Traits of a good ecologist : What do ecologists think ?

Much is inferred, but little is actually known, about how ecologists view themselves and what they value as scientists. We investigated these attitudes as part of a survey of non-student, U.S.-addressed members of the Ecological Society of America. The part of the survey reported in this paper requested ratings of 15 possible traits of a good ecologist. These traits embodied values associated with religious principles, aesthetics, caring for nature, and epistemic competence (via professional proficiency). The survey was completed by 20% of the potential respondent pool of 6,083 ecologists. The resulting trait scores were analyzed with respect to age, gender, level of education, nature of employment, primary domain of inquiry, principle method of analysis, and source of greatest professional satisfaction. Respondents ranked traits associated with professional proficiency highest and religious principles lowest. Reactions to ethical and aesthetic traits were mixed. With increasing age, rankings for ethical and aesthetic traits increased while scores for some aspects of epistemic proficiency declined. Gender effects were few. Ecologists with BA/BS or MA/MS degrees scored aesthetic and pro-social traits higher and epistemic proficiency traits lower than did PhD degree-holding respondents. Ecologists employed by government, business, or non-profits had uniquely different scoring spectra from one another and from the entire pool. Characterization of values held by ecologists classified by various domains of inquiry, primary methods of analysis, or sources of greatest professional satisfaction was complex. Particularly intriguing was that those who worked in the ecosystem domain, who used modeling and meta-data analysis as methods, or who received the most professional satisfaction from these same activities, exhibited strong support for several epistemic proficiency-related traits, while being relatively negative about traits in the aesthetic and nature-caring categories. These results put some empirical substance and conceptual structures behind perceptions that distinct subcultures exist in U.S. ecology. These results are relevant in three ways. First, they provide a framework by which individual ecologists may become aware of the motivations that drive them, the values that shape their attitudes, and the source and meaning of their professional priorities. Second, these results challenge mentors to be mindful of how they guide young ecologists in the development of their own values and priorities. Third, these results inform the leadership of professional organizations of the range and variation of positions that their member hold and that are relevant to initiatives the leadership may which to promote.


The question
What are the traits of a good ecologist?Ecologists might generally agree on the traits they would expect for a good tax accountant, physician or attorney.They would expect relevant skills, professional experience and ethical performance.But what equivalent traits might they claim are necessary or desirable for being a good ecologist?''Good,'' of course, is a complex judgment and opinions are likely to vary among ecologists depending on their individual perspectives.Their perspectives are linked to their experiences, education, training and the nature of their particular practice of ecology.Furthermore, their professional practices are linked to their domains of proficiency, scientific objectives, and research methods (Reiners and Lockwood 2009).Given this diversity of practices, to what extent do ecologists share normative opinions on what constitutes a good ecologist?In what ways might differences in these opinions relate to personal and professional circumstances or reflect factional subgroups?
How do we define ''ecologist?''Ecologists come in many stripes so that no single definition exists.Essentially anyone who studies the interaction between biota and the environment can claim to be an ecologist.Self-identification is probably the only universally defensible definition, and membership in the Ecological Society of America is a conservative definition for U.S. ecologists-conservative because many who would regard themselves as ecologists are not members of that organization.
The objective of this study is to determine the normative values and patterns of variation on what constitutes a good ecologist held by established, U.S. ecologists.We posit that selfawareness of our own value systems will make us better scientists.We also submit that ecology's professional institutions need to be aware of their members' values as leaders attempt to promote educational, research, and advocacy programs.

What traits?
Our objective was addressed through a survey described below.But, traits as used in this research must first be defined and discussed.
What do we mean by traits?''Trait'' has specific meanings in different scientific fields-genetical, behavioral, psychological, sociological and political.In the psychological realm there is a complicated set of relationships between value, attitude, trait and character (Rokeach 1973).Our use of the term is derived from that system but follows common usage of being ''a distinguishing quality.'' For this research, we composed a list of 15 traits, several of which were meant to be analogous to those one might seek in evaluating other professionals, e.g., an accountant, a physician, or a lawyer.These are listed below in the order in which they were presented to respondents.
1. Has an appreciation for the beauty of nature.

Has a broad training in the various
ecological sub-disciplines.3. Believes that nature is intrinsically valuable.4. Believes that humans have caused serious damage to the natural world.5. Believes that we are morally obligated to preserve aspects of the natural world.6. Enjoys being outdoors.7. Possesses scientific objectivity.8. Has a background in the fundamentals of science philosophy.9. Has had rigorous mathematical training.10.Seeks to be of service to society.11.Has a sense of duty to future generations.12. Values knowledge for its own sake.13.Possesses a religious foundation for valuing nature.14.Has a passion for scientific inquiry.15.Has a personal affinity for the entity, process, or place he/she studies.
An individual's determination of value falls into the axiological domain of philosophy which primarily pertains to aesthetic, ethical and religious beliefs.Some of the traits in our list are obviously of these types (e.g., enjoys being outdoors or has a sense of duty to future generations).However, some traits may be valued for their relevance to conducting scientific inquiries and making important contributions to human knowledge.As such, even traits of an apparently more functional nature-the mental tools that help us infer causation, interaction and generalization-pertain to the value we ascribe to knowledge and can also be appreciated in ways similar to aesthetics (Wimsatt 2007).We derive emotional gratification from resolving a difficult question, and we find beauty in elegant experimental designs and equations (Chandrasekhar 2010).We take all of the listed traits as reflecting an axiology but posit that some have greater value for some individuals than for others.Before implementing the survey, we anticipated that even within a particular discipline like ecology, perceptions as to what is of value (''cognitive goals'' according to Laudan (1984)) would vary among practitioners.
The listed traits can be classified into three types of values.First, some are overtly aesthetic in nature.These include: 1 (has an appreciation for the beauty of nature), 6 (enjoys being outdoors), and 15 (has a personal affinity for the entity, process, or place of study).Though this study is exploratory, we suspected that an ecologist whose primary method is field observation would be likely to value ''being outdoors'' more than would an ecologist primarily involved in data-mining.Additionally, we expected that a modeler who sees beauty in a programmatic code may not score the trait ''appreciation for the beauty of nature'' as very important.Thus, both outdoor experiences and programmatic codes have aesthetic value but would be scored differently by different ecologists.
Second, some of these traits are ethical in nature.These include: 3 (believes nature is intrinsically valuable), 4 (believes that humans have caused serious damage to the natural world), 5 (believes that we are morally obligated to preserve aspects of the natural world), 10 (seeks to be of service to society), 11 (has a sense of duty to future generations), 13 (possesses a religious foundation for valuing nature).We expected that ecologists would vary in their acceptances of these ethical statements.For example, some might value attitudinal trait 3 without adopting action-oriented trait 10.
Regarding trait 13 ( possesses a religious foundation for valuing nature), philosophers might classify this as a moral, rather than ethical, value (the former being grounded in cultural practices, the latter being grounded in reason), but we set this distinction aside.We anticipated that some ecologists might possess either a biblical sense of stewardship on one hand, or a sense that nature is to be exploited for human welfare, on the other.
We recognize that these ethics-related traits might be viewed by some ecologists as the raison d'être for ecology.Others, though, may believe these traits should be detached from professional life-suppressed in the interest of maintaining one's scientific objectivity, as well as retaining one's scientific credibility (Voosen 2012, Reiners et al. 2013).We can imagine that a certified applied ecologist would value some of these ethics-related traits whereas theoretical ecologists would take them as being in conflict with the higher value of objectivity.
Third, some traits primarily reflect epistemic value via professional proficiency.That is, scientists value knowledge (or episteme, which philosophers take to be justified, true beliefs), with professional abilities being a means to the acquisition of knowledge.Relevant traits were: 2 (has a broad training in the various ecological disciplines), 7 (possesses scientific objectivity), 8 (has a background in the fundamentals of science philosophy), 9 (has a rigorous mathematical training), 12 (values knowledge for its own sake), and 14 (has a passion for scientific inquiry).We expect that virtually all respondents would score trait 7 (objectivity) very high.In fact, in science, objectivity is viewed as fundamental to acquiring knowledge which is taken to be intrinsically good, whereas subjectivity leading to bias corrupts the search for truth.In fact, though, few respondents were likely to have appreciated the complex meaning of scientific objectivity or that few of us possess it to the extent that most of us believe we do (Longino 1990).The survey was not the venue, however, to provide a philosophical tutorial.Thus, it is reasonable to assume that respondents shared a commonsense notion of objectivity as unbiased, disinterested experience that precisely corresponds to a mind-independent world.
Trait 8 (has a background in the fundamentals of science philosophy) may be the most ambiguous of all we posed.Our experience has been that most ecologists know something about science philosophy but are unaware of what they do not know, which in many cases is considerable.Science philosophers themselves are partly to blame for this through their obscure arguments and arcane focus using physics as the ideal (Gasper 1991, Reiners andLockwood 2009).A related obstacle is that some ecologists interpret science philosophy as purely aesthetic-and ethicbased axiology, such as presented in otherwise excellent literature such as Walden (Thoreau 1854), Silent Spring (Carson 1962) or A Sand County Almanac (Leopold 1949).
We expected diverse but predictable reactions to traits 2 (has a broad training in the various ecological disciplines) and 9 (has a rigorous mathematical training).That is, we anticipated that responses were likely to depend on the respondent's breadth of training in the first case and the individual's mathematical proficiency in the second case.It is for reasons like these that we believed that a broad range of values could be expected for the traits presented in our survey.

The survey
We conducted a survey of part of the membership of the Ecological Society of America (ESA) to learn what ecologists think are traits of a good ecologist.ESA is the single largest ecological organization in the U.S., composed of over 8,000 members.Not all scientists identifying themselves as ecologists are members of ESA, however.Thus, our sample represents the views of those who seek membership in an organization central to ecology; other ecologists may affiliate with no scientific society or with organizations focused on peripheral areas such as biological oceanography, range management or earth system science.Because cultural norms vary with national membership, we restricted the survey to those members with addresses in the United States.Of course, not all ecologists living in the U.S. are native-born and some might be temporary residents.We also wanted responses from ecologists who had perspectives associated with employment.For this reason we also restricted the survey to non-student members.These considerations delimited our sample frame to a total of 6,083.
Our survey was distributed via email in one The survey requested information on personal and professional characteristics of each respondent that could be related to axiological traits.Age, gender, highest degree and employment are obvious choices for qualities that might influence values.In addition, we sought to discriminate among ecologists in terms of how they practiced ecology.The domain of inquiry-organismal, community, biosphere, etc.-is one obvious axis of classification given the variety of more specialized journals, specific funding sources and subdisciplinary meetings.Partially orthogonal to domain is the primary method of research-e.g., field observations and lab experiments-by which ecologists investigate nature.In some cases, domain and method are strongly connected (e.g., biosphere domain and remote sensing method), but in others cases methods can apply across domains (e.g., modeling is a method in all domains).The last property we included was what respondents found most satisfying in their professional lives-activities that may not be consonant with what they do in their employment.
The Following these questions, the respondents were asked to evaluate the 15 traits of a good ecologist described above with a continuous slider tool that ranged from 0-100.Descriptive markers were placed above the gradient as follows : harmful, 10; risky, 30; irrelevant, 50; helpful, 70; vital, 90.Results are presented as averages.

Statistical methods
We used OLS regression (IBM Corp 2010 [SPSS version 19]) to measure the effects of our independent variables (see above) on trait evaluations.Only effects significant at the p , 0.05 level are reported.The regression models used are specified in the results.As indicated in the introduction, we held that each trait primarily reflected aesthetic, ethical or epistemic (professional proficiency) values.The survey data permitted cross-correlation of responses to identify classes of traits analogous to those we initially described.We sought trait clusters with exploratory factor analysis to examine joint variation in trait ratings.For this we used the maximum likelihood extraction method with oblimin rotation.The resulting groupings were tested for statistical coherence with the Cronbach alpha statistic (Cronbach 1951).

Properties of the respondent pool
Data on the seven variables characterizing the respondent pool are themselves interesting and are analyzed in another paper (Lockwood et al. 2013).For the purposes of this paper, pool characterization is limited to the properties listed in Table 1.Salient features are: (1) the median and mean respondent ages were in the 50-59 year v www.esajournals.orgclass, (2) females represented 32.6% of the respondents, (3) female representation was inversely correlated with age, (4) the pool was dominated by PhD degree holders, (5) the majority source of employment was academia with government a distant second, (6) frequencies of responses within primary domain of inquiry, primary method of analysis, and most satisfying activities were broadly but unevenly dispersed (Lockwood et al. 2013).

Collective ranking for traits
Overall scores for the 15 traits are presented in ascending rank order but are numbered in the order in which they were presented to respondents (Fig. 1).All but number 13 (religious foundation) received a score greater than 50%, signifying that the average respondent felt that the traits had positive qualities for being a good ecologist.The rating for trait 13 was significantly lower at p 0.05 than all others.Pairwise differences between average scores for all traits are significant (based on non-overlapping 95% confidence limits) except for 1 versus 3, 4 versus 5, 8 versus 12, 10 versus 11, and 11 versus 15.

Trait groupings into classes
Iterative combinations of rotation and extraction methods led to a four-factor solution to trait groupings that we have titled aesthetic values, nature values, pro-social values and epistemic values (Table 2).In effect, what we took to be ethical values prior to the study were broken down into values pertaining to nature (environmental ethics) and pro-social (interpersonal ethics)-as well as a conceptual outlier.Trait 13 (possesses a religious foundation for valuing nature which we initially included in the ethical category) failed to correlate with any other traits, thereby lowering coherency statistic for its most related group.We isolated trait 13 as a special case unrelated to any of the others (Table 2).

Age and gender effects
Age and gender effects on favorability ranking for each trait were tested with the regression: Data from the 80s and above age-class were eliminated from this and other tests because of the small number of respondents in that class (n ¼ 14).Statistical results of this and other tests are v www.esajournals.orgprovided in the Appendix.Significant effects are presented in here in graphical form.Reference to ''positive effects'' or ''negative effects'' in the following descriptions of results should not be interpreted as respondent groups categorically favoring or opposing particular traits.Rather, these effects represent statistically higher or lower degrees of support than registered by the main body of respondents.They are positive or negative relative to the entire pool.
With age, there is an increase in favorability for all six aesthetic and nature valuing traits, one pro-social valuing trait, and two professional values; while there is a decrease in favorability toward religious foundations, science philosophy and mathematical proficiency (Fig. 2).Gender significantly affects only two traits.Compared with males, females were more positively inclined toward the intrinsic value of nature and more negatively toward science philosophy.
In this test and all others, linear regression models produce low adjusted R 2 values.Individual variable effects are significant but account for a small portion of the total variation of trait ratings.Nevertheless, significant effects are strik-ing as exemplified by the increasing importance of having a sense of duty to future generations with age (Fig. 3).

Degree and employment effects
The effects of degree and employment were tested with the following model: Age and degree were shown in previous analyses to be correlated with employment, and were therefore included in the model.BA/BS and MA/MS degree effects are evaluated with respect to the PhD degree as the PhD is the majority degree (84%).Likewise, employment categories are tested with respect to academic employment, the majority category (61%).
Respondents with BA/BS degrees are more favorably disposed to traits 5, 3 and 1-aesthetic and ethics valuing traits-than are PhD degree holders (Fig. 4), but less favorably disposed toward mathematical proficiency and broad training-more epistemic traits.MA/MS degree holder effects mostly align with those of BA/BS   Respondents working in the government sector are more negatively inclined than are academics to traits in the aesthetic, nature and epistemic categories (15, 4 and 7) (Fig. 4).Significant effects of corporate or privately employed respondents were concentrated within the epistemic category of traits.They placed relatively more value on traits 2 and 7 (broad training and objectivity), but less on traits 12 and 14 (value of knowledge and passion for inquiry) than did academics.Non-profit employees also placed less value on trait 12 than did academics.

Primary domain of inquiry effects
Because respondents could choose more than one domain (see Table 1 for domain options), we were able to evaluate the influence of every domain independently of the others.This option produced 2,032 respondent choices.Concerned that the term ''domain'' might mean different things to academic ecologists than to those otherwise employed, we ran this analysis twice, once with the total pool and separately for academics alone.The base model for the total pool is: where age, gender and employment act as controls for domain of inquiry.The model for the academic pool is: Within the total pool there was only one positive effect in the aesthetics category, none in the nature category, two in the pro-social category, and seven, shared by all domain groups, in the epistemic category (Fig. 5).Landscape ecologists registered all the positive effects in the aesthetics and pro-social categories.Respondents in the ecosystem domain were less favorably inclined toward recognizing beauty in nature trait as an attribute of a good ecologist.Respondents in the community domain assessed mathematical proficiency less positively than did others.
The academic subset, comprising 61% of total pool, had an even more focused concentration of positive effects within the epistemic category of traits (Fig. 5).Only the landscape domain group showed a positive effect toward a pro-social

Primary method of analysis effects
Respondents were permitted two choices of primary method for self-characterization resulting in a total of 2,142 choices for testing for the association of method with scoring of traits.Tests for the entire pool were performed with the following model: For academics as a respondent subset, the model was: The total respondent pool registered 17 signif-icant effects, 11 positive and six negative, indicating considerable variability among respondents identifying with different method categories (Fig. 6).There were only two positive effects regarding aesthetic and pro-social valuing traits and none for nature values.Thus the preponderance of positive effects was directed toward the epistemic proficiency traits.All of the negative effects were manifest by members of three groups-lab experiments, meta-analysis and modeling-with respect to aesthetic and nature valuing traits.
The academic subset had fewer significant effects than the total pool (10 versus 17), but the patterns of preference were similar.The four positive effects reflected support for epistemic proficiency traits (12, 2, 14, 7).All six of the negative effects were directed by members of the lab experiment, meta-analysis and modeling domain groups toward aesthetic-, nature-, and pro-social traits, consistent with the total respondent pool results.

Primary source of professional satisfaction effects
Respondents could, but did not necessarily, indicate two choices for source of professional v www.esajournals.orgsatisfaction which led to a total of 2,321 selections.See Table 1 for distribution and order of choices of satisfaction.With the inclusion of teaching alternatives, it was especially necessary to do a separate test with the academic subset of the pool.
The regression model used for testing all respondents was: Fig. 6.Significant effects for ecologists as members of primary method of analysis groups on traits scores.Left panel, total pools of respondents; right panel, academic respondents.Solid arrows arrows are positive effects; dashed arrows are negative effects.See Table A4 in the Appendix for statistics for these effects.
v www.esajournals.orgFor academics only, the testing model was: Within the total respondent pool, six positive and 12 negative effects were significant (Fig. 7).Four of the six positive effects were directed toward aesthetic and pro-social trait categories, all from those who most enjoyed working with students, being involved in oral communication or conducting field work.Eight of the 12 negative effects were for aesthetic-or nature-valuing or religious foundation traits of which six were from respondents favoring data analysis as their most satisfying activity.Respondents identifying oral communication and ''other'' activities as their most satisfying professional activity showed negative effects with respect to scientific objectivity.
The academic subset of respondents registered fewer significant effects than did the entire pool (seven versus 18) indicating that many of the effects came from non-academically employed respondents who composed only 39% of the respondent pool.Four of these seven were negative effects, three toward aesthetic-and nature-valuing traits, and the only significant negative effect toward religious foundations from those most satisfied with designing experiments.

The survey
The objective of this research was to discover what professionally established, U.S. ecologists think are traits of a good ecologist.Choices of traits were directed, not spontaneously devised by the respondents.It is likely that respondents would have volunteered different traits than those we presented, but the heterogeneity of such responses would have been challenging to evaluate in a systematic manner.Moreover, we desired responses to traits covering a range of motivational and conditioned attitudes that WAR and JAL, as ecologists, had experienced in our careers (Reiners and Lockwood 2009).Data gathered by this survey represent opinions of professionally established or retired, U.S. ecolo-gists sufficiently committed to ecology to be members of ESA.The 20% response rate (1,215 respondents) of this delimited population is well above the sample size recommended for either continuous or categorical data (Bartlett et al. 2001).

Properties of the respondent pool
Distributions of respondent properties are summarized in Table 1; an analysis of these data is provided by Lockwood et al. (2013).Except for gender, we cannot compare the distributions of our respondent properties with those of the entire membership, as such data are not available.With regard to gender, however, our response rate for females is 32.6% which conforms well with ESA datum of 32.1% for female membership.
Several distributional properties of the survey population are critical to interpretation of response data.First, the youngest age-class (20-29 years) of ESA members included few individuals partly because many members are students who were purposely excluded from the survey.Second, the !80 age-class is very small and was not included in the analyses.Third, some proportion of the 60-69 and 70-79 age-classes are retired so that their self-categorization may represent how they practiced ecology before retirement, or their responses were registered in ''other'' categories.Fourth, there is almost surely bias of unknown magnitude by virtue of selfselection by members who completed the survey.Fifth, the terminology used in this survey (e.g., ''trait,'' ''good,'' ''value'') is subject to semantic and philosophical interpretation.Definitions were not provided in the survey instrument so that meanings were taken to reflect colloquial usage.

Collective valuation of traits of a good ecologist
The ranking for all traits by the total respondent pool in Fig. 1 is the best collective expression of the sentiments of established (employed or retired) ecologists in the U.S. Fig. 1 illustrates that respondents scored highest those traits we associate with epistemic proficiency.All six in that class (7, 14, 2, 12, 8, 9) are the most highly ranked (1st to 6th, respectively).
The aesthetically-related traits (1, 6, 15) are generally the second ranked set (7th, 9th and 10th, v www.esajournals.orgrespectively).We were surprised that aesthetically-related traits outranked pro-social and naturerelated traits.We would have thought that responsibilities to society and an intrinsic caring about nature would have been more highly valued than aesthetic considerations among ecologists.In this respect, respondents did not match our preconceived stereotypical values.Perhaps these relatively higher scores are motivated by emotions.Alternatively, appreciation of the beauty of nature, enjoying being outdoors, and having a personal affinity for the entity one studies are considered to be assets to professional proficiency (e.g., ecologists who enjoy nature are more likely  A5 in the Appendix for statistics behind these effects. v www.esajournals.orgto show devotion to the pursuit of knowledge).Philosophers are divided as to the importance of cognition in aesthetic judgment, although Carlson (2002) has persuasively argued that background knowledge about the origin and function of ecological entities and processes is vital to aesthetic appreciation.
The relatively low ranking of pro-social traits (11 and 10 ranked 11th and 12th, respectively) indicates that ecologists regard basing one's work on obligations to society as a dubious trait of a good ecologist.Surprisingly, nature-related traits 3, 4, 5 were ranked even lower (8th, 13th and 14th, respectively).We will discuss later how particular subgroups among the respondents consistently rated nature-related and societallyrelated traits relatively negatively as traits of a good ecologist.
The least favored trait is 13 (possesses a religious foundation for valuing nature).In fact, with a rating of less than 50%, it is actually considered a negative trait by this population.While this trait's average score is low (44.2 on a scale of 0 to 100), it is not entirely dismissed.Religiously based values must be important to some ecologists.We might expect a particularly high variance around the mean score for this trait, but the 95% confidence level (1.06) is similar to confidence levels for other traits.
Clearly, objectivity is highly valued by most ecologists.The average score, however, is not 100.Those with a MA/MS and those in corporate or private positions (Fig. 4), and those whose greatest satisfaction is from written (academics) or oral communication, or ''other,'' scored scientific objectivity lower than the bulk of the respondent pool (Fig. 7).A negative scoring need not signify that objectivity can be dismissed or sacrificed for some other value, however.It simply may indicate that from the point of view of professional satisfaction, objectivity as a practice is not as valued as other activities such as service to future generations or to contemporary society.

Age and gender effects
The increased appreciation of aesthetic, nature and pro-social traits with age (Fig. 2) is exactly contrary to the rankings for the respondent population as a whole (Fig. 1).Two explanations might be: (1) a broader, humanistic understand-ing of ecology develops with age and experience, or (2) these kinds of traits were more prominent in ecological education in decades past.If the former explanation is true, we can predict that a more humanistic outlook will be maintained as ecologists age.If the latter is true, ecologists will increasingly focus on epistemic proficiency as the population ages.
While scoring for humanistic values increased with age, the potentially allied trait of religious foundations ( 13) declined (Fig. 2), indicating that older ecologists did not equate religious values with secular aesthetic, pro-social and nature values.
Valuation of two epistemic traits declined with age: mathematical proficiency and science philosophy.It is tempting to infer but difficult to know whether older ecologists were less mathematically competent in the past.It is also possible that older ecologists become increasingly disillusioned with mathematically based understandings of nature.As for science philosophy, older ecologists may have had less science philosophy in their training, or if they did, they never found it useful or grew to discount its value.Taken together, we can only speculate on whether these trends are driven by changes in viewpoint with age, or are historical artifacts of the times in which careers developed (Hofer andPintrich 1997, Inglehart 2000).The answer to this is key to whether the distribution of ecologists' attitudes will remain stationery over time or will change as the generations turn over.
Gender is inversely related to age (Table 1) as we expect from the historical trend of women entering this science.Thus, it is natural to question whether age effects are not, at least in part, gender effects.According to these data, they are not.Only two significant effects were attributable to gender when controlled for age.Females valued trait 3 (intrinsic value of nature) more, but valued science philosophy less, than did males.The lower ranking of philosophy by females accords with the low representation of women in philosophy in general, an intriguing question outside of ecology (Haslanger 2008;Saul, in press).Perhaps we can celebrate the paucity of gender differences as evidence that we are successfully educating males and females equally.Gender-related differences do occur, however, in other aspects of the overall survey (Lockwood et al. 2013, Reiners et al. 2013).

Academic degree and employment effects
Respondents with BA/BS or MA/MS (3% and 13% of the pool, respectively) showed markedly different responses from those with PhD degrees (Fig. 4).Both BA/BS and MA/MS ecologists are more favorably disposed toward aesthetic, nature and pro-social traits while relatively unsympathetic toward epistemic traits.We speculate that these significant responses align with the interests and professional functions that BA/BS ecologists are likely to have, along with the lack of exposure to values associated with the research-intensive PhD degree.We also venture that individuals with the lower degrees, who are generally younger than PhD respondents, are still endowed with more idealistic attitudes toward ecology and the environment than may be true for more jaded individuals with the PhD degree.It might also be the case that non-PhD ecologists are not aggressively pursuing tenure and professional advancement, so they may have a less myopic and professionalized view of science.
The valuation pattern for ecologists of different employment sectors other than academia is dominated by negative reactions (six of eight significant effects) to the full range of traits (Fig. 4).Ecologists employed in government placed lesser value than did academicians on aesthetic and nature traits and had less use for pursuing knowledge for its own sake.This group seems to have a more utilitarian view of good traits in an ecologist.The corporate/private group was neutral on non-epistemic traits but leaned positively toward objectivity, passion for inquiry, and broad training.The negative effect on knowledge for its own sake seems inconsistent with the passion for inquiry, but perhaps represents the entrepreneurial interest in problem-solving for particular issues.Overall, the preferences of the three, non-academic employment groups (government, corporate/private, non-profit) were consistent with mission-driven, rather than curiosity-driven science.

Primary domain of inquiry effects
Analyses of respondents with respect to primary domain of inquiry yielded three inter-esting outcomes (Fig. 5).The landscape/aquatic/ marine group differs from all others in the number and breadth of significant effects, all positive for both total pool and academics alone.This domain group, consisting of only 25% of total respondents, generated a disproportionate number of significant differences, accounting for 56% of the effects.This is the sole group with positive effects for aesthetic, pro-social, and epistemic values (but not nature) among all domain groups.The over-representation of this domain group may be caused by its mixed composition of broadly differing kinds of ecologists.On the other hand, such diversity might be expected to increase variation and thereby obscure differences from the main trend.Possibly, respondents in this group are more frequently involved in environmentally relevant activities, requiring them to think more about aesthetic and pro-social values than do ecologists in the other domain groups.
The second outcome is that in both the total and academic sets, members of the ecosystem domain believed that seeing beauty of nature ( 1) is not a trait of a good ecologist.This negative effect may only mean that ecosystem ecologists view the trait as dispensable.but it might mean something stronger, such as compromising to scientific objectivity.
The third interesting outcome is that ecologists associating themselves with the community domain scored mathematical proficiency (9) significantly lower than did the bulk of respondents.Mathematics has traditionally been a strong element of at least some aspects of community ecology, although perhaps less so than for population ecology, members of which were neutral on this trait.It may be that those who know the most about mathematics are skeptical as to its value.
Overall, different kinds of ecologists in the total pool have more positive inclinations toward epistemic traits than they do for ethics and aesthetics.From a technical point of view, this is desirable, but we might ask whether ecologists would benefit from thinking as much about why they practice ecology as how they do so.It would seem that understanding one's own motivations would engender better judgment of the purpose and value of these actions.
Trends in academic pool responses are similar v www.esajournals.orgto those of the entire pool.The former pool has a higher proportion of PhD-degree-bearing members (95% compared with a low of 39% for corporate/private to a high of 84% for ''other'').
On one hand, we might expect that academicians, most of who teach, would pay more attention to aesthetics, nature and pro-social values in classroom settings, but the results do not show this to be the case.On the other hand, academic researchers, generally with higher degree levels and possibly more devoted to curiosity-driven research, might exhibit stronger propensities toward objectivity and other epistemic traits.And so we posit that basic, or curiosity-driven, science conducted by academics does not entail the ethical and societal dilemmas of applied research.

Primary method of analysis effects
Dividing respondents as a function of primary method of research led to different patterns of outcomes (Fig. 6).Looking first at the total pool of respondents, there are three general trends.First, ecologists mainly doing field observation uniquely, and not surprisingly, scored enjoying outdoors-an aesthetic value-as a positive trait.This same subset also positively scored service to society and knowledge for its own sake, suggesting that this is a relatively socially oriented, idealistic subset of ecologists.
The second salient outcome of the entire pool responses is the strong support by all other method groups for epistemically related traits.In fact, 11 out of 13 positive responses were for traits in this category.However, there were no positive effects for any societal or nature-related traits.
The third major result in this set is the relatively negative responses associated with members of the lab experiment, meta-analysis and especially modeling groups toward aesthetic and nature-valuing traits.Ecologists associated with these three groups represent 14%, 8% and 10%, respectively of the total method choices in the survey.Because some individuals chose two methods, there is some overlap by single respondents among these three classes.Those associating with modeling in particular, registered lower ratings for having a personal affinity for the entity studied, for believing that nature was intrinsically valuable, and for having a moral obligation to preserve aspects of the natural world.This suggests that these ecologists are the most extreme outliers in describing what makes a good ecologist.Modelers value broad training more highly than do other methods groups, but it is evident from the modelers' negative responses that this breadth does not include the non-epistemic values.
Results for the academic pool (right panel Fig. 6) have fewer significant effects but the trends reflect those of the entire pool.In spite of their greater teaching responsibilities, the broader issues of aesthetic and ethical connections with nature and society are not considered very important by academic ecologists involved in modeling, lab experiments and meta-analysis.

Primary source of professional satisfaction effects
The final survey question set divided ecologists in terms of which professional activity provided them with the most personal satisfaction.The alternatives offered for self-classification differ from what ecologists studied or how they conducted their work, and take better account of the activities of non-research and non-academic ecologists (Table 1).These results give us quite a different assortment of preferences (Fig. 7).Within the total respondent pool, those preferring to do field work value two of the aesthetic traits, while those most fulfilled by data analysis regard all the aesthetic traits as negative for a good ecologist.The negative effects associated with data analysis enthusiasts extend to all three of the pro-social values as well.In fact, this group views none of the traits any more positively than the total pool mean.It appears that those most fulfilled by data analysisincluding those in the academic pool-represent a pronounced outlier group in the same way that modelers do when classed by primary method (Fig. 6).It would be interesting to know more about this subset of ecologists.
It is not surprising to learn that ecologists most enjoying working with students or performing oral communication see service to society as a trait of a good ecologist.In terms of valued traits, this group is the antithesis of those most enjoying data analysis.
The outcomes in the academic respondent pool are a little different.There are fewer significant effects, probably because academics comprised 61% of the total pool and thus heavily weighted the main trends.We might expect that those most enjoying field work would also think that an affinity for that which one studied would be a positive trait, while those most enjoying program management would see service to society as a positive trait.It is interesting that those most enjoying experimental design should be the only subset, aside from older ecologists, who registered a significant negative effect with respect to religious foundations.We can only speculate that those most inclined to uncovering direct causal relationships should be most adverse to religiously based explanations.
Significant effects based on professional satisfaction mostly align with effects based on domain or method.Traits considered by those gratified by field work are enjoyment of the outdoors and an affinity for the entity studied (6,15).Those working with students place high value on service to society (10).If it is fair to characterize those who most enjoy designing experiments and analyzing data as dealing with nature in a more abstract manner, it seems consistent that they should put little value on the aesthetic and ethical traits.
We note that mathematical proficiency (trait 9) received no positively or negatively significant responses.In other tests, this trait was scored negatively among BA/BS and MA/MS degree respondents (Fig. 4) and by community domain respondents (Fig. 5), but scored positively by modeling method respondents, as one might expect.Similarly, facility with science philosophy had significantly positive effects registered by members of the population, landscape and ecosystem domain groups (Fig. 5) and by the modeling method group (Fig. 6) but had no significant effects from ecologists categorized in terms of satisfaction.Just as vitamins are good for health but onerous to take, mathematics and science philosophy may be valued traits for a good ecologist when framed in terms of doing science, but are not so highly valued in context of professional satisfaction.

Ecological practices and ecological subcultures
We tacitly accept variation in the practices of ecology.Different ecologists subscribe to different periodicals and attend different meetings as suit their professional and personal interests.In academic departments, we deliberately recruit for diversity in domains and methods, and sometimes for demographic variables.On the other hand, we may attempt to convince ourselves that in spite of these differences, certain convictions are held by a majority and unify the discipline.We are not aware that anyone has tested that belief.This faith seems most often rhetorically expressed by organizational leadership-a rhetoric necessary, perhaps, for projecting disciplinary integrity or coherence.However, given the differences across groups in terms of what constitutes a good ecologist, such unity seems more prescriptive than descriptive.
We might assume that certain traits have ''motherhood'' properties that would be widely shared while others would be viewed divergently.The overall ranking of traits in Fig. 1 tells us the degree to which particular traits are embraced.It does not tell us about the divergence internal to the population of American ecologists.More specific analyses show that valuations of these traits vary with age, although little with gender, and with the level of professional education and mode of employment.More complex manifestations of diversity are revealed in the questions classified by primary domain of inquiry, primary method of analysis, and greatest professional satisfaction.Here, divergences become manifest as every trait generated both significant positive and negative judgments by at least one category of respondent.
One way to evaluate divergences is to identify a core ''subculture'' around which other subcultures radiate divergently along multiple axes.If we choose the largest group in terms of primary method of inquiry, it would be those whose principle method was field observation (40% of all methods choices).This group showed a broad range of valuations including positive responses to service to society, enjoying outdoors, and valuing knowledge for its own sake.Perhaps this is the prototypical ecology group that inspires the rhetoric of unity.Of course, the larger the group, the more influence it has on the overall regression, and the less likely it will be exposed as a significant effect on the slope.In contrast, ecologists whose principle methods were lab experiments, modeling or meta-data analysis were frequently negative toward traits not directly related to epistemic proficiency.At first glance this might be thought to reflect a value-free ideal of science.However, such proficiency and objectivity are clearly values themselves, reflecting the valuing of scientific knowledge itself.
This analysis was not designed to identify ecological subcultures.Nevertheless, our results suggest that the discipline of ecology is comprised of rather strongly differentiated subgroups of varying and as-of-yet undefined coherency in a multidisciplinary space of multiple professional criteria.

Relevance of this work
The primary purpose of this work is to determine what U.S. ecologists think are traits of a good ecologist, but also to probe the relationship between values and ecologists' personal and professional characteristics.Our results illustrate the extent and nature of this association.We realize that our understanding of ecologists' preferences are constrained by those traits and options offered by the survey.Other investigators might have chosen different traits and ways of partitioning ecological practices.But no one else has done so, therefore this is what we know at present.Partial knowledge is preferable to complete ignorance.So, why is this information important?
We propose that our findings are relevant in at least three ways.First, we hope this work will stimulate individual ecologists to reflect on and clarify their own motivations, values and professional priorities.They may find that they have never considered the direct question of what makes a good ecologist, but their contemplations may stimulate some useful self-awareness about why they chose ecology as a career, whether those influences still prevail, whether half-forgotten or subliminal values unknowingly influence present career decisions and attitudes, and whether they remain sufficiently confident in their values to retain them, or, to even promote them among others.
The second point of relevance stems from the first.How should we handle issues of value in the broader education of young ecologists, or mentor career development of junior colleagues?Is it responsible for us to just to assure technical proficiency, impose a stricture about objectivity, but then let the other axiological aspects of our professional lives take care of themselves?If a mentor does not deliberately instruct on these aspects of the scientific life, that mentor does so by omission.From other data in our survey, we can attest to the fact that ecologists are conflicted about whether or how to incorporate their scientific expertise into issues of environmental relevance (see also Blockstein 2002, Sisk et al. 2011).In fact, values insert themselves whether we are conscious of them or not (Lautensach 2005, Pokallus et al. 2011).It is better to realize that our values may be influencing what we think are purely empirical inputs as we develop our research questions and draw our scientific conclusions (de Melo-Martin and Intemann (2012).We maintain that educators and mentors ought to provide intentional guidance in preparing young ecologists for their professional lives (e.g., Oppenheimer 2011, Reiners andLockwood 2009).
Finally, these results are relevant to those who lead ecological institutions like the host of our survey, the Ecological Society of America.The relatively low rating of pro-social traits (10th and 11th among 15) compared with aesthetic and epistemic traits suggests that many ecologists are not as motivated or enthusiastic about the environmental activism and advocacy as may be reflected in ESA's societal literature (see also Reiners et al. 2013).Some ecologists might even have some reservations about the last element of ESA's Code of Ethics which reads: ''. . .and to facilitate the application of ecological science in the management of ecological systems.''Very probably, many would cavil with the view of Nelson and Vucetich (2009) adopted in a recent SEEDS (ESA's program: Strategies for Ecology Education Diversity and Sustainability) statement (Bonilla et al. 2012).
''Nevertheless, advocacy cannot be classified as inappropriate for scientists, since as citizens first and scientists second, scientists have the responsibility to share their knowledge in a transparent manner for the betterment of society''.
Rhetoric notwithstanding, the leadership probably has no illusions about there being conceptual unity of ecology or wholesale agreement on extending ecological expertise to resolve environmental issues.Strong (2008) has illustrated the complicated dilemmas faced by v www.esajournals.orgecologists who care about their scientific integrity yet feel a responsibility to nature and society.Some ecologists enthusiastically embrace well-considered advocacy roles in environmental issues (e.g., Meyer et al. 2010, Pace et al. 2010), while others are more cautious (e.g., Lackey 2007).Such insight appears in ESA's sophisticated guidebook to policy engagement (Ecological Society of America 2011) which states, ''Increasingly, in addition to their research and educational pursuits, ecologists are choosing [our emphasis] to become involved in policy at the local, state, regional, national and international levels'' (p.4).By recognizing that some ecologists are choosing to become involved, ESA wisely accepts the reality that some will not.
Our results affirm that many ecologists do not embrace those traits we have classified as having aesthetic, nature-valuing, pro-social or religious relevance.Some may not choose to become involved because of concerns for the appearance of compromised objectivity (Voosen 2012), while others may eschew engagement for different reasons.Our data suggest that if the attitudes of modeling, meta-analysis and experimental subsets of the membership were to become more common in the future, the discipline might come to reflect a more mechanistically-and analytically-committed flavor while environmentally-and societally-responsible propensities might decline.The results of our study suggest that as ESA seeks to develop organizational priorities and programs, it is advised to explore further the values of the membership.Such understanding is critical for forming realistic expectations of support, for gaining enthusiastic participation in programs, and for fostering dialogues regarding the role of the ecologists in the changing social environment of our time.

Fig. 1 .
Fig. 1.Ranking and scores of 15 traits for being a good ecologist from the entire respondent pool.Error bars are 0.95 confidence intervals.See introduction for full text on traits.Bars for numbers 5, 4 and 3 are pattern-coded as nature valuing traits; bars 10 and 11 as pro-social traits; bars 15, 6 and 1 as aesthetic traits, and bars 9, 8, 12, 2, 14 and 7 as epistemic value traits.Trait 13, the sole religious principles trait, is unfilled.See text for explanation of trait categories.

Fig. 2 .
Fig. 2. Significant differences for increasing age and for female responses compared with male for individual traits.Traits are organized according to their membership in value categories.Religious foundations is placed between nature values and societal values as a seeming best fit.Solid arrows in the left column represent positive significant effects and dashed arrows on the right represent negative significant effects.See TableA1in the Appendix for statistics behind these interactions.

Fig. 3 .
Fig. 3. Relationship of age and gender on the ranking of trait 11 (sense of duty to future generations).Error bars are 0.95 confidence limits.

Fig. 4 .
Fig. 4. Significant differences for BA/BS degree holders compared with PhDs for individual traits on the left, and differences for three different employment sectors relative to academia on the right.Solid arrows represent positive significant effects and dashed arrows represent negative significant effects.See TableA2in the Appendix for statistics behind these effects.

Fig. 5 .
Fig. 5. Significant differences for ecologists as members of domains of inquiry groups.Results from the total pool are on the left; results from ecologists employed in academia on the right.Solid arrows represent positive significant effects and dashed arrows represent negative significant effects.See TableA3in the Appendix for statistics behind these effects.

Fig. 7 .
Fig. 7. Significant effects for ecologists as members of greatest professional satisfaction groups on the traits scores.Left panel, total pools of respondents; right panel, academic respondents.Solid arrows are positive effects; dashed arrows are negative effects.See TableA5in the Appendix for statistics behind these effects.

Table 1 .
A synopsis of main results for the first six questions of the set describing the respondent pool of 1815 complete responses.mean were in the 50-59 year class.Only 2.3% were in the 20-29 class, 3.9% in the 70-79 class, and 1.2% in the 80 and above class.Classes ranging from 30 to 69 composed 92.6% of the pool.Gender Males composed 67.4% and females 32.6% of the pool.No one selected ''other.''Gender and age are inversely correlated, with females composing 76% of the small 20-29 age class, 18.8% of the 60-69 age class and 6.4% of the small 70-79 age class.
Two choices were permitted but not always taken.In ascending order: other, 5.1% of cases; program management, 8.2%; oral communication of findings, 11.8%; classroom teaching, 15.5%; design and implementation of experiments, 17.7%; working with individual students, 22.1%; written communication of findings, 23%; data analysis and synthesis, 37.7%; field work, 50.2%.Note that the two teaching choices applied only to those respondents employed in academia.