Short-Term Effects of Cattle Browsing on Tree Sapling Growth in Mountain Wooded Pastures

Browsing by livestock has been identified as an important factor preventing tree regeneration in wooded pastures. Two field experiments were performed to investigate the effects of cattle browsing on tree sapling growth in a mountain-wooded pasture. Two size classes (ca. 12–17 cm and 41–59 cm) of each of 4 species (Picea abies, Abies alba, Acer pseudoplatanus and Fagus sylvatica) were exposed to zero, low and high grazing intensities. We measured the proportion of saplings browsed and the effect of browsing on growth. A mowing treatment within the zero grazing intensity treatment showed no significant effect on sapling growth. One percent of the large saplings but 25% of the small saplings escaped browsing. Saplings were better protected when surrounded by taller vegetation. The proportion of saplings browsed was not significantly different among species although evergreen tree saplings lost a larger proportion of biomass than deciduous species when browsed. Under grazing, average current-year shoot production and total aboveground biomass of all species were significantly reduced. Browsing effects tended to be smaller at the lower grazing intensity. Because the deciduous species were least reduced in aboveground growth, especially under the low grazing intensity, we conclude that at least in short-term, deciduous species are less affected by cattle browsing.     

Ungulate browsing in winter reduces the growth of <Emphasis Type="Italic">Fraxinus</Emphasis> and <Emphasis Type="Italic">Acer</Emphasis> saplings in subsequent unbrowsed years

Browsing by ungulates has become a hotly debated issue in many European mountain forests in the past century. Acer pseudoplatanus L. and Fraxinus excelsior L. are broadleaved tree species that are preferentially browsed by roe deer (Capreolus capreolus) in central Europe. We analyzed growth data from shaded saplings of both tree species to quantify the extent to which height growth after game browsing is reduced in subsequent, unbrowsed years in forest stands. Sixty saplings of F. excelsior and A. pseudoplatanus from forest stands at Albisriederberg (Switzerland) were available to us that had been dissected into pieces that then were split in the middle for counting tree rings and assessing ungulate damage. We fitted the von Bertalanffy growth equation to these height growth data and included a reduction factor for winter browsing. Both tree species showed significantly reduced height growth in unbrowsed years after one to several browsing events in winter, and this effect increased with the number of browsing events. Saplings with a high growth rate showed a higher growth reduction. After winter browsing, height growth of A. pseudoplatanus saplings was less affected in unbrowsed years than that of F. excelsior saplings. We conclude that browsed saplings of these species in forest stands are not able to compensate browsing-induced height loss, but that height differences between browsed and unbrowsed saplings probably increase over time. A comparison between our analysis and the parameters estimated using equations published by Eiberle for predicting age at 130 cm height suggests that our parameter values are rather conservative estimates of the growth reduction effect after winter browsing. Neither F. excelsior nor A. pseudoplatanus show a distinct pattern in browsing-induced growth reduction with respect to soil moisture, nutrient level, and altitude. We thus conclude that our results are likely to be valid for a wide range of forested sites.     

Coarse woody debris can reduce mammalian browsing damage of woody plant saplings in box‐gum grassy woodlands

The critically endangered box‐gum grassy woodlands of south‐east Australia face numerous threats including the failure of woody plant regeneration caused by over‐browsing. In the Australian Capital Territory, over‐browsing of tree and shrub saplings is likely caused by dense populations of Eastern Grey Kangaroo (Macropus giganteus) found in many nature reserves free of livestock. One possible way to protect these saplings is using coarse woody debris (CWD) as a browsing deterrent. We tested this idea by planting palatable Red Stemmed Wattle (Acacia rubida) saplings among manually applied CWD, among naturally fallen CWD, and in the open, in five woodland reserves. We recorded the proportion of saplings browsed, the number of weeks to first browsing and the browsing severity (sapling height lost). Applied CWD protected saplings from being browsed only at relatively low‐to‐moderate kangaroo browsing pressure (as measured by faecal pellet counts). At relatively high browsing pressure, the probability of a sapling being browsed among applied CWD was 100%, similar to the probability in the open treatment (no CWD). Natural CWD, in contrast, provided some protection even at high browsing pressures. Time to browsing was most affected by browsing pressure, although CWD cover also had an influence. Browsing severity was similar between the three treatments and was only affected by browsing pressure. These results indicate that without protection, palatable woody plant saplings have a high chance of being browsed by kangaroos in woodland reserves, and therefore, some protection is needed for successful regeneration. The CWD being applied to reserves has a limited capacity to protect regenerating saplings. If more protection is wanted a CWD structure more resembling natural fallen timber should be used. This could be done by artificially placing branches around plantings. However, the most important action to facilitate regeneration is to manage kangaroo populations to reduce overall browsing pressure.     

Effects of deer browsing on the early stage of pyrogenic succession on Miyajima Island,southwestern Japan

The early stage of forest regeneration on Miyajima Island, southewstern Japan, was studied for three years after a fire in 1984, with respect to the effects of deer browsing. The regeneration patterns of woody species, in terms of the biomass increment and browsing damage to plants, were classified into the following three groups: non- or rarely browsed species with little increment of biomass, heavily browsed with little increment, and heavily browsed with large increment. The aboveground biomass increased from 0.4 kg to 2.7 kg per 100 m² during three years, and was less than half of the biomass obtained in other burnt pine forests on the adjacent deer-free islands. About 28% of the plant biomass was consumed by deer. The effects of browsing on forest regeneration were heavier at lower than that at the higher altitudes. On this island, deer browsing seems to have an important effect on vegetational succession in the burnt areas.     

The role of moose <Emphasis Type="Italic">Alces alces</Emphasis> L. in boreal forest – the effect on ground beetles (Coleoptera,Carabidae) abundance and diversity

The indirect effect of moose Alces alces browsing on ground beetle's abundance and diversity was investigated by pitfall trapping in a mixed coniferous forest in Vestfold County (59°19′ N, 9°50′ E, Norway), during the summer of 2002. Three areas with different browsing pressures, ranging from non- to medium- and heavily browsed were chosen and dry weight of bilberry Vaccinium myrtillus and humidity at ground level were measured in the three locations. As predicted, the gradient analyses showed that browsing by moose influenced the composition of carabid fauna, and that browsing intensity and humidity covaried with the most important gradient in carabid species composition found across the three locations. Species that live in light stands with rather dry soil, were more often captured in the browsed areas, whereas shade tolerant and hygrophilous species were more abundant in the non-browsed area. The carabid abundance increased significantly with increasing browsing pressure. According to our predictions, the diversity at trap level (α-diversity) was higher in the highly browsed area. Conversely, species turnover (β-diversity) decreased with browsing intensity. On the other hand, the rarefaction analysis showed that the regional species richness (γ-diversity) was considerably higher in the medium browsed area than in the heavily browsed one, which is consistent with the intermediate disturbance hypothesis. This study indicates that moose, by reducing the bilberry that constitutes the field layer in summer, affect carabid species composition and might be capable to reshape the whole ecosystem in our study area by a cascade effect.     

Disentangling the causes of damage variation by deer browsing on young Thuja plicata

Long‐lived trees experience different levels of damage due to mammalian herbivores. To untangle the mechanisms that underlie this variation, we combined chemical with dendrochronological analyses to study variation in browsing on Western redcedars (Thuja plicata) on Haida Gwaii (British Columbia, Canada). Since the last glaciation, Haida Gwaii forests had lacked large herbivore browser until Sitka black‐tailed deer (Odocoileus hemionus sitkensis) were introduced at the beginning of the 20th century. Dendrochronology yielded information on radial growth and plant annual responses to environmental stresses including herbivory. Secondary metabolite content and plant nutritional quality provided insights into proximate causes of food choices made by herbivores. We sampled lightly‐ and heavily‐browsed young trees at four sites: three clear‐cut sites with high browsing pressure and one old‐growth forest site where browsing pressure had, until recently, been lower. Heavily‐browsed young trees had lower concentrations of secondary metabolites and were of lower nutritive value than lightly‐browsed trees at all sites. Under high browsing pressure, tree growth patterns suggested that all young trees were initially severely browsed until some trees, currently scored as lightly‐browsed, started to escape deer. At the old‐growth site, both lightly‐ and heavily‐browsed trees tended to have lower overall average secondary metabolite concentrations than those of all other sites, a trend possibly related to greater canopy closure. Lightly‐browsed trees were older than heavily‐browsed ones which resulted, during the period of lower browsing pressure, in higher growth rate and a same pattern of change in growth from one year to the next year. This suggests that, under low browsing pressure, selection of young trees related to chemical defense was weak and that growth differences due to other factors than browsing could be expressed. Under strong browsing pressure, however, all young trees had equally low growth rates until trees with better genetic potential to produce effective defenses were able to escape deer. This suggests that selection by deer could occur on a long‐lived tree.     

Associational refuge in practice: can existing vegetation facilitate woodland restoration?

Herbivores can dramatically diminish revegetation success, but associational refuge theory predicts that neighbouring plants could hinder browsing of planted seedlings. The key to strategic restoration using associational refuge is to define which patch variables are effective against the appropriate herbivores, at multiple scales, and to understand which stages of the foraging process these variables disrupt. Our study aimed to test the capacity of existing vegetation to act as associational refuge for planted seedlings by affecting search, detection and consumption decisions, and more generally influence herbivore foraging patterns. We conducted a field trial with free‐ranging, mammalian herbivores and nursery‐raised, native tree seedlings. We quantified seedling browsing damage over time in relation to a suite of existing patch variables at two spatial scales (100 m² and 4 m²). After two months, 78% of seedlings were browsed, suffering mean foliage loss of 90.5%. Focal seedlings were almost exclusively consumed by swamp wallabies Wallabia bicolor, an abundant generalist browser. Once a swamp wallaby investigated a seedling, the probability of consumption was high (86%). At the large scale, browsing of seedlings was delayed in patches with lower canopy cover and fewer browsed plant species. Seedlings in fern‐dominated patches escaped browsing for longer than those in grass‐dominated patches. At the small scale, browsing was delayed with higher cover of understorey vegetation. Associational refuge was provided by vegetation with characteristics, and at spatial scales, consistent with disrupted search and detection of focal seedlings by herbivores. Thus strategic placement of seedlings in existing vegetation – based on understanding which herbivore species is responsible and how it responds to vegetation – can take advantage of associational refuge during restoration. However, given rapid seedling detection by herbivores, associational refuge may be inadequate in the long‐term under high browsing pressure unless high absolute numbers of seedlings are planted among refuge.     

Bryophyte Species Diversity in Secondary Forests Dominated by the Introduced Species Spathodea campanulata Beauv. in Puerto Rico

The introduced tree species Spathodea campanulata (Bignoniaceae) forms novel forests in Puerto Rico, these having emerged after the abandonment of fields in the mid‐20th century and resulting in forests with a new species composition. We assessed bryophyte species richness in these novel forests and sought correlations with geological substrate, past land use, forest edge and patch area, forest structure, elevation, microhabitat diversity, tree species richness, and microclimatic conditions. Transects were established (edge and forest interior) in nine moist forest patches dominated by Spathodea in north‐central Puerto Rico. These Spathodea forest patches ranged from 0.6 to 9 ha. ANOVA, Chi‐square, correlation, and cluster analyses were used in data analyses. We found 57 bryophyte species. There was a significant difference in bryophyte richness among patches. Those on karst exhibited highest bryophyte richness due to microhabitat diversity, past land use, and shorter hydroperiods. Alluvial sites scored lowest in bryophyte species richness, and forest structure was important for bryophyte communities on these sites. Significant differences in temperature, relative humidity, and light intensity were observed between edge and forest interior. These appeared important for establishing bryophyte species cover but not richness and composition. Microhabitat diversity, patch area, and forest age were more related to bryophyte species richness than elevation, exposed edge, and tree species richness, regardless of geologic substrate. Collectively, Spathodea patches were similar to mature forests on the Island with respect to bryophyte species richness and composition. Novel Spathodea forests have conservation value due to their habitat suitability for bryophyte communities.     

Browsed twig environmental DNA: diagnostic PCR to identify ungulate species

Ungulate browsing can have a strong effect on ecological processes by affecting plant community structure and composition, with cascading effects on nutrient cycling and animal communities. However, in the absence of direct observations of foraging, species‐specific foraging behaviours are difficult to quantify. We therefore know relatively little about foraging competition and species‐specific browsing patterns in systems with several browsers. However, during browsing, a small amount of saliva containing buccal cells is deposited at the bite site, providing a source of environmental DNA (eDNA) that can be used for species identification. Here, we describe extraction and PCR protocols for a browser species diagnostic kit. Species‐specific primers for mitochondrial DNA were optimized and validated using twigs browsed by captive animals. A time series showed that about 50% of the samples will amplify up to 12 weeks after the browsing event and that some samples amplify up to 24 weeks after browsing (12.5%). Applied to samples of natural browsing from an area where moose (Alces alces), roe deer (Capreolus capreolus), fallow deer (Cervus dama) and red deer (Cervus elaphus) are sympatric, amplification success reached 75%. This method promises to greatly improve our understanding of multispecies browsing systems without the need for direct observations.     

The benefits of being in a bad neighbourhood: plant community composition influences red deer foraging decisions

Diet selection by mammalian herbivores is often influenced by plant community composition, and numerous studies have focused on the relationships between herbivore foraging decisions and food/plant species abundance. However, few have examined the role of neighbour palatability in affecting foraging of a target plant by large mammalian herbivores. We used a large-scale field dataset on diet selection by red deer Cervus elaphus in Fiordland National Park, New Zealand to: (1) estimate the palatability of native forest plant species to introduced deer from observed patterns of browse damage; and (2) examine whether intraspecific variation in browsing of plants can be related to variation in the local abundance of alternative forage species. Overall, 21 of the 53 forest species in our dataset were never browsed by deer. At a community level, plants were more likely to be browsed if they were in a patch of vegetation of high forage quality, containing high abundances of highly palatable species and/or low abundances of less-palatable species. Our findings suggest that deer make foraging decisions at both a coarse-grain level, selecting vegetation patches within a landscape based on the overall patch quality, and at a fine-grain level by choosing among individual plants of different species.     

Development of Tree Regeneration in Fern-dominated Forest Understories After Reduction of Deer Browsing

Dennstaedtia punctilobula (hay‐scented fern) can act as a native invasive species in forests in eastern North America where prolonged deer browsing occurs in stands with partially open overstory canopies. Ferns dominate the understory with a 60‐cm tall canopy, with little regeneration of native tree species. It has been hypothesized that, once established, ferns may continue to inhibit tree regeneration after deer browsing has been reduced. To test this hypothesis, we documented the pattern of recovery of the tree seedling understory in plantations of Pinus strobus (white pine) and Pinus resinosa (red pine) on the Quabbin Reservation watershed protection forest in central Massachusetts, where after 40 years of intensive deer browsing the deer herd was rapidly reduced through controlled hunting. Dense fern understories occur on nearly 4,000 ha of the predominantly oak–pine forest. Three years after deer herd reduction, stands with the highest density fern cover (77% of plots with>90% cover) had significantly fewer seedlings at least 30 cm in height, compared with stands with lower fern density, and those seedlings consisted almost entirely of Betula lenta (black birch) and white pine. Height growth analysis showed that black birch and white pine grew above the height of the fern canopy in 3 and 6 years, respectively. In contrast, two common species, Fraxinus americana (white ash) and Quercus rubra (red oak), grew beneath the dense fern cover for 5 years with height growth less than 5 cm/yr after the first year. A study of spring phenology indicated that the ability of black birch to grow through the fern canopy might have been due to its early leaf development in spring before the fern canopy was formed, in contrast to oak and ash with delayed leaf development. Thus, the ferns showed differential interference among species with seedling development after reduction of deer browse.     

Does the strength of facilitation by nurse shrubs depend on grazing resistance of tree saplings?

Facilitation of tree regeneration by nurse shrubs that offer protection against large herbivores is an important driver of wood-pasture dynamics. Here we asked whether the response to facilitation by nurse shrubs depends on the grazing resistance of the protégé saplings. We experimentally tested the protective effects of the thorny Rosa rubiginosa on browsing frequency, survival, and biomass change of saplings of two species-groups, presumably differing in grazing resistance: the coniferous Abies alba and Picea abies and the deciduous Acer pseudoplatanus and Fagus sylvatica saplings. The saplings were planted under and outside (1.5 m) planted nurse shrubs, under zero, low and high grazing intensity. In total, 1920 young saplings were transplanted to 60 blocks and followed for 1 year.Although the number of saplings browsed did not differ between species-groups, the coniferous saplings showed lower resistance to cattle browsing (i.e. lower survival and growth rates) than the deciduous saplings. The less resistant coniferous saplings benefited significantly more from nurse shrubs than the more resistant deciduous species in terms of growth of the surviving saplings, but not in terms of overall survival. This was likely due to herbivory on the nurse shrubs causing incidental browsing on protégé saplings and differences in biomass off-take. At high grazing intensity facilitative effects of the nurse shrubs decreased, especially for the coniferous species.These results have important management implications for the endangered wood-pastures in Western Europe. For a sustainable management and conservation that allows tree recruitment, grazing intensity should remain low to best promote facilitation processes for all tree species, but in particular for the less resistant conifer saplings.     

Does forest restoration using taungya foster tree species diversity? The case of Afram Headwaters Forest Reserve in Ghana

The taungya agro‐forestry system is an under‐researched means of forest restoration that may result in high tree diversity. Within a forest reserve in Ghana, the forest core and its surrounding Teak‐ and Cedrela‐taungya on logged, cropped and burned land were mapped with ALOS satellite imagery. Native trees, seedlings and saplings were enumerated in 70 random, nested plots, equally divided between forest and taungya. The native tree regeneration was assessed by species richness (SR), Shannon‐Wiener Index (SWI), Shannon Evenness Index (SEI) and species density (SeD) for seedlings, saplings and trees separately and combined and subsequently correlated with canopy covers (CC) in taungya. As anticipated, the taungya diversity was lower than the forest diversity but higher than reported from nontaungya exotic plantations. In the forest, the diversity of native trees increased from seedlings through saplings to trees. The reverse was found in the taungya. Taungya seedling diversity was not significantly different from the forest, while the sapling and tree diversity were significantly lower. Weak correlations of CC with SR, SWI, SEI and SeD were found. Our results suggest the need for treatment to maintain the tree diversity beyond the seedling stage in the taungya.     

Suppression of Regeneration in New Zealand Mountain Beech Forests is Dependent on Species of Introduced Deer

We compared the impacts on forest regeneration of introduced sika (Cervus nippon) and red (Cervus elaphus) deer in New Zealand. Plot data were used to compare mountain beech (Nothofagus solandri var. cliffortioides) regeneration between a region with sika deer, and four regions without sika deer. All regions surveyed had red deer present. In the region where sika deer had been present for more than a decade, there was evidence of poor mountain beech seedling regeneration. In the four regions without sika deer, there was evidence of a strong regenerative response at stands with low occupancy by trees. When compared to larger deer species, sika deer have a digestive morphology allowing greater dietary versatility, which may result in them impeding forest regeneration where red deer do not. In contrast to mountain beech, some small-leaved shrub species may have been competitively advantaged by intensive browsing from sika deer. This is contra to a current view that small-leaved shrub species with interlacing branches were able to tolerate browsing from extinct ratite birds, but not introduced deer. Sika deer have been introduced into countries where other deer species are indigenous, such as Canada, Denmark, Great Britain, Czech Republic, Ireland and the United States of America. Because of their dietary advantage, sika deer may have a greater potential to impede forest regeneration and competitively exclude larger deer species, particularly at low basal area sites where impacts on tree regeneration are likely to be greatest.     

Landscape of fear in Europe: wolves affect spatial patterns of ungulate browsing in Białowieża Primeval Forest,Poland

Large carnivores can either directly influence ungulate populations or indirectly affect their behaviour. Knowledge from European systems, in contrast to North American systems, on how this might lead to cascading effects on lower trophic levels is virtually absent. We studied whether wolves Canis lupus via density‐mediated and behaviorally‐mediated effects on their ungulate prey species influence patterns of browsing and tree regeneration inside the Bia?owie?a National Park, Poland. Browsing intensity of tree saplings (height class < 150 cm), irrespective of tree species or forest type, was lower inside a wolf core area (50.5%) where predator presence is highest, than in the remainder of the wolf pack’s home range (58.3%). Additionally, browsing intensity was reduced when the amount of coarse woody debris (CWD), which can act as a ?ungulate escape impediment?, increased (within 5‐m radius) inside the wolf core area. No relationship existed outside the core area. As a result, the proportion of trees growing out of herbivore control increased more strongly with increasing amount of CWD inside compared to outside the wolf core area. This suggests that next to direct effects of wolves on ungulate density caused by a higher predation pressure inside the core area, risk effects are important and are enhanced by habitat characteristics. These results indicate that behaviorally‐mediated effects of predators on prey can become more important than density‐mediated effects in affecting lower trophic levels. This is the first study we are aware of, that shows CWD can create fine‐scale risk effects on ungulates with the potential for cascading effects of large predators on patterns of tree regeneration for a European forest system. This knowledge broadens the discussion on how the impact of large predators on ecosystem functioning depends on the physical landscape, by illustrating these effects for a system which largely contrasts in this respect to the North American systems.     

Effects of Deer Browsing, Fabric Mats, and Tree Shelters on Quercus rubra Seedlings

We examined the effects of deer browsing, fabric mats, and tree shelters on the growth and survival of Quercus rubra (northern red oak) seedlings planted as part of a reforestation project in southcentral Minnesota. Browsing by white-tailed deer occurred on 68.6% of the seedlings growing without tree shelters, whereas only 3.6% of the trees were browsed by mice or voles and 2.5% were browsed by rabbits. Fabric mats, used to control competition from herbaceous plants for roughly half of the seedlings included in the study, had a detrimental effect overall. Seedlings grown with mats had a greater frequency of deer browsing and a greater chance of dying than seedlings grown without mats. Stem height for seedlings browsed by deer was less overall than for nonbrowsed seedlings, although this pattern varied with use of fabric mats and plot location. The use of plastic tree shelters effectively prevented deer browsing and reduced the mortality rate from 34.6% to 3.2%. Our results indicate that fabric mats should not be used in restoration projects with large deer populations. They also suggest that planting seedlings away from existing forest edges and using seedling protection measures such as tree shelters will increase seedling survival and growth in future restoration projects.     

Introduced red deer reduce tree regeneration in Pureora Forest, central North Island, New Zealand

This study uses data from repeatedly measured forest monitoring plots (20 × 20 m) (n = 32) and nine ungulate exclosures (paired fenced and unfenced plots; 20 × 20 m) to show the effects of introduced ungulates on tree regeneration in Pureora Forest Park, central North Island, between 974 and 2002. Results show that introduced ungulates, particularly red deer (Cervus elaphus scoticus), have suppressed the regeneration of hardwood species such as Coprosma grandifolia, Elaeocarpus dentatus, Griselinia littoralis, Melicytus ramiflorus, Schefflera digitata and Weinmannia racemosa. These species were only common as saplings and small trees in the complete absence of ungulate browsing. The results of this study suggest that red deer will need to be culled to low densities to assure regeneration of palatable tree species in Pureora Forest.     

Evidence of a spatial auto‐correlation in the browsing level of four major European tree species

1. The contribution of spatial processes to the spatial patterns of ecological systems is widely recognized, but spatial patterns in the ecology of plant‐herbivore interactions have rarely been investigated quantitatively owing to limited budget and time associated with ecological research. Studies of the level of browsing on various tree species reported either no spatial auto‐correlation or a small effect size. Further, the effects of disturbance events, such as hurricanes, which create large forest openings on spatial patterns of herbivory are not well understood.
2. In this study, we used forest inventory data obtained from the federal state of Baden‐Württemberg (Southern Germany) between 2001 and 2009 (grid size: 100 × 200 m) and thus, after hurricane Lothar struck Southern Germany in 1999. We investigated whether the browsing level of trees (height ≤ 130 cm) in one location is independent of that of the neighborhood.
3. Our analyses of 1,758,622 saplings (187.632 sampling units) of oak (Quercus), fir (Abies), spruce (Picea), and beech (Fagus) revealed that the browsing level is characterized by a short distance spatial auto‐correlation.
4. The application of indicator variables based on browsed saplings should account for the spatial pattern as the latter may affect the results and therefore also the conclusions of the analysis.

     

Tree fern trunks facilitate seedling regeneration in a productive lowland temperate rain forest

Seedling regeneration on forest floors is often impaired by competition with established plants. In some lowland temperate rain forests, tree fern trunks provide safe sites on which tree species establish, and grow large enough to take root in the ground and persist. Here we explore the competitive and facilitative effects of two tree fern species, Cyathea smithii and Dicksonia squarrosa, on the epiphytic regeneration of tree species in nutrient-rich alluvial forests in New Zealand. The difficulties that seedlings have in establishing on vertical tree fern trunks were indicated by the following observations. First, seedling abundance was greatest on the oldest sections of tree fern trunks, near the base, suggesting that trunks gradually recruited more and more seedlings over time, but many sections of trunk were devoid of seedlings, indicating the difficulty of establishment on a vertical surface. Second, most seedlings were from small-seeded species, presumably because smaller seeds can easily lodge on tree fern trunks. Deer browsing damage was observed on 73% of epiphytic seedlings growing within 2 m of the ground, whereas few seedlings above that height were browsed. This suggests that tree ferns provide refugia from introduced deer, and may slow the decline in population size of deer-preferred species. We reasoned that tree ferns would compete with epiphytic seedlings for light, because below the tree fern canopy photosynthetically active radiation (PAR) was about 1% of above-canopy PAR. Frond removal almost tripled %PAR on the forest floor, leading to a significant increase in the height growth rate (HGR) of seedlings planted on the forest floor, but having no effects on the HGRs of epiphytic seedlings. Our study shows evidence of direct facilitative interactions by tree ferns during seedling establishment in plant communities associated with nutrient-rich soils.