Effects of Invasive Mammal Control on Tree Communities

Essay, Pages 10 (2432 words)

Views

Abstract

The presence of invasive mammals is known to influence the structure and composition of plant communities, making the effect of mammal control in New Zealand forests a topic of long-enduring interest. We assessed the effects of mammal control on plant communities, by measuring the number of tree species, tree evenness and density in plots with mammal control and without. Our results, generated by ANOVA tests in R, did not indicate significant variance in these three factors between mammal control and non-control areas.

Don't use plagiarized sources. Get your custom essay on

“ Effects of Invasive Mammal Control on Tree Communities ”

Get custom paper

NEW! smart matching with writer

Therefore, it remains unclear whether mammal control in the Aongatete forest impacts the distribution of the tree communities at the sampling sites. These results were surprising, as we expected greater variation in tree communities associated with the impact of invasive mammals in the area. The results pose a greater question about the effectiveness of invasive mammal control in New Zealand forests and how management strategies could be altered to maximise their impact on tree communities.

Introduction

New Zealand is a marginal fragment of Gondwanaland.

The long isolation of New Zealand has resulted in considerably diverse flora, derived from ancient species that were present when New Zealand formed a part of Gondwanaland or arrived during subsequent isolation.

Although some browsing herbivores, such takahe (Notornis mantelli) and native pigeon (Hemiphaga novaeseelandiae) had an influence on the vegetation, New Zealand flora developed in the absence of browsing mammals until the arrival of humans.

The consequences of deliberate and accidental human introduction of fauna and flora species into areas outside their natural ranges are important research themes in ecology.

The implications of introduced animals on native vegetation is of particular importance to ecologists in New Zealand. The so called "noxious-animal" problem of New Zealand is often summarized as follows:

New Zealand flora evolved in the absence of native terrestrial mammalian herbivores, making them highly vulnerable to damage from browsing and grazing.
Populations of non-native fauna, introduced mainly during the nineteenth century, grew explosively, due to the lack of competitors and predators.
Browsing and grazing by introduced species has had devastating impacts on native vegetation, often resulting in depletion of plant cover and accelerated erosion (Veblen & Stewart, 1982).

Table 1. Wild exotic mammals (including feral domesticated animals) presently found in New Zealand (Veblen & Stewart, 1982).

Table 1 continued

Due to the repeated introduction of most wild animals, they rapidly spread and increased in numbers, becoming gradually regraded as pest. The Australian brush-tailed possum (Trichosurus vulpecula) for example, was liberated several hundred times at over 100 locations and has consequently colonized nearly the entire range of native forests in New Zealand. Aongatete forest, located in the Kaimai Mamaku Forest Park of the North Island of New Zealand is no exception. Introduced pest species present in the forest include possums, goats (Capra hircus), pigs (Sus scrofa), mice (Mus musculus), stoats (Mustela ermine) and deer (Cervus). These animals trample the forest, as well as eat small trees, seeds and leaves of growing trees.

The Aongatete Forest project aims to restore the wildlife and plant life to a part of the Kaimai Mamaku forest, and demonstrate the benefit of pest control. In 2006, the project initiated the control of rats, possums and stoats over an area of 140 ha, which has subsequently extended to 500 ha (less than 2% of the park) under partnership with Royal Forest and Bird Society. To date, volunteers have cut 55 kilometers of bait lines and put up 1200 bait stations. However, reinvasion of pests from the surrounding forest puts constant pressure on pest control efforts.

Although there is no doubt that the mammalian pest species present in Aongatete have had a significant impact on the forest ecosystem, the nature and magnitude of their effects are not easily determined. This study aims to investigate the effects of mammal control on the tree communities of Aongatete forest. We predicted (i), higher species richness, (ii) greater tree evenness and (iii) greater tree density in areas with mammal control.

Materials and Methods

Study location description

Sampling was conducted at Aongatete forest, a remnant of lowland forest, located in the Kaimai Mamaku Conservation park, near Tauranga, New Zealand, on the 16th and 17th of March, 2019. The moderate climate and regular rainfall has led to a great variety of flora in the region. The vegetation at this location is mixed broadleaf/podocarp, dominated by a variety of native trees, including rimu (Dacrydium cupressinum), kahikatea (Dacrycarpus dacrydioides), miro (Prumnopitys ferruginea), tanekaha (Phyllocladus trichomanoides), matai (Prumnopitys taxifolia) and totara (Podocarpus totara). In 2006, the Aongatete Forest Project initiated pest control over an area of 140 hectares in the forest, with the aim to restore the wildlife and plant life to this part of the Mamaku Conservation Park. Over the years the area has been extended and pest control continues to date.

Sampling design

56, 10m x 10m plots, distributed from ~200 m to ~400 m in altitude were set up both in areas of invasive mammal control and in no-control areas. Plot placement guidelines included a minimum of 1 meter from forest edge, paths and fences, as well as no streams or seepages.

Figure 1. Map of sampling sites in Aongatete forest, over a range of altitudes (~200 m to ~400 m), both within invasive mammal control areas (filled circles) and in no-control areas (open circles).

Field sampling methods

For each sampling plot, woody tree and tree fern species with a minimum diameter at breast height (DBH) (1.35 m) of >2.5 cm were identified and recorded. The diameters of these species were measured using DHB tape to give accurate indications of their circumferences and hence basal areas. Measurement guidelines for a variety of tree shapes, following during the vegetation survey, are presented in Figure 2.

Figure 2. Measurement guidelines for trees with a variety of shapes.

Statistical analysis

Data collected in the vegetation survey was analysed using R statistic. A number of ANOVA test were run, with mammalian presence as the predicted variable and tree species richness, tree evenness and tree density as the response variables. Assumptions of normality were tested for each response variable, by creating a histogram of each ANOVA model's residuals. The histogram was used as an indication of how normal our data was.

Results

The aim of this project is to investigate the effect of mammal control on tree communities in Aongatete forest. We expect (i) tree species richness, (ii) tree evenness and (iii) tree density to be higher in areas with mammal control, compared to areas without control. Our hypothesis on tree species richness and tree evenness were based on the knowledge that through selective browsing, many pest species, including possums, reduce diversity and accentuate a strong bias towards specific plant species that are more palatable. Furthermore, our hypothesis on tree density was based on our knowledge that the mammalian pest species present in Aongatete eat small trees, seeds and leaves of growing trees. Therefore, we would expect greater vegetative growth and hence tree density in areas with less mammals. These hypotheses were tested using an ANOVA model in R.

Hypothesis (i), which predicted higher species richness in areas with mammal control, was not supported by the data (Fig. 1). The F value produced by the ANOVA test is relatively small (1.4969), indicating low variability in species richness between mammal control and no-control areas relative to the variability within each group (Table 1). The large p-value obtained (0.2265) further indicates no significant relationship between mammal status and tree species richness in the surveyed plots (Table 1). We found that there was only more tree species in non-mammal sites compared to mammal sites. Therefore, our hypothesis, that tree species richness would be higher in areas with mammal control, was not supported.

Figure 1. Species richness in mammal control ('no mammals') and non-control ('mammals') areas.

Table 1. Analysis of variance for tree species richness, tree evenness and tree density.

Df Sum sq. Means sq. F value Pr(>F)

Tree species richness 1 6.307 6.3072 1.4969 0.2265

Tree evenness 1 0.02464 0.024638 1.4969 0.2265

Tree density 1 21.2 21.184 0.2759 0.6015

Hypothesis (ii), which predicted greater tree evenness in areas with mammal control compared to areas without control, was also not supported by the data collected (Fig. 2). The F value produced by the ANOVA test is very low (1.4969), indicating low variability in tree evenness between mammal control and no-control areas relative to the variability within each group (Table 1). The large p-value (0.2265) again indicates no significant relationship between mammal status and tree evenness in the surveyed plots (Table 1). We found only a difference in tree evenness between mammal sites and non-mammal sites. Therefore, our hypothesis, that tree evenness would be greater in areas with mammal control, was not supported.

Figure 2. Tree evenness in mammal control ('no mammals') and non-control ('mammals') areas.

Finally, hypothesis (iii), which predicted greater tree density in areas of mammal control, was not supported by the data collected (Fig. 3). The F value produced by the ANOVA test is very small (0.2759), indicating low variability in species richness between mammal control and no-control areas relative to the variability within each group (Table 1). The large p-value obtained from this data (0.6015) indicates no significant relationship between mammal status and tree density in the surveyed plots (Table 1). We found only more tree species in mammal sites compared to non-mammal sites. Therefore, our hypothesis, that tree density would be greater in areas with mammal control, was not supported.

Figure 3. Tree density in mammal control ('no mammals') and non-control ('mammals') areas.

Discussion

Cross-continent human movement has greatly increased the transportation and establishment of non-native fauna in places where they can have detrimental effects on native plant and animal species (Anagnostakis, 2001). Forest dwelling mammals, like those present in Aongatete forest, have been introduced to New Zealand over the past 220 years; prior to which such mammals were absent from New Zealand landscapes (Wardle, Baker, Yeates, Bonner & Ghani, 2001). New Zealand forest ecosystems, having therefore evolved in the absence of browsing and grazing mammals, present a classical example of how introduced browsing mammals under certain conditions can upset the natural stability of wildlife habitats either temporarily of indefinitely (Howard, 1964).

Howard (1964) suggest a primary factor that helps explain why introduced species have, and continue to have, detrimental effects on certain habitats. He notes that since New Zealand's flora evolved without the presence of browsing or grazing mammals, they have little innate resistance to heavy browsing pressures. Natural selection did not have the opportunity to either eliminate highly palatable plants that could withstand the selective feeding pressure of the introduced mammals or to favour those species that were browse-resistant or unpalatable to browsing animals (Holloway, 1960). Howard (1964) remarks that in many forest ecosystems, a new equilibrium of the animal-vegetation-soil complex has developed, where unpalatable, browse resistant plants have adequately replaced those destroyed by browsing mammals.

A study conducted by Wardle et al. (2001), sampled 30 long-term fenced enclosure plots in indigenous forests throughout New Zealand in order to evaluate ecosystem-level impacts of introduced browsing mammals. They reported that browsing mammals significantly altered plant community composition by reducing vegetative density in the browse layers, especially that of palatable species and consequently promoting other, less palatable species. Owen and Norton (1995) support these findings, claiming that the selective browsing of many mammal species reduces diversity and accentuates a strong bias towards unpalatable biomass. New Zealand's Environmental Reporting Series (2015) carried out a survey of 874 (20m x 20m) plots across New Zealand forests between 2002 and 2014 to determine the impact of introduced browsing mammals, in particular possums, goats and deer, on the mortality and recruitment rates of tree species either preferred or avoided by the mammals. They reported that, for tree species preferred by possums and goats, the number of trees that died exceeded the number of newly established trees.

Comparatively, for tree species avoided by goats and possums, the number of newly established trees exceeded the number of tree deaths. They consequently express the concern that tree species preferred by invasive mammals, such as possums and goats, may become locally extinct and nationally much rarer than palatable species (New Zealand's Environmental Reporting Series, 2015). Salmon (1975) also reports that since as early as the 1940s, considerable concern for the deterioration of forest ecosystem has been caused by extensive tree mortality due to the presence of introduced browsing mammals. Forest deterioration results in accelerated erosion in forest ecosystems, increasing the risk of flooding of adjacent pastoral lowlands. Salmon (1975) claims that browsing of the foliage in the canopy layer by possums is commonly accepted as the cause of many cases of widespread tree mortality. Hence the New Zealand Forest Service strongly encourage trapping and poisoning of these mammals in order to prevent further deterioration of New Zealand forest ecosystems.

Furthermore, changes in the composition of forest vegetation may have indirect effects on animal species in the environment, limiting habitat availability and food sources. Owen and Norton (1995) echo this concern, therefore highlighting the considerable long-term benefits pest control could have for the ecosystem health of New Zealand forests.

Although there is no doubt that introduced browsing mammals have had extensive effects on native vegetation, determination of the magnitudes of these effects is surprisingly difficult (Thomas & Stewart, 1982). For example, while it is known that red deer (C. elaphus) cause shifts in understory species composition of beech (Nothofagus) forests, their long term effects on regeneration of dominant tree species remain unclear. Similarly, while the possums are known to increase the rate of tree mortality in certain forests, other processes may also be contributing to tree mortality in the same area. Consideration of the effects of these introduced mammals on New Zealand's native forest vegetation therefore illustrates the difficulties of distinguishing between animal-induced changes and other types of vegetation changes.

Our results did not show a strong correlation between invasive mammal control and increased tree density, tree evenness and tree species richness. The reasons for this are uncertain, however, mammal control in the area began relatively recently and the area of control has been extending over subsequent years since 2006. In some areas control began even more recently. Continued surveying and analysis of the area may begin to show a greater correlation between mammal trapping and forest ecosystem health. Furthermore, areas with mammal control are not necessarily void of all invasive species. These trapping areas are not isolated ecosystem sanctuaries or "islands", meaning there would be continual reinvasion of pests from surrounding forest. A greater correlation between the studied variables would be expected in forest sanctuaries, completely isolated from surrounding forest.

To improve the study, other aspects of ecosystem health could also be considered alongside our existing variables. For example, it is known that the presence of invasive browsing mammals not only has aboveground consequences, but also belowground consequences. They are also know to extensively affect the abundance of birds and other animal species in some areas. Finally, browsing level vegetation should also be considered in future surveys, including herbaceous ground cover and trees with DBH.