Protection and Management Plan for the Millbrook Marsh Nature Center

ivyrule.gif (786 bytes)

75

Recommendations

Ecological

Hydrology and Aquatic Habitats

• Adoption of best management practices (BMPs) for stormwater in the Spring Creek watershed is critical for the long term health of the Millbrook’s streams and wetlands. Thus, we recommend that adoption of BMPs for stormwater be considered by the surrounding municipalities and University.

• The community should strive to eliminate any additional stormwater flows and work toward reducing existing storm surges into the marsh. Enhanced infiltration and detention of stormwater in the developed portions of the watershed are paramount to protect Millbrook Marsh.

• We recommend that a shallow wetland/pond be constructed on Farm 12 in the near future to complement the adjacent aquatic habitats on site and enhance interpretive opportunities.

Stormwater Management. Stormwater runoff into the streams and wetlands of the Millbrook Marsh Nature Center will continue to have negative environmental impacts as the State College area becomes more urbanized. There is evidence that high stream flows have altered the morphology of Slab Cabin Run through downcutting of the stream bottom. In addition, heavy sediment loads originating in upstream urban and agricultural areas may be depositing abnormally high loads on the wetland during flood events. This could potentially alter the composition of the plant community within the site.

The macroinvertebrate communities of 5 of 6 sampling stations reflected moderate degradation, which is probably due to impaired water quality associated with this urban runoff. Besides the usual mixture of heavy metals and petroleum derivatives in stormwater entering Thompson Run, there is a substantial sediment load in the stream originating from the ditch that conveys stormwater to the Duck Pond. Stabilization of this ditch, as planned, will greatly benefit the aquatic communities in the wetland complex. 

Shallow Wetland/Pond Construction. The creation of a shallow pond and fringing marsh would provide an intense aquatic study site in the appropriate Transition-Buffer Zone, relatively free from the constraints evident in the more dispersed natural wetlands beyond ("K" on Map 9). A pond would also be consistent with the heritage of farm ponds in the region. Pond studies for school groups and researchers alike could be accommodated structurally by a small deck, limestone slabs, or geotextile-underlain gravel. Native plantings could mimic a range of plant associations: floating and emergent hydrophytes, shrub-scrub and, along the north side, overhanging trees. If connected via a planted corridor to the adjacent woodlot and wetland, one could expect a variety of riparian and upland related fauna. Migratory waterfowl and waterbirds, too, might use the pond on a seasonal basis, thus further enhancing opportunities for close observation of wildlife.

Several pond locations have been considered. Initially, it seemed logical to position the pond in the northeast field of Farm 12, adjacent to the small wetland (#56) and hydrologically connected to the Bathgate Spring Run. Upon further analysis, however, it has become evident that this could disrupt surficial flow which may be partially feeding the wetland. A second, equally strategic position would be in the field to the rear of the barnyard. Preliminary indications of a wet depression suggest a mid-point position between the two fence lines. If this site proves sufficient to sustain a shallow marsh pond, there would be excellent potential to link this habitat with an extended boardwalk and other pathways.

76

Pond size and morphology should reflect objectives of high biodiversity and water quality and low maintenance. Generally, a pond 0.1 ha (0.25 ac) in surface area will accommodate a suitable range of breeding macroinvertebrates, selected fish species, herptofauna and several species of migratory waterfowl. Water budget calculations will need to consider infiltration, evaporation, transpiration and storm flows (in and out) to ensure adequate turnover and good water quality. In general, 25 to 50 percent of the water surface area should be between 0.5-lm deep (2-3 feet)(Brittingham 1991). Shallowly sloping, edges for amphibian and reptile access and shorebird feeding, subsurface shelves to support aquatic plants, a deeper hole for overwintering fish, and a waterfowl nesting island and nesting structures could all be incorporated into the overall pond design (USDA-SCS 1982, Brooks et al. 1993).

Vegetation

• We recommend a proactive program to control invasive species, including the active reintroduction of native plant associations.

• A Vegetation Management and Restoration Plan should be prepared for the restoration and creation of specific, desirable native plant communities and wildlife habitats. This can be accomplished, in part, by encouraging natural succession and planting desirable native species as needed.

Native Species. Native plant species, along with aquatic habitats, form the basic building blocks for a healthy, functional biotic environments in the Millbrook Marsh Nature Center. As elaborated in section on wetland and vegetation, native plant communities on site still exhibit some characteristics of species richness and integrity, particularly wetland areas. However, human-induced disturbances have significantly fragmented and impoverished most of the site. Agricultural and landfill activities around the fringes of the marsh have created a situation of ecological stress, and disturbed hydrological regimes are prevalent in all corners of the site. Invasive plants, as discussed above, have in some areas of the site entirely replaced native plant communities. Thus, the "management" of native species is more a matter of directing interventions that seek to ameliorate those conditions which limit the ability of nature to heal itself. This will require re-establishment of more natural hydrologic regimes and protection of existing soil resources. The recommendations are discussed in greater detail below.

Management of Invasive Plants. Because a range of natural resource values are at stake, it is wise to begin assuming responsibility for management and control of invasive plants in Millbrook Marsh. Three approaches can be identified. For comparative purposes, several of the species of primary concern in Millbrook Marsh will be addressed in each ("A" on Map 9).

1) Direct management primarily involves application of physical, chemical and biological interventions to the stand of invasive plants. Simple physical removal is one apparent solution, but will quickly prove unsatisfactory because of the shrub’s ability to resprout from any remaining root stock. More frequently, managers have advocated direct management through chemical means: the application of a non-persistent systemic herbicide, such as Roundup®, to cut stems to effect a complete kill. Biological interventions—usually the introduction of a herbivorous insect originating from the same source as the invasive host plant—are often considered where physical and chemical approaches have shown to be ineffective, for example, with purple loosestrife.

Directing ecological succession is a common secondary management effort, often entailing the inter- or under-planting of robust, early successional native woody and herbaceous species. The objective is to out-compete the invasive plant. For example, after an initial round of honeysuckle eradication, riparian species such as ash (Fraxinus spp.), willow (Salix spp.) or red-osier and gray dogwood (Cornus spp.) can be interplanted to prevent the honeysuckle from re-sprouting through canopy closure and root competition.

77

2) Indirect ecological management is gaining momentum as a more holistic approach to dealing with aggressive, non-indigenous plants. It stresses the importance of early plant detection, modification of human activity, and manipulation of system attributes as components of an ecological approach to modifying plant invasion (Hobbs and Humphries 1995). Unfortunately, the existing disturbance regime in Millbrook Marsh is already well-established: flashy hydrology, landfilled areas, a ready seed source from surrounding suburban landscapes, and a degree of on-site trampling. For example, there is little precedent for successful indirect management of honeysuckle where extensive thickets are already well-established. A literal interpretation of this approach would see the wholesale removal of deposited fill and the reestablishment of pre-disturbance landform, soil and hydrologic regimes. Indications are that this would essentially result in the restoration of the fen environment.

Once control and native revegetation tactics prove successful, indirect ecological management becomes more feasible. For example, people and deer management would reduce trampling and exposure of an open seed bank, thus limiting the potential for invasives such as garlic mustard to become re-established.

3) The do nothing option is a management option that is gaining renewed interest, albeit with ecological overtones this time around. Eastman (1995) argues that there is no definitive solution to the challenge of invasive species and, therefore, we should tolerate them as best we can. He cites community self-regulating mechanisms (herbivory, diseases, etc.) for their ability to moderate invasive dominance over the long term. The invasive plant is eventually integrated into a stable but changed ecosystem.

This perspective would be pragmatic in its recognition of Millbrook Marsh as a disturbed ecosystem within a quickly urbanizing watershed. On the other hand, this stance can be seen as fatalistic - it denies the real progress made in restoration ecology over the last decade. And it belies an inherent pessimism: that any human intervention is, in the long run, bound to fail. Finally, it under-emphasizes the extensive damage invasive plants have inflicted on many stressed ecosystems across the continent.

We would suggest that the application of new ecological knowledge within an adaptive management approach - where monitoring and responsive-but-cautious interventions are essential to both long-term success and a heightened understanding of ecosystem dynamics - is the most appropriate approach for Millbrook Marsh.

Priorities for control or removal of invasive species should be directed at those that pose the greatest ecological threats, namely those that:

• replace key indigenous species;

• substantially reduce indigenous species diversity, particularly with respect to the species richness and abundance of conservative species;

• significantly alter ecosystem or community structure or functions;

• persist indefinitely as sizable sexually reproducing or clonally spreading populations;

• are very mobile and/or are expanding locally (SER 1994).

Continuous monitoring over a period of several years may be necessary to check for the spread of small clumps and/or recurrences after implementation of control measures. Aerial photography and field measurements are essential in building a geo-referenced data set on abundance and density.

78

Tracking with a global positioning system (GPS) should be considered for its ability to pinpoint even subtle expansions or contractions of plant populations. The principles and priorities noted above should be considered in advance of any action in the field. The most environmentally benign approach should be used whenever possible.

Control Measures by Species. Honeysuckle - Light infestations may be cleared by hand with a shovel or grubbing hoe, provided the entire root is removed. However, since honeysuckle tends to stock its own prodigious seed bank, any soil disturbance is likely to be reflected in multiple sprouts the following spring. As noted previously, a single cut is inadvisable because of the plant’s capacity to resprout. If a long-term commitment is made, a severe initial cut plus on-going clearing of seedlings and resprouts may be effective (Luken and Goessling 1994).

Severe infestations may be controlled by application of glyphosphate herbicide. Glyphosphate, like all biocides, should be used with restraint and applied carefully. However, should other means prove ineffective, spot application can be used with minimal impact on ecosystem functions, since glyphosphate is a low-toxicity herbicide that does not bioaccumulate (Monsanto 1994, Evans and Eckardt 1995). Applied to cut stumps late in the growing season usually results in a high mortality rate. Glyphosphates are a non-selective, affecting all green vegetation, so broadcast spray application is not recommended in sensitive areas.

Prescribed burning has been used to control honeysuckle. Generally, bums conducted during the growing season will top-kill shrubs and inhibit new shoot production. These bums will also favor warm-season grasses and perennial forbs increasing species diversity and productivity. Treatment of any severely infested areas usually requires management for a period of three to five years to inhibit the growth of new shoots and eradicate target plants (Williams 1995). Obviously, burning should be conducted with great care and full disclosure to the surrounding community before proceeding.

Multiflora rose - Active control may be necessary where meadows and natural succession are to take place on Farm 12 and around the periphery of the marsh. Periodic mowing and light grubbing are common for small or juvenile infestations, but soil disruption should be kept to a minimum. More mature and entrenched stands will require other means. Cutting to the main stem and direct application of a glyphosphate herbicide during peak growth seems to be the most popular and effective means of control (Evans and Eckardt 1995). Considering the typical regenerative abilities of the rose family, monitoring of control measures is important.

Garlic mustard - Establishment of garlic mustard should be a primary concern in any areas above-gradient of the wetland slated for revegetation, boardwalk or trail construction, and non-saturated soils subject to foot or vehicular traffic. Control has included careful hand-pulling prior to seed ripening, but severe infestations are most often treated with herbicides. Since it is a biennial, its first year basal rosette can be treated with a glyphosphate in the late fall, although follow-up treatment is usually required. Prescribed burning early in the growing season has also been used with some success but, again, follow-up treatment of new sprouts are required. In his search for biological controls, Blossey (1997) has noted that all of these methods are most effective only in the short term. Furthermore, some infested areas on Millbrook Marsh may be mixed within a fire intolerant community.

Autumn olive - Because of Autumn olive’s high seed germination rate and ability to resprout if cut or burned, there are few effective control methods at the present time besides herbicide application Spot application of glyphosphate to cut stumps in early fall is effective. Other managers have used basal injections of triclopyr at very low concentrations, with good success (Eckardt 1995).

Common buckthorn - Recommended control methods for buckthorn are similar to Autumn olive, due to high seed dispersal rate vigorous resprouting following top removal. Burning and hand pulling are generally not considered effective; there has been some success winter girdling, as well as underplanting of hardwood saplings to effect light competition (Converse 1995). The most effective means of control is the spot application of glyphosate herbicide to stumps early in the Fall. Glass (1994) has found low-volume basal bark treatment with triclopyr (Garlon 4®) 97 percent effective.

79

Purple Loosestrife - An effective control for purple loosestrife has yet to be found. Early detection of the plant is important as small populations are more successfully controlled than large, entrenched populations. For small stands, uprooting by hand is often recommended. Pulling should be completed before flowering so as not to scatter seed, and plant fragments should be minimized. Burning is the preferred method of disposal. Follow-up treatments are often necessary on new plants which sprout from seed persisting in the ground. Digging plants out is not recommended as this creates disturbance, which may favor the spread of the species. Other conventional means — mowing, burning, and herbicide application— have proved expensive and environmentally degrading (Blossey 1997). Treatment with Rodeo® has often been used to control second and third generation seedlings which often follow initial eradication.

Cornell University’s Center for Biological Control of Non-Indigenous Plant Species is spearheading a campaign using natural biological enemies to prevent purple loosestrife from expand its range. The objective of this biocontrol program is to restore the self regulatory potential of plant-insect interactions, which, in the case of purple loosestrife requires the use of European insects. Four host-specific insect species (two weevils and two leaf-eating beetles) have been approved by USDA for release in infested areas. The strategy is to achieve long-term control, not complete eradication (Blossey 1997). Note, however, that Eastman (1995) contradicts these findings, contending that European results using biological control have been unimpressive. A more detailed control program, including monitoring suggestions, based on work by Blossey (1997) is available from the authors.

Japanese knotweed - Because this plant is poised just upstream along Slab Cabin Run, monitoring should key into this species. Establishment of a few individuals can be prevented by manually removing the entire plant. Small stands may be controlled by repeated cutting, which may need to be supplemented by native revegetation once growth has been reduced (Seiger 1995). The City of Toronto (1997) is conducting controlled test plots, wherein volunteers hand-pull the weed and revegetate the disturbed soil with fast-growing willow (Salix) saplings. 

Native Plantings and Field Management

Upland Forest Restoration. The regeneration of small, interconnected forest stands is crucial to the overall long-term integrity of Millbrook Marsh. Enhanced forest habitat, improved ecological transition from cultural landscapes to wetland settings, strengthened visual buffers and spatial definition, moderated microclimate for human activities—all build a compelling case for the expansion of forest cover.

From a landscape ecology perspective, an increase in the size of the Farm 12 woodlot complex, to perhaps double its existing area, would result in much enhanced forest habitat. This strategy would effect a better balance of ecosystems; currently, "edge" habitats and related generalist species dominate the site. Woody plantings shown as "G" and "H" on Map 9, would serve to link the Niebel’s woodlot with the pond, providing a conduit for wildlife and "animating" barn based nature studies. Plantings in the back field ("I" on Map 9) would serve to bolster the size of the Niebel woodlot, serve as an added buffer to outdoor activities, and create a series of outdoor "rooms" that are more human-scale than the current setting. The result is a new naturalistic environment for both humans and wildlife characterized by a high level of variety and choice.

80

Another area of key interest is the old grove of Bur oak along Slab Cabin Run. These few specimens are over mature and seedlings and saplings are not present, possibly due to browsing pressure by deer. Thus, it is not too soon to consider a replacement plan of planting similar native seedlings and saplings in their vicinity. Since this grove has been an attractant to local users for years, the proposed boardwalk and path is designated to bring visitors to this point. Keeping this scenic place intact will preserve a unique feature of the site and tie it into other old growth trees found at Slab Cabin Park.

A number of techniques can be considered to establish wooded growth. One effective but materially intensive method, combines the planting of a closely-spaced nurse species of hybrid poplar (Populus spp.) with heavily mulched underplantings of native hardwood seedlings and saplings. Canopy closure is quick, reducing weed competition and limiting the affects of rodent girdling. After several years, the poplars are culled and the near-caliper sized hardwood trees are ready to mesh canopies (HSW 1989). Squirrels and birds help to recruit acorns and other seeds from nearby sources. Human volunteers can perform a similar function.

Other methods include the use of translucent plastic sapling shelters (Tubex®), which guard against rodent girdling, provide a protective microclimate, and somewhat limit weed competition and deer browsing. These are popular locally and across the northeast in reforestation projects involving oak (Quercus spp.) and hickory (Carya spp.). "Nucleation" entails the planting of clumps of larger caliper trees which are ready to produce seed; these are then permitted to regenerate outward from the center of the clump. Conventional planting of saplings in small mulched tree pits may also be tried, however, mortality is often high due to impacts from weeds and rodents and inhospitable microclimatic conditions. It is recommended that a Vegetation Management and Restoration Plan be prepared to provide an in-depth, phased approach to reforestation and other plant communities described below.

Sustainable Pastures and Native Meadows. Map 9 shows the recommended landscape management approach for Farm 12 pastures that are to be adaptively re-used as open space. Two basic vegetation covers are shown ("F" and "H"), each requiring conversion from the current dominant non-native cool-season grasses: F - new warm-season grass pastures along the cultural "front face" of Farm 12; and, H - native mixed grass/wildflower meadows in open areas of the rear fields.

The latter provides an ecological transition habitat between pasture and woodland, and provides greater opportunity for wildlife-human interaction by obscuring movement of visitors. Aesthetically somewhat more unkempt than grass pastures, these plant communities are best located back from Puddintown Road. The native pastures, as noted below, may provide for both light grazing and hay production, while requiring almost none of the fertilizers and biocides typical of non-native crops.

Although the scenario of native prairies and meadows may seem inconsistent with the image of dominant forests in central Pennsylvania, there are historical precedents of persistent tall grass dominated landscapes in this region. The existence of a prairie community - the Great Plain of Penn’s Valley - was determined by Losensky (1961). His hypothesis was that these ecosystems were relics of a vast prairie peninsula extending eastward from Ohio that formed during a lengthy warm, dry climatic period several thousand years ago. It apparently covered an area of several hundred hectares up until the time of settlement in the late 1700s. The area supported a number of species normally associated with prairie habitats of the Midwest, including native warm season grasses such as grama grass (Bouteloua curtipendula) panic grasses (Panicum spp.), Kalm’s brome (Bromus kaimii), little bluestem (Schizachyrium scoparium), Indian grass (Sorghastrum nutans), and big bluestem (Andropogon gerardiO. The last two species have been noted as excellent forage plants (Brown 1979); in fact, the feed value of warm-season grasses is generally comparable to that of non-native cool season grasses (Ernst 1996). Representative herbs in the Great Plain included whorled and green milkweeds (Asclepias verticillata and virid flora), longfruited anemone (A. cylindrica) and hoary vervain (Verbena stricta) (Losensky 1961).

81

Planting of native pastures should occur in late spring once soils have reached at least 55, F. The conventional approach entails seeding with a no-till drill preceded with herbicide application to control existing perennial pasture grasses. A more benign, but less instantaneous, method entails what is known as "successional restoration" - the seeding of scarified, unplowed pasture with conservative native species. This method can take up to 5 years to yield results, since it relies on the building of below-ground root biomass as native prairie species prepare to out-compete non-native grasses and weeds (Packard 1994). The proposed mixed meadow will require the introduction of herbs, resulting in heightened biodiversity and providing for enhanced wildlife and insect cover and food. Native wildflower and grass seeds may be mixed and applied with a broadcast seeder. Where a higher intensity of color and more immediate habitat value is needed rooted plugs can be installed in "drifts". Plugs should be contract grown in advance, using seed or cuttings collected from local sources.

Prescribed fires are the most ecologically appropriate means of managing native meadows and grasslands. Controlled fires every several years also bum off duff and reduce the probability of a natural fire outbreak. If fire is selected as a management tool, it is recommended that a fire management strategy be developed in conjunction with the local fire officials and neighbors, and be incorporated into the recommended Vegetation Plan.

Mowing at a height of 15 to 20cm (6-8 in) is a fall-back method of keeping woody succession at bay and ensuring persistent herbaceous growth. A late winter or early spring cut is preferred for several reasons: it coincides with the seeding period of warm-season grasses, it stretches winter habitat availability, and it maintains the beauty of tall grass meadows as long as possible (Lambert 1992). Mowing should be scheduled to minimize interference with breeding grassland and forest-edge birds; after July 15 is a reasonably safe date, although later is even better.

The detailed Vegetation Management and Restoration Plan should include a strategy for establishment, maintenance and monitoring protocols for proposed native pastures and meadows. Related issues such as soil impoverishment, seed collection of local genotypes (within a 15 to 20 mile radius), and on-site propagation using volunteer help should also be an integral part of the implementation process. 

Old Field Succession. Perhaps one of the most dynamic ecological lessons to be learned on site is one that requires relatively little human effort and resources: old field succession, or natural change. Map 9 ("I") designates several areas for transition from pasture/woodlot edge to herbaceous and woody old field. A variety of worthy educational themes would be available:

underlying geologic and soil processes, nutrient and energy cycling, climatic and microclimatic influences, the role of birds and other wildlife, etc. Larger ecological debates—for instance, "balance-of-nature" versus "flux-of-nature~~ views— are neatly encapsulated in any landscape undergoing successional change. Young students would be especially likely to benefit from "growing up" with successional ecosystems on site.

Areas designated as successional old field are sited adjacent to forested areas so as to provide a continuum of plant communities from pasture to woodlot. This also serves to canopy cover in wooded communities, and helps strengthen the ecological corridor that links the proposed nature center with the existing woodlot. Successional areas are generally intended to move through all stages over the next century or so. Three to four decades of growth should result in early successional forest. Representative species might include ash (Fraxinus spp.), elm (Ulmus spp.), red cedar (Juniperus virginiana) , bigtooth aspen (P. grandidentata), black cherry (Prunus serotina) and gray birch (Betula populifolia), and red maple (Acer rubrum), among others. As the site program, facilities, and landscape setting are tested through time, a clearer image of spatial requirements will be formed. Active management (ecological disturbance) may then be required in some areas to "freeze" or recycle succession in order to accommodate other objectives of the Center.

82

Riparian Vegetation. Planting of streambank vegetation provides riparian habitat, improves bank stability, enhances water quality, shades the stream and contributes organic detritus to the benefit of aquatic organisms (Davis et al. 1994). Generally, the riparian edge along all streams on site should be well-vegetated and vertically stratified. Streamside plantings can be used in conjunction with careful regrading to bio-remediate bank failures evident primarily along Slab Cabin Run. The Bathgate Spring Run on Farm 12 proper should likewise have a vegetated buffer strip; it should be primarily herbaceous, to ensure the continuity of open space in this area. Open reaches along the larger streams may be revegetated without the use of fencing, but narrower sites, such as Bathgate Spring Run, should be defined with woven wire fencing to delimit the extent of mowing/prescribed burning of adjacent fields and to keep any grazing livestock well back from open water. Since the list of appropriate riparian plants is a long one, a native species palette should be determined during the preparation of detailed planting plans (Banski et al 1996). Seed, rooted cuttings and containerized plants should be obtained from on-site or locally-available stock whenever possible.

Hedgerows. Most hedgerows on site are relatively free of woody plants, indicative of recent ongoing maintenance. Generally, the plan envisions that natural succession would be permitted along fence lines. This would permit reduced mowing and spraying, enhance edge habitat, define outdoor activity spaces and screen undesirable views. Appropriate native woody species could be installed to speed up screening where it may be most desirable; species could include the native cherries (Prunus spp.), hawthorne (Crataegus spp.), native roses (Rosa spp.), raspberries (Rubies spp.), sumac (Rhus typhina), viburnums (Viburnum spp.), serviceberry (Amelanchier spp.), red mulberry (Morus rubra), aspen (Populus tremuloides) and red cedar (Juniperus virginiana). Since fences represent perches from a bird’s perspective, these strips will have to be monitored for invasive species such as honeysuckle, privet (Ligustrum spp.) and Multiflora rose, among others.

Specialized plantings. Strategic plantings for wildlife habitat can focus activity in key areas. One important location for such installations would be the habitat demonstration area shown on Map 9 "L". For example, nectar-producing plants can be concentrated to attract butterflies (old field and edge habitat would be required nearby to provide plant hosts to accommodate caterpillar feeding and reproductive activities). Planting of native evergreens and creation/installation of snags (standing dead trees) would attract insectivore birds, provide winter cover, and offer opportunities for wildlife observation, particularly in concert with the feeding stations.

Wildlife and Fisheries

Most wildlife species respond to habitat structure, primarily vegetation, so a "build/plant it and they will come" strategy should work. Several habitat projects have already been discussed, including the shallow wetland/pond, expanding woodlot, and native meadows. These areas can be enhanced by providing suitable nesting and denning structures. The following items are appropriate for consideration at the Millbrook Marsh Nature Center.

Osprey restoration. Osprey frequently migrate through the Spring Creek watershed, spending time at downstream fish hatcheries, but no recent nests have been reported. Removal of antiquated and unused utility lines could make artificial snags available for perching and nesting. The best location would be in the center of the marsh near the confluence of Slab Cabin and Thompson Runs. This is a centrally located and highly visible area, but affords the birds adequate protection from human disturbance. A closed-circuit video camera (proposed in a later section) could provide visitors in the barn with a "bird’s eye" view.

83

Barn owl restoration. The existing silos on site provide a unique opportunity to restore a pair or two of barn owls to the area. This species has not fared well in urbanizing settings as lawns have replaced fields, and malls have replaced barns. Since visitor access to the tops of the silos is unlikely despite their appeal as observation towers, they provide a perfect nesting site for these nocturnal raptors. The meadow vole population is very high in the surrounding fields, so one of their preferred prey species is readily available.

Brook trout restoration. Thompson Run undoubtedly supported native brook trout sometime in the past. The large inputs of groundwater from Thompson Spring, Bathgate Spring, and other springs along this reach help maintain cool water temperatures in the summer. The establishment of a wild brook trout population would require removal of the existing brown trout and rainbow trout from Thompson Run, which could be easily accomplished by repeated use of electrofishing gearL Probably, it would be necessary to close Thompson Run to fishing, or at least prohibit harvest, because brook trout are vulnerable to intensive fishing pressure and their presence may attract anglers. Native brook trout from Buffalo Run or Galbraith Gap (both locations within the Spring Creek watershed) could be transplanted into Thompson Run. Periodic surveys would be needed to monitor the brook trout and remove encroaching brown trout.

Nest boxes for other birds and mammals A bluebird nestbox trail could be established in the fields within the Transition-Buffer areas of the site along existing and proposed paths. The central portion of the wetland and riparian corridors could support several wood duck nestboxes, and this species has been observed frequently during migration. A purple martin house would be appropriate in the vicinity of the barn, providing not only opportunities for observation, but an opportunity to discuss the role of insectivorous birds in controlling flying insects, particularly mosquitoes (which, by the way, are not much of a problem within Millbrook Marsh). Bats play a similar role in insect control, so encouraging summer maternity colonies to use bat houses located on buildings or large trees can be beneficial and educational (Williams-Whitmer and Brittingham 1996). Plans for building artificial nesting structures can be found in Hassinger (no date).

Bird feedin2 area. Watching birds feed is one of the most common and enjoyed interaction between people and wildlife. We suggest that over time, a bird feeding area be established for year-round viewing. The most likely place in the vicinity of the barnyard at the back of the barn where it could be incorporated with the structures and plantings in the Habitat Demonstration Area (Map 9, "L") and perhaps relate to the orientation of a new building. It could be modeled after similar arrangements at Shaver’s Creek and Sapsucker Woods (Cornell University), with large viewing windows, bird identification charts, and microphones that transmit sounds through speakers to the viewers inside. The inclusion of shrubs and flowers appealing to hummingbirds and butterflies would expand the activity into the summer season, when attracting birds to feeders is usually more difficult.

Shallow wetland/pond and natural streams. A number of habitat features could be incorporated into these aquatic habitats, including basking logs for turtles, fish habitat structures, artificial nesting burrows for belted kingfishers, and the like. There is almost no limit to the number of improvements that can be made. This in itself should instill some caution in managers of the facility, so that the site does not become a cluttered demonstration area. Rather each design element should be blended into the whole fabric of the Center.

Dead and down woody material. Many wildlife species benefit from, and in some cases rely upon dead and down woody material for nesting, denning, and foraging. As trees and shrubs die or are cut down for various reasons, the material should be used. Standing dead trees, or snags support the breeding and foraging activities of dozens of birds and mammals and should be left standing provided they do not pose a safety hazard for visitors. Likewise, downed woody material such as logs and branches, provide cover for amphibians, small mammals, and their prey. As shrubs and brush are cleared for other management activities, brush piles can be constructed for wildlife cover.

84

Long narrow piles (2-3 m (6-10 ft) wide and 5-10 m (15-30 ft) long) or circular mounds (2-3 m (6-loft) in diameter are recommended. Partially buried piles of large logs and stones could serve as a hibernaculum for hibernating reptiles.

Hunting, trapping. and fishing. Local residents and visitors have engaged in these consumptive recreational activities in the past, and thus, there are cultural and historic connections to be considered. However, because the uses of the Millbrook Marsh Nature Center have shifted toward non-consumptive recreation and education, some concerns have been raised regarding the continuation of hunting, trapping, and fishing at the site. We recommend that consumptive activities not be immediately excluded, and that a dialogue with consumptive and non-consumptive users be initiated by the Centre Region’s Parks and Recreation Department. Involvement of the Pennsylvania Game Commission and Pennsylvania Fish and Boat Commission in these discussions is recommended. In addition to providing the benefits of outdoor recreation to the participants, consumptive activities provide a means to manage wildlife and fish populations at Millbrook Marsh should it be necessary to control excessive numbers (e.g., deer, raccoon, fox) or nuisance animals (e.g., burrowing by muskrat, flooding by beaver).

Hunting poses the greatest possible danger to visitors. A requited safety zone of 150 yd (450 ft) from buildings, legally limits the discharge of firearms to a relatively small area on the east side of the site along upper Slab Cabin Run. Once the boardwalk is in place in this portion of the site (see Map 8), hunting in this area may be ill advised. If allowed in the future, perhaps hunting should be strictly regulated through a permit system with limits on the number of individuals and the length of the season. Trapping, primarily for aquatic and riparian furbearers, poses less of a safety concern to visitors, particularly if only conibear-style and padded leg-hold traps are allowed. Although the monetary value of furs is minimal for an area this size, the option of using recreational or professional trappers to manage problem wildlife should be maintained. Again, a strictly regulated permit system seems appropriate. Anglers pose the least threat to visitors, and probably form the majority of consumptive users. Anglers should be encouraged to engage in only the most ethical and responsible practices. They could serve as a source of information through creel surveys and personal interviews. Once the boardwalk is installed, it seems appropriate that hunter, trappers, and anglers use it for access only, conducting their consumptive recreational activities away from this pedestrian path.

On to Cultural Recommendations