Keywords

1 Introduction

The nation behaves well if it treats the natural resources as assets which it must turn over to the next generation increased; and not impaired in value. – Theodore Roosevelt (1910)

Thousands of tired, nerve-shaken, over-civilized people are beginning to find out that going to the mountains is going home; that wildness is a necessity; and that mountain parks and reservations are useful not only as fountains of timber and irrigating rivers , but as fountains of life. – John Muir (1901, p. 15)

US national parks are firmly engrained into the culture of this country. Since the first national park was established in 1872, they have been central to conservation, education, and recreation in the United States (Sellars 2009). Within the US national park system lie diverse and iconic landscapes such as the geological formations at Arches National Park in Utah, the glacier-carved peaks of Glacier National Park in Montana, and the River of Grass in the Florida Everglades (Byerly 1996), as well as iconic species like the bison (Bison bison) of Yellowstone, the towering sequoia (Sequoiadendron giganteum) in Kings Canyon National Park in California, and the sleepy American alligator (Alligator mississippiensis) of the Everglades. National parks mean different things to different people (Mayo 1975). To some they are a safe haven for plants, animals, and wilderness. Even though they may not visit, the mere existence of such places, animals, and wilderness is righteous, invigorating, and quintessential to the culture of the United States. To others, summer’s arrival means packing the family car and setting out on the open road to participate in an American rite of passage. Yet for others, these parks represent places where not only species can get away from development and civilization, but they are also places where people can immerse themselves in sights, sounds, and smells of wild nature (Runte 1997). Regardless of different perceptions of US national parks, their popularity has never been greater with a record high of 330 million people visiting in 2016 (NPS Stats, https://irma.nps.gov/Stats/Reports/National). Furthermore, US national parks are an international destination with over 13 million international visitors in 2015 (about 35.4% of total visitors, U.S. Travel Association).Footnote 1

While the US national parks set aside land for some of the most awe-inspiring landscapes and species in North America, simple park designation does not ensure the preservation of such natural resources (Parrish et al. 2003). National parks face many of the same threats to biodiversity as non-protected areas, including non-native species invasion (Allen et al. 2009; Stohlgren et al. 2013; Abella 2014). This is exemplified by the greater than 3700 non-native plant species that have been identified in US national parks (Allen et al. 2009). While the majority of non-native species are not considered “major” threats to US national parks (Hiebert and Stubbendieck 1993), there are textbook examples of non-natives that have become invasive and are greatly altering ecosystems. The invasive white pine blister rust (Cronartium ribicola) has greatly increased tree mortality across the Western United States including within national parks (Van Mantgem et al. 2009). The exotic pest hemlock woolly adelgid (Adelges tsugae) is an aggressive invader in eastern North America that threatens to eliminate eastern hemlock (Tsuga canadensis) from its range, which includes 21% of all parks in the US national park system (Abella 2014). Eastern hemlock is a foundation species, which means that it determines community composition (i.e., which species are present) by physically structuring ecosystems and modulating ecosystem processes (Ellison et al. 2005). As a result, the loss of hemlock has cascading effects on food web structure (Baiser et al. 2013), arthropod assemblages (Sackett et al. 2011), and avian community composition (Tingley et al. 2002). The list of harmful invaders also includes species such as the invasive Burmese python (Python molurus bivittatus), which has caused severe mammal declines in Everglades National Park (Dorcas et al. 2012; McCleery et al. 2015).

It is clear that non-native species can have detrimental effects on the ecosystems they invade. At larger scales (e.g., continental), non-natives can also fundamentally alter patterns of biodiversity. As non-native species establish in new locations and either replace or join native species , ecological communities become more similar in their composition (Olden and Poff 2003). As a result, non-native species play a major role in the process of biotic homogenization (McKinney and Lockwood 1999; McKinney 2004; Qian and Ricklefs 2006). Biotic homogenization has implications for US national parks that are valued for their uniqueness and representation of local flora and fauna.

While the homogenization of plants and animals across US national parks undermines the biological distinctiveness of these important resources, it may have implications beyond simply changing the species that park visitors encounter. Non-native species carry a set of traits and evolutionary history to the invaded locale. If non-native species become widespread and reach multiple locales, they introduce similar traits and evolutionary histories to the invaded assemblages that can result in functional and/or phylogenetic homogenization. Furthermore, because successful non-natives are often nonrandom in regard to taxonomy (Lockwood 1999; Cadotte et al. 2006; Pysek et al. 2017) and tend to possess similar trait values (Pyšek and Richardson 2008; Blackburn et al. 2009), disparate locales that receive different species can still functionally and/or phylogenetically homogenize (Baiser and Lockwood 2011). These two aspects of homogenization are important yet understudied components of global change (Olden et al. 2004) and can give contrasting information regarding changes in biodiversity (Monnet et al. 2014). Critically, functional and phylogenetic homogenization may hold the key to predicting the ecosystems of the future because they quantify the functional and evolutionary potential of ecosystems to adapt to changing conditions.

Here we assess phylogenetic and taxonomic homogenization of bird and vascular plant assemblages in US national parks. We use species lists for over 240 park units to explore how non-native species have altered biodiversity patterns, asking if non-native species have led to an increase in taxonomic similarity (i.e., taxonomic homogenization) and phylogenetic similarity (i.e., phylogenetic homogenization) across all US national parks and across only those parks within the same ecoregion.

2 Methods

2.1 Study Sites

The National Park Service (NPS) of the United States is actively assessing and managing resources for more than 270 park units through its Inventory and Monitoring (I&M) Program (Fancy et al. 2009). The I&M Program performed an inventory of species occurrence in the 2000s for these parks with the help of local taxonomic experts. During these inventories, species’ provenances (native or non-native) were assigned and then carefully reviewed before merging them into the NPSpecies database (https://irma.nps.gov/NPSpecies/). We used the occurrence and provenance of vascular plants and birds from the NPSpecies dataset because these two groups are well studied and likely have the most complete and accurate records. We only kept records listed as “present” and with a clear provenance as either “native” or “non-native.” For birds, we also removed species that were tagged as “vagrant” and then cross-validated all other species in terms of their provenance with an up-to-date avian invasion atlas dataset (GAVIA; Dyer et al. 2017). We extracted a list of successful introduced species (tagged as “breeding” or “established”) to the United States from GAVIA. We then verified non-native species in the NPS dataset with this list. For both plants and birds, we removed parks from Hawaii since they are far away from other parks and have not been classified into ecoregions (more details in the ecoregions subsection). In total, we obtained a list of 14703 plants from 241 park units and a list of 729 birds from 244 park units. For more details about the dataset, see Li et al. (2018).

2.2 Phylogenies

In order to assess changes in phylogenetic similarity across US national parks, we situated the species extracted above with previously published phylogenetic trees. Phylogenetic trees detail the evolutionary history of a given set of organisms based on their DNA. For plants, we built a phylogeny using Phylomatic v4.2 (Webb and Donoghue 2005) and the supertree “zanne2014,” which is an up-to-date phylogeny for >32,000 species based on 7 genes and maximal likelihood approximation (Zanne et al. 2014).

For birds, we used the phylogeny generated by Jetz et al. (2012). They provided 10,000 phylogenies using Bayesian analyses and documented their detailed methods at http://birdtree.org. In this study, we randomly selected 100 phylogenies since these are very similar and thus gave qualitatively the same results regarding phylogenetic diversity (Baiser et al. 2018). We then used the average results from these 100 phylogenies for all avian analyses in this study.

2.3 Ecoregions

National parks are widely distributed across the United States (Fig. 17.1). Because homogenization is likely to be scale dependent, we chose to explore changes in taxonomic and phylogenetic similarity within and across ecoregions (i.e., continental scale). Ecoregions are areas where the type and nature of environmental resources, land use, and vegetation are generally similar (Omernik 1987; Omernik and Griffith 2014). North America has been divided into 15 broad level I, 50 level II, and 182 level III ecoregions (https://www.epa.gov/eco-research/ecoregions-north-america). For this study, we focused on the 15 broad level I ecoregions. We downloaded the ecoregion shapefiles from the US Environmental Protection Agency (EPA) website and then clipped these shapefiles to just include those that are in the United States (Fig. 17.1). Because Hawaii is not mapped in the shapefile, we excluded Hawaiian parks from our analyses, resulting in 254 parks in total (240 in the conterminous states and 14 in Alaska). When a park was located on the boundary between two ecoregions, we put that park in both ecoregions when conducting analyses at the ecoregion scale. We removed three ecoregions (water, tropical wet forests, and southern semiarid highlands) from our analyses because they have few parks (< 8), resulting in 8 ecoregions in the lower 48 states and 4 ecoregions in Alaska (Fig. 17.1). Because there are only 14 parks within Alaska, we did not conduct ecoregion level analyses for them.

Fig. 17.1
figure 1

The distribution of level I ecoregions across the United States. Dots represent national parks used in this study

Full size image

2.4 Data Analysis

We calculated pairwise taxonomic and phylogenetic distances for all unique combinations of parks. For taxonomic pairwise distance, we chose the commonly used Jaccard index, β j = (b + c)/(a + b + c), where a is the number of species observed in both parks and b and c are the number of species unique to each of the two parks being compared. The Jaccard index takes on a value of 1 when no species are shared between two assemblages (parks) and 0 when both assemblages (parks) have the same exact species composition . For phylogenetic pairwise beta diversity , we used mean pairwise distance (MPD, Webb et al. 2002), which is independent of species richness (Bello et al. 2016). For each species in one park, MPD finds the average phylogenetic distance to each species in the other park and then calculates the mean of these values. Lower values of MPD indicate that two assemblages are phylogenetically similar, while larger values indicate phylogenetic differentiation.

We calculated pairwise distances between all parks for native species only and then for native and non-natives together. To test whether pairwise phylogenetic and taxonomic similarity changed significantly after including non-native species, we conducted nonparametric paired Wilcox tests at both continental and ecoregion levels. If the average pairwise distance is lower after including non-native species, it suggests that non-native species have homogenized national parks in their taxonomic or phylogenetic composition and vice versa. Because parks in Alaska are geographically distant from the lower 48 states, we analyzed Alaskan parks separately. We further visualized phylogenetic and taxonomic similarity, for each park, by calculating its average pairwise distance with all other parks within the same ecoregion (within-ecoregions) and with all other parks that are not in the same ecoregion (across-ecoregions). Because within-ecoregions and across-ecoregions patterns were qualitatively similar, we only presented the figure for across-ecoregions below.

3 Results

3.1 Changes in Taxonomic and Phylogenetic Similarity at the Continental Scale

At the continental scale, non-native species homogenized the taxonomic and phylogenetic composition of plant assemblages across US national parks (Table 17.1, Figs. 17.2 and 17.3). This indicates that on average, parks across the United States are receiving a similar suite of non-native invaders and that these non-natives have similar evolutionary histories. The latter can be due to the establishment of the same species across parks or the establishment of different invaders that are evolutionarily closely related. Parks in the Northwestern United States had the highest levels of homogenization (Fig. 17.3). High levels of phylogenetic homogenization were found in the Pacific Northwest, Great Lakes region, and the Northeastern United States, while lower levels of phylogenetic homogenization were found in the Southeastern United States (Fig. 17.3).

Table 17.1 Overall changes in taxonomic and phylogenetic composition across all parks in the conterminous United States and Alaska
Full size table
Fig. 17.2
figure 2

Changes in taxonomic similarity vs phylogenetic similarity in plant and bird communities found within US national parks. Negative values indicate homogenization, whereas positive values represent differentiation

Full size image
Fig. 17.3
figure 3

Changes in taxonomic and phylogenetic similarity after including the presence of non-native species within national parks. Each point represents the changes in the average pairwise taxonomic or phylogenetic similarity between one park and all other parks not within the same ecoregion. For each panel, values were scaled to have mean zero and standard deviation one before calculating the changes via distance(native + exotic) – distance(native). Negative values indicate taxonomic/phylogenetic homogenization, while positive values indicate differentiation

Full size image

Bird assemblages at the continental scale have homogenized in taxonomic composition but a lower degree that plant assemblages across parks (Table 17.1, Figs. 17.2 and 17.3). In addition, bird assemblages became less similar in their phylogenetic composition (i.e., phylogenetic differentiation) across parks. This indicates that even though similar non-native birds are establishing across parks, these birds are increasing the phylogenetic distances between park bird assemblages. The highest levels of avian taxonomic homogenization occurred in the southwestern and mid-Atlantic portions of the United States (Fig. 17.3). High levels of avian phylogenetic differentiation were found in parks across the United States; however, parks along the coast of the Southeastern United States and the Gulf of Mexico tended to have lower levels of phylogenetic differentiation (Fig. 17.3).

3.2 Changes in Taxonomic and Phylogenetic Similarity Within Ecoregions and Alaska

Comparing parks within the same ecoregion, both plant and bird taxonomic and phylogenetic composition showed the same responses to non-native species presence as they did across the continent. For plants, all ecoregions experienced both taxonomic and phylogenetic homogenization (Table 17.2). Higher levels of taxonomic homogenization were observed in ecoregions concentrated in the Western United States (e.g., Mediterranean California, Marine West Coast Forest, Northwestern Forested Mountain ecoregions; Table 17.2; Fig. 17.1). Lower levels of taxonomic homogenization and some cases of differentiation were found in parks in the southeastern portion of the Eastern Temperate Forest ecoregion (Table 17.2; Fig. 17.1). Phylogenetic homogenization was highest in the Mediterranean California, Marine West Coast Forest, and Northern Forest ecoregions.

Table 17.2 Changes in bird and plant species and phylogenetic composition within national parks located across ecoregions
Full size table

Bird communities in most ecoregions experienced phylogenetic differentiation and taxonomic homogenization (Table 17.2). The exceptions to this pattern were the Temperate Sierras ecoregion, which showed taxonomic and phylogenetic homogenization (with the later not showing statistical significance), and the Northern Forest ecoregion which showed significant taxonomic and phylogenetic differentiation (Table 17.2). Parks within the Eastern Temperate Forests and Great Plains ecoregions had the highest levels of taxonomic homogenization, while high levels of phylogenetic differentiation were spread across several ecoregions (Table 17.2). Across parks in Alaska, we observed significant biotic differentiation in bird taxonomic and phylogenetic composition (Table 17.1). Plant assemblages taxonomically differentiated and phylogenetically homogenized across Alaskan parks (Table 17.1).

4 Discussion

Overall, our results show that bird and plant assemblages in US national parks are becoming more similar in their taxonomic composition due to the establishment of non-native species. Non-native plants have also increased the phylogenetic similarity of plant assemblages across US national parks, while non-native birds have decreased the phylogenetic similarity of bird assemblages . These results hold at the continental and ecoregion scale. Interestingly, parks in Alaska do not conform to same taxonomic pattern and are becoming less similar in composition to one another for both birds and plants. Our results of taxonomic homogenization of birds and plant assemblages across both regional and continental scales mirror those of a meta-analysis of homogenization studies conducted worldwide (Baiser et al. 2012).

Our finding that taxonomic and phylogenetic homogenization are not necessarily coupled shows the importance of exploring multiple aspects of homogenization (Baiser and Lockwood 2011; Monnet et al. 2014). Bird assemblages experienced phylogenetic differentiation in spite of taxonomically homogenizing. This can result from at least two processes. First, it is possible that the suite of invaders that are taxonomically homogenizing a set of assemblages (e.g., birds in US national parks) have different evolutionary histories (i.e., do not have close phylogenetic relationships). As a result, a set of relatively large pairwise phylogenetic distances are added to a set of assemblages. Second, the suite of non-native species that are establishing across a set of locales may have very different evolutionary histories from the native community. Thus, pairwise distances between natives and non-native are relatively large, on average, compared to pairwise distances between natives. Both of these scenarios result in larger pairwise distances across assemblages which is the hallmark of phylogenetic differentiation.

Our results indicate that park visitors are more likely to see the same non-native species when visiting different national parks both within the same ecoregion and across the entire continental United States. For example, park visitors are highly likely to encounter the non-native plant curly dock (Rumex crispus) which occurs in 161 national park units or the non-native, invasive European starling (Sturnus vulgaris), which occurs in 204 national parks. How the homogenization of bird and plant assemblages alters the perception and recreational use of these important resources is likely dependent on the experience of individual park visitors. While some park visitors take a laissez-faire approach to invasive species in national parks, others recognize the threat they pose and advocate for management (Sharp et al. 2011). Visitors who come in contact with an invader may have greater concern regarding the potential threats of such species (Harvey et al. 2016). Although seeing the same non-native species at different parks due to taxonomic homogenization may not influence some park visitors, the implications of non-natives for iconic species (e.g., the loss of hemlock due to the non-native invasive woolly adelgid) may affect park visitation. For example, the decision to visit Yellowstone National Park in Wyoming and Montana was not greatly influenced by the presence of the non-native lake trout (Salvelinus namaycush) (Cherry and Shogren 2001). However, the prospect of a decrease in the opportunity to view birds of prey and grizzly bears (Ursus arctos horribilis) due to the replacement of Yellowstone cutthroat trout (Oncorhynchus clarki bouvieri) by lake trout did influence visitor’s proclivity to visit the park (Cherry and Shogren 2001).

In our analysis, we focused on the role of non-native species in homogenizing bird and plant assemblages in US national parks using presence data. However, both species extinctions or extirpations and changing abundances can contribute to patterns of homogenization (Olden and Rooney 2006). The conceptualization of biotic homogenization as “a few winners replacing many losers…” (McKinney and Lockwood 1999) underscores the role of extinction in tandem with the spread of non-native species in the homogenization process. The Earth is experiencing extinctions at an unprecedented rate (Pimm et al. 1995). Because species that have small ranges have a greater risk of extinction (Manne and Pimm 2001; Pimm et al. 2014), it is likely that species unique to specific regions will be lost, especially when considering large spatial extents (e.g., continental or global). Thus, considering the loss of species (through extinction or extirpation) unique to each national park will likely increase the degree in taxonomic homogenization . Furthermore, the fact that extinction tends to target evolutionarily distinct species (Isaac et al. 2007) suggests that extinctions will also lead to phylogenetic homogenization .

Considering species abundance is also an important facet of documenting biotic homogenization . Anthropogenic change can lead to increases in native species that respond well to anthropogenic disturbance (Chace and Walsh 2006) and decreases (but not extirpations) in species that do not. Similarly, some non-natives may occur in high abundances across a set of locales and consequently have a greater homogenizing effect than those that are at low abundances. While quantifying homogenization based on species abundances is ideal, obtaining abundance data for a large number of locales at continental or global spatial extents is prohibitive.

National parks are essential to conservation and recreation in the United States and globally. However, these protected areas face many of the same threats to biodiversity as non-protected areas, including climate change (Hansen et al. 2014; Monahan and Fisichelli 2014; Rodhouse et al. 2016), pollution (Fakhraei et al. 2016), and poaching (Hilborn et al. 2006), and non-native species invasion. These threats are synergistic, show no sign of abating, and have the potential to completely alter ecosystems (Hansen et al. 2014). In order to “… turn over natural resources to the next generation increased; and not impaired in value” as President Theodore Roosevelt said at the beginning of the twentieth century, management of non-native species is crucial.