra Ghana, an exceptional outbreak occurred in 1972/73 [8].

Mass outbreaks of insects are well documented, such as the recent Brown locust (Locustana pardalina) outbreak in the Karoo, SA [9], and the recent African armyworm (Spodoptera exempta) outbreaks in SA [10,11], said to be the worst on record. However, the conditions leading to massive outbreaks (gregarious phases) compared to solitary or low levels of insects are not clearly known [12]. More information on the conditions favouring outbreaks is required to model and predict outbreaks more effectively.

In March of 2025 a massive defoliation event due to A. lienardi caterpillars was observed in Hluhluwe-iMfolozi Park (HiP) in KZN, SA. The climatic conditions of the previous five years were explored to determine factors contributing to the mass outbreak observed.

2. Materials and Methods

2.1. Study Area

Hluhluwe-iMfolozi Park is a protected area occurring in KZN, SA (Figure 1). It is 89,597 ha in extent and occurs in the savanna biome, a habitat of the big five game. It has a subtropical climate with distinct wet and dry seasons. In the summer months the temperature can reach 30–38 °C, and most of the rainfall occurs during this season. Minimum temperatures in winter drop to 7–10 °C.

2.2. A. lienardi Biology

The colour and markings of the moth exhibit large colour–pattern variation in the forewing [2]. Adults can survive up to 171 days [7] and may migrate. In Nigeria, the species can breed throughout the year [13], whereas in South Africa, breeding is confined to the summer season. Oviposition takes place when climatic conditions are favourable [4] and eggs are laid singly, at night [7]. Incubation takes 2.5–4 days. The larva is a semi-looper caterpillar (Figure 2) with considerable variation in the larva ground colour, with the colouring varying on the host plant and hence the diet of the caterpillar. There are six larval instars, ranging from 16 to 48 days [7]. The larva pupates on the soil surface amongst debris or just beneath the surface of the soil. A soil-coated cocoon is woven, and pupation varies from 13 to 51 days.

2.3. Methods

The defoliation of primarily Tamboti trees (Spirostachys africana) was observed in the Park on the 2nd March 2025 (Figure 3). Sentinel satellite imagery [14] was assessed for the previous five summer seasons (2020–2025) to determine if any other mass defoliation events had taken place during that time. A cloud-free false colour composite (bands 5, 6, and 7 (Vegetation Red Edge)) from March 2021 [14] was used to manually delineate the Tamboti vegetation associations (Figure 4) to compare to the defoliated areas in 2025. The defoliated areas on the satellite imagery were visually assessed and compared to known areas of defoliation based on the field visit on the 2nd March 2025.

Climate data was obtained from Visual Crossing Corporation [15] for Hluhluwe town approximately 25 km away from the site, as well as some climatic data from the research station in the protected area. The daily maximum temperature, minimum temperature, mean temperature, relative humidity, precipitation, and solar radiation variables were measured from 1 September 2019 to 31 March 2025. The data was aggregated into monthly variables. A T-test was performed to compare outbreak conditions to non-outbreak conditions. Time lagged variables (prior seasons’ conditions) were considered.

3. Results

The defoliation event was clearly visible on the colour satellite imagery. Cloud-free satellite imagery for the entire park was not available, hence the study area was limited to an area that was largely cloud-free. Defoliation was not yet visible on the imagery of the 8 February 2025 (Figure 5), but extensive defoliation was visible by the 5 March 2025. The Tamboti vegetation associations were the primary areas of defoliation, but defoliation did spread to neighbouring vegetation, and the caterpillars were also observed on Schotia brachypetala. By the 17 April 2025, the vegetation had largely recovered. Satellite imagery for the last five years was examined but no similar large defoliation events were observed on the imagery during this period.

Rainfall in early spring (September 2024) was high (149 mm), roughly 4–5 times the September rainfall of previous years (Figure 6). This drove up humidity levels, reaching 80% versus 72–75% in prior years. Rainfall tapered off in November 2024, creating a dry spell, reaching 45 mm of rain compared to the November average of 100 mm, except for November of 2021, which saw 33 mm of rain. December summer rainfall returned to normal. February 2025 rainfall was good (200 mm) but not excessive compared to the January (261 mm) and February (308 mm) rainfall in 2021 and February (536 mm) rainfall in 2023, as measured at the research station within the reserve. The late summer rainfall of 2025 was persistent and distributed in nature, with rain occurring on 24 out of 28 days in February 2025, and 19 days in January 2025, indicating near-continuous moisture. There was steady rain and cloud cover without the destructive storm extremes of 2021 and 2023.

December 2024 was also warm, with a daily average temperature of 26 °C, which was 1–2 °C higher than previous Decembers. Nights were warm, at 22.2 °C, compared to 19–21 °C in previous Decembers. Temperatures in January and February 2025 were also warm, especially at night. In February 2025, nighttime temperatures stayed at around 23 °C, which was 1–3 °C higher than previous years. The combination of frequent rain and warmth kept relative humidity high during the outbreak. Cloud cover was high compared to drier years. By March 2025 rainfall began to taper, and temperatures started to ease, but by then the defoliation by the caterpillars was already complete.

The six-month lag of mean maximum temperature was significantly higher in outbreak years compared to non-outbreak years (outbreak mean 25.71 °C compared to the non-outbreak mean of 24.23 °C, T-statistic 3.90, p = 0.0046), suggesting warm conditions (July–August). No other three-month and six-month lagged variables showed significant differences between outbreak and non-outbreak years.

The winter period (June–August 2024) was warmer and wetter than preceding years. July and August 2024 mean temperatures were 18.2–19.7 °C, ~1–2 °C above the 2019–2023 average for those months. Winter rainfall in June 2024 was 74 mm compared to <20 mm in non-outbreak years.

4. Discussion

The early spring rains likely triggered an early flush of vegetation for the Tamboti trees, which are deciduous. The early summer dry spell would have stressed or slowed plant growth and depressed populations of predatory insects [16]. The warm December–February period, especially its night temperatures, likely accelerated insect metabolic rates, leading to accelerated insect development, increased activity, and potentially greater consumption [17,18,19]. Warm nights create a favourable scenario for nocturnal moth activity. Achaea moths are active at night and higher temperatures allow them to fly and mate more frequently [16]. The warm moist conditions would have supported vegetation and insect growth. The warmer and wetter winter conditions likely promoted insect larval survival [20,21], limiting cold- and desiccation-related deaths, leading to a larger potential starting population. Humid conditions during January and February 2025 favoured caterpillar survival, preventing their soft bodies or eggs from drying out.

The extreme rainfall events in the summer of 2021 were due to Tropical Storm Eloise, and in 2023 due to a La Niña event, resulting in extensive flooding and a national state of disaster being declared in 2023 [22]. Extreme precipitation events such as these are likely to dislodge insect eggs and wash away caterpillars.

Webb [4] observed that spring and early summer dry conditions limit the Achaea caterpillars, but when late, good summer rains do fall, there is a good supply of young leaves available for the caterpillars, coincident with warm night temperatures favourable for egg laying. However, in Zimbabwe, outbreaks were more prevalent after heavy rain and following wet summers [7]. Ossowski [23] indicated that outbreaks occur every 8–10 years, usually following late summer rain, but our results did not support this, and reported outbreaks in the literature point to more frequent outbreaks under suitable conditions.

South Africa experienced the worst African armyworm infestation on record in February 2025 [11]. The outbreak occurred in the summer rainfall areas of Mpumalanga, Gauteng, KZN, North West, Limpopo, and the Free State. Other caterpillar outbreaks have also been recorded, including lawn caterpillars, looper and semi-loopers, and fall armyworm. The explosion of caterpillars is attributed to favourable climatic conditions allowing for swarming. Specifically, the summer rainfall season received a lot of early-summer rain, followed by a mid-summer dry spell and significant rainfall from January onwards, combined with warm temperatures, similar to the conditions favouring the Achaea outbreak. Webb [4] noted that the outbreaks of A. lienardi were synchronised with outbreaks of armyworm. Sokame et al. [12] developed a system dynamics model for predicting African armyworm occurrence and population dynamics. In the solitary phase, life cycles unfolded more gradually, whereas in the gregarious phases there were more intense egg laying and faster development rates. Optimal climatic conditions included temperature, humidity, rainfall, and solar radiation; however, the importance of these factors varied according to location in Africa.

Changing landscapes may be favouring insect outbreaks. Asogwa et al. [24] observed that the moth only became a pest on cashews when large plantations were established. Natural climatic cycles may favour outbreaks in suitable conditions, but in future, climate change could result in phenological mismatches, adults living for shorter periods, a shorter period of oviposition, or fewer eggs [21].

Natural predators of A. lienardi include birds [7] such as red-winged starlings, rock kestrels, yellowbilled kites, hadeda ibis and Indian (common) myna, Karoo prinias, and Cape sparrows, as well as monkeys. Ossowski [25] recorded the larval parasite (Exorista fallax) and two predatory species of Pentatomidae (Hemiptera), Glypsus maestus and Macrorhaphis spurcata, and the sphegid wasp Ammophila beniniensis as predators. Glypsus conspicuous was observed attacking the larvae.

In natural systems, no control is necessary, and the affected vegetation recovers within a few months. Given that A. lienardi is a fruit-sucking moth, Moore [5] recommends that no action be taken against them in orchards unless they become a nuisance in pack-houses. Leston [8] reported that the moths became a nuisance during an exceptional outbreak in 1973 in Ghana. The moths are not attracted to ordinary light but are attracted to white ceilings, i.e., reflected light [7]. In commercial plantations, Govender [26] advocates an integrated pest management approach for general pests, with restrictions on the use of insecticides and biological control. Control of the larvae is impractical given the wide range of food plants and the scattered nature of the larvae [27]. Loss of citrus crops could be avoided by cultivating varieties that produce crops outside of the major larval periods, or the early harvesting and storage of fruit [27]. The use of pesticides should be managed judicially to avoid negative impacts on non-target species.

This study has identified two new host plants, Spirostachys africana and Schotia brachypetala, thereby advancing our knowledge on the ecology of the species. Future research should include monitoring and a more detailed model such a system dynamic model to better predict outbreaks and the species population dynamics, and to address the deficiencies of this short-term, correlative study. Additional collections from museums could be good resources for biological modelling to indicate if outbreaks are unusual or a normal part of population dynamics [28]. Citizen science data such as that in iNaturalist will be immensely useful for future modelling. Future research should employ long-term climatic modelling (>20 years), population genetics, and predator-exclusion experiments.

5. Conclusions

Multiple climatic factors acted in concert to create ideal conditions for A. lienardi survival in the Hluhluwe-iMfolozi Park over the last five years. The evidence suggests that it was the combination and timing of factors, rather than any single extreme factor, that made the 2025 outbreak so severe. This included the mild and wetter winter of 2024, promoting higher overwintering survival in the insects. The early spring precipitation likely triggered some early leaf flush, but the dry spell in November delayed further flush to later summer, resulting in the synchronisation of host plant phenology and the Achaea life cycle. The early dry summer period may have reduced predator insect populations due to less food/prey, meaning fewer natural enemies were around when the Achaea caterpillars began multiplying. The late summer rainfall with near continuous light and moderate rains maintained high humidity, maximising caterpillar survival rates and plant vigour. The warm nights boosted the insects’ rate of development and reproduction.

This underscores how sensitive ecosystem dynamics are to climate variability. From a management perspective, the warning signs of an outbreak could potentially be detected, such as a warm wetter winter, or an early summer drought followed by good later rains, along with high summer night temperatures. Outbreaks should be monitored and related to changes in land use and potential climate change impacts.